25 years of Landscape Ecology: Scientific Principles in Practice

Part 1, Plenary sessions, Theme 1-5

Proceedings of the 7th IALE World Congress

8 – 12 July Wageningen, The Netherlands
July, 2007

Editors: R.G.H. Bunce, R.H.G. Jongman, L. Hojas and S. Weel

ISBN 978-90-78514-02-2

This publication should be cited as follows:

Bunce, R.G.H., Jongman, R.H.G., Hojas L. and Weel S. (Eds) 2007. 25 years Landscape
Ecology: Scientific Principles in Practice. Proceedings of the 7th IALE World Congress 8 – 12
July Wageningen, The Netherlands, IALE Publication series 4

Published by IALE,
This publication has been made possible with support of the Dutch Ministry of Agriculture,
Nature and Food Quality, The Dutch Ministry of Spatial Planning and Environment, Alterra,
Wageningen UR, ArgoBecas (Spain)
This publication is available from IALE:http://www.landscape-ecology.org
Price: € 40 for members including handling and mailing
Price: € 50 for non members including handling and mailing

Cover drawing: Bob Bunce

The cover drawing is a model landscape designed to show the Congress themes as follows:
Theme 1: The landscape contains policy issues e.g. wind power , urbanization and tourism.
Theme 2: Urban areas, industrial development and motorways are all present.
Theme 3: Agriculture dominates in the centre of the drawing with the mtotrway causing
fragmentation Urban development and tourism are also involved.
Theme 4: There is a large lake with a river flowing into it.
Theme 5:Monitoring is needed to follow changes taking place.
Theme 6: The mountains have tourist pressures and High Nature Value land.
Theme 7: Many changes are taking place e.g. the construction of a ski resort.
Theme 8: The power plant and the wind farm show the impacts of global energy demand.
Theme 9: Many planning issues are included e.g. urban expansion.
Theme 10: The motorway cuts through forests on the mid-slopes.
Theme 11: Forests are being conserved in priority areas

Cover design: Silvia Weel

Printed by Ponsen and Looijen bv Wageningen, The Netherlands

No part of this publication may be produced or published in any form or by any means or
stored in a database or retrieval systems without written permission of IALE

IALE assumes no liability for any losses resulting from the use of the research results or
recommendation in this publication.

2
Content

Introduction ..........................................................................................................................17
Plenary lectures....................................................................................................................19
The role of landscape ecology in planning and decision-making. A.N. van der Zande,............... 21
Landscape Ecology and Agriculture. J. Baudry, S. L. Poggio ...................................................... 23
Landscape ecology and north-temperate forests. K.J. Kirby........................................................ 25
Landscape ecology and urban development. J. Iverson Nassauer 27
Future options of Landscape Ecology: Development and research. Xiuzhen Li ......................... 29
Theme 1. Landscape, stakeholders, land use and policy ................................................31
1.1 Symposium 1: Globalisation and the sustainability of agricultural landscapes.............33
Agricultural liberalisation, multifunctionality and the WTO: competing agendas for the future of
Europe’s farmed landscapes. C.A.Potter ..................................................................................... 33
US federal agricultural policy in the context of world trade: Linking global change to local
landscapes. J. I. Nassauer ........................................................................................................... 35
Globalisation and the local agricultural landscape: Change patterns and policy developments in
three OECD countries. J. Primdahl .............................................................................................. 36
New urban-rural relationships and the peri-urban landscape; new approaches in Dutch spatial
planning. M.C.Hidding, M.Pleijte................................................................................................... 38
Urban agro-activities in Asian mega-cities. M. Yokohari .............................................................. 40
The future role of agriculture in a differentiated countryside: example of a typology of rural areas
in Portugal. T.Pinto-Correia .......................................................................................................... 42
Local landscape consequences of macro scale policy reform. S R Swaffield ............................. 44
Vulnerability of forested landscapes and landscape changes in Slovenia. D. Hladnik, J. Pirnat. 46
1.2 Symposium 22: The Young Landscape Ecologist........................................................48
The avifauna of agricultural land mosaics: how do the structural properties and countryside
elements of productive landscapes affect bird occurrence? A. Haslem and A. F. Bennett ......... 48
Here today, gone tomorrow? Quantifying the decline in occurrence of Malleefowl (Leipoa
ocellata) in Western Australia. B.C. Parsons, J.C. Short and J.D. Roberts ................................. 50
Fragmentation and plant dispersal capacity in Dutch wetlands. H. Soomers, M.J. Wassen, P.A.
Verweij .......................................................................................................................................... 52
How are landscape and environmental factors related to forest tree species richness in a
Mediterranean context? O. Torras, A. Gil-Tena and S. Saura ..................................................... 54
A farmer typology for modelling multiple land-use change at the regional level. D. Valbuena, P.
H. Verburg, A. Bregt , A. Ligtenberg............................................................................................. 56
How landscape ecology concepts can be applied to make policy for landscape reclamation. F.
Azari-Dehkordi, D. Minai-Tehrani, N. Mashayekhi-Kerahroodi, N. Khazaee ............................... 58
1.3 Open Session 4: Biodiversity Conservation and Agriculture ........................................60
Eco-structure cartography in a Mediterranean Delta. R. Pérez Campaña, A. Matarán Ruiz, L.M.
Valenzuela Montes ....................................................................................................................... 60
Arresting woodland bird decline in Australian agricultural landscapes: potential application of the
European agri-environment model. S.E. Park, S.J. Attwood, M. Maron, S.J. Collard, K. Reardon-
Smith............................................................................................................................................. 62
Are there more biologically and economically effective ways to protect aquatic biota in
agricultural landscapes? B.R. Davies, J. Biggs , P.J. Williams , S. Maund 2, S. Thompson ......... 64
Landscape ecology at the service of the policy maker: evaluating the Swiss agri-environmental
measures. F. Herzog, W. Richner , T. Walter............................................................................... 66
The role of habitat networks in guiding integrated land-use planning in agricultural landscapes.
M. Smith, J.W. Humphrey............................................................................................................. 68
An up-to-date cost benefit analysis of English agri-environment schemes: their impact at the
landscape scale and the cost of adequate monitoring. P.D. Carey, R.F. Pywell ......................... 70
Spatial Ecology Models and Conservation Targeting in the UK. R.D.J. Catchpole ..................... 72
Can agri-environmental schemes reduce habitat isolation and enhance biodiversity? The
example of butterflies in Switzerland. S. Aviron, P. Jeanneret, B. Schüpbach, F. Herzog .......... 74
1.4 Open Session 5: Sustainability and agriculture ............................................................76
Silvoarable Agroforestry for Europe: environmental and economic performance. J.H.N. Palma,
A.R. Graves, P.J. Burgess, Y. Reisner, F. Herzog ....................................................................... 76
Shifting cultivation or agroforestry? The dilemma of selecting a livelihood model for the forest
dwellers and preserving landscape ecology in north-east Himalayas. M. Choudhury ................. 78
Large-scale biomass production and potential effects on farmland habitats and related
biodiversity. B.S. Elbersen, P. Carey............................................................................................ 80
Greenhouse planning criteria based on landscape ecology: a case study in the South of Spain.
A. Matarán Ruiz, and L.M. Valenzuela Montes, ........................................................................... 82
Multifunctionality of landscapes – rural development, landscape functions and their impact on
biodiversity. N. Guiomar, J. P. Fernandes.................................................................................... 84
Landscape configuration, vegetation condition and ecosystem services in cotton agro-
ecosystems in southern Queensland, Australia. A.P.N. House, N.A. Schellhorn, S.D. Brown,
F.J.J.A. Bianchi ............................................................................................................................. 86
Provision of non-commodities in agricultural production at farm and landscape level in the MEA-
Scope case study area Mugello (Italy). S. Uthes, C. Sattler, A. Ciancaglini, G. Piani and G.
Lombardi ....................................................................................................................................... 88
1.5 Open session 6 Biodiversity, management, policy and stakeholders...........................90
Identification of Conservation Opportunities: a study case of management and conservation of
biodiversity in rural landscapes based on field research in the Central Andes of Colombia. F. H.
Lozano-Zambrano, J. E. Mendoza , E. Jiménez, P. Caycedo-Rosales, W. Vargas .................... 90
Investigation of the impact of herbivores on a Mongolian steppe vegetation and the implications
for conservation. M. A. van Staalduinen, M. J. A. Werger............................................................ 92
Analysis of recent ecosystem transformation processes in two sectors of the buffer zone of
Colombia's Pisba National Natural Park (1955 - 2001). N.A. Ciontescu Camargo...................... 94
Conservation priorities for threatened Yatsu valley landscapes in central Japan: Evaluation
based on ecological value and vulnerability. K. Takahashi, K. Hara and K. Short....................... 96
The rural policy as a tool for the natural resource management. D. Franco ................................ 98
A computer aided, modular, spatial support tool for the systemic land-use management toward
environment conservation and socio-economic development. M.F. Falcetta, F. Attorre, C.
Corugati ...................................................................................................................................... 100
The effects of the attitude concept, public education campaigns, social discourse and socio-
demographic factors on water conservation behaviors in residential landscapes. B.J. Andersen
.................................................................................................................................................... 102
Catchment management in lowland England: an integrated, multifunctional approach. C. Stoate
.................................................................................................................................................... 104
Integrated river basin management plan: public involvement and stakeholder participation. Z.
Gulbinas...................................................................................................................................... 106
1.6 Open Session 8: Landscape and Nature Perception .................................................108
From re-discovering to dealing with nature. The Netherlands as an example. S. Kost ............ 108
Do people experience landscapes as they state they do? Stated and observed landscape
preference assessment on site. M. Sevenant, M. Antrop........................................................... 110
Integrating geographic information into scenic assessment of a catchment basin .................... 112
S. Yamashita and N. Nakamura ................................................................................................ 112
1.7 Posters .......................................................................................................................114
Improved Decision-Making for Food Security, Income and Environmental Services through
Modeling Cropping Systems and Optimizing Land Use Options: A Case Study of Taita Hills,
Kenya, D.W. Odede-Oremo......................... 114
The multifunctional character of the landscape in areas of intensive agriculture: towards a
sustainable planning strategy. J. Ruiz, G. Domon ..................................................................... 115
Understanding graziers’ decision-making in the Great Barrier Reef catchment to improve
environmental, social & economic objectives. I. Bohnet ............................................................ 116
Do woods help biological control of cereal aphids by hoverflies? F. Arrignon, A. Ouin, B. Bouyjou,
M. Deconchat, C. Monteil, J.P. Sarthou ..................................................................................... 117
Five centuries of landscape change in Colombia: a conservation perspective. A. Etter............ 118
Environmental analysis and diagnosis of recreational zones in protected areas of Madrid (Spain)
C. Rozas, B. Martín, N. Roblas, C. Muñoz & L. Jiménez ........................................................... 119

4
Easy evaluation of the abundance of carnivores through habitat use map, in a Brazilian savanna
region. M.C. Lyra-Jorge, M.C. Ribeiro, V.R. Pivello ................................................................... 120
Relationships between the naturalness of the landscape and the nature conservation status of a
particular habitat in it. E. Illyés, Z. Botta-Dukát, F. Horváth, Zs. Molnár, J. Bölöni, I. Pándi ...... 121
Adjacency analysis: a method to evaluate landscape conservation status. L. Zavattero, D.
Smiraglia, M. Marchetti & C. Blasi .............................................................................................. 122
Estimating the effect of protected lands on the development and conservation of their
surroundings. R.I. McDonald, C. Yuan-Farrell, C. Fievet, M. Moeller, P. Kareiva, D. Foster,
T.Gragson, A. Kinzig, L. Kuby, C. Redman ................................................................................ 123
Identification of sites for mitigation banks using multiple ecological and social criteria. N.Hope,
M.C. Fessey, N. Bailey, S. Thompson........................................................................................ 124
A conceptual model for restoration site selection based on a review of reserve selection
procedures. J. Imanishi, A. Imanishi, Y. Natuhara, Y. Morimoto................................................ 125
Preservation of sand scenery and native plants conservation in the coastal dunes, the landmark
of Tottori, Japan. D. Nagamatsu, S. Tominaga .......................................................................... 126
Landscape structural analysis at human scale in a Mediterranean nature park. L. Hadar, Z.
Henkin, I. Noy-Meir and H. Muklada........................................................................................... 127
Impact of agriculture on water erosion and new organization of landscape. I. Šimonides, T.
Hrnčiarová................................................................................................................................... 128
Using public inquiries in the validation of expert based landscape assessment: a case study in
Southern Portugal: Castelo de Vide. H. Menezes, T. Pinto-Correia, I. L. Ramos...................... 129
Saxon cultural landscapes: what amphibians would like most? Getting baseline data for the
management of a woodpasture. K. Öllerer, T. Hartel ................................................................. 130
Integrating livelihoods and multiple biodiversity values in landscape mosaics: a new global
initiative. P.K. Koponen , J-L. Pfund, M. van Noordwijk ............................................................. 131
Arasbaran forest ecological conditions and studying to establish forest park (case study: Ilgineh
Chay basin). M. Akbarzadeh, S. Babaei Kafaki.......................................................................... 132
Representing biodiversity: Spatial analysis of multi-taxon species composition in the
Netherlands. A. Barendregt, M. Schouten.................................................................................. 133
Indigenous landscape conception, management and conservation: lessons for bird conservation
from Mexico. L. Cabrera ............................................................................................................. 134
Soft-system approach to stakeholder participation in landscape planning & management. L.F.
Cassar, S. Morse, G.H. Griffiths ................................................................................................. 135
The Influence of Agricultural Roads and Water Channels on Farm Landscape Structure in
Taiwan. Y. C. Chen , C. P. Chang, Z. F. Wu .............................................................................. 136
Modelling the impact of agriculture policy changes on farmland birds. T.K. Gottschalk, V. Wolters
.................................................................................................................................................... 137
The West Weald Landscape Project, south east England. R.T. Howorth, T. Whitbread ........... 138
Resolving landuse conflicts through pedogeomorphic knowledge of farmers in Hararghe
highlands, Ethiopia. D. Kassahun............................................................................................... 139
Naturalistic grazing in modern British landscapes. K.J. Kirby .................................................... 140
Are traditional and new planted hedgerows part of the sustainable development of agricultural
landscapes?. A. Lotfi, F. Burel.................................................................................................... 141
The spatial plan as instrument for protection of the underground water basin. S. Hadzi Pecova,
P. Penev, K. Taleska .................................................................................................................. 142
What is the role of riparian areas set aside from production in a large-scale industrial tropical
plantation landscape in Riau, Sumatra for vegetation and primates? P. Koponen, R. Nasi, J.G.
Poulsen, M. Buitenzorgy, W. Rusmantoro.................................................................................. 143
One step back, one step forth: agricultural expansion and landscape change trajectories in
shifting cultivation system of Southern Cameroon. V. Robiglio, F. L. Sinclair............................ 144
Rationalising biodiversity conservation in dynamic ecosystems (RUBICODE project). P.A.
Harrison, R. Bugter, RUBICODE partners.................................................................................. 145
EucaLand European Culture expressed in Agricultural Landscapes. G. Pungetti, A. Kruse ..... 146
Landscape context affects the abundance & diversity of bees on annual crops in Europe. G.
Carré, B.E. Vaissière, R. Chifflet, N. Morison, R. Bommarco, S.G. Potts, S.P.M. Roberts, G.
Rodet, J. Settele, I. Steffan-Dewenter, H. Szentgyörgyi, C. Westphal, M. Woyciechowski and P.
Roche.......................................................................................................................................... 147
Landscape organization and plant biodiversity in an intensively used agricultural region of central
France. F. Di Pietro..................................................................................................................... 148
Theme 2: Urban Environment and Transport ..................................................................149

5
2.1 Symposium 2: Efects of roads and traffic on wildlife populations and landscape
functions ...........................................................................................................................151
Effects of wildlife passages on the viability of badger populations in the Netherlands. E.A. van
der Grift, J. Verboom .................................................................................................................. 151
Species level differences in the road effect zone of a major motorway for anuran populations in
Ontario, Canada. F.Eigenbrod, S.Hecnar, and L. Fahrig ........................................................... 153
Effects of traffic infrastructure on patch dynamics. H. Reck, D. Lorenzen, J. Schrautzer.......... 155
Effects of road network density on abundance and road mortality of wildlife populations. I. A.
Roedenbeck................................................................................................................................ 157
How does the configuration of road networks influence the degree to which wildlife populations
are affected by roads? J. A.G. Jaeger........................................................................................ 159
Road effects on wildcats on different spatial scales. N. Klar, M. Herrmann, S. Kramer-Schadt 161
A management dilemma – when habitat for threatened species occurs primarily along and
across roads. R. van der Ree, C. Stewart, S. Cesarini, S. McCall, A. Hahs.............................. 163
Impacts of roads and development on functional landscape connectivity for wildlife in eastern
Collier County, Florida, USA. D. J. Smith, R. F. Noss, M. B. Main............................................. 165
Methodological and epistemological constraints on the estimation of the effects of roads on
ecosystem function. S. Findlay ................................................................................................... 167
Multi-scale analysis of wildlife crossing structures effectiveness in Spain. C. Mata, I. Hervás, J.
Herranz, J.E. Malo, F. Suárez. .................................................................................................. 169
Road effect on fragmentation of forests and agricultural landscapes: examples from Italy. E.
Padoa-Schioppa, M.C. Poggesi, L. Bottoni ................................................................................ 171
Landscape variables determining movements and road crossing by ungulates. J.E. Malo, C.
Mata, F. Suárez. ......................................................................................................................... 173
Reflecting on the Big Picture for Road Ecology, Transportation, and Society. R.T.T. Forman.. 175
Missing knowledge to reconcile growing road networks to viable (meta)populations. C.F.
Jaarsma, F. van Langevelde ...................................................................................................... 177
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and
shrinking urban landscapes..............................................................................................179
Variety matters! The planners’, lawyers’ and the landscape ecologists’ view on land consumption
combined in an interdisciplinary approach. H. Wittmer, H. Nuissl, D. Haase............................. 179
Landscape impact assessment for multifunctional land uses. K. Helming, T. Kuhlman, D. M.
Wascher, M. Perez-Soba, S. Sieber, P. Tabbush, O. Dilly, H. Bach, K. Tscherning, B. König, H.
Wiggering.................................................................................................................................... 181
Monitoring Urban Sprawl in Germany - Towards a GIS-based Measurement and Assessment
Approach. S. Siedentop.............................................................................................................. 183
A green infrastructure strategy to enhance the adaptive capacity of restructuring city regions to
climate change: the case of Greater Manchester. J. Handley, S. Gill, R. Ennos, S. Pauleit...... 185
Urban restructuring in Central Eastern Europe – does the late phase of transition still affect the
environment? A. Vaishar ............................................................................................................ 187
Too much urban green? the challenge of shrinkage in cities for green space, recreational
ecosystem services and public acceptance of related policy strategies. S. Schetke, D. Haase, J.
Breuste........................................................................................................................................ 189
2.3 Symposium 4: Applying landscape ecological principles in urban environments.......191
The efficacy of the urban-rural approach in studying the ecology of human settlements.M. J.
McDonnell, A. K. Hahs............................................................................................................... 191
Beetle assemblages along urban to rural gradients: an international comparisonJ. Niemelä ... 193
The response of species to urban landscape patterns.U.M. Mörtberg....................................... 195
Urban soil sealing – key indicator for urban ecological functionality and ecological planning.J. H.
Breuste........................................................................................................................................ 197
Suburban habitats and their role in the urban-rural habitat network: Point of local invasion and
extinction? R.B. Blair .................................................................................................................. 199
Suburban business and industrial sites: an unexplored opportunity for biodiversity conservation?
R. Snep, W. Timmermans, V.Kuypers........................................................................................ 201
Biodiversity in urban habitat patches: habitat quality matters.J.P. Sadler, P.G. Angold ............ 203
Patterns of exotic species invasion in fragmented native grasslands along urban-rural gradients
in South Africa and Australia.S.S. Cilliers, N.S.G. Williams & F.J. Barnard ............................... 205
Predicting land cover change and avian community responses in rapidly urbanizing
environments.J.A. Hepinstall, M. Alberti, J. Marzluff .................................................................. 207
2.4 Workshop on current and future research in urban ecology.......................................209

6
Preservation of pastures as parts of the urban green infrastructure at Järvafältet, Stockholm: risk
of damage caused by man? C. Florgård .................................................................................... 209
Urban semi-natural vegetation and the ecological functioning of urban green space at the
landscape scale.A. Millard .......................................................................................................... 211
Can biological traits help to explain changes in plant communities along a rural-urban gradient?
J. Vallet, H. Daniel, V. Beaujouan, F. Rozé................................................................................ 213
The role of the landscape on the distribution pattern of animals along a gradient of
urbanisation.S. Croci, A. Butet, P. Clergeau ............................................................................. 215
Modelling bird species distribution and abundance in urban areas.C.H. Nilon,and P.S.
Warren . ...................................................................................................................................... 217
Spatial variation of soils and vegetation structure in urban landscapes of the USA.R.V. Pouyat,
I.D. Yesilonis, W.C Zipperer, K. Schwarz, D. Nowak ................................................................. 219
Rethinking urban waters from an ecosystem services perspective.D.G. Gledhill, P. James , D.H.
Davies ......................................................................................................................................... 221
Understanding everyday urban biodiversity – interpreting the ‘intrinsic value’ of biodiversity
through human perception and mutualism.M. Gyllin .................................................................. 223
Applicability of metapopulation theory to biodiversity conservation efforts in Greater Manchester.
A.E. Kazmierczak, P. James ...................................................................................................... 225
Modelling, assessing and monitoring urban socio-ecological systems: The new challenge of
shrinkage and perforating cities for urban green and nature conservation.D. Haase ................ 227
The New Zealand urban ecology problem and some resolutions.C.D. Meurk........................... 229
Social functionality of urban green on the example of Karachi/Pakistan.M.M. Anwar ............... 231
Wild in the city: urban lakefill becomes nature’s refuge. A case study of Tommy Thompson Park,
Toronto, Canada. K. Landman, Y. Cardoso, W. Kehm .............................................................. 232
2.5 Open session 23: urban ecology and greenspace .....................................................234
Shifting the urban landscape paradigm – the ecosystem engineering and design approach.A.
Stokman...................................................................................................................................... 234
Escaped plants from garden into fallow lands in urbanizing rural area: influence of local versus
landscape factors.A. Marco , S. Oliveau T. Dutoit, M. Deschamps-Cottin, J-F. Mauffrey, V.
Bertaudière-Montes .................................................................................................................... 236
Towards strategies for the development of urban green spaces.C. Smaniotto Costa, J. Mathey,
B. Edlich...................................................................................................................................... 238
Landscape Ecological Indicators to Evaluate Urban Greenland.W.-C. Su, C.-Y. Chang........... 240
Floristic patterns and mediterranean exurban landscapes: Relative importance of economic
geography and landscape structure in functional diversity.E. Dumas, C. Napoleone, G. Geniaux,
T. Tatoni...................................................................................................................................... 242
The use of herbaceous vegetation to promote biodiversity in urban parks.J.C. García-Albarado,
N. Dunnett................................................................................................................................... 244
Effects of land-use intensity and forest patch complexity on avian diversity in an urbanized
tropical island landscape.M. Suárez-Rubio, J.R. Thomlinson .................................................... 246
The evaluation of the vegetation component of the urban landscape of Milan.V. Ingegnoli, S.
Bresciani ..................................................................................................................................... 248
Managing feral exotic pets: decision-making process revisited.V. Servais, P. Teillac-Deschamps,
R. Lorrillière, V. Delmas, A.C. Prévot-Julliard............................................................................. 250
2.6 Open Session 24: The urban Landscape...................................................................252
The Crisis And Rehabilitation Opportunities Of Urbanization In Urban Fringe Rural
Settlements－A Case Study Of Taoyuan Region, Taiwan.M. Kuo............................................. 252
Investigating the relationship between the tree cover connectivity of the metropolitan regions
across the eastern United States and the number of breeding occurrence of the three forest bird
genera.S. Kato ............................................................................................................................ 254
The spatial relation between “urban green” and “surface sealing” in semi-urban areas.S. Meeus,
H. Gulinck ................................................................................................................................... 256
Land-cover classification of Chongqing City and its surroundings using Landsat data.Y. Zhao, M.
Tomita, K. Hara, M. Fujihara, Y. Yang, L. Da ............................................................................. 258
Linking physical landscape character with open space aesthetics.Z. Wang, J.I. Nassauer, D.
Brown.......................................................................................................................................... 260
Applying a landscape ecological approach in the analysis of prospective development for low-
rise housing in Tyumen suburbs (Tyumen region, Russia).N.R Rakhimova.............................. 262
2.7 Posters .......................................................................................................................264

7
Change analysis for planning sustainable landscape in the Urban Biosphere Reserve of Rome.C.
Blasi, M. Marta , G. Capotorti, M. Marchese............................................................................... 264
Woody biomass in semi-urban landscapes.S. Meeus, H. Gulinck ............................................. 265
Plant trait patterns: differences between urban and rural areas? S. Knapp, I. Kühn, S. Klotz... 266
Landscape change in the municipality of Rome (1954-2001).R. Frondoni, B. Mollo, C. Blasi... 267
Apomictic species, apophytes and anecophytes as elements of phytodiversity of urban-industrial
regions and their applicability as indicators.G.H. Loos............................................................... 268
Development of trees and soils of the forested area in Expo’70 Park 30 years after reclamation.
T. Sasaki, Y. Morimoto, J. Imanishi ............................................................................................ 269
Vegetation cover where there is none: the third dimension of trees over sealed surfaces and their
influence on the city environment. A. Walz................................................................................. 270
Landscape ecological study on urban green spaces in Hanoi, Vietnam. U.D Pham, N. Nakagoshi
.................................................................................................................................................... 271
Settlement and urban areas along motorways: Changes in area and composition.K. Müller, C.
Steinmeier , M. Küchler .............................................................................................................. 272
Ecological research on Riga city street greenery.G. Čekstere, A. Osvalde, O. Nikodemus ...... 273
Restoration of multifunctional forests in north-west suburban area of Milan: scenario and
proposals. E. Padoa-Schioppa, A.B. Digiovinazzo, L. Bottoni ................................................... 274
Urban influence in bird communities. A case in three neighbourhoods of Buenos Aires city,
Argentina. P.V. Perepelizin , A.M. Faggi .................................................................................... 275
Landscape-ecological conditions for the development of Bratislava. T. Hrnčiarová.................. 276
Socio-cultural influences on urban ecosystems: A case study of Tommy Thompson Park. Y.
Westerveld Cardoso, K. Landman.............................................................................................. 277
Feasibility of trail development along secondary urban watercourses in Tucson, AZ, USA. M.
Livingston, J. E. Jones, J. L. Patton ........................................................................................... 278
Development of a future scenario model of compact growth in the Charlotte, North Carolina
region of the U.S. D.H. Yankee, M.H. Mehaffy, E.R. Smith ....................................................... 279
Effects of traffic noise on acoustic communication in birds. K.M. Parris .................................... 280
Human-Nature Symbiosis -- Landscape Ecological Planning for the Renewal of Suburban Park.
N. Yen, F. C. Yu.......................................................................................................................... 281
Searching for participatory planning in the urban-rural fringe in Spain. V. Hernández-Jiménez, B.
Ocón, D. Pereira ......................................................................................................................... 282
Green space strategy (Dublin). D. Sgouros................................................................................ 283
Landscape fragmentation of urban green spaces in Hong Kong. Y. H., Tian, C. Y., Jim .......... 284
Theme 3. Ecological Networks, fragmentation and connectivity ..................................285
3.1 Symposium 5: Ecological infrastructure: theory and application ................................287
Green Infrastrucure for Cities: The Spatial Dimension. J.Ahern, S. Kato .................................. 287
Why patches of fragmented woods in the city? - Contemporary uses of satoyamas in Japanese
urban areas as a source of renewable resources. M. Yokohari, T. Terada ............................... 289
Green infrastructure: a strategic approach for landscape ecology in urban areas.P. Pellegrino291
Patch Dynamics and Urban Design. S.T.A. Pickett, B.P. McGrath, M.L. Cadenasso, J.M. Grove
.................................................................................................................................................... 293
Ecological Infrastructure and Landscape Urbanism. J. Koh....................................................... 294
Design Ahead of Time: Urban Growth Pattern Based on Ecological Infrastructure. K. Yu, D. Li
.................................................................................................................................................... 295
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow.
Does the matrix matter? Landscape structure and population genetic architecture of amphibians.
A.G. Nicieza, M. Choda, S. García, and D. Álvarez ................................................................... 297
What can landscape genetics contribute to landscape ecology?. V. H. Dale, G. M. Crutsinger, A.
Classen ....................................................................................................................................... 299
Molecular genetic testing of landscape ecological hypotheses: the dynamics of amphibian
species distributions. J. W. Arntzen............................................................................................ 301
Inferring the influence of landscape on roe deer gene flow using connectivity estimates based on
a weighted combination of several landscape features. A. Coulon, J.F. Cosson, N. Ray, A.J.M.
Hewison ...................................................................................................................................... 303
Influences of landscape features on gene flow of Martes Americana in northern Idaho, USA .. 305
T. N. Wasserman, S.A. Cushman, M.K.Schwartz ...................................................................... 305
Corridor network connectedness does not necesseraly enhance biological connectivity: the case
of the landscape genetics and pollen flows of Primula vulgaris in a hedgerow network landscape.
P. Campagne, A. Baumel, L. Affre, P. Roche, T. Tatoni ............................................................ 307

8
A comparison of current methods for linking genetic variation to landscape heterogeneity using
simulated data. N. Balkenhol, L. P. Waits .................................................................................. 309
3.3 Symposium 16: Animals on thermal landscapes........................................................311
Global climate model data for the Predictive Modeling in Seascapes: Pelagic Seabird-Habitat
associations revisited. F. Huettmann.......................................................................................... 311
Amphibian physiology and landscape permeability: Modeling thermal landscapes to predict
amphibian movements. P.E. Bartelt, R.W. Klaver, W.P. Porter, D.S. Pilliod, C.R. Peterson, A.L.
Gallant......................................................................................................................................... 313
Thermal heterogeneity and river fish distribution in a groundwater-influenced high desert
landscape. C.E. Torgersen, D.S. Bateman, D.P. Hockman-Wert, M.D. McSwain, R.E. Gresswell
.................................................................................................................................................... 315
The role of thermal landscapes in resource selection by black-tailed deer. J.G. Kie, R.T. Bowyer
.................................................................................................................................................... 317
3.4 Open Session 9: Ecological Networks .......................................................................319
A forest habitat network approach to directing native woodland improvement and expansion on
the Argyll islands, Scotland. D G Moseley, R Worrell, B Black, D Ray ...................................... 319
Landscape management approach – vital to conservation in India. P.K. Mathur ...................... 321
Designing a general plan for reserved territories based on landscape cover analysis. M.Y.
Puzachenko, I.A. Onufrenya, D.N. Kozlov, Y.G. Puzachenko ................................................... 323
3.5 Open Session 10: Landscape modelling and population ecology ..............................325
Absent or non-detected? an application of occupancy models in fragmented landscapes. A.
Mortelliti, L. Boitani ,.................................................................................................................... 325
The importance of movement behavior in setting landscape connectivity. B.J. Goodwin.......... 327
Effects of landscape complexity in agent-based population models. J. Nabe-Nielsen, R. Sibly, C.
J. Topping, M. C. Forchhammer, K. Sudharsan ......................................................................... 329
The use of Mahalanobis Typicalities to model the distribution of species with different response
curves to environmental gradients. F. Sangermano, J. R. Eastman .......................................... 331
An automated procedure for the bulk re-processing of species range polygons. J.R. Eastman, F.
Sangermano ............................................................................................................................... 333
Testing the ecological significance of gradients in habitat quality for fauna in Australian
savannas. B. Price, C. McAlpine, S. Phinn, D. Pullar and J. Ludwig ......................................... 335
Testing the performance of landscape structure variables as predictors of biodiversity: a case
study from Dadia NP, Greece. S. Schindler, V. Kati, K. Poirazidis ............................................ 337
The interest of modeling the effect of habitat loss and fragmentation on population dynamics at
different nested observation scales. J.B. Pichancourt, P. Auger, F. Burel................................. 339
Patchy responses to a patchy landscape: the spatial structure of habitat selection. S.J. Mayor,
D.C. Schneider, J.A. Schaefer, S.P. Mahoney ........................................................................... 341
3.6 Open Session 11: Landscape modelling and bird populations ..................................343
Conservation of understory birds in fragmented landscapes: relative importance of habitat cover
and configuration. A. C. Martensen & J. P. Metzger .................................................................. 343
Understory birds and forest cover: a strong association in a highly forested region in the Atlantic
forest, Brazil. C. Banks, J.P. Metzger........................................................................................ 345
Modelling the influence of landscape structure on the incidence pattern of bird species living in
fragmented Brazilian Atlantic forest. D. Boscolo, J.P. Metzger, K. Frank. ................................. 347
Airborne Laser Scanning data provide three-dimensional information for analyses of woodland
bird habitats. H. Hashimoto, J. Imanishi and Y. Morimoto ......................................................... 349
Effects of landscape patterns and resource distribution on frugivorous birds in the Central Andes
of Colombia: tracking ecological thresholds. J. E. Mendoza, G. H. Kattan, J. de Jong ............. 351
Continent-scale patterns in temporal dynamics of avian assemblages. M. Maron, P.K. Dunn and
A. Apan ....................................................................................................................................... 353
Response of two diurnal raptors, the Common Buzzard (Buteo buteo) and the Eurasian Kestrel
Falco tinnunculus), to agricultural intensity in three landscape units of Western France. N.
Michel, V. Comor, Y. Rantier, Y. Delettre, F. Burel, A. Butet ..................................................... 355
Effects of biogeographical factors on bird species sensibility to habitat fragmentation in the
Atlantic Rainforest. R. G. Pimentel, E. Hasui, M. C. Ribeiro, J. P. Metzger ............................... 357
Effects of forest landscape composition and structure on forest bird species richness in the
Mediterranean at two different spatial scales. A. Gil-Tena, O. Torras, S. Saura ....................... 359
Relative and independent effects of habitat amount, fragmentation, matrix quality, and road
density on forest bird diversity. A.C. Smith, L. Fahrig, C. Francis .............................................. 361
Birds respond to large landscapes more than to “islands”. M.A. Cunningham, D.H. Johnson .. 363

9
Landscape dynamic effecting bird diversity in an Atlantic Forest fragmented region: an evidence
for time lag response. A. Uezu and J.P. Metzger ....................................................................... 386
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect
populations .......................................................................................................................365
Landscape change and extinction debt: measuring long-term change in the mammal fauna of
forest fragments. A.F. Bennett, A. Haslem & R. Mac Nally ........................................................ 365
Pond or landscape characteristics – which is more important for common toads (Bufo bufo)? A
case study from central Romania. T. Hartel, S. Nemes, L. Demeter, K. Öllerer ........................ 367
Changes in Female Grizzly Bear (Ursus Arctos) Homerange Configuration in the Multi-use
Albertan Rocky Mountain Foothills, Canada. J. Linke, A. Pape, J. Cranston, G. McDermid, M.
Hall-Beyer, S.E. Franklin, G. B. Stenhouse................................................................................ 369
Space use in desert areas by guanaco (Lama guanicoe) and its seasonal dependence on the
most productive patches. P. Acebes, J.E. Malo, F. Suarez, C.E. Borghi, S.M. Giannoni & J. Traba
.................................................................................................................................................... 371
Small mammals in fragmented Atlantic forest landscapes - importance of patch size in
landscapes with different amounts of remaining habitat. A.A. Buen, J.P. Metzger, R. Pardini.. 373
European rabbit (Oryctolagus cuniculus L.) abundance at a regional scale: controlling factors. A.
Saldaña, G. García-Salgado and S. Rebollo.............................................................................. 375
Arthropod assemblage responses to agricultural intensification in heterogeneous landscapes –
local testing of global patterns. S.J. Attwood, M. Maron, A.P.N. House, C. Zammit.................. 377
Influence of habitat quality and landscape structure on the distribution of Maculinea teleius and
Maculinea nausithous. N. Örvössy, Á. Kőrösi, P. Batáry, L. Peregovits .................................... 379
Effect of landscape structure on Collembola communities in 30 oilseed rape fields. P. Querner,
A. Bruckner, T. Drapela, D. Moser, J. G. Zaller, T. Frank .......................................................... 381
Landscape structure and small mammal distribution – the importance of matrix quality for the
connectivity of a fragmented Atlantic forest landscape. F. Umetsu, R. Pardini, J.P. Metzger ... 382
Assessing functional connectivity by simulating migration areas for amphibians: a tool for
conservation management of landscapes. A. Janin, J.P. Léna, C. Delacourt, P. Allemand, P. Joly
.................................................................................................................................................... 384
3.8 Open Session 13: Habitat Fragmentation and Mitigation ...........................................386
Using a consistent habitat network approach to reduce fragmentation at site to national scales.
A. E. Eycott, K. Watts ................................................................................................................. 388
Combining ecological networks on different spatial scales leads to synergy. C.J. Grashof-
Bokdam, M. van Adrichem, J.M. Baveco, M. van de Berg, P. Chardon, R. Jochem, H.A.M.
Meeuwsen, P. Schippers, M. van der Veen, J. Verboom, C. Vos .............................................. 390
Habitat availability and connectivity for a focal species (jaguar) in the Yungas region, Argentina.
D. Somma, P. Perovic, R. Jongman, H. van Lier, A. Cerezo, R. van Lammeren ...................... 392
Harvesting: a keystone process in cover type evolution at landscape level? Fourteen landscape
histories in Quebec (Canada). E. Alvarez, L. Bélanger, L. Archambault and F. Raulier............ 394
At which spatiotemporal landscape scales does habitat amount explain the local abundance of
species? H. Paltto, A. Moilanen, B. Nordén ............................................................................... 396
Modelling the effects of elevation and topographical-based behaviour on inter-patch dispersal.. J.
Alderman, S.A. Hinsley............................................................................................................... 398
Importance of riparian corridors for native fauna – small mammals in forest patches, matrix
habitats and linear structures in an Atlantic forest fragmented landscape. L. Naxara, R. Pardini
.................................................................................................................................................... 400
Ecological Infrastructure: The Concept, Development and Application. H. L. Liu, D. H. Li and
X.L.Han ....................................................................................................................................... 402
Impacts of landscape fragmentation on breeding habitats of Oriental White Stork in the
Sangjiang Plain, China. Liu HongYu, Li Zhaofu ......................................................................... 404
A graph-based methodology for integrating habitat connectivity in landscape conservation
planning: application to the capercaillie in Catalonia (NE Spain). L. Pascual-Hortal and S. Saura
.................................................................................................................................................... 406
Refining the fragmentation paradigm: the over-riding influence of land-use history on present day
ecosystems in fragmented agricultural landscapes. PG Spooner, ID Lunt .............................. 408
3.9 Posters .......................................................................................................................410
Seed rain in rainforest fragments: relating seed and fragment features. F.M. Jesus, P.R. Pivello,
S,T. Meirelles, J.P.Metzger...... 410Agro-environmental structures in the countryside. Their role in
local bird movements within the Sile River Natural Park. D. Tocchetto, M. Borin...................... 412

10
Applying the bidirectional cost-distance theory to the connectivity analysis of forest bird species
richness in a fragmented landscape in the Andean mountains. J. Aubad, C. Martinez. H. Gulinck.
M. A. Rodriguez. ......................................................................................................................... 413
Direct and indirect effects of habitat heterogeneity on flower-visiting insects and associated
species interactions. T. Diekötter, K.J. Haynes, D. Mazeffa, T.O. Crist ..................................... 414
Habitat Networks in England. R.D.J. Catchpole ......................................................................... 415
Prioritizing forest restoration based on late-seral habitat connectivity. W.H. Richards .............. 416
Developing ecological networks – a functional approach. M.C. Fessey, N. Bailey.................... 417
A multiscale framework for analysis of species and population distribution and persistence in
heterogeneous landscapes. M. L. Lorini, V.G. Persson, I. Garay and J. Xavier-da-Silva ........ 418
Multiple-scale analysis of species filtering processes and their determinants. L. Blank, Y. Carmel
.................................................................................................................................................... 419
Influence of inter-habitat gap size on the dispersal pattern of Xiphorhynchus fuscus (Aves,
Passeriformes, Dendrocolaptidae) in the Brazilian fragmented Atlantic forest. D. Boscolo,
C.Candia-Gallardo, M. Awade and J.P. Metzger........................................................................ 420
The role of connectivity on species-area relationship: evidences from birds in the Atlantic
Rainforest. E. Hasui, M. C. Ribeiro, A. C. Martensen, R. G. Pimentel, J. P. Metzger................ 421
Species richness and frequency of large and medium mammals related to landscape parameters
in an agro-forested region (São Paulo State, Brazil). G. Ciocheti, L.R. Tambosi, M.C. Lyra-
Jorge, M.C. Ribeiro, V.R. Pivello. The importance of patch size, connectivity and annual variation
to understanding the effects of fragmentation on leaf litter frogs in the Atlantic Rainforest. M.
Dixo, J.P. Metzger ...................................................................................................................... 423
Predicting badger (Meles meles) sett suitability across England and Wales. T.R. Etherington. 424
Scale does matter! N.C. Brouwers, A. Newton........................................................................... 425
Distribution of large terrestrial mammals in an Atlantic Forest fragmented landscape: Does
mesopredators benefit from deforestation? K.D. Espartosa and R. Pardini .............................. 426
Effects of landscape composition and physical characteristics of the land on the biological control
of aphids N. Roullé, E. Lucas, G. Domon, J. Ruiz....................................................................... 427
Consequences of habitat availability and land-use change for the population genetic structure of
a grassland bush-cricket. S. I. J. Holzhauer, V. Wolters ............................................................ 428
Prediction of future habitat distribution for song birds under climate change in Switzerland. B.
Schröder, J. von dem Bussche, R. Revermann, H. Schmid, R. Spaar and N. Zbinden............ 429
Dispersal traits and fragmentation of the habitats. M.O. Vandewalle, L. Jönsson, T. Reitalu, M.T.
Sykes, K. Hall, H. C. Prentice, S. Lavorel and E. Garnier .......................................................... 430
NELI: A landscape index of effective habitat availability. R. Fraser, D. Pouliot, I. Olthof........... 431
Long distance dispersers and displacement kernels: effect on resource utilization. T. Lindström,
N. Håkansson, L. Westerberg, U. Wennergren .......................................................................... 432
Synchronization and noise colour: a threat to endangered species? F. Lögdberg, U. Wennergren
.................................................................................................................................................... 433
Habitat quality assessment and evaluation based on remote sensing – a case study for red kite
(Milvus milvus). A. Buschmann, C. Kleinn.................................................................................. 434
Preliminary study of ecological connectivity in the region Veneto (Italy). G. Caravello, C.Gallo 435
Habitat correlates and distribution pattern of the Tiger (Panthera tigris) in the Terai Arc
Landscape: a case for landscape ecology in surrogate conservation in India. K. Ramesh, A.J.T.
Johnsingh, Q. Qureshi, A. David, G.S. Rawat, S.P. Goyal ........................................................ 436
Setting corridor network priorities for Bogotá’s main ecological infrastructure. F. Remolina .... 437
Estimating species richness of cerrado (Brazilian savanna) carnivores by using landscape
structural metrics in predictive GARP models. L.R. Tambosi, M.C. Lyra-Jorge, M.C. Ribeiro, G.
Ciocheti, V.R. Pivello .................................................................................................................. 438
Landscape fragmentation as an indicator of coastal landscape quality: an application along the
Apulian coast (southern Italy). F. Minunno, V. Leronni, C. Tarantino, M. Mininni, P Mairota..... 439
Process-based connectivity metrics for conservation corridors. B.Rayfield, A.Fall, M.-J.Fortin 440
Seed flow from a dry calcareous grassland community into the adjacent landscape. J.C. Bolli,
H.H. Wagner, H. Bugmann, Ch. Scheidegger, P.J. Edwards..................................................... 441
The influence of temperature and irradiation on the behaviour of butterflies. H. Malinowska, A.
Cormont ...................................................................................................................................... 442
Plant species co-occurrence patterns on different spatial scales in dry, semi-natural grasslands.
T. Reitalu, H. C. Prentice, M. T. Sykes, M. Lönn, L. Jönsson, K. Hall........................................ 443
Analysis on Landscape Fragmentation in the Coastal Area of Pearl River Estuary Based on RS
and GIS. Z. F. Wu, J. Cheng, C. P. Chang and H.F Wan .......................................................... 444

11
Formation of necessary ecological conditions for sustainable socio-economic development in
Armenia. E. Shahbazyan ............................................................................................................ 445
Increased Human Management Negatively Affects Beetle (Coleoptera) Species Richness in
Swiss Cities. T. Sattler1, 2, P. Duelli3, M.K. Obrist3, F. Bontadina2, R. Arlettaz2, M. Moretti1 ...... 446
Experimental deliberate releases of genetically modified plants near NATURA 2000 sites – using
buffer zones to simplify the risk assessment. A. Benzler, U. Euler ............................................ 447
Genetic factors in metapopulation survival - introduction to a PhD-project. M.M.P. Cobben, P.
Arens, J. Verboom, M.J.M. Smulders ......................................................................................... 448
Spatial analysis of landscape patterns and their relevance for large mammal conservation in the
dry-deciduous forests of Central India. A. Paliwal, V.B. Mathur................................................. 449
Indicative map of the Pan-European Ecological Network in Western Europe. R.H.G. Jongman,
I.M. Bouwma, A. van Doorn........................................................................................................ 450
The Development of a Site Inventory for Triturus cristatus in Cheshire. J. Hollinshead, A.P. Hull
.................................................................................................................................................... 451
Landscape effects on anuran pond occupancy in an agricultural countryside: barrier-based
buffers predict distributions better than circular buffers. F. Zanini, A. Klingemann, R. Schlaepfer
and B. R. Schmidt....................................................................................................................... 452
Theme 4 Ecohydrology, water and Rivers .......................................................................453
4.1 Symposium 8: Landscapes and rivers .......................................................................455
Landscape characteristics and aquatic habitats. E.A. Steel, D. Jensen, B.E. Feist, M.B. Sheer,
I.A. Lange, A. Odle, R. Brannom, M. Danielsdottir ..................................................................... 455
Altered river landscapes in the alpine region: patterns of land use as a crucial factor for the
status of the aquatic environment. M. Poppe, S. Muhar, A. Melcher, C.Trautwein.................... 457
Management of nutrient fluxes in large river basins – the River Danube as an example. M.
Zessner ....................................................................................................................................... 459
Impacts of climate and land use change scenarios on water quality of stream. C. P. Tung, Y.J.
Chen, Y. P. Lin, N. M. Hong ...................................................................................................... 461
How do landscape scales influence riverine fish in Europe? S. Schmutz, C. Trautwein ........... 463
Landscape characteristics and Pacific salmon distribution: consistent patterns across
watersheds. B.E. Feist, E.A. Steel, G.R. Pess, D. Jenson......................................................... 465
Limiting factors controlling the spatial distribution of redband trout (Oncorhynchus mykiss
gairdneri) and their implications across a basin. L. F. Madriñán, S. White, H. W. Li and G.
Giannico...................................................................................................................................... 467
Relating fish assemblages to environmental patterns at three multistate scales. R.M. Hughes,
A.T. Herlihy, and J.C. Sifneos..................................................................................................... 469
Impact of barriers on aquatic species composition in Japan. M. Fukushima, S. Kameyama, M.
Kaneko, K. Nakao....................................................................................................................... 471
The influence of population dynamics and landscape condition on pacific salmon (Oncorhynchus
spp.) re-colonization. G.R. Pess, T. Quinn, R. Hilborn, K. Kloehn ............................................. 473
Multiscale analysis of the relationship among land use cover and streams water quality in the
Venice lagoon watershed. L. Favero, A. Zuin, E. Mattiuzzo, G. Zanetto, P. Ghetti, D. Franco . 475
Aquatic ecosystems in the core Cerrado: environmental impacts and conservation. F. L. Tejerina
Garro........................................................................................................................................... 477
Rivers and roads controlled by, and central players in, urban regions. R.T.T. Forman ............. 479
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for
catchment water management .........................................................................................481
Approaches to Link Microbial Structures and Processes at the Microscale to Ecosystem
Functioning. A. Ogram, R. Corstanje, M. Christman .................................................................. 481
Progress and challenges in demonstrating riparian buffer effects on nutrient discharges from
whole catchments. D. E. Weller, M. E. Baker, T. E. Jordan ....................................................... 483
Linkages between surface and subsurface hydrology and ecological functioning of mangrove
ecosystems in Ft. Pierce, Florida. D.F. Whigham, C.E. Stringer, M.C. Rains, J.T.A. Verhoeven,P.
Baas, P.J.M. van der Ven ........................................................................................................... 485
From the micro- to the macroscale: effects of microbial Fe(III) and sulfate reduction on element
fluxes in acidic peatlands. K. Küsel ............................................................................................ 487
Spatio dynamic modelling of the hydrological processes to investigate the functionality of
nutrient-retaining landscape features. P. Merot, P. Durand ....................................................... 489
The Operational Landscape Unit concept as a guidance for landscape restoration and water
quality enhancement. J.T.A. Verhoeven, M.B. Soons ................................................................ 491
Scaling up denitrification in heterogeneous landscapes. G. Pinay, T.P. Burt ............................ 493

12
Hotspots for denitrification and nitrous oxide emission in riparian zones. M.M. Hefting, R.N. van
den Heuvel.................................................................................................................................. 495
Learning from the past: the impact of human-induced changes to landscape (dis)connectivity on
biophysical fluxes. G. Brierley..................................................................................................... 497
4.3 Open Session 17: Ecohydrology, rivers and wetlands ...............................................499
Quantifying lost opportunities in conservation: an evaluation of current and proposed scenarios
to protect the Pantanal wetland in Brazil. R. Lourival, H. McCallum, G.Grigg , C.Arcangelo, R.
Machado, H.Possingham........................................................................................................... 499
The relationship between floodplain sedimentation and the succession of rare riverine grasslands
in the Netherlands; implications for conservation strategies. H.P. Wolfert, G.J. Maas, B.
Makaske, P.W.F.M. Hommel, B.S.J. Nijhof ................................................................................ 501
The spatial organisation of savanna landscapes within drainage networks. C. J. Cullum ......... 503
Influences of changing scale of modelling land use on simulating hydrological components in
watershed land use planning assessments. Y.P. Lin, P.J. Wu, N.M. Hong ............................... 505
Simulated annealing with landscape metrics to optimally simulate landscapes in watershed
landscape planning and management. Y.P. Lin, M.H. Chen...................................................... 507
A recent landscape evolution of the Shiyang river basin---patterns, changes and causative
factors. L. Lu, X. Li, E.S. Kang, T. Wang.................................................................................... 509
Response of ecological process after water deliveries to dry watercourse of lower reaches of the
Tarim River Basin. Chen Y.N, W.H. Li, Y.P. Chen, C.C. Xu, X.M. Hao .................................... 511
Site selection in dynamic landscapes: a probabilistic model to protect hydroseres in the Pantanal
wetland. M.Drechsler, R.Lourival, H.P. Possingham .................................................................. 513
Indicators for the linkage of forest, river, village and ocean ecosystems. Y. Natuhara, A. Imanishi,
K. Imai......................................................................................................................................... 515
4.4 Posters...................................................................................................................517
Landscape models predict distribution and assess protective status of freshwater fishes. K. M.
Mattson, P. L. Angermeier, S. D. Klopfer ................................................................................... 517
Landscape Ecology in the Pantanal, integrating science and society. R.H.G. Jongman, M. van
Eupen, B. Makaske, C.R. Padovani ........................................................................................... 518
Channel and vegetation changes in the Sharda River - conservation implications for swamp deer
in Kishanpur Wildlife Sanctuary, Uttar Pradesh, India. N. Midha, P. K. Mathur ......................... 519
From catastrophe to conservation: Protecting the pond complex of Tommelen, Belgium. J.
Hollinshead, T. de Bie, A.Hul...................................................................................................... 520
A research agenda for the Pantanal and the Upper Paraguay River basin. R.H.G. Jongman, J.M.
Leeuwestein, P. Girard ............................................................................................................... 521
Landscape scale influences on the streams habitats and biota: the riverine system of Natisone
(Italy). R. Zorza, M. Sigura, G. Oriolo, P. Bonfanti, G. Honsell .................................................. 522
Corridor effect of the spatial changes of landscape patterns in arid areas: a case of the river
corridor areas in the middle and lower reaches of Tarim River. Zhou Huarong, Xiao Duning &
Zhou Kefa ................................................................................................................................... 523
A landscape ecological study of the Ganges-Brahmaputra-Meghna delta basin, Bangladesh. T.
Byomkesh, N. Nakagoshi, M. S. Rashid..................................................................................... 524
Quick scan system analysis as a basis for ecological restoration measures. J.C.J de Hoog, M.H.
Jalink........................................................................................................................................... 525
Changing Landscape Mosaic of the Mid-Paraíba do Sul River Valley: Geo-Hydroecological
Responses to Eucalyptus Growth A.M. Sato, L.G.G. Vianna, R.C.G. Almeida, A.S. Avelar, A.L.
Coelho Netto ............................................................................................................................... 526
Analysis of the quality of the riparian forest as a bioindicator in mediterranean river basins
monitoring. S. Sànchez Mateo, M. Boada Juncà ....................................................................... 527
Anthropogenic interference on fish assemblages, Meia Ponte River, Goiás, Brazil. M. Pazete de
Oliveira, F. L. Tejerina-Garro ...................................................................................................... 528
Interaction between land use and fish assemblage in the River João Leite basin, Goiás, Brazil.
A.M. Dias, F.L. Tejerina-Garro.................................................................................................... 529
Composition and spatial distribution of the fish assemblage in the tributaries of the Ribeirão João
Leite sub-basin, Goiás, Brazil. A. P. Fialho, L. C. Gomes, F. L. Tejerina-Garro. ....................... 530
Reconstruction of paleoenvironment to understand involved processes on landscape
development; an important tool for long term research and management at the Valdivian urban
wetlands (Chile) Rubilar, H.; Jessel, B. Rojas, C.; Ramírez, C. ................................................. 531
The application of an artificial wetland on the treatment of river water. D.F. Juang, P.C. Chen,
Y.C. Lee ...................................................................................................................................... 532

13
A human-modified ecological hotspot: Jabbul salt lake (Syria). F. Turkelboom, Z. Masri, A. Saeed
, B. Kasmo .................................................................................................................................. 533
Influence of groundwater system on landscape ecology: a case study in Beijing, China. P. Yun,
N. Nakagoshi, G. Huili, X. Zaiping .............................................................................................. 534
Theme 5: Monitoring at the Landscape scale..................................................................535
5.1 Symposium 10: Monitoring at the landscape scale ....................................................537
Monitoring patterns of land cover change in cultural Irish landscapes using landscape metrics
and SPOT 5 imagery. S.J.P. McKenzie, A. Cooper, T. McCann................................................ 537
The NILS programme, monitoring the Swedish landscapes for biodiversity assessment. Allard, A,
Esseen, P-A, Glimskär, A, Ståhl, G, Sundquist, S and Inghe, O................................................ 539
Small-scale extraction of non-timber forest products in developing countries – a framework for
ecological monitoring. A. Fröde .................................................................................................. 541
Monitoring landscape changes in the Netherlands. A.J.M. Koomen, W. Nieuwenhuizen, D.J.
Brus, L.J. Keunen, G.J. Maas, T.N.M. van der Maat, T.J. Weijschede. ..................................... 543
Why is strategic conservation monitoring so rare in Europe? R.G.H. Bunce, R.H.G. Jongman 545
Indicators to Monitor changes and conservation in Natura 2000 sites: A focus on drivers,
pressures and states. S. Luque, J. Pino, C. Basnou, R. Swetnam, C.A Mücher, G. Hazeu,
Halada L. , F. Gerard .................................................................................................................. 547
Towards a standardized biodiversity assessment approach in topographically complex
landscapes. G. Hofer, F. Herzog, H.H. Wagner, R.G.H. Bunce, P.J. Edwards ......................... 549
Landscape monitoring of spain based on air photo interpretation of a stratified network of land
samples. R. Elena-Rosselló, J. Manuel García del Barrio, F. Bolaños, V. Gómez-Sanz, M.
Ortega, C. García-Feced, S. González ...................................................................................... 551
Understanding the causes and consequences of landscape change: lessons from the British
Countryside Survey monitoring programme. S.M. Smart, S. Petit, D.C. Howard, L.C. Maskell,
L.R. Norton, R.C. Stuart, R.G.H. Bunce .................................................................................... 552
A landscape integrated analysis and classification: application to a river valley in central Spain.
A. Patrono, A. Saldaña López, A. Gómez Sal ............................................................................ 554
Optimising thematic resolution of landscape metrics for biodiversity monitoring in rural Europe
D. Bailey, R. Billeter, S. Aviron, M. Parracchini, O. Schweiger and F. Herzog........................... 556
The use of species compositional dissimilarity to assess the effectiveness of ecological
classification schemes. B.E. Lawson, S. Ferrier, G. Wardell-Johnson, D.V. Pullar, R.J.S. Beeton
.................................................................................................................................................... 558
Large scale land cover classification systems – a pragmatical appraisal. C. S. Cruz, J. P.
Fernandes, N. Guiomar, T. Baptista, M. J. Mateus, ................................................................... 560
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping ........562
European Landscape Typologies as a Reference Base for Data on Cultural Heritage and
Traditional Knowledge Systems in the Mediterranean. T. Kizos ................................................ 562
From Landscape Character Assessment to Sustainable Design – Tools for Planning and Policy
D. Wascher, P. Smeets, W. Timmermans, V. Kuypers .............................................................. 564
Risk and vulnerability of landscape identity. É Konkoly-Gyuró, S Jombach .............................. 566
Comparative review of European national and international landscape classifications with regard
to future applications, G. Groom, D. Wascher, S. Mücher ........................................................ 568
The role of landscape science in higher education in Europe: the ATLAS experience. G.B.M.
Pedroli, H. Palang, M. Bogers .................................................................................................... 570
Defining and mapping the landscapes of Italy. C. Blasi, D. Guida, V. Siervo, M. Paolanti, L
Michetti, G. Capotorti, D. Smiraglia ............................................................................................ 572
Developing a science-based approach for delineating natural heritage system in Southern
Ontario. D. Puric-Mladenovic, S. Strobl...................................................................................... 574
Habitat opportunity mapping at the landscape scale in the UK G.H. Griffiths, I.N. Vogiatzakis. 576
The assessment of landscape diversity for the purposes of spatial organisation of tourism: the
case study of Brest region. I. I. Schastnaya ............................................................................... 578
Landscape characterisation in Belgium: integration of different scale levels and analysing
temporal differences. V. Van Eetvelde, M. Antrop...................................................................... 580
Open Session 2: Landscape modelling and earth observation ........................................582
Problem of characteristic space scale of landscape processes. A.V.Khoroshev, G.M.
Aleshchenko, K.A.Merekalova.................................................................................................... 582
Using between-patch boundaries parameters for conservation status assessment on coastal
dune ecosystems. M.L. Carranza, S. Feola, A. Acosta, A. Stanisci........................................... 584
Landscape-dynamical analysis: studies in the North-Western Russia. G. A. Isachenko .......... 586

14
Spatial modelling of landscape patterns derived from land use and land cover changes. J. P.
Fernandes, N. Guiomar .............................................................................................................. 588
The application of satellite imagery to identify landscape structure. C.A. Mücher, C.C. Vos, C.
Renetzeder, T. Wrbka, M. Kiers, M. van Eupen and R. Bugter................................................. 590
Identification of geosystem formative factors on the basis of field and remote sensing data. D.N.
Kozlov, M.Yu.Puzachenko, M.V. Fediaeva, Yu.G.Puzachenko ................................................. 592
Integrated Remote Sensing Monitoring for Environmental Changes of a Solid Mineral Deposit
Area (Diamond Deposit. Arkhangelsk District. Russia). T.V.Orlov............................................. 594
Landscape mapping: does the scale lie within? F. Tolle, J.-C. Foltête ...................................... 596
Investigation of spatial resolution in remote sensing of fragmented environments for ecological
models. A.M. Lechner, S.D. Jones, S.A. Bekessy...................................................................... 598
5.4 Open Session 15: Landscape metrics and geostatistics ............................................600
Ecological Aspects of Landscape Fragmentation Measurements. P. Csorba ........................... 600
Determining patch-level metrics in naturally disturbed and managed forest landscapes:
implications for enhancing ecological attributes and processes. R. Lafortezza, R.C. Corry , R.D.
Brown, G. Sanesi........................................................................................................................ 602
Mathematical models of landscape patterns, A.S Viktorov ........................................................ 604
Statistical challenges integrating Landsat time series data with forest inventory data for
characterizing forest disturbance and regrowth in the U.S. G.G. Moisen, S.P. Healey, R.E.
Kennedy, S.L. Powell, W.B. Cohen, S.N. Goward, J.G. Masek, C. Huang................................ 606
Display of the basic functional properties of landscape cover on the basis of the remote
information for maintenance of initial landscape planning stages. A.N. Krenke, Y.G.Puzachenko
.................................................................................................................................................... 608
The landscape disparity index: an ecologically weighted measure of the landscape diversity. P.
K. Roche ..................................................................................................................................... 609
Are scaling functions for landscape pattern metrics really accurate and useful? S. Saura and S.
Castro ......................................................................................................................................... 611
AMOEBA algorithm to identify natural clusters of mixed forest in Poland using CLC2000 data. T.
Edman......................................................................................................................................... 613
How important is the third dimension for the analysis of spatial patterns of landscape structure?
U. Walz, S. Hoechstetter, N.X. Thinh ......................................................................................... 615
Approaches to revealing landscape spatial pattern based on analysis of landscape between-
component relations. K.A. Merekalova....................................................................................... 617
Use of a bioclimatic classification for a large-scale application of a biogeochemistry model. A.D.
Shkaruba, V.V. Kireyeu, I.P. Usava............................................................................................ 619
5.5 Posters .......................................................................................................................621
How to recognize a landscape boundary? – Finding appropriate scale for the ecosystem
management –N. Tokiwa, J. Morimoto, F. Nakamura ................................................................ 621
Using graphs and graph theory to describe and analyze connectivity in linear habitat networks
A.Jellema, P.F.M. Opdam, W.A.H. Rossing ............................................................................... 622
ATLAS-Action for training in Land Use and Sustainability, B. Pedroli, M.Bogers ...................... 623
Landscape character assessment using region growing in GIS. A. Jellema, D.J. Stobbelaar,
J.C.J. Groot, W.A.H. Rossing .................................................................................................... 624
The use of remote sensing data for modeling differing levels of grassland improvement within the
Welsh landscape. J. Breyer, K. Medcalf..................................................................................... 625
Application of System Approach towards Small-Scale Mapping and Classification of
Geographical Landscapes for the Purposes of Melioration and Agriculture. A.A. Nikiforova .... 626
Biophysical surveys integration and multifunctional gis management for running sustainability-
conscious agro-based projects of the millennium challenge account programs. E. Amamoo-
Otchere, Foster Mensah ............................................................................................................. 627
Structural thresholds at landscapes, across different habitat amount and aggregation levels. M.C.
Ribeiro, J.P. Metzger, A.C. Martensen ....................................................................................... 628
Incorporating Local Landscape Knowledge into the Cultural Heritage Mapping Process. E.A.
Moylan ........................................................................................................................................ 629
Identification of Functional Landscapes in Catalonia (NE Spain). A case-study for the High
Pyrenees Natural Park. J.M. Serra, M. Ninyerola, J. Cristóbal .................................................. 630
Application of BioHab in the Israeli mediterranean zone: a test of the habitat sampling method L.
Olsvig-Whittaker, M. Walczak, D. Rotem, Y. Magal, S. Amir, A. Perevolotsky, I. Tauber, R. Einav,
A. Zahavi, Y. Carmel, M. Sternberg, R. Frumkin, T. Achiron-Frumkin ....................................... 631
Landscape Types of the Czech Republic. D. Romportl, T. Chuman, Z. Lipský ......................... 632

15
Habitat Guide of the Landscape Ecological Habitat Mapping of Hungary. J. Bölöni, A. Kun, Z.
Molnar, E. Illyés .......................................................................................................................... 633
Programme for planned biodiversity studies: ecologically robust and cost-effective site-based
ecological data for multi-purpose research and management. J-M Hero, W.E. Magnusson , B.E.
Lawson........................................................................................................................................ 634
Evaluation of the vulnerability to floods on the state of Tabasco, Mexico. L. Gama, C. Zequeira-
Larios, L. Giddings, M. Soto-Esparza, A. Galindo-Alcantara and A. Hernadez-Moralez ........... 635
Evaluation of farm woodlands as agri-environmental schemes in rural landscapes. M. Sigura, A.
Marelli ......................................................................................................................................... 636
Characterisation of the rural landscape. Application in a case study of the plain of North Eastern
Italy. M. Sigura, A. Treleani, P. Bonfanti..................................................................................... 637
Analysis of spatial structure of landscape cover classes of sub-basin Balsas, south of Maranhão
state, Brazil. L. Barreto, M.C. Ribeiro, E. Pontes, A. Veldkamp................................................. 638
Assessment of a land change model using a three-dimensional matrix at multiple scales. S.
Peethambaram, .G. Pontius Jr ................................................................................................... 639
Delineation of landscape units at diverse scales using moving windows for heterogeneity
analysis. E. Diaz-Varela, R. Crecente-Maseda .......................................................................... 640
Prediction of soil properties in paddy rice landscapes using terrain data and satellite information
as indicators. K. Sumfleth, R. Duttmann..................................................................................... 641
Object-based classification of rural landscapes using remotely sensed data of various resolutions
K. Hara, N. Kamagata, T. Ishitsuka, M. Tomita .......................................................................... 642
Stewardship and monitoring of conservation lands. P. August, J. Coit, D. Gregg, R. Friday .... 643
Multi-scale landscape pattern analysis for h-resolution imagery of tropical dry forest: integration
of object-oriented approach and topographic data. J.A. Gallardo-Cruz and J.A. Meave ........... 644
Evaluation of landscape characteristics derived from digital elevation models. M. V. Ravibabu, P.
Negi, K. Jain................................................................................................................................ 645
Using conditional transit cost to identify multiple dispersal routes: an example from the Brazilian
Atlantic forest. N. Pinto ............................................................................................................... 646
Classification of spatial unites for the forested Korean mountainous landscape using DEM:
Comparison of GIS analyzed data with landscape and memory of people. J. Kwon, J.H. Oh, J.H.
Shin, Y.K. Kim, C.Y. Park ........................................................................................................... 647
Decision tree and vegetation indices assisting land use classification in a heterogeneous
landscape, a case study in Brazil. S. Weel, M.E. Schaepman, J.G.P.W. Clevers ..................... 648
Classification of the landscape of Huelva (Andalusia, Spain) using multivariate methods. J.
Alcántara Manzanares, J. M. Muñoz Álvarez, J. Quijada Muñoz, J. M. Moreira Madueño........ 649
Environmental Impact Monitoring Based on Ecological Carrying Capacity. W. C. Su, C. Y.
Chang, K. C. Sung...................................................................................................................... 650
Satellite remote sensing for monitoring ecological integrity of Canada’s national parks D. Pouliot,
I. Olthof, R. Fraser, A. Clouston, S. Wang, W. Chen, J. Poitevin, D. McLennan, J. Kerr, M.
Sawada ....................................................................................................................................... 651
Author list 650

16
Introduction
The present volume contains the abstracts of the 7th International Association of Landscape
Ecology (IALE) World Congress. The Congress celebrates the first 25 years of IALE, which
was founded in a divided world by western and central European participants in the
Symposium on Landscape Ecological Problems in Piestany, Slovakia in 1982. IALE initially
developed in North America and Europe but has now extended to all continents and to many
countries, as is evident from the approximately fifty national representatives in the present
volume. IALE continues to expand and new regions are being set up, for example in 2005
and 2006 in Brazil and Argentina
Two other books will be launched at the Congress, one by Landscape Europe and one by
the Dutch landscape ecological organization (WLO). The present volumes serve a different
purpose, which is to bring together in one document the full range of topics covered by
modern landscape ecology and to serve as a guide to the oral presentations and posters of
the Congress. They will also be useful after the Congress as a source of reference material
on subjects and authors..
We have done our best in the time available to get the book into a consistent format but
inevitably with such a large number of abstracts the standards are not as high as in a
scientific journal or report. However, we considered it to be more important to include all the
abstracts that were sent in rather than being selective. We hope you find them as useful as
previous collections of Congress abstracts
The Congress of is being held in Wageningen and Ede, The Netherlands and represents a
great opportunity for landscape scientists from the whole world to share their ideas. This
book of abstracts presents an overview of the work that will be presented at the congress.
The Congress focuses on the scientific principles of landscape ecology and their practical
applications to conservation, land and water management, as well as their relationship with
land use planning, both now, and in the future. Many abstracts show the increasing
importance of remote sensing and Geographic Information Systems, spatial statistics and
modeling. The Congress also demonstrates the way in which landscape ecology is playing
an increasingly important role in spatial planning for landscape and biodiversity objectives.
Landscape ecology can be defined as the holistic understanding of the relationships between
ecological components of landscapes, including the impact of human activities through
planning and management. The underlying motivation for many landscape ecologists is that
their research should lead to potential applications for social benefit. Within landscape
ecology there is also a strong recognition of the role of man in the functioning of many
landscapes as well as in their past development.
The Scientific Committee invited IALE members to propose Symposia and 25 were selected
from those who reacted. These Symposia form the core of the Congress and the organisers
accepted the challenge to coordinate programmes in their field. Many additional papers and
posters were also forwarded to the Organising Committee. All of these contributions were
accepted, although many needed editing before being included in the book of abstracts. The
Organising Committee first examined whether the contributions could be assigned to
existing symposia, but then looked for common topics and scheduled these to
complementary Open Sessions. In addition about 250 posters were also sent in, which were
subsequently linked to eleven overarching themes, which form the basic structure of the
Congress programme. The abstracts of all contributions are included in this book , with two
pages for oral presentations and one page for posters. In most cases the full posters are also
included in the CD-ROM in the back of the book. The eleventh theme consists of posters
describing Landscape oriented EU-Life projects.

The Congress programme has been structured as far as possible to ensure that related
Symposia and Open Sessions are not in conflict. Participants can therefore choose to listen
to presentations with minimal conflicts of interest, although inevitably some problems will
occur. The Workshops have also been integrated within the programme, although at first,
because of the number of participants, this was not possible. As the posters are linked to the

17
themes, they will be set up under these headings in sections of the hall in the Congress
venue so that participants can see which posters they wish to visit.

The Congress and the book of abstracts have been made possible with internal support
from the Environmental Science Group of Wageningen UR, The Dutch Ministry of
Agriculture, Nature and Food Quality, the Dutch Ministry of Environment and Spatial
Planning, The Dutch Royal Academy of Sciences, Argos Becas in Salamanca Spain, CTA,
the Wageningen Research Schools PE&RC and WIMEK-SENSE, ESF-Eurodiversity, the
province of Drente, the Dutch State Forestry and the Dutch Society for Nature Conservation
(Natuurmonumenten). The town of Wageningen will also give a reception at the start of the
Congress. Finally, the Wageningen Congress Organization Bureau (OBW, Yvonne van
Hezik and Ingrid Luitse-Looijen) not only helped with the organization for two years before
the Congress but also arranged the social programme.

Bob Bunce, Rob Jongman, Lorena Hojas and Silvia Weel

Wageningen
12-5-2007

18
Plenary lectures

Plenary lectures

19
Plenary lectures

20
Plenary lectures

The role of landscape ecology in planning and decision-making

A.N. van der Zande,

E-mail: [email protected]

Land Use Planning Groups, Wageningen University, The Netherlands

Since the realisation of the Ecological Network started in 1990, ecology is a part of decision-
making in the Dutch legislation. But since, problems to solve in society are becoming more
complex. The complexity partly originates by EU and international legislation, the water
directive and Kyoto protocol, but also by interference of economic and social problems. The
development in agriculture influences the development of the rural area and the possibilities
of the Ecological Network and Natura2000. The recent knowledge about climatic changes is
adding to that.
Development in landscape ecology is not following this development (van der Zande, 2007).
There are different reasons that development in landscape ecology and complex problems in
society are not developing parallel.
There is not one simple factor to declare the differences in development, scientists,
policymakers and society, all are taking their share.
When we analyse where (landscape) ecology is adding to policymaking or to public
discussions, we come to a variety of experiences.
The Netherlands Environmental Assessment Agency reports annually about the state of
nature and environment in the Netherlands. Usually this results in some public or political
awareness, but has not yet resulted in strong political actions or changes in research
questions.
These reports are usually descriptive and building on existing knowledge and usually doesn’t
have breaking news. Reports influencing political discussions usually originate from
international background; Millennium assessment and the IPPC reports on climatic changes
are examples.
In recent years universities and research institutes are able to add in discussions on very
specific topics in decision making on national level. Examples in the Netherlands are for
instance policy making on NH4 in the environment of Natura2000 areas. Localisation of
Natura 2000 areas. Landscape ecology is used in a descriptive way or on very specialised
issues like NH4 models.
More fundamental research is usually done in a European or international context, for
instance ecology in relation to climate changes. Public and political discussions following
usually years after scientists discovered the phenomena and therefore do not often add
directly in national decision-making.

The most innovative development in landscape ecology is happening in community of

practise while implanting the national legislation on the local and regional level. An example
is given
Community of practise
To compensate environmental problems subsidies is granted for owners to restore
ecosystems. To help owners expert groups are active per ecosystem type. These expert
groups include scientists and also representatives of nature management organisations,
ngo’s and water boards and policy makers. Selection of people is based on knowledge of
the specific ecosystem of the nature reserve where experiments take place. Scientists are
form different background, depending of the ecosystem ecologists, hydrologist, soil scientists
etc. These expert groups recognize actual field problems more quickly and are to innovate
new management methods. In the past 15 years, dozen of nature sites have been restored
and more than hundred rare of even long lost species have returned to the treated sites.

21
Plenary lectures

Recently are comparable expert group is established for meadow birds. Scientists, farmers,
nature management organisations, ngo’s, local policymakers are included in this expert
groups.
So developed knowledge moves from the expert group into the organisations from which
people were part of the group. Spreading knowledge by “own “people works relatively easy,
but needs attention.

More innovation of landscape ecology in the near future is necessary. Implementation of

Natura 2000 and of the Water Directive combined with climatically changes in a heavily
populated and industrial society, as the Netherlands needs all the creativity and innovation
we can think of. We have to learn from history how innovation can occur. Many innovations
develop in a non-institutional environment, where unexpected disciplines and persons meet
each other. Community of practise is an example of a non-institutional instrument, proven to
be a working instrument but not necessary the only one.
For the future an important target is to realise a better communication between the problems
in the society, the world of policymakers and scientists. Beside communication, a targeted
process, where using adequate knowledge during the process of decision-making is
absolutely necessary.
To realise a more knowledge intensive and more innovative policymaking, a number of
things have to change.
Scientific institutions should reward researcher for research applications, not only for
scientific status.
Working with real expectations on both sites. For policy makers, complex problems cannot
be solved by science in a couple of month.
Scientists have to accept that society defines the problems with they should solve, not single
issue but with all the different disciplines necessary. A common developed research agenda
could be one of the instruments.
Learning from the community of practise, more flexible networks on specific topics should be
created to help the national and international policymakers.

22
Plenary lectures

Landscape Ecology and Agriculture

J. Baudry1, S. L. Poggio2
1
INRA (National Institute for Agronomic Research), SAD Paysage, Rennes, France.
e-mail: [email protected]
2
Facultad de Agronomía, Universidad de Buenos Aires, Av. San Martín 4453, (C1417DSE)
Ciudad de Buenos Aires, República Argentina

Introduction
Agriculture is a major factor driving landscape changes at a global scale. It has been
so for millennia, centuries and even shorter periods in some regions .Contrasting historical
backgrounds mean that landscapes have different structures and ecological characteristics.
Currently, changes in the ecology of agricultural landscapes are progressively driven by
global factors such as international commerce and technological advances. We propose to
explore those dynamics at different scales taking the Rolling Pampas of Argentina and the
Atlantic bocage of Western Europe as case studies. These two areas are not only currently
linked by the flow of agricultural commodities from Argentina to Europe, but also by the
cultural heritage that built Argentinean agriculture and rural landscapes due to European
immigration during the XIXth century. They also offer strong contrasts in terms of agricultural
history and vegetation before the arrival of agriculture. The Pampas were strictly grassland
vegetation without trees, while forest was the dominant land cover in Western Europe.

A diversity of farming and landscapes situations

In the Atlantic bocage, farms are diverse, even though they are mostly oriented
toward stock farming. The amount of permanent grassland and hedgerow density is an
important difference as is the amount of fertilization, which increases with the amount of
arable fields and decreases with more hedgerows. Farm size varies from 20 to 100 ha.
The farms of the rolling Pampas have become increasingly specialised in the
production of crops during the 1990’s. Formerly farms were more diverse, including cattle
and more crop species. Nowadays few mixed farms remain and are mainly restricted to the
lowlands. Landscapes are characterised by fields ranging from 30 to 100 ha, which are
surrounded by a complex network of linear elements. Furthermore, abandoned small patches
near settlements and the outskirts of small towns and cities, are also characteristic features
of Pampean landscapes.

The driving forces

In the 1950’s, the pressure in some regions (e.g. the bocage of Brittany) came from
the willingness of young farmers to stay on farms. This created a major pressure on the land
and led to more intensive land use involving higher stocking rats and fertilization). Starting in
the 1960’s the European Common Agricultural Policy encouraged farmers to produce more
so that the Community could become self-sufficient. Many farmers left to provide labour to
industries, allowing farm enlargement and further increases in productivity. The WTO is
becoming a major agent trying to regulate the world commerce of agricultural commodities
and to change agricultural policies toward more liberalization and less public subsidies in
Europe. In contrast agriculture in Argentina is under a free market regime with high tax rates
on the export of agricultural commodities.
Agricultural and environmental policies have been, and still are, important tools in
shaping agricultural systems and the landscapes they produce. Labour is becoming a scarce
resource in Western Europe and now influences the maintenance of landscape elements
such as hedgerows and field margins In Argentina, unlike Western Europe, there was an
important migration of people from rural areas to urban areas during the 1940’s and 1950’s.
An important process of agricultural expansion started during the 1970’s, in order to increase
foreign currency income through the export of grain crop commodities.

23
Plenary lectures

The pressures
Similar pressures prevail in Argentina and Western Europe, with the three dominant
types being 1) changes in land cover (increasing area of arable land at the expense of
permanent grassland; fewer crops), 2) changes in landscape structure (hedgerow removal,
filed enlargement) and 3) changes in land management (higher fertilization and usage of
pesticides; less varieties of a given crop). The use of herbicides has also been extended
from fields to field boundaries and road verges that were formerly refuge habitats for species.

The ecological impacts

The ecological impacts have been extensive in Europe, with losses of biodiversity, at
least in certain taxa such as birds and butterflies at the European scale with other declines at
the landscape scale. Nowadays, habitat fragmentation is caused by practices such as
pesticide as much as land cover change. Eutrophication of terrestrial and aquatic habitats is
also a major cause of biodiversity loss. Mineral fertilization is now however a minor impact.
The main cause is organic nitrogen coming from pig and cattle slurry; these animals being
largely fed in some areas by imported feedstuffs e.g. from Argentina..
Landscape homogenisation and habitat loss are the most important ecological
impacts of the agricultural intensification in the Pampas, negatively affecting associated
biodiversity. Field enlargement, fencerow removal, and the permanent replacement of
pasture land by crops, has reduced remnant habitats for native flora and fauna. Furthermore,
agriculture intensification has changed the seasonal dynamics of ground-cover and
productivity affecting food and shelter availability for wildlife.

The policy responses

Policy responses are very different in the EU and Argentina, because of the historical
factors referred to above. The EU has a long tradition of subsidizing agriculture, starting in
the 1960’s and 1970’s to increase production and the efficiency of the sector. Since the
1990’s a minor proportion of the subsidies has been shifted toward agri-environmental
schemes aimed at reducing pressures on the land and restoring landscapes.
Since the exportation of agricultural commodities is the most important foreign income
of Argentina, policies are strongly oriented to increase agricultural productivity promoting
crop breeding and the adoption of new technologies. Unlike the EU agri-environmental
schemes, Argentina lacks of clear policies aiming to reduce the negative impact of
agriculture on land and to restore rural landscapes.

The role of landscape ecology

The development of landscape ecology, often originating from landscape design and
management questions, has been essential for understanding the causes and mechanisms
of ecological change in agricultural landscapes. The issues of habitat fragmentation,
diminishing landscape permeability, interactions between crops and their surroundings were
all made possible by the emergence of the concepts and methods of landscape ecology. The
discipline is also very helpful as it strongly emphasizes the linkages between social and
ecological systems and has strong spatial and temporal components.
There are still many themes to develop. In Europe pastoral, bocage landscapes are more
studied than cereal prairies; the later harbour different types of species, some of which are
very valuable. The complementarity of landscapes at national and continental scales is not
well understood. In addition there is the compatibility of landscape requirements for different
types of species as well as a further understanding of the relationships between biodiversity
and water quality. The intention should be to progress towards a more comprehensive
understanding of landscape functions, in both ecological and socio-technical terms.

24
Plenary lectures

Landscape ecology and north-temperate forests

K.J. Kirby

Natural England, Northminster House, Peterborough, PE1 1UA, UK,.

e-mail: [email protected]

Introduction

Anton Chekov in his play Uncle Vanya (c1889) refers to the clearance of Russian forest
and the consequent loss of species. Worldwide, forest cover continues to decline, mainly
through conversion to agricultural land. However, in the north-temperate zone the decline is
less and in Europe forest cover is increasing (FAO, 2006). Locally the scale of reforestation
can be substantial: in the Eastern United States forestry cover had dropped to about 25%
cover in the 19th century, but grew back so it is now 60-90% as farms were abandoned by
settlers who moved west (Foster and Aber, 2004). In Europe alpine meadows are scrubbing
up as traditional grazing is abandoned. In Britain forest cover has risen from about 5%
c1900 to 12% (Forestry Commission, 2002); the National Forest (c520km2 in the middle of
England) has gone from 6% to 17% forest since 1991 (www.nationalforest.org).

Changes to the pattern of patches as used by different species do not necessarily reflect
those of the forest cover as a whole. Some species are confined to coniferous forest, some
to broadleaved; within broadleaved forest some plants are restricted to acid soils, some to
base-rich ones; within base-rich broadleaved woodland some invertebrates are associated
with old decaying trees, others with the ground flora of canopy gaps.

Forests are often associated with stability, but for the species in them conditions may be
far from stable. A patch of ground moves from being an open gap, through a dense thicket
stage, to being a patch of mature or over-mature trees. For a species associated with a
particular stage, eg a butterfly of open glades, the number and distribution of suitable
patches change over time. The classic metapopulation model, with stochastic loss from
suitable patches, needs to be super-imposed on a shifting suite of patches.

To what extent however are forest species limited by fragmentation in practice (Bailey,
2006)? Forest species differ considerably in their colonisation abilities, from large herbivores
and birds that may migrate over thousands of kilometres each year to woodland plants that
appear to spread at only a metre or two a year (eg Brunet and Oheimb, 1998). The most
sedentary species pose a particular challenge: if rates are so low how did the species spread
back through Europe after the last glaciation: are we missing the rare long-distance
dispersal, is a dispersal agent missing, or have species dispersal characteristics changed?

Drivers of change in temperate forests and their surroundings

Within the cultural landscapes and forests that characterise much of Europe, and parts of
the temperate zone elsewhere, changes in forest management practice are leading to
changes in the extent and distribution of different forest age classes (patch types). For
example shifts from coppice management to high forest systems lead to a reduction in
proportion of open space (Hopkins and Kirby, in press).

Across much of the temperate zone rising deer populations have led to changes in the
structure and composition of forests. Increased movement of deer between forest patches
may help disperse some seeds, but for other species the changes lead to reduced
populations and hence increased risk of stochastic extinction events and reduced dispersal.

25
Plenary lectures

Changes in landscape diversity through removal of other semi-natural habitats may

increase the effective isolation of forest species populations through reduced landscape
permeability. Major developments such as roads create additional barriers to species
movement. On farmland use of pesticides and fertilizers has increased, leading to impacts
on the vegetation at woodland edges. More general eutrophication and acidification effects
have been detected in woods, presumably from the effects of atmospheric deposition
(Thimonier et al., 1994). However forest nutrient levels may be increasing because fewer
nutrients are removed, for example as wood-fuel (forest biomass in Europe is increasing)
(Hofmeister et al., 2004).

Climate change effects have been found in woods (Kirby et al. 2005) and models
produced on potential future species distributions. There are however uncertainties as to
how individual species responses will be affected by interactions with other species through
competition, predation, expansion of alien species, changes in herbivory etc.

Policy and practice responses

Managing to maintain species richness within existing sites may be the most effective
short-term response to enhance the future spread of forest species through a landscape.
Policies and incentives can be developed to promote traditional-type management (in cultural
landscapes), control of high deer populations, and buffering of small woods against the
effects of adjacent agricultural operations.

Agri-environment schemes and forest grants can be used to encourage the maintenance
and creation of features that improve the permeability of the landscape to forest-species
movement. Opportunities should be taken to incorporate trees and woodland associated
with urban and transport developments, as part of sustainable drainage systems, in public
access land, and as part of greenways (Jongman and Pungetti 2004).

We must also recognise that through climate change our forests will alter – we cannot
keep all as it is, but we should aim to develop a forest landscape of at least equal value.

References
Bailey, S. (in press). Increasing connectivity in fragmented landscapes: an investigation of evidence
for biodiversity gain in woodlands. Forest Ecology and Management
Brunet, J. & Oheimb, G.Von (1998) Migration of vascular plants to secondary woodlands in southern
Sweden. Journal of Ecology 86: 429-438.
FAO (2006) Global forest resources assessment 2005. FAO, Rome.
Forestry Commission (2002) National inventory of woodland and trees: Great Britain. Forestry
Commission, Edinburgh.
Foster, D.R. & Aber, J.D. (2004) Forests in time. Yale University Press, New Haven and London.
Hofmeister, J., Mihaljevic, M. & Hosek, J. (2004) The spread of ash (Fraxinus excelsior) in some
European oak forests: an effect of nitrogen deposition or successional change. Forest Ecology and
Management 203: 35-47.
Hopkins, J.H. &Kirby, K.J. (in press) Ecological change in British broadleaved woodland since
1947. Ibis .
Jongman, R. & Pungetti, G. (2004) Ecological networks and greenways. Cambridge University
Press, Cambridge.
Kirby, K.J., Smart, S.M., Black, H.I.J., Bunce, R.G.H.,Corney,P.M.& Smithers,R.J. (2005) Long
term ecological change in British woodland (1971-2001). Peterborough: English Nature.
Thimonier, A., Dupouey, J.L., Bost, F. & Becker, M. (1994) Simultaneous eutrophication and
acidification of a forest in north east France. New Phytologist 126: 533-539.

26
Plenary lectures

Landscape ecology and urban development

J. Iverson Nassauer

e-mail: [email protected]
School of Natural Resources and Environment, University of Michigan, USA.

While each of the plenary themes addresses landscapes that are affected by human
activities, the theme of urban development puts us squarely in the realm of culture and
intentional landscape change. Furthermore, this theme confronts us with the challenging
global trend of an increasingly urban population -standing at more than 50 per cent living in
urban areas today. In this urbanized world of culturally constructed landscapes, landscape
ecology offers key insights for urban development by:
• reminding us of fundamental material realities that exist with or without culture:
biogeochemical processes, ecological functions, and spatial patterns,
• investigating these material realities in the context of cultural phenomena, and
• demonstrating the interdisciplinarity that has been recognized as essential for
understanding urban ecosystems.

I frame this review of by characterizing several advances in our understanding of landscape

ecology and urban development. These include:
• similarities and differences between the landscape ecology of urban landscapes and
other landscape types,
• spatial patterns of urban development and related biogeochemical and ecological
processes of urban and peri-urban landscapes,
• ecological functions and services of urban landscapes,
• cultural interpretations of and human effects upon these patterns, functions, and
services.
The underlying assumption of my framework is that cultural interpretation of landscapes both
initiates and limits certain spatial patterns, disturbance regimes and dynamics of landscape
change, and that, conversely, these patterns, regimes, and dynamics affect the quality of
biogeochemical and ecological processes and ecological services.

27
Plenary lectures

28
Plenary lectures

Future options of Landscape Ecology: Development and research

Xiuzhen Li 1,2
1
Institute of Applied Ecology, Chinese Academy of Sciences, P.O.Box 417, Shenyang
110016, China. 2State Key Laboratory of the Estuarine and Coastal Research, East China
Normal University, Shanghai 200062, China.
e-mail: [email protected]

The last 25 years have seen the rapid progression of landscape ecology from Europe
to the whole world. One of the driving forces of this development is the need of scientific
knowledge to guide human activities towards a sustainable future. In countries where spatial
heterogeneity is fundamentally important, landscape ecology can play an active role with its
integrated ideas derived from a variety of theoretical backgrounds.
The top 10 hot topics of Landscape Ecology were addressed by Wu and Hobbs (2002), with
Turner (2005) also describing the present “scientific status” of the discipline. How about the
future development and research fields of Landscape Ecology?
It is impossible for us to ignore the past and start our future research completely
independently. Some of the points will still be valid in the future, for example, “Integrating
humans and their activities into Landscape Ecology”, and “Optimization of landscape
pattern”. As Turner (2005) pointed out, Landscape Ecology should continue to push the limits
of understanding of the reciprocal interactions between spatial patterns and ecological
processes and seek opportunities to test the generality of its concepts across systems and
scales. However, there should be new focuses in the future, corresponding to the new needs
of human being, as well as the theoretical and methodological development of Landscape
Ecology itself. In this contribution I will only be able to outline a few ideas:
1) Marine Space Ecology: in comparison with the predominately terrestrially oriented
landscape ecological studies, phytoplankton and its associated habitats in the marine
space have yet to be investigated, due to difficulties in data collection and the
uncertainty of the study object. Whether or not the methods used for terrestrial
landscape ecology can be transferred into the marine environment is also under
consideration. However, the need for large scale ecological studies in marine or
seascape systems is still increasing, due to the ever-expanding human demand for
resources.
2) Large scale ecological response of the nival zone to global changes: at high altitudes
and latitudes, global climate change and ubiquitous human disturbance have caused
major changes to the landscape and its functions. The surface layers are very sensitive
to changes in temperature, soil moisture, and physical stress. Vegetation productivity,
species composition, decomposition process, and hydrological process will also change
correspondingly. Since population density is relatively low, the ecological studies that
have been conducted in these regions are fewer in comparison with temperate zones.
3) Multi-regional Ecology: bird-flu around the world reminds people that different regions
can be affected by the same diseases due to migration of the avi-fauna. The habitats
for birds in different regions need to be considered, both for scientific reasons and for
disease control. Similar problems have been introduced by the intensification of
international communication among different regions and populations: for example
SARS, exotic species invasion, and the extinction of endemic species. More attention
and emphasis should be paid to multi-regional ecological relationships.
4) Disaster prevention and control: with rapid economic growth, the annual average world
economic losses associated with natural disasters has increased exponentially since
1960s (UNDP, 2004). The new millennium even sees ever more devastating disasters
such as the Tsunami in South-east Asia (December, 2004), and the Hurricane Katrina
in the US (August, 2005). Landscape Ecology can play an active role in planning
development choices for disaster risk reduction, as well as for post-disaster
reconstruction.

29
Plenary lectures

5) Natural resource management and bio-protection: although Landscape Ecology has

played an active role in providing theoretical and methodological ideas for natural
resource management, much can still be done in this field of research with newly
developed techniques. For example, remotely sensed data with increased resolution
and improved transmitter sensors can greatly help us to track animals and strategic
habitats. The quality and quantity of natural habitats can also be measured more
precisely.
6) Integration of visual landscapes and ecological landscapes: evaluation of the visual
aesthetic quality of the landscape provides a new perspective for the discipline, which
complements to its traditional ecological research. We might need more knowledge
about aesthetics to develop this line of research. So far pioneering studies have been
done with the help of representative photographs, at different scales and periods, in
combination with remotely sensed images and aerial photos.
7) The role of human beings in landscape evolution: including construction of buildings,
roads, farms; management of pastures and forest. The different types and intensities of
human activities have resulted in the multiple directions and complex intensities of
flows, from population to materials and energy as well as between source and sink
landscapes. The landscape as the habitat of human beings, or Homo sapiens, is often
less well understood than the habitat of many wild animal species.
As the study of “ecology at the human scale” and a transdisciplinary science, Landscape
Ecology should continue to serve as the “decision support system” for sustainable
development of the human race. Throughout the world, landscapes are being altered more
rapidly, more extensively, and more profoundly than at any point in human history (IALE
mission statement). Landscape ecologists should play a more active role in comprehensive
land use planning and the development of sound land use policies in the future.

References
Bastian, O. (2001) "Landscape Ecology: Towards a unified discipline?" Landscape Ecol. 16: 757-766
Tanner, J.E. (2006) “Landscape ecology of interactions between seagrass and mobile epifauna: The
matrix matters” Estuarine, Coastal and Shelf Science 68(3-4): 404-412.
Tress, G., Tress, B., Fry, G. (2005) "Clarifying integrative research concepts in landscape ecology"
Landscape Ecol. 20: 479-493
Turner, M.G. (2005) “Landscape ecology in North America: Past, present, and future: Landscape
ecology” Ecology 86(8): 1967-1974.
Wu, J., Hobbs, R. (2002) "Key issues and research priorities in landscape ecology: An idiosyncratic
synthesis" Landscape Ecol. 17: 355-365

30
Theme 1. Landscape, stakeholders, land use and policy

Theme 1. Landscape, stakeholders, land use and policy

31
Theme 1. Landscape, stakeholders, land use and policy

32
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

1.1 Symposium 1: Globalisation and the sustainability of agricultural

landscapes

Agricultural liberalisation, multifunctionality and the WTO: competing agendas for

the future of Europe’s farmed landscapes

C.A.Potter
Imperial College London – Centre for Environmental Policy, London, United Kingdom.
e-mail: [email protected]

Introduction

The international policy context is becoming increasingly important in any discussion of

farmland sustainability and landscape change (Potter, 2006). An emerging infrastructure of
trade rules, subsidy codes and dispute procedures means that the governance of farming
activities through domestic public policy is more and more being decided in forums like the
World Trade Organisation (WTO). The Agreement on Agriculture which concluded the
Uruguay trade round already sets boundaries on what national governments can do in terms
of offering subsidies to farmers and we have seen changes to the pattern of farm support in
the European Union (EU) that have been a direct response to the politics of the more recent
round of negotiations under the still to be concluded Doha trade round. This paper argues
that we can expect to see a much more fundamental reform of agricultural support in the
years ahead as signatories to any eventual Doha agreement seek to reduce the trade
distorting effects of farm policy and strengthen the principle of market rule in agriculture. The
implications of this liberalisation of agricultural markets for the pattern of farming and for
landscape change are likely to be profound.

Competing policy models

A specific concern of recent debates has been how best to ensure the continued provision
of public environmental goods in a much more neoliberal context. The paper suggests that
there are two competing policy models on offer here. The first, aligned with the concept of
‘multifunctionality’, is generally pessimistic that sufficient farmers or levels of farming activity
would be possible at world market prices to sustain the sort of managed countryside that
European citizens appear to value and wish to see continue. A key assumption here is that
nature conservation and landscape goals need to be achieved as a by-product of agricultural
production. The policy implication is that large numbers of farmers need to be retained on the
land in order to continue producing an essentially managed countryside. Trading partners,
however, see multifunctionality as an excuse for continued high (and trade distorting) levels
of agricultural support. The difficulty for advocates of a multifunctional model is being able to
separate their interests in public goods from the rent seeking motivations of a farm lobby
anxious to defend their traditional policy entitlements.
An alternative model is that of the ‘public goods’ approach to sustainable farmland
management. This shifts the focus away from farmers and their occupancy of land towards
the environmental outputs that need to be achieved. Critically, landscape outcomes are
viewed as potentially separable from the activity of farming and thus capable of being paid
for and secured under a (presumably very extensive) system of agri-environmental contracts
and private agreements that pay by results and outcomes. Largely in line with the decoupled
model favoured by the WTO, such arguments have the support of many of the EU’s trading
partners. However, questions remain about the delivery mechanisms available and the
degree to which these can be made fit for purpose (Dobbs and Pretty, 2005). Administrative
costs tend to be high and there are difficulties in targeting and delivering geographically
differentiated outcomes. At the same time there is concern about the sustainability of a set of
landscapes that would increasingly depend on government schemes for their existence and
construction. Far better, and more authentic, say the multifunctionalists, to retain the idea of

33
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

farmed landscapes as working environments, underpinned and sustained by the activities of

the individuals and families seeking to gain a livelihood there.
The paper concludes by suggesting that, rather than being a debate about farmland
sustainability as such, the essential focus of recent international discussion in the context of
Doha has been how far the multifunctionality of European agriculture could be put at risk by a
long term restructuring of farming due to increased world market exposure. For many, this
debate has a causal emphasis that the often rather vaguely prescribed notion of
sustainability lacks. It draws attention to the difficulty of reconciling globalising tendencies
with local attributes and processes and has offered something of a counter to the dominant
narrative of market rule as a governing discourse. The question of how far governments have
a responsibility to sustain the countryside by protecting the incomes of farmers is still one
that remains unresolved after over a decade of contestation and debate.

References
Dobbs, T and Pretty, J (2005) Agri-Environmental Stewardship Schemes and Multifunctionality,
Review of Agricultural Economics, 26, 2, 220-237
Potter, C (2006) Competing narratives for the future of European agriculture: the agri-environmental
consequences of neoliberalization in the context of the Doha round, The Geographical Journal,
172, 190-196

34
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

US federal agricultural policy in the context of world trade: Linking global change
to local landscapes

J. I. Nassauer

School of Natural Resources and Environment, University of Michigan, USA.

e-mail: [email protected]

American agricultural policy is being set in a new, more explicitly global context by world
trade policy, particularly negotiations occurring within the World Trade Organization (WTO),
which is pushing governments of the developed world away from subsidizing the production
of commodities or protecting markets but not necessarily requiring withdrawl of other forms of
farm income support. Interestingly, this global context may create opportunities for federal
policy to support American farm income by promoting a wide array of ecologically beneficial
local landscape changes in land cover and management. In America, global trade policy is
linked to possibilities for local landscape changes through federal agricultural policy.

As new American federal agricultural policy is considered over the next decade, American
policymakers are unlikely to discontinue massive federal subsidies to privately owned and
operated farms (over the past decade, an approximate average of US $20 billion/year), even
if global trade policy presses them to do so. The structure of the United States Senate gives
great power to relatively sparsely populated agricultural regions which benefit enormously
from these subsidies. However, policymakers may find it politically viable to replace some
existing commodity and production related payments with payments for environmental
improvements to agricultural landscapes. Changing the target of subsidies to explicitly
supporting farmer incomes as well as environmental outcomes rather than supporting farmer
incomes through payments related to commodity production could allow policymakers to
continue payments to agricultural regions in a way that benefits the broader tax-paying public
by addressing pressing increasingly well-known environmental degradation from existing
agricultural practices (e.g., hypoxia of coastal waters, contamination of local wells for potable
water, drawn-down of regional aquifers, dramatic reduction of wildlife habitat).

In addition, some new agricultural enterprises (e.g., production of biofuels) could affect
environmental degradation, either increasing loss of habitat and downstream water quality
degradation or enhancing the quality of these environmental benefits. As new agricultural
practices, enterprises, and land uses (like reserves and wetlands) are implemented in the
agricultural landscape matrix, policy directing resulting landscape patterns could aim to also
assure that new agricultural landscapes are immediately recognized as valuable by the
broader American public – providing even more tangible rewards for subsidies. Such an
approach could be developed by adapting conservation policies and practices that are
already familiar to American farmers and policy makers – from the reserve lands approach of
the highly successful Conservation Reserve Program (CRP) to the working lands approach
begun under the Conservation Security Program (CSP) under the current farm law. By
providing agricultural subsidies for a broad array of environmental benefits, American
agricultural policy could address real, pressing environmental problems, continue agricultural
subsidies in a way that may be more broadly acceptable to the American electorate, and, at
the same time, conform with expectations of global trading partners to move away from
commodity subsidies.

35
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

Globalisation and the local agricultural landscape: Change patterns and policy
developments in three OECD countries

J. Primdahl

Danish Centre for Forest, Landscape and Planning, University of Copenhagen, DK-1958
Frederiksberg C.
e-mail: [email protected]

Introduction
The local agricultural landscape is increasingly being affected by central policy decisions
and worldwide events – through the opening of markets, through technological
developments, through changing urban-rural relationships, and through changes in public
policy interventions. Two international policy agendas are affecting agricultural landscapes –
the open market agenda institutionalised by WTO and the sustainability agenda
institutionalised through the UN’s program for sustainable development. Although the first
one is based on centralised decisions taken at national and international level as opposed to
the sustainability agenda which is filtered down to the regional and local levels – the two
agendas eventually meet in the local landscape. And often they do not meet in any
‘symmetrical’ or ‘balanced’ way.
The decisions related to agricultural production for the market taken by the farmers are
decisions connected to more or less closely to global networks in “space of flows” whereas
the decisions made for the local landscape as a living place – as a “space of place” – are not
any more closely linked to production. It is argued that the functionality of the agricultural
landscape and the changes in landscape patterns are increasingly being affected by the way
the two policy agendas are integrated (Figure 1).

Case studies
Based on case studies in New Zealand, Portugal and Denmark current change patterns
are analysed in relation to the two agendas and the two spatial dimensions, ‘space of flows’
and ‘space of place’. New Zealand and Denmark represent two countries both with highly
developed agricultural sectors closely linked to world food markets, whereas Portugal and
Denmark represent two contrasting EU member states affected by the same “common
agricultural policy” unlike New Zealand which more or less abandoned a subsidy based
agricultural policy more than 20 years ago. In each of the three countries two case study
areas has been selected - one with relatively good conditions for arable and livestock
farming, the other with more poor conditions and a more unstable landscape history. The
role of productivist agriculture, counter urbanisation, tourism and public policy interventions
including nature agricultural policy, environmental policy, nature conservation and physical
planning are analysed and dominating driving forces and overall change patterns are
identified. At the level of concrete landscape decisions, the different roles of “producer’s”
versus “property owner’s” is analysed by use of detailed data on recent landscape changes
and the farmer’s perception of the changes occurring in the local landscape.
Similarities and contrasts between and within the countries and the landscape types
are discussed in the framework of globalisation and sustainable development. Finally, some
implications for multifunctional landscape research is presented and ways towards more
integrated and collaborative landscape policies are proposed.

36
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

Figure 1. Two international policy agendas affecting the local agricultural landscape, the
WTO’s open market agenda and the UN’s sustainable development agenda

37
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

New urban-rural relationships and the peri-urban landscape; new approaches in

Dutch spatial planning.

M.C.Hidding1, M.Pleijte2
1
Land Use Planning Group, Wageningen UR, Wageningen, The Netherlands.
e-mail: [email protected]
2
Alterra, Wageningen UR, Wageningen, The Netherlands.

Planning in the face of growing complexity

Urban-rural relationships have been radically changed in the last few decades. The idea that
town and country are to be considered as ‘separate entities’ (Hidding et al. 2000) loses its
validity more and more. Instead, the interwovenness of both is stressed. This is also reflected
in the domain of spatial planning. Here, the growing interwovenness of town and country
goes together with the rise of new policy discourses, institutional settings and planning
strategies. Table 1 summarizes some characteristics of spatial planning for town and country
as ‘separate entities’ and as a ‘complex whole’ respectively.

Table 1. Characteristics of spatial planning for town and country as ‘separate entities’ and as
a ‘complex whole’
Town and country as Town and country as a ‘complex
‘separate entities’ whole’
Leading policy Town and country as Regionally differentiated policies,
discourse for town and separate entities (compact according to the situation and
country cities and green open potentials of the area
spaces)
Leading plan concepts Stressing ‘place’ Stressing the position and
characteristics potentials of the area in the ‘space
of flows’ (Castells, 1995)
Institutional settings for Separate institutional settings Intermingling of institutional
planning and on local, provincial and settings for planning and
development of town national levels development of town and country
and country on the regional level
‘Normal’ practices Urban planning and Area-based policies through
development on the cooperation of actors from state,
municipal level: combination market and civil society, focussed
of ‘planning by permission’ on the realisation of ‘integrated’
and ‘facilitating urban land projects
development’
Rural planning and
development: through ‘land-
consolidation’ on area-level
Governance style Hierarchical relations Multi-actor, multi-level and multi-
between national, provincial sector governance.
and local planning levels
Dominant planning Physical engineering, in Social engineering; bringing
style accordance with an age-long together different parties with their
tradition (The Netherlands as respective assets (powers, ideas,
a ‘man-made country’) experiences, land, money, etc)

The position of agricultural landscapes

Although The Netherlands are a strongly urbanised country, the major part of the land
is used for agriculture and so, agricultural landscapes still predominate. In the postwar

38
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

period agricultural landscapes have been transformed into modern agricultural production
landscapes on a large scale, fit for a highly productive agriculture. Today, the societal claims
on agricultural landscapes go far beyond agricultural production as such, due to the
sustainability agenda and growing urban influences. The strong interwovenness of town and
country in The Netherlands is an important factor in the dynamics of these landscapes. 1.The
position of the most urbanized parts of the country in the ‘space of flows’ offers also
favourable conditions for highly industrialized agro-complexes, especially horticulture and
intensive husbandry. 2. The demand for ‘green services’ by urban dwellers is strong and
growing. 3. The claims on agricultural land in behalf of urban and infrastructural development
are large, especially in highly urbanized regions; here agricultural landscapes are a
threatened category. 4. Agricultural landscapes represent differentiated ‘price landscapes’
for urban investors. Urban developers are increasingly active on the rural land market, taking
strategic positions with regard to possible future urban developments.

New planning approaches for peri-urban areas

From the eighties of the 20th century on a new planning approach has been evolving
on a regional level, which also applies to peri-urban areas. Its main characteristics today: 1.
Area based, exceeding the borders of individual municipalities. 2. Governance oriented,
bringing actors from state, market and civil society together, each with their respective assets
(powers, knowledge, ideas, experiences, land, money, etc) 3. Oriented towards the
introduction of new or better qualities in the area, through combining red, green and blue
functions according to a coherent view. 4. Combining the assets of strategic planning and
project development. 5. Linking sectoral investments to strategic spatial plans. 6. Strongly
market-oriented.

Case studies

This planning approach is further explained through two case studies: ‘Duivenvoorde
corridor’ and ‘Groningen Meerstad’. Both cases include agricultural landscapes in peri-urban
areas. The ‘red for green strategies’ (‘red’ pays for the improvement of ‘green’ qualities) as
followed in both cases are illustrative for the market-oriented character of this planning
approach. In both cases landscape qualities – existing or to be created– were seen as a
strategy to strengthen the position of the area and the enterprises involved in the space of
flows.

Reflection

The cases illustrate that the future of agricultural landscapes in peri-urban areas
highly depends on their landscape value, as defined by urban dwellers and project
developers. Profit driven ‘red for green strategies’, often launched under the heading of
‘sustainable development’, can be harmful to existing values of agricultural landscapes. The
domination of market rationality in recent planning strategies raise questions about the
definition of the public domain in todays spatial planning, and so about the position of spatial
planning itself.

References
Hidding, M., Needham, B & Wisserhof, J. (2000) Discourses of Town and Country. Landscape and
Urban Planning 48, 121-130
Castells, M. (1995) The rise of the network society, Blackwell, Oxford.

39
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

Urban agro-activities in Asian mega-cities

M. Yokohari

Graduate School of Frontier Sciences, The University of Tokyo

Kashiwanoha, Kashiwa, Chiba, 277-8563, Japan
e-mail: [email protected]

Shrinking cities

The declining birth rate and one of the world’s most rapidly ageing populations has
meant that cities in Japan are starting to shrink for the first time in their history. There has
been an increase in the number of patches of abandoned land within these cities because
they are shrinking in terms of their populations, economies and land areas. These
abandoned patches of land are particularly prevalent in new towns, which were hurriedly
built-up during the 1960s and 1970s. New towns have long-since deteriorated into “old”
towns, populated by residents that moved there in the 30s and 40s and never moved out.
Finding a use for the abandoned urban land that lies interspersed among the tenement
blocks of the new towns is an urgent task.

Agriculture in the city

To solve the problem of increasing abandoned land, understanding the traditional

structure of the Japanese urban fringe area is necessary. During the Edo period (i.e. pre-
1868) urban fringe areas in Japan were characterized by a mixture of developed areas and
agricultural areas. Edo, former Tokyo, was not an exception. Edo was already one of the
largest cities in the world with over one million populations at the beginning of the 18th
century. Despite such enormous accumulation of population Edo was a city with agriculture.
A number of varieties in vegetable on market today are given their name by the area in Edo
where they were grown during Edo era, e.g. Komatsu-na (variety of Chinese cabbage) and
Nerima-daikon (variety of radish).
Fujii, et al. (2002) estimated the land use of Edo in 1850s by using various
documents and maps and identified that over 40% of the the land area of Edo was for
agricultural use. Another feature which characterizes the land use of Edo is the mixture of
urban and agricultural land uses in its fringe areas. By calculating the join value of
agricultural land patches Fujii, et al. (2002) estimated that the area between 4km to 6km from
the core of Edo was the area where series of small and segmented patches of agricultural
lands remaining inside residential neighborhoods were identified. The agriculture on each
area not only provided nutrition, but each area also accepted organic wastes from the
surrounding residential areas.

Ecological functions of urban agricultural areas

Today, agricultural areas in the city may fulfill a variety of modern ecological
functions. These functions could range from reducing the summer heat of surrounding
residential areas, to controlling floods and improving the aesthetic quality of an area
(Yokohari, et al, 1994). From the consumers’ point of view, urban agriculture may also have
a number of benefits. Because of food-scares in Japan, Japanese consumers are extremely
concerned about obtaining safe food. If the fruit and vegetables they consume are grown
locally consumers may have more confidence in them. These consumers may therefore be
willing to pay a premium for such food, which will support local farming in the area.
If a recycling system that accepts wastes and reduces the amount of waste each
household produces were implemented at the same time, it could help to realize the ideal of
a sustainable city. Hirohara, et al. (2002) identifies that such a recycling system may be
realized in a residential neighborhood with remaining patches of agricultural land, where

40
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

organic wastes from surrounding houses are brought to agricultural lands as a material for
organic fertilizer, and vegetables grown on the fields will be feed-backed to houses which
provided wastes.

Agro-activities by urban residents

Agricultural lands in residential neighborhoods are not only for farmers. Today, rapidly
growing demands on agricultural activities can be found among urban residents in Japanese
urban fringe areas. Many of those who had been involved in industrial and commercial
businesses and retired at around age 60 are now choosing agriculture as their second
career, resulting in new agro-activities. Namiki, et al. (2006) surveyed such new activities
practiced by urban residents in the suburbs of Tokyo and identified that the characteristics of
those activities were represented by their intermediate character in the cultivated area per
person, smaller than traditional farming but far larger than backyard gardening, and the
number of grown vegetables; less than backyard gardening but more than traditional farming.
Agriculture in Japan is sharply deteriorating. Although it has been regarded that the
revitalization of agriculture is an urgent task for the food security of Japan, raising the self-
sufficient rate of food, now less than 40%, cannot be expected only by encouraging
agriculture in a traditional manner. Agro-activities by urban residents in urban fringe areas
are still a minor portion, resulting in a very limited amount of the land area that may be
cultivated. However, by realizing the reality that the average age of farmers now on the
fields are already nearly 70 and most of their following generations are not willing to succeed
farming, we have to encourage new agro-activities by urban residents by admitting that
urban residents are indeed one of the very limited successors of Japanese agriculture.
Land use in the fringe of Asian mega-cities should not be achieved by a simple application
of Western urban planning concepts, but by an application of a fresh approach, which
regards agro-activities as a vernacular element of the area (Yokohari, et al, 2000) The
integration of agro-activities in the urban fabric is one of the key issues for bringing a new
order to the urban fringe landscapes of Asian mega-cities.

References
Fujii, M, Yokohari, M and Watanabe, T. (2002) Identification of the distribution pattern of farmlands in
Edo. City Planning Review Special Issue, 37: 931-936 (in Japanese with English abstract)
Hirohara, T, Yokohari, M, Kato, Y, and Watanabe, T. (2002) A study of the small-scaled material
cyclic between urban and rural and land uses in urban fringe areas of Tokyo. Journal of Japanese
Institute of Landscape Architecture, 65 (5): 889-892 (in Japanese with English abstract)
Namiki, R, Yokohari, M, Hoshi, T, Watanabe, T, and Amemiya, M. (2006) Agro-activities by urban
residents on the farmlands of urban promotion area. Journal of Rural Planning Association Special
Issue, 25: 269-274 (in Japanese with English abstract)
Yokohari, M., Brown, R.D, and Takeuchi, K. (1994) A framework for the conservation of rural
ecological landscapes in the urban fringe area in Japan. Landscape and Urban Planning, 29: 103 -
116
Yokohari, M, Takeuchi, K, Watanabe, T, and Yokota, S. (2000) Beyond Greenbelts and Zoning: A
new planning concept for the environment of Asian mega-cities. Landscape and Urban Planning,
47: 159-171

41
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

The future role of agriculture in a differentiated countryside: example of a

typology of rural areas in Portugal

T.Pinto-Correia

Dept.Landscape and Biophysical Planning, University of Evora, 7000 Evora, Portugal

e-mail: [email protected]

Introduction
Throughout Europe, the role of farming as the private provider of public goods and
services increasingly valuated by society is today generally acknowledged. As defined by
OECD, the joint production of commodity and non-commodity outputs is the main
characteristic of the multifunctional agriculture that corresponds to the new European model.
Nevertheless, in the turn towards rural development concerns, multifunctionality as an
attribute of rural space has emerged. As stated by C.Potter (2004), such territorial approach
is justified by the fact that farming is no longer the economic activity supporting the rural
economy, but rather the opposite, and it has the clear advantage of being able to be
exploited by a wider community of stakeholders than the farmers alone. The production
function is thus combined and integrated with other functions, displayed mostly at the
landscape level, with more or less economic value, but also with social and environmental
values (Wiggering et al 2006).
The situation is nevertheless not the same in all European regions, and even within
countries there is a clear differentiation of the countryside between various regions, which by
all means is getting strengthened in the transition towards post-productivism (Murdoch et al
2004). In some regions there is a productivist orientation and production has a dominant
economic role, while others will need to be supported on other functions to survive
economically and socially, or may be best suited to environmental functions alone. The
multifunctionality challenge is not the same in these separated types of regions. And even
within the regions which will depend on their capacity to provide non-production functions,
there are specific potentialities and limitations to be taken in consideration. The vocation of
the rural territories is different, and thus also the functions they are able to support.
In order to face the challenges raised by new demands, the management of rural
landscapes needs to be anchored in an understanding and evaluation of this differentiation.
This has to be reflected in strategies as well as in policies and related instruments.
One of the characteristics of the differentiated development of rural areas means there
are, on one side, no clear-cut relations between marginalization, abandonment and the loss
of dynamics and diversity, and, on the opposite, no clear relation between a dynamic farming
and a dynamic countryside. Therefore, a more detailed and separated analysis of each of the
processes occurring, and of their drivers and consequences, is required.

The Portuguese example

Portugal is a good example of these various processes going on, as it is naturally a
country of biophysical contrasts and diversified landscapes, and it has both very dynamic
and very fragile rural areas, areas with a strong farming sector and others with a very
traditional and non competitive agriculture. In this country, a range of recent studies is
showing an increasing differentiation in the development of rural areas, which develop at
present along diverse trajectories. Some regions are developing very dynamically, but this
can happen with or without a strong role of agriculture. Other regions are affected hard by
marginalization processes, that raise concern as they are often related to socio-economical
decay, the loss of cultural landscapes (Meeus et al., 1990), land abandonment and
decreasing biodiversity, and also physical desertification.
Separating the various components of the countryside may lead to a more clear analysis of
the processes going on. This has been applied in a study undertaken in 2006 for the Ministry
of Agriculture, in Portugal, which aimed at assessing the differentiated characteristics and
dynamics of the Portuguese rural territory.

42
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

Methodological Approach
The analysis focus separately on three dimensions: 1) the land use/land cover, based
on Corine data; 2) the farming sector and its production and structure, based on Agricultural
Statistics; and 3) the community characteristics and behaviour, based on the Population
Census. Specific indicators, concerning data from 1990 and 2000, at the level of the
municipality, have been used for a cluster analysis, for each of the three dimensions. This
cluster analysis resulted in the classification of the municipalities following their behaviour in
each dimension. Combining the three analysis, it was possible to identify different broad
vocations of the rural space, and the role that farming could have in the future for the
multifunctionality of the landscape. Accordingly, the municipalities have been grouped in
types, pre-defined as ideal types. These types consider the characteristics and trends of the
rural landscapes and the functions (commodity and non commodity) they may secure, but
also the role that farming may play, if the transition of perspectives from sectoral to territorial
is fulfilled.

Results and Discussion

This presentation will discuss the concept of multifunctionality of the rural areas, and of the
role of farming for these functions. Furthermore, it will present the possible methodological
approach towards the identification of the different types of rural areas in Europe, based on
the identification of ideal types, supported on the analysis of selected indicators.
Furthermore, the ten different types identified for Portugal, as well as their distribution by
municipalities, will be presented, and examples in some of the types will be discussed. This
examples will stress the potential future functions to be integrated at the landscape level in
each specific context.
This paper is based on what was a first approach to an understanding of the differentiation of
the rural territory in Portugal, that needs to be further developed in detail. It shows
nevertheless that not all rural areas, even those that are excluded from a possible
competitive agriculture, should be considered in a similar way in strategic guidelines and
policies, as they have different opportunities and threats in relation to the dynamics of the
community, alternative economic activities, landscape character or environmental quality. It
also shows that a territorial approach to agriculture may be the key for the maintenance of
the sector in many areas where production by itself will soon not be viable any more.

References
Murdoch J., Lowe P., Ward N. and Marsden T., 2004. The Differentiated Countryside. Studies in
Human Geography. Routledge, London.
Potter C., 2004. Multifunctionality as an agricultural and rural policy concept. In:Brouwer F. (Ed.).
Sustaining Agriculture and the Rural Environment. Governance, Policy and Multifunctionality.
Advances in Ecological Economics. Edward Elgar. Pp. 15-35.
Wiggering H., Dalchow C., Glemnitz M., Helming K., Muller K., Schultz A., Stachow U. And
Zander P., 2006. Indicators for multifunctional land use – linking socio-economic requirements
with landscape potentials. Ecological Indicators, 6: 238-249

43
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

Local landscape consequences of macro scale policy reform

S R Swaffield

Lincoln University, PO Box 84 Lincoln University, Canterbury, New Zealand

e-mail: [email protected]

Introduction

Analysis of change in a range of agricultural landscapes following macro scale policy

reform in New Zealand reveals landscape ecological consequences of trade globalization.
Using existing data from recent studies, the analysis is based upon five dimensions of
landscape structure and function - land use, land tenure, vegetation cover, water
management, and landscape management. Factors that may be critical in sustaining
agricultural landscape systems under a global free trade regime are identified.

Aotearoa- Britain of the South

New Zealand was first settled by Polynesian migrants, and became part of an
international trading system following colonization by Europeans in the early 19th century. For
the majority of the 20th century it was effectively an agricultural outpost of Britain, trading
commodities for imported manufactured goods. However, although part of a world economy,
it was not yet fully globalised in the contemporary sense of the word [Baragwanath et al
2003], and was highly regulated until the 1980s [Easton 1997].
The bio-geographical consequences of these historical migration and trading
relationships have been profound [Holland 2000]. New Zealand has a high level of
endemism, with a unique biota. Human induced fires, hunting, introduction of grazing animals
and numerous other exotic plant and animal species, combined with extensive forest
clearance to create agricultural land, have caused dramatic changes in its landscape ecology
[McGlone 1989]. Most lowland forest, wetlands and scrublands were replaced by exotic
agricultural grasslands, shelterbelts, and plantations, with consequential dramatic effect on
indigenous flora and fauna [Pawson and Brooking 2002].

The NZ Experiment in trade reform

UK entry into the European Common Market in 1972 precipitated a significant decline
in the terms of trade for NZ [Easton 1997]. When combined with the oil shocks and growing
internal political tensions, by 1984 the country was in socio-economic crisis. The response
was to liberalise the economy [Kelsey 1995]. In a few short years NZ became one of the
most open economies in the world and its society, economy and landscape have rapidly
become globalised [LeHeron and Pawson 1996].

Landscape Change

Using a regional case study, five dimensions of landscape change during this period
of globalization are analysed. The analysis is based upon different parts of the Canterbury
Region, which extends from the watershed of the Southern Alps eastward across extensive
outwash plains to the outlying volcanic remnant known as Banks Peninsula.
Removal of agricultural production subsidies and development grants in the mid
1980s was a major shock to the predominant system of dryland pastoral agriculture and has
over time lead to increasing differentiation of land use regimes. Some farms have specialized
and intensified within vertically integrated global commodity systems; others have diversified,
with many new lifestyle farms drawing income from urban sources, as the landscape has
become a focus of consumption rather than production. Liberalisation of controls on land

44
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

subdivision has enabled fragmentation of land tenure, particularly in locations suitable for
urban commuting, or in areas of high landscape quality such as the coast and hills. There
has also been amalgamation of titles in some locations, where land use economics require
larger farms, and where the City Council has purchased land for recreation, conservation and
stormwater management. Changes in tenure and land management have led to
differentiation in vegetation cover, exemplified by the removal of mature exotic shelterbelts
and plantations on the plains to create dairy pasture, and the reversion of marginal hill
country to indigenous bush. Intensification has also created conflicts over allocation of
groundwater, with intense debate over both the causal relationships involved, and the future
management options. Finally, shifts in land use and tenure have diversified landscape
management regimes. There is a growing disjuncture between urban and rural values, and
between different types of land management. In some locations, new institutions have
emerged to encourage cooperation and collaboration. They face the challenge of spatially
integrating adjoining land uses that are functionally dislocated, and linked to diverse global
markets.

Discussion

Globalization and ‘opening up’ of the NZ economy over the past two decades has
increased local landscape differentiation, both in structure and function. Reforms have
enabled increased economic efficiencies in agriculture [Johnston 2001], and stimulated
significant rationalisation and differentiation of land use. However there has been an
‘implementation gap’ in environmental management [Barnett and Pawling 2005] as local and
regional government has struggled to manage the landscape consequences of change. The
NZ experiment reveals that trade reform must be matched by enhanced local knowledge,
and improved management capacity. New institutions are needed to ‘sustain’ agricultural
landscape systems within the context of the global economy.

References
Baragwanath L; McAloon J; Perkins H. 2003. Globalisation and New Zealand: anchoring the
Leviathon in a Regional Context. New Zealand Geographer 59[2]:16-26
Barnett J; Pawling J. 2005 The environmental effects of NZ free market reform Environment
Development and Sustainability 7(2):271-289
Easton B. 1997 In Stormy Seas:The Postwar New Zealand Economy Otago University Press,
Dunedin.
Holland P. 2000 Cultural Landscapes as Bio-geographical experiment : A New Zealand Perspective.
Journal of Biogeography 27(1):39-43
Johnston RWM. 2001. Agricultural reforms and their international implications. Institute of Economic
Affairs, London.
Kelsey J. 1995.The New Zealand Experiment: A world model for structural adjustment? Auckland
University Press, Auckland.
Le Heron R; Pawson E 1996 Changing Places : New Zealand in the Nineties. Longman Paul,
Auckland.
McGlone MS. 1989 The Polynesian settlement of NZ in relation to environmental and biotic changes.
New Zealand Journal of Ecology 12:115-129
Pawson E Brooking T [eds] 2002 Environmental Histories of New Zealand Oxford University Press,
Melbourne.

45
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

Vulnerability of forested landscapes and landscape changes in Slovenia

D. Hladnik, J. Pirnat

University of Ljubljana, Biotechnical Faculty, Department of Forestry and Renewable Forest

Resources, 1000, Ljubljana, Slovenia;
e-mail: [email protected]

Slovenia is 60% covered by forest. Forests and forested landscape types are therefore
among the most significant visual characteristics of the country. Present rural landscapes are
now facing developing of intensive agriculture in suitable areas and lack of human activity,
resulting in marginalisation of agriculture and spontaneous afforestation in other areas
(Hladnik, 2005). Forestry turned into close to nature orientation in the second part of the 20th
century, yet former orientation into the forests with artificial dominance of Norway spruce is
still significant in some areas.

Teuffel has estimated (Teuffel et al., 2004) that there are 6-7 million ha of forests with
artificial dominance of Norway spruce in Europe, with 4-5 millions on sites of original
broadleaved species. Norway spruce on such sites turns out to be highly sensitive to
drought, bark-beetles, storms (Larsen, 1995) and has negatively influenced the soils.

Factors which help us understand a landscape are its structure, functioning and changes
occurring in it. For all the three factors, the relationship between natural and social factors is
of significance. The latter are the most important factors that affect the changes in the
cultural landscape. An assessment of the landscape structure has been carried out at the
karst region of Slovenia on the basis of Slovenian forest maps, data on the present land use,
an assessment of forest cover of the region at the end of the 18th century and data from the
19th century Franciscean land register.

It was estimated that natural factors mainly affected the forms of past settlement and that
humans left indelible imprints on the landscape of the region at least 200 years ago. The
physiographic factors do not result in a sharp delineation that would justify the differences
between preserved and changed forested landscapes.

A model of landscape functioning was established based on comparison of recent changes

in respect to native spruce and fir forests as opposed to forests with artificial dominance of
Norway spruce stands that have resulted from high artificial energy inputs in the past. All
artificial energy (human labour, machine work including depreciation of machinery and spare
parts needed for felling, skidding) and material inputs needed for silvicultural and protective
measures, were calculated as “artificial energy input” needed for functioning of managed
forests (Pirnat, 1995).

When the spatial model was designed, the temporal differences were assessed in the last 15
years – the period of the most significant changes in the landscape structure. The vegetation
indices were collected using satellite images Landsat TM from years 1992, 2000, 2005 to
show the areas of stable landscape structure and the areas of marginalisation. Intensive
forests with artificial dominance of Norway spruce turned out to be most energy consuming
and especially demanding in terms of silvicultural and protective measures. The result show
that forest areas with former highest artificial energy input (0.92 MJ/m2/year) turned out to be
the most unstable part of the forested landscapes in the last decade.

Within a climate change scenario approach, the possibility of extreme weather conditions is
increasing, resulting in an extreme vulnerability of changed forest stands with artificial
dominance of Norway spruce in the lowlands.

46
Theme 1. Landscape, stakeholders, land use and policy
1.1 Symposium 1: Globalisation and sustainability of agricultural landscapes

References
Hladnik, D. (2005) Spatial structure of disturbed landscapes in Slovenia. Ecological Engineering 24:
17-27.
Larsen, J.B. (1995) Ecological stability of forests and sustainable silviculture. For. Ecol. Manage. 73:
85-96.
Pirnat, J. (1999) Natural and Artificial Energy Flows in Forest and Agricultural Landscapes of
Kočevsko. Geografski zbornik 39: 29 - 50.
Teuffel, K, v. Heinrich, B. Baumgarten, M. (2004) Present distribution of secondary Norway spruce
in Europe. H. Spiecker, E.Klimo, J.P. Skovsgaard, H. Sterba. K.v. Teuffel (Eds). Norway spruce
conversion – options and consequences. European Forest Institute Research Report 18. Brill NV,
Leiden, Boston, pp. 63-96

47
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

1.2 Symposium 22: The Young Landscape Ecologist

The avifauna of agricultural land mosaics: how do the structural properties and
countryside elements of productive landscapes affect bird occurrence?

A. Haslem and A. F. Bennett

Landscape Ecology Research Group, School of Life and Environmental Sciences, Deakin
University, 221 Burwood Highway, Burwood, Victoria 3125, Australia.
e-mail: [email protected]

Introduction

The global expansion of human-dominated environments means that productive

landscapes have an important role in future biodiversity conservation. Two key features of
agricultural environments affect their potential contribution to conservation: the structural
properties of landscapes (Andrén, 1994) and the types, and heterogeneity, of countryside
elements present within the production zone (Benton et al., 2003). Yet few studies have
incorporated both considerations at the level of the entire landscape: this was our aim in a
study of bird occurrence within an agricultural region in Australia.

Methods and Results

Birds were sampled in all landscape element types within 27 agricultural land mosaics,
each 1 x 1 km (100 ha) in size. Study mosaics were selected to incorporate variation in two
factors: native vegetation cover and number of different landscape elements. Seven types of
landscape element were recognised: native vegetation, linear vegetation, non-indigenous
tree plantation, scattered paddock trees, pasture, natural wetland, and farm dam. Data
collected from 90 bird point counts conducted in each mosaic (15 fixed sampling locations x
6 survey rounds) was used to determine species richness and incidence values for each
mosaic as a whole. Each study mosaic was also classified on the basis of its structural
properties. Eight independent variables described different aspects of the extent and
diversity of landscape elements, the spatial configuration, and the landscape context (in
terms of native vegetation cover) of mosaics.
Regression analyses were used to investigate the relationship between mosaic properties
and the richness of all species, and of three habitat-association groups (woodland, open-
tolerant, open-country) of birds, in study mosaics. Overall bird richness was strongly related
to measures of mosaic diversity (number of different landscape elements/mosaic) and
landscape context (surrounding cover of native vegetation) while sub-groups of birds
responded more strongly to measures of mosaic composition (cover of preferred element
type). For example, woodland birds were positively related to the extent of native vegetation
within mosaics while open-tolerant and open-country species showed a closer relationship to
the extent of countryside elements (e.g., pasture and scattered trees) in mosaics.
Ordination analysis revealed that variation in bird assemblages across mosaics followed a
gradient in tree cover extent (native vegetation and tree plantation) within mosaics and native
vegetation extent in the surrounding landscape. To test whether this community gradient
resulted from variation in the ecological and biological characteristics of birds, all species
were classified by six life-history traits widely identified as responding to habitat
fragmentation. Species incidence in relation to three trait groups (nesting substrate, foraging
substrate and clutch size) showed significant variation along the community composition
gradient.

Discussion and Conclusions

48
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

Our results indicate that properties of mosaic composition, diversity, configuration and
broader landscape context are important influences on bird occurrence in our study region.
The relative effect of these mosaic properties on species richness varied between different
types of birds as defined by their habitat requirements. Our findings are consistent with
those of related studies: species reliant on natural vegetation are often strongly affected by
its extent and configuration within agricultural landscapes (McGarigal and McComb, 1995;
Radford et al., 2005) while broader taxonomic assemblages are more strongly related to the
heterogeneity of these systems (Böhning-Gaese, 1997; Atauri and de Lucio, 2001). We
found, as have others (e.g., Benton et al., 2003), that countryside elements make an
important contribution to the heterogeneity of agricultural landscapes.
The importance of wooded vegetation to the birds of agricultural landscapes was further
confirmed by the strong influence of tree cover extent (native vegetation and tree plantation)
on bird community composition in study mosaics. The alignment of tree plantation with
native vegetation on this tree cover gradient appears to support the hypothesis that
structurally complex countryside elements may enhance the connectivity of agricultural
landscapes (Renjifo, 2001).
Our results highlight the particularly important role of two features of agricultural
landscapes in affecting the conservation value of these systems: native vegetation cover and
the heterogeneity of landscape element types. In summary, we found that: 1) mosaic
properties affect the richness and composition of bird communities in agricultural landscapes;
2) the influence of mosaic properties on bird occurrence differs between types of birds; 3)
countryside elements make an important contribution to the heterogeneity, and bird richness,
of productive environments; and 4) individual land-owners, by preserving/enhancing wooded
vegetation cover and maximising landscape heterogeneity, can make a real contribution to
sustaining rich bird assemblages in Australian agricultural landscapes.

References
Andrén, H. (1994) Effects of habitat fragmentation on birds and mammals in landscapes with different
proportions of suitable habitat: a review. Oikos 71: 355-366.
Atauri, J. A. & de Lucio, J. V. (2001) The role of landscape structure in species richness distribution
of birds, amphibians, reptiles and lepidopterans in Mediterranean landscapes. Landscape Ecology
16: 147-159.
Benton, T. G.; Vickery, J. A. & Wilson, J. D. (2003) Farmland biodiversity: is habitat heterogeneity
the key? Trends in Ecology and Evolution 18: 182-188.
Böhning-Gaese, K. (1997) Determinants of avian species richness at different spatial scales. Journal
of Biogeography 24: 49-60.
McGarigal, K. & McComb, W. C. (1995) Relationships between landscape structure and breeding
birds in the Oregon Coast Range. Ecological Monographs 65: 235-260.
Radford, J. Q.; Bennett, A. F. & Cheers, G. J. (2005) Landscape-level thresholds of habitat cover for
woodland-dependent birds. Biological Conservation 124: 317-337.
Renjifo, L. M. (2001) Effect of natural and anthropogenic landscape matrices on the abundance of
Subandean bird species. Ecological Applications 11: 14-31.

49
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

Here today, gone tomorrow? Quantifying the decline in occurrence of Malleefowl

(Leipoa ocellata) in Western Australia.

B.C. Parsons1,3, J.C. Short2 and J.D. Roberts3

1
CSIRO Sustainable Ecosystems, Private Bag 5, Wembley WA, 6913, Australia.
email: [email protected]
2
Wildlife Research and Management Pty Ltd, Kalamunda WA, 6926, Australia.
3
School of Animal Biology M092, University of Western Australia, Crawley WA 6009,
Australia.

Introduction
The Malleefowl (Leipoa ocellata) is one of many terrestrial fauna species that has
experienced a decline in its distribution and abundance in Australia (Burbidge and McKenzie
1989; Recher and Lim 1990; Garnett and Crowley 2000). It is listed as endangered at the
national level and the national recovery plan (Benshemesh 2000) suggests a contraction of
45% in the western portion of its range. Studies in south-eastern Australia (Frith 1962,
Priddel and Wheeler 1999) have reported rapid and widespread declines of Malleefowl in
response to agricultural development and the introduction of exotic predators. These
threatening processes operate in Western Australia also and so there is concern that similar
declines may have occurred. However, previous estimates of decline have been restricted to
detailed studies of individual populations (e.g. Priddel and Wheeler 2003) or crude broad-
scale estimates based on expert opinion and anecdotal evidence (Benshemesh 2000). As a
consequence, the pattern and extent of decline of Malleefowl across Australia are not well
understood.
There is a need to more accurately quantify the decline of Malleefowl in Western Australia
and assess the relative contributions of various processes that threaten Malleefowl
persistence. Currently, a wealth of presence-only data exists for the species and despite the
recognised shortcomings of this type of data (Austin 2002; Wintle et al. 2005); it has potential
to answer key questions about the status of Malleefowl. This study addresses the following
questions:
ƒ Can we use presence-only data to quantify species decline?
ƒ Has the Malleefowl suffered a range contraction in Western Australia, and more
specifically, within the Western Australian wheatbelt?
ƒ Is there a relationship between changes in the range of Malleefowl and landscape-
scale environmental predictors?

Methods
We sought to quantify changes in the range of Malleefowl in Western Australia by basing
our methods on those used in the national recovery plan and applying them to an expanded
dataset of community collected sightings data dating from 1839 to 2006. We also collected
presence-absence data for the Western Australian wheatbelt, allowing us to more rigorously
examine decline. Areas where Malleefowl populations were apparently stable were
compared with those where numbers had declined, and various landscape attributes were
considered as possible causes of change. Issues relating to scale, bias and false absence
were also explored.

Results
Our results suggest that Malleefowl have suffered a range contraction within Western
Australia but this contraction is smaller than previously claimed. In contrast to south-eastern
Australia, Malleefowl appear to have persisted across much of their former range within the
agricultural landscapes of south-west Western Australia, with declines largely confined to the
western margin of the species’ former range. The contraction in the range of Malleefowl was
correlated with the number of years since commencement of agricultural activity and also

50
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

with land clearing. Analysis of sightings data also showed that false absences had the
potential to exert substantial influence on estimates of species decline.

Conclusion
The Malleefowl has suffered less of a range contraction in Western Australia than previously
claimed, particularly within the Western Australian wheatbelt. This conclusion is in part due
to the compilation of a larger dataset than was previously available to researchers assessing
this species. However, it is also likely that the species has persisted within large areas of the
wheatbelt because agricultural development has been relatively recent (< 50 years ago)
compared with areas in eastern Australia. As a result, these landscapes are of higher quality
in that they are less fragmented and the remaining habitat has had less exposure to agents
of degradation such as grazing, altered fire regimes and weed invasion. It is probable that
the decline of Malleefowl is likely to continue unless these processes are mitigated. The
correlation between Malleefowl decline and land clearing suggests that the future extinction
of Malleefowl from highly fragmented areas within the wheatbelt is likely, particularly given
the presence of exotic predators.
This study was limited to making broad estimates of Malleefowl decline using relatively
coarse-scale spatial data but nevertheless represents a step forward in the use of presence-
only data to assess species decline. Presence-only data can provide a useful starting point
for understanding patterns of decline, provided an appreciation of bias within the data (e.g.
false absence) is incorporated into the analysis. Collection of absence data via community
survey is an effective way to account for such bias on a broad scale, particularly when
studying a high-profile iconic species such as the Malleefowl. However, collection of spatially
and temporally structured presence-absence data across a range of environmental gradients
is necessary to provide rigorous assessments of population trends in decline over time.

References
Austin, M.P. (2002) Spatial prediction of species distribution: an interface between ecological theory
and statistical modelling. Ecological Modelling 157: 101-118.
Benshemesh, J. (2000) National Recovery Plan for Malleefowl, Department for Environment and
Heritage, Adelaide.
Burbidge, A.A. and McKenzie, N.L. (1989) Patterns in modern decline of Western Australia's
vertebrate fauna: causes and conservation implications. Biological Conservation 50: 143-198.
Frith, H.J. (1962) Conservation of the mallee fowl, Leipoa ocellata Gould (Megapodiidae). CSIRO
Wildlife Research 7: 33-49.
Garnett, S. and Crowley, G. (2000) The Action Plan for Australian Birds 2000, Environment Australia,
Canberra.
Priddel, D. and Wheeler, R. (1999) Malleefowl conservation in New South Wales: a review. R.W.R.J.
Dekker, D.N. Jones, and J. Benshemesh (Eds). Proceedings of the Third International Megapode
Symposium, Nhill, Australia, December 1997. Zoologishe Verhandelingen 327, pp. 125-141.
Priddel, D. and Wheeler, R. (2003) Nesting activity and demography of an isolated population of
Malleefowl (Leipoa ocellata). Wildlife Research 30: 451-464.
Recher, H.F. and Lim, L. (1990) A review of current ideas of the extinction, conservation and
management of Australia's terrestrial vertebrate fauna. Proceedings of the Ecological Society of
Australia 16: 287-301.
Wintle, B.A., Elith, J., and Potts, J.M. (2005) Fauna habitat modelling and mapping: a review and
case study in the Lower Hunter Central Coast Region of NSW. Austral Ecology 30: 719-738.

51
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

Fragmentation and plant dispersal capacity in Dutch wetlands

H. Soomers1, M.J. Wassen1, P.A. Verweij2

1
Environmental Sciences, Faculty of Geosciences, Utrecht University – Heidelberglaan 2,
3584 CS Utrecht, the Netherlands.
e-mail: [email protected]
2
Science, Technology and Society, Faculty of Science, Utrecht University – Heidelberglaan
2, 3584 CS Utrecht, the Netherlands.

Introduction

Wetlands fulfil important ecological functions as regulators of worldwide water regimes

and carbon-, and oxygen cycles (Best et al., 1993) and there biodiversity is often large
(Mitsch and Gosselink, 1993). In the Netherlands, plant diversity is especially large in the
groundwater dependent, nutrient poor floating fens, belonging to the community of Caricion
davallianae (Best et al., 1993). However, wetlands are declining rapidly worldwide (Finlayson
and Moser, 1991), and only 80 hectare floating fen is left in the Netherlands (Beltman et al.,
2001).
Many studies investigated the causes of deterioration of these low productive fens (e.g.
Barendregt et al., 1995; Bootsma and Wassen, 1996). However, most studies only
considered abiotic processes, ignoring the importance of seed dispersal for plant occurrence.
In fragmented habitats, such as the Dutch wetlands, dispersal capacity can become limiting,
resulting in a decreasing connectivity of the remaining subpopulations and thereby affecting
the persistence of plant species (Hanski and Gyllenberg, 1993).
We hypothesise that habitat fragmentation has negative effects on floating fen plant viability,
in addition to changes in site factors. To gain more insight in both the effect of abiotic factors
and of habitat fragmentation on floating fen species occurrence, several datasets comprising
data on site factors, plant species composition and spatial configuration of plant populations
in the Dutch Vecht river plain were subjected to a logistic regression analysis.
To gain more insight in seed dispersal and colonization capacity of the considered fen plant
species in a fragmented area, several field experiments were carried out. As soil moisture is
strongly positively related to the percentage of hydrochorous species (Ozinga et al. 2004), it
is expected that dispersal by water plays an important role in the wet fen ecosystems.
Therefore, focus in this study is on hydrochory.

Methodology

Logistic regression analysis

To gain insight to the key-factors that determine plant presence, a forward stepwise
logistic regression analyses was performed with abiotic variables and variables related to
fragmentation as independent factors and presence/absence of plant species as dependent
factor.
Abiotic variables considered in the analysis were Electro-Conductivity (EC25) pH,
groundwater table, soil temperature, nutrients (K+, NO3-, NH4+, H2PO4-, Ptot) and ions (Ca2+,
Mg2+, Na+, Cl-, HCO3-, SO42-, Fetot, Altot, Mntot, S2-, SiO2, Cutot, Pbtot and Zn2+) in the
groundwater samples. Both the first order and second order polynomials were candidate
predictors.
Patch area and perimeter, number of populations within 500 metres, distance to the
neighbouring hill ridge and height are variables related to habitat fragmentation or spatial
configuration that were used as independent variables in the analysis.
For the present study, two floating fen species (Carex diandra and Menyanthes trifoliata) and
one fen meadow species (Lychnis flos-cuculi) were selected.

52
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

Seed dispersal experiments

To assess the dispersal and colonization capacity of the considered fens species, painted
seeds were released on fens and in water bodies, and recaptured with nets in the water and
astroturf mats on the banks. Results of these experiments are not available yet, as the study
is ongoing.

Results

Logistic regression analysis

Preliminary results show a highly significant positive relation between the number of
populations within 500 meters of each sampling point and the probability of occurrence of C.
diandra and M. trifoliata. Furthermore, highly significant positive (C. diandra and M. trifoliata)
and optimum (L. flos-cuculi) relations between plant occurrence and Ca2+ content were
found. Other significant predictors for plant occurrence were the area of the suitable habitat
(L. flos-cuculi, positive relation), temperature (M. trifoliata, positive relation) water table (C.
diandra and M. trifoliata, optimum relation), Cl- content (L. flos-cuculi, positive relation) and
Altot content (L. flos-cuculi, negative relation).

Discussion

The fact that the number of populations within 500 meters of a location (C. diandra and M.
trifoliata), and the size of the habitat patch (L. flos-cuculi) are positively related to the
occurrence of the species is in line with metapopulation theory (Hanski and Gyllenberg,
1993) and indicates the negative effect of fragmentation on plant species viability. The
positive relation of occurrence of all three species with Ca2+ content indicates the
dependency of the species to discharge of calcareous groundwater. This endorses the
efforts of nature managers to restore the original groundwater flow with seepage water
discharging near the hill ridge.

References
Barendregt, A; Wassen, M.J. & Schot, P.P. (1995) Hydrological systems beyond a nature reserve,
the major problem in wetland conservation of Naardermeer (the Netherlands). Biological
conservation 72: 393-405.
Beltman, B; van den Broek, T; Barendregt, A; Bootsma, M.C. & Grootjans, A.P. (2001)
Rehabilitation of acidified and eutrophied fens in the Netherlands: effects of hydrologic
manipulation and liming. Ecological engineering 17: 21-31.
Best, E.P.H; Verhoeven, J.T.A. & Wolff, W.J. (1993) The ecology of The Netherlands wetlands:
characteristics, threats, prospects and perspectives for ecological research. Hydrobiologia 265:
305-320.
Bootsma, M.C. & Wassen, M.J. (1996) Environmental conditions and fen vegetation in three lowland
mires. Vegetatio 127: 173-189.
Finlayson, C.M. & Moser, M. (eds.) (1991) Wetlands. Facts On File Ltd., Oxford
Hanski, I. & Gyllenberg, M. (1993) 2 General Metapopulation Models And The Core-Satellite Species
Hypothesis. American Naturalist 142:17-41.
Mitsch, W.J. & Gosselink, J.G. (1993) Wetlands. Second edition. Van Nostrand Reinholt, New York.
Ozinga, W.A; Bekker, R.M.; Schaminee, J.H.J. & Van Groenendael, J.M. (2004) Dispersal
potential in plant communities depends on environmental conditions. Journal of Ecology 92: 767-
777.

53
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

How are landscape and environmental factors related to forest tree species
richness in a Mediterranean context?

O. Torras, A. Gil-Tena and S. Saura

Department of Agroforestry Engineering. ETSEA. University of Lleida, Spain.

e-mail: [email protected]

Introduction

Factors determining patterns of biodiversity have received considerable attention from

scientists in recent years. Different hypothesis have been proposed to explain latitudinal
gradients of richness; in particular, correlations between global patterns in species richness
and climate are widely known, and climatic factors as predictors of plant distributions have
been analysed by a large number of studies. However, biodiversity patterns are strongly
influenced by multiple variables, both biotic and abiotic, at multiple scales (Huston, 1994). In
this context, factors operating at the landscape level could be related with vegetation spatial
distribution at different scales, although there is a considerable lack of knowledge in this
respect, particularly in the Mediterranean areas. From a landscape ecology approach, effects
of landscape structure on abundance and distribution of organisms can be explored (Turner,
1989), and could be analysed as a factor explaining biodiversity. In this sense, forest
landscape fragmentation has been shown to be in some cases a determinant of biodiversity
loss, and shape complexity is also considered a key landscape pattern characteristic in this
respect. This study intends to provide further insights into the understanding of
Mediterranean tree species richness patterns from a landscape ecology perspective,
investigating to what degree forest landscape structure can explain forest tree species
richness distribution in a Mediterranean region at the scale of 100 km2.

Material and methods

Our case study was carried out in Catalonia (NE Spain), located within 42º53’00’’N and
40º31’23’’N and with a total extension of 32,098 km2, where 38% of the territory is covered
by forests. The main forest tree species are Pinus halepensis, Pinus sylvestris and Quercus
ilex. The dominant climate is Mediterranean, although a great contrast in altitude and a
complex relief favours climatic diversity at the micro scale.
The analysis was performed in a grid of UTM 10 x 10 km cells, considering only those
cells with at least 90% of its area falling within the territory of Catalonia and excluding those
not covered by forest area, which resulted in a total of 278 cells.
The recent Spanish Forest Map at scale 1:50,000, developed within the Third Spanish
National Forest Inventory (Ministerio de Medio Ambiente, 1997-2006), was the information
source used for obtaining the tree species richness and the rest of forest landscape variables
related to composition (forest area, forest canopy cover, forest development stage, land
cover diversity) and configuration (fragmentation, shape irregularity). Environmental variables
included topographic (elevation, slope, aspect) and climatic information (temperature,
precipitation, radiation) and were obtained from different official sources. Spatial factors were
also analysed through the geographic coordinates of the centre of each UTM 10 x 10 km cell.
We evaluated how much of the variation in species richness could be attributed
exclusively to landscape structure compared to other spatial and environmental factors
(topography and climate) through partial linear regression (Legendre and Legendre, 1998)
and the variation partitioning method proposed by Bocard et al. (1992). Then, we conducted
a new multiple linear regression by backward stepwise method focusing on the effect of
forest landscape characteristics on tree species richness, in order to explore which of them
were more relevant to explain tree species richness at the scale of study.

54
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

Results and discussion

The whole set of variables considered explained a considerably large amount of total tree
species richness variation, nearly 60%. The largest fraction (28%) resulting from the variation
partitioning corresponded to the join effects of the three types of variables, which indicates
that both environmental and forest landscape factors present similar spatially structured
patterns highly influencing species richness. Among the pure fractions, forest landscape
factors were those that most contributed to explain tree species richness (15.2%), much
more that the pure effect of environmental (2.3%) and spatial factors (1.8%). This result
highlights the importance of forest landscape variables at this scale of analysis (10 x 10 km).
The forest landscape regression model identified several factors related to forest
landscape composition (forest canopy cover (FCC), forest area and land use diversity) as
those explaining the largest amount of tree species richness variation, and revealed that a
high FCC is more related to tree species richness than the amount of forest area itself.
Composition variables were followed, with a lower importance in the regression model, by
shape metrics (minimum circle circumscribing index, area-weighted shape index and density
of shape characteristic points), agreeing with previous studies showing that shape complexity
is a potentially valuable measure of plant species richness (Moser et al., 2002, Saura and
Carballal 2004). On the contrary, no fragmentation variables were selected in the model.
We believe that these results may provide valuable guidelines for the analysis of forest
diversity and for forest landscape management in the Mediterranean. However, considering
the scale dependence of ecological patterns, we recognise that further research at other
spatial scales is also needed.

Acknowledgments

The Spanish Forest Map for Catalonia was provided by the Dirección General para la
Biodiversidad (Ministerio de Medio Ambiente, Spain). This study was funded by the
IBEPFOR project (CGL2006-00312/BOS, Ministerio de Educación y Ciencia, Spain). A. Gil-
Tena benefited from a predoctoral research grant from the Catalan Government.

References
Bocard, D; Legendre, P. & Drapeau, P. (1992) Partialling out the spatial component of ecological
variation. Ecology 73: 1045-1055.
Huston, M.A. (1994) Biological diversity: the coexistence of species on changing landscapes.
Cambridge University Press, Cambridge.
Legendre, P. & Legendre, L. (1998) Numerical ecology, 2nd edn. Elsevier, Amsterdam.
Ministerio de Medio Ambiente (1997-2006) Tercer Inventario Forestal Nacional. Dirección General
para la Biodiversidad, Madrid.
Moser, D.; Zechmeister, H.G.; Plutzar, C.; Sauberer, N.; Wrobka, T. & Grabherr, G. (2002)
Landscapes patch shape complexity as an effective measure for plant species richness in rural
landscapes. Landscape Ecology 17: 657-669.
Saura, S. & Carballal, P. (2004) Discrimination of native and exotic forest patterns through shape
irregularity indices: an analysis in the landscapes of Galicia, Spain. Landscape Ecology 19: 647-
662.
Turner, M.G. (1989) Landscape ecology: the effect of pattern on process. Annual Review of Ecology
and Systematics 20: 171-197.

55
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

A farmer typology for modelling multiple land-use change at the regional level

D. Valbuena, P. H. Verburg, A. Bregt , A. Ligtenberg

Wageningen University and Research Centre. Environmental Sciences Group. Wageningen.

The Netherlands.
e-mail: [email protected].

Introduction

Land-use patterns have been shaped by the spatial and temporal interaction between
people and the environment (Moss 2000; Veldkamp and Verburg 2004). Although this
interaction is affected by many biophysical and socio-economic factors and processes, the
accumulation of individual human decisions is the main direct cause of current land-use
change processes. These decisions are taken based on many different aspects such as:
individual knowledge, experience and attitudes, family and social structure, the environment
(e.g. soil, water availability, climate and slope), resources (e.g. savings, loans, subsidies,
number and size of farms), previous land-use type, the degree of uncertainty (about the
environment and the market), and policies (Gasson 1973; Willock et al. 1999). The intricate
interactions of all these aspects, together with the lack of data, makes it impossible to take
into account all human decisions and actions that are continuously reshaping the landscape
structure. The aim of this study is to simplify the representation of the diversity of land-use
decisions, and in particular those related to implement non-agricultural practices, by using
the concept of farmer typology.

Selected parameters

A typology is an abstract way to simplify and order reality (McKinney 1950; Jollivet 1965).
The criteria to construct a typology, as well as to evaluate it, primarily depend on the
objectives of its implementation (Escobar and Berdegué 1990). For instance, a study of how
policies affect nature conservation at the regional level requires a different approach than a
study of what are the production systems in a village.

Because several typologies have been already well defined and implemented, this study
aims only to adjust one of them by combining three main aspects related to the farmer:
perceptions, personal and socio-economic characteristics, and the surrounding landscape.

Farmers’ perceptions
Whether a farmer implements non-agricultural practices partially depends on his/her
perception of the role of agriculture (Guillaumin et al. 2004). Farmers with a positive
perception of nature and landscape conservation and society’s demands are more likely to
participate in conservation programmes and to develop recreational facilities than those who
do not consider these aspects important.

Personal and socio-economic characteristics

Decisions on (multiple) land-use are also influenced by personal (e.g. age, educational
level, family size) and socio-economic characteristics (e.g. farm size, main agriculture
activity, economic and technical resources, etc) of the farmer (Willock et al. 1999). For
example, land-use decisions of a young farmer who practice intensive pig farming might
differ from those of a farmer who practice dairy farming in an extensive way.

Spatial location: landscape characterization

Human land-use decisions affect the structure and dynamics of the landscape. At the
same time, the nature of these landscape properties also influences farmers’ land-use

56
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

possibilities and decisions. For instance, a farmer whose holding is located in an area
dominated by intensive agriculture does not have the same possibilities than one whose farm
is located in a landscape with small-scale farming and with some nature areas. Thus, a
landscape characterization was spatially defined.

By simplifying the diversity of land-use decisions with a farmer typology, it is not only
possible to understand better the complexity of land-use change processes, but also to
include partially farmers´ decisions in land-use modelling at the regional scale, which in turn
can improve the relevance of such tools in spatial planning and policy-making processes.

References
Escobar, G. and J. Berdegué (1990). Conceptos y metodologías para la tipificación de sistemas de
finca: la experiencia de RIMISP. Tipificación de Sistemas de Producción. G. Escobar and J.
Berdegué. Santiago de Chile, RIMISP: 13-43.
Gasson, R. (1973). "Goals and Values of Farmers." Journal of Agricultural Economics 24(3): 521-542.
Guillaumin, A., D. Bousquet and A. Villaret (2004). Multifonctionnalité de l'agriculture: demandes
locales et attitudes des agriculteurs. Les Cahiers de la Multifonctionnalité. C. Laurent and J.
Rémy.
Jollivet, M. (1965). "D'une méthode typologique pour l'étude des sociétés rurales." Revue Française
de Sociologie VI: 33-54.
McKinney, J. C. (1950). "The role of constructive typology in scientific sociological analysis." Social
Forces 28(3): 235-240.
Moss, M. R. (2000). "Interdisciplinarity, landscape ecology and the `Transformation of Agricultural
Landscapes'." Landscape Ecology 15(3): 303-311.
Veldkamp, A. and P. H. Verburg (2004). "Modelling land use change and environmental impact."
Journal of Environmental Management 72(1-2): 1-3.
Willock, J., I. J. Deary, M. M. McGregor, A. Sutherland, G. Edwards-Jones, O. Morgan, B. Dent,
R. Grieve, G. Gibson and E. Austin (1999). "Farmers' Attitudes, Objectives, Behaviors, and
Personality Traits: The Edinburgh Study of Decision Making on Farms." Journal of Vocational
Behavior 54(1): 5-36.

57
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

How landscape ecology concepts can be applied to make policy for landscape
reclamation

F. Azari-Dehkordi1, D. Minai-Tehrani2, N. Mashayekhi-Kerahroodi1, N. Khazaee1

1
Graduate Faculty of Environment, University of Tehran, Tehran-Iran. PO Box 14155-6135,
e-mail: [email protected]
2
Department of Biology, Faculty of Science, Bio-Research Lab, Shahid Beheshti University,
Tehran-Iran.

Introduction

Almost two thirds of Japan is not habitable due to mountainous topography or forest
coverage (FAO, 2002). On the other hand, population density in this country is quite high
(MIAC of Japan, 2005) and needs a very complex road-network for transportation. For
example, till 1999 there was 129,190 km of constructed rural-road category in Japan that
was actually increased into 278,000 km (Nihon Rindo Kyokai, 2001). This network of rural
roads provides a highly fragmented rural and forest landscapes.
We have introduced a landscape reclamation approach in rural Japan for policy making
to reduce landscape fragmentation (Azari Dehkordi, 2005). The study site was decided as
Haizuka Dam Watershed, with 21,130 ha that is located in Eastern Hiroshima Prefecture,
Japan. In this area, ten sub watersheds were delineated as working units, according to river
system ordering of Strahler (1964) (Fig. 1).
Japan
Fig. 1: Study area, 10 sub-watersheds, in
Gono River

Sea

Eastern Hiroshima, Japan

Hiroshima City
2
The road (scale; 1:25,000) and vegetation
Haizuka
Dam
1
stream
3 layers were overlaid to compute fragmentation for
flow
10 the study area. Then, the Number of Patches
mountain
4
(NumP) in the fragmented vegetation layer was
ridge
9 plotted against the length of each type of roads in
8
5 the whole area. It was clear that Type 5 road has
7
6 the highest N
correlation with landscape
fragmentation (Table 1). The Type 5 road is
1 0 1

categorized as rural roads, (some times forest

kilometer

roads) with 1.5 m width on Japanese topographic maps.

Table 1: Number of patches was plotted against types of roads

Type 5 Road
Type 4 Road Type 3 Road
R square 0.82 0.79 0.58
P-value 0.81 0.25 0.01

No Rural Roads: A scenario approach

In a scenario approach in the working units, all roads of Type 5 were removed and the
fragmented patches were combined together in the fragmentation layer using ArcMap8.3®.
Then, for the 10 sub-watersheds, spatial statistic analysis was conducted to compute NumP
(number of patches), median, and mean of patches in two scenarios of before and after
landscape reclamation (Table 2).

58
Theme 1. Landscape, stakeholders, land use and policy
1.2 Symposium 22: The young landscape Ecologist

Table 2: Scenario approach for landscape reclamation removing Type 5 road category.
(NumP: Number of Patches; Pr: Statistics of the patches at present; RmR5: Statistics of
patches removing 5th type of roads; MedPS: Median Patch Size; MPS: Mean Patch Size;
TLA: Total Landscape Area.)

Landscape metrics
Study
Units

TLA
(ha)
NumP reclaimable MedPS reclaimable MPS
Pr RmR5 (NumP) Pr RmR5 (MedPS) Pr RmR5
1 558 396 162 1.3 1.6 0.3 3.9 5.4 2152
2 472 300 172 1.4 2.0 0.6 4.6 7.3 2190
3 867 632 235 1.2 1.7 0.4 3.4 4.6 2922
4 441 323 118 0.9 1.2 0.3 3.0 4.1 1317
5 700 425 275 1.0 1.1 0.1 2.8 4.6 1968
6 716 443 273 1.0 1.3 0.3 3.5 5.6 2487
7 172 98 74 1.2 1.5 0.3 2.9 5.2 505
8 928 654 274 1.3 1.5 0.2 3.2 4.5 2958
9 1035 704 331 0.9 1.1 0.2 2.8 4.2 2940
10 687 443 244 0.9 1.2 0.3 2.4 3.8 1671

Discussion: SLOSS and Policy Making for Landscape Reclamation

Several major ecological values of large patches are known. Some advantages are
known for small patches (Forman, 1995). Arguments will remain between ecologists
concerning SLOSS (Single Large Or Several Small) patches. However, for rural landscape
reclamation in Japan, compilation of several small patches are more reasonable as many
kilometers of rural access roads are already remote and covered by stones or debris.
Given these facts, our preliminary findings show that for example, removing the Type 5
roads will lead to a reduction of NumP in sub watershed 9 from1035 to 704 (saving 331
patches), or the median of the patches will increase from 1.4 ha to 2.0 ha and the mean has
an increase of 4.6 to 7.3 in sub watershed 2. Therefore, policy making for landscape
reclamation in Japan should concentrate on the Type 5 roads, which can be prioritized within
working units according to the landscape ecology metrics.

References
Azari Dehkordi, F. (2005) Environmental Impact Assessment of Dams in Japan and Iran: a landscape
ecological modeling perspective, PhD Dissertation. Hiroshima University, Hiroshima 358 pp.
FAO (2002) State of the World's Forests, FAO, Rome.
Forman, R.T.T. (1995) Land Mosaics: the ecology of landscapes and regions, Cambridge University
Press, New York.
MIAC (Ministry of Internal Affairs and Communications) of Japan (2005) Statistical Handbook of
Japan, Retrieved 2007-01-26 from www.stat.go.jp/English/data/handbook/c02cont.htm.
Nihon Rindo Kyokai (2001) New forest road construction – approach from the experience, Nihon
Rindo Kyokai (in Japanese), Tokyo, 217 pp.
Strahler, A.N. (1964) Quantitative geomorphology of drainage basin and channel networks. V.T.
Chow (ed). Handbook of Applied Hydrology, McGraw Hill, New York, section 4-11.

59
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

1.3 Open Session 4: Biodiversity Conservation and Agriculture

Eco-structure cartography in a Mediterranean Delta

R. Pérez Campaña, A. Matarán Ruiz, L.M. Valenzuela Montes

Dep. of Urban and Spatial Planning, University of Granada, Edificio Politécnico, Laboratorio
de Urbanística y Ordenación del Territorio, Campus Universitario de Fuente Nueva,
Granada. Spain.
e-mail: [email protected].

Introduction

In this article we present an innovative zoning of the Motril Local Council in the Guadalfeo
Delta (south of Spain, coast of Granada) based on an eco-structure, that could be defined as
a physical support for the environmental fluxes and processes essential to maintain the main
functions of the ecosystem in the context of multifunctional landscapes and spatial
development. The main objective of this planning proposal is to define a multifunctional
strategy leading to an Agrarian Park (Diputación de Barcelona, 2003) for an agricultural plain
with great environmental values (the so called Vega) articulated by this eco-structure but
threatened by the spatial grown of greenhouses, industries, urban areas and tourism as
elsewhere of the Mediterranean Coast (Terra, 2002) (Imbroglini, 2003) (Matarán, 2005).

Zoning proposal

On the basis of the agricultural habitats and the main structural elements (rural roads,
irrigation nets, fences and water courses) we propose a zoning cartography (Figure 1)
including new planning tools that are more efficient to regenerate degraded areas, to
reinforce the eco-structure and to increase the multifunctionality:

- Protect and promote agrarian farms against urban expansion.

- Improve agrarian landscapes that constitute and produce fundamental eco-structure
functions, such as: habitats, corridors, buffer or sinks.
- Generate transition areas and diverse landscapes based on the coexistence of different
agrarian typologies together with urban open spaces and even other land uses.
- Generate agrarian open spaces or with a low rate of urban occupation.
- Considerate the carrying capacity of the area and of any farming.

As seen in Figure 1 we are proposing a new territorial model for the Vega of Motril, where
most of the planning efforts are focused on the central and western areas, because there we
found the highest environmental values (e.g. biodiversity) and the most important conflicts of
urban expansion and subsequent agricultural land abandonment. The proposal for the
eastern areas is trying to reduce the saturation of greenhouses (Matarán, Aguilera and
Valenzuela, 2005) and the reinforcement of corridors (e.g. rivers and litoral). We have also
proposed most of the interventions along our interpretation of the main corridors of the eco-
structure. Finally, this zoning strategy must be developed in a more detailed scale within the
involvement of stake holders and citizens in the participative councils and management
structure of the proposed Agrarian Park.

60
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

References
Diputación de Barcelona. (2003). “Plan Especial de Protección y Mejora del Parque Agrario del Bajo
Llobregat”. Diputación de Barcelona.
Imbroglini, C. (2003). “Le infrastrutture ambientali. Matrici del progetto territoriale”. Quaderni dei
Dipartimenti di Architettura e Urbanistica di Pescara nº 14.
Matarán Ruiz, A. (2005). La valoración ambiental-territorial de las agriculturas de regadío en el litoral
Mediterráneo: el caso de Granada. PhD Thesis. University of Granada.
Matarán, A.; Aguilera, F. y Valenzuela, L.M. (2006). Modelling future landscapes: greenhouse
expansion in the Mediterranean coast. In Meyer, B.C. (Ed.), 2006. Sustainable Land Use in
Intensively Used Agricultural Regions. Landscape Europe. Alterra Report No.1338, Wageningen.
Terra. (2002). “The percolating urban plan”. Terra SRL

61
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

Arresting woodland bird decline in Australian agricultural landscapes: potential

application of the European agri-environment model

S.E. Park¹, S.J. Attwood², M. Maron², S.J. Collard², K. Reardon-Smith²

¹CSIRO Sustainable Ecosystems, 203 Tor Street, Toowoomba, Queensland, 4350, Australia.
e-mail: [email protected]
²Australian Centre for Sustainable Catchments and Faculty of Sciences, University of
Southern Queensland, Toowoomba, Queensland, 4350 Australia.

Introduction

Temperate Australia’s wheat/sheep zone and much of Western Europe have both
experienced dramatic declines in native bird populations associated with agricultural
landscapes, with many species exhibiting range contractions and greatly reduced abundance
(Ford et al. 2001; Donald et al. 2002). We present a comparison of European farmland and
Australian woodland bird declines and a critique of the recent strategies for addressing
declines. Finally we offer an evaluation of the European agri-environment model, as
represented by England’s Environmental Stewardship scheme, as a potential policy
mechanism for delivering targeted on-farm management for declining Australian woodland
birds and their habitat through providing financial incentives and support to participating
landholders.

Context

In both Australia and Europe, recent bird diversity declines in agricultural landscapes
have been attributed to a loss of habitat heterogeneity, resulting from moves towards
broadscale agriculture and homogenous management. However, there are at least two key
differences in the nature and cause of declines in the two regions. The declining species at
threat in Australian agricultural regions are largely woodland specialists, whereas in Europe
they are species dependent upon centuries-old traditional management of semi-natural
habitats. The former depend upon natural systems of mid to late successional stages; the
latter upon intermediate levels of intervention indicative of early to mid successional stages
(Sutherland 2004).
The distinction is largely due to differing land use histories of these regions. In Western
Europe, large areas of indigenous vegetation were converted to farmland over a long time-
period, dating from prehistoric times (Donald et al. 2002); many species have either adapted
to the changes or followed the gradual spread of agriculture from open habitats. In Australia,
the conversion to farmland is recent, ongoing and rapid, leaving scant opportunity for
woodland species to adapt to the new, more open environments.

Current conservation models

Biodiversity decline is engendering widespread scientific and government concern in both

regions. Increasing recognition of the problem is evidenced by recent changes in land use
policy, including a shift from production subsidies to agri-environmental payments in Europe
(Donald et al. 2002) and the introduction of legislation to police land-clearing in Australia.
While Australia’s conservation efforts have historically focused on the establishment of
conservation reserves, there is increasing pressure to address biodiversity conservation
priorities on private lands. Approaches have to date centred on one-off, short-term schemes,
administered through a wide variety of government and non-government organisations and
often relying on voluntary landholder involvement. Payments are generally one-off leveraged
contributions towards the direct financial cost of the capital works (landholders are often
expected to make an in-kind contribution to the work). Whilst this demonstrates landholder

62
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

willingness to undertake environmentally beneficial activities on their land, there are ample
indications that many are unable to do so for lack of financial resources (Cocklin et al. 2006).
These issues have been addressed to a considerable extent by the European model of
agri-environmental payments. In England, the Environmental Stewardship (ES) scheme is
national government managed and regionally delivered by one dedicated agency, the Rural
Development Service. One of the explicit, underpinning objectives of the scheme is
biodiversity conservation, with the stated target of “reversing the long-term decline in the
number of farmland birds by 2020” (Gregory et al. 2004). Payments consist of both one-off
reimbursement for capital works and ongoing ‘income-foregone’ payments for the loss of
intensive production land to more extensive practices.
Economically the stewardship model offers ten-year management agreements and allows
for much greater ongoing financial support for landholders. Organisationally it has a
dedicated agency and trained staff to administer, advise and police. Ecologically it has
targeted management options developed by ecological authorities, ongoing consultation with
external experts and organisations and specifically designed monitoring protocols.

Stewardship in Australia?

Vickery et al. (2004) state that agri-environment schemes may represent the only
currently available mechanism to reduce declines in farmland biodiversity over large areas.
As such, the ES model may appear an appropriate model for achieving Australia’s
environmental goals.
However, agri-environment schemes operate at enormous cost and, in the European
case, are potentially vulnerable to changes in EU funding and WTO regulation. There is also
considerable discussion as to whether such schemes actually deliver the biodiversity benefits
that they purport to (Kleijn and Sutherland 2003).
The applicability of the model to the unique environments of Australia is a matter for
debate. It is nevertheless clear that many of the mechanisms of bird decline in agricultural
landscapes could be addressed by the development of a highly targeted, well-designed,
economically robust and nationally consistent scheme that addressed matrix management
and large-scale habitat restoration as well as remnant protection and provided significant and
ongoing support to facilitate land holder adoption.

References
Cocklin, C., Dibden, J. & Mautner, N. (2006) From market to multifunctionality? Land stewardship in
Australia. Geographic Journal 172: 197-205.
Donald, P.F., Pisano, G., Rayment, M.D. & Pain, D.J. (2002) The Common Agricultural Policy, EU
enlargement and the conservation of Europe’s farmland birds. Agriculture, Ecosystems and
Environment 89: 167-182.
Ford, H.A., Barrett, G.W., Saunders, D.A. & Recher, H.F. (2001) Why have birds in the woodlands
of Southern Australia declined? Biological Conservation 97: 71-88.
Gregory, R.D., Noble, D.G. & Custance, J. (2004) The state of play of farmland birds: population
trends and conservation status of lowland farmland birds in the United Kingdom. Ibis 146: 1-13.
Kleijn, D. & Sutherland, W.J. (2003) How effective are European agri-environment schemes in
conserving and promoting biodiversity? Journal of Applied Ecology 40: 947-969.
Vickery, J.A., Bradbury, R.B., Henderson, I.G., Eaton, M.A. & Grice, P.V. (2004) The role of agri-
environment schemes and farm management practices in reversing the decline of farmland birds
in England. Biological Conservation 119: 19-39.
Sutherland, W.J. (2004) A blueprint for the countryside. Ibis 146: 230-238.

63
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

Are there more biologically and economically effective ways to protect aquatic
biota in agricultural landscapes?

B.R. Davies1, J. Biggs 1, P.J. Williams 1, S. Maund 2, S. Thompson

1
Pond Conservation c/o School of Life Sciences, Oxford Brookes University, Gipsy Lane,
Headington, Oxford OX3 0BP, UK.
e-mail: [email protected]
2
Syngenta Crop Protection AG, 4002, Basel, Switzerland

Introduction

There is much policy and legislation pertaining to freshwater ecosystems in the UK,
including drinking water standards, flood defence, abstraction restrictions and nature
conservation legislation. Ideally these measures should cumulatively result in high quality
freshwater ecosystems, rich in aquatic biodiversity. In reality, policy and legislation has
concentrated on attaining water chemistry targets, and there is little evidence that these
measures have translated into improved aquatic biodiversity on the ground. Indeed overall,
the indication is that the UK’s aquatic biodiversity has levelled off or declined in recent years.
The expansion and intensification of agriculture is seen as one of the major influences on
the degraded state of much aquatic biodiversity worldwide, caused largely by a combination
of the hydro-physical alteration and loss of waterbodies and increased sediment, nutrient and
pesticide loadings in runoff (e.g. Allan et al., 1997; Cresser et al., 2000; Woltemade, 2000).
These impacts have been widespread, partly due to the scale of agriculture which, in the UK
dominates almost 80% of land cover in England and Wales (Defra, 2002; Brown et al., 2006).

UK Agri-Environment schemes and their protection of aquatic biota

Detrimental effects from agriculture are mitigated for through agri-environment measures,
such as the UK’s Environmental Stewardship scheme, which is designed to protect both
terrestrial and freshwater habitats, the historic environment and promote public access
through a range of tiered measures. The extensive scale of agriculture means that, in
practice, it is these measures which should deliver the majority of broad-scale benefits for
aquatic biodiversity in England and Wales. One of the most widely employed measures for
the protection of waterbodies is the use of buffer strips. Although buffer strips have been
very well investigated and it is broadly agreed that they remove agricultural pollutants, the
widths required to do so have been much debated with over 25 m frequently being
recommended. Such widths are much greater than those suggested under Environmental
Stewardship, which offers 2 m to 6 m field margin buffers or over 10 m surrounding in-field
ponds, implying that at least a 100% increase in buffer widths may be needed to ensure that
they are effective. More broadly, evidence from several other countries has indicated that,
buffer zones alone may not be sufficient, with declines in aquatic biodiversity being observed
when the proportion of a waterbody’s catchment under agriculture exceeds 30% to 50% or
sometimes less (Wang et al., 1997; Fitzpatrick et al., 2001; Allan, 2004; Donohue et al.,
2006).
Therefore, it is not surprising that recent evidence has, so far, often shown a lack of
realisation of the desired benefits from agri-environment schemes, with research showing the
actual enhancement of farmland biodiversity to be limited and in some cases negative
(Whitfield, 2006). In eastern England the impacts of diffuse agricultural pollution have been
identified as so severe that it has been suggested that they can only be addressed by a
complete halt to farming in this area with the conversion of extensive areas of arable land to
semi-natural uses (Clover, 2006). This implies that current solutions provided by agri-
environment schemes are unlikely to be optimal and has led to a call for ‘smart’ solutions to
ensure that agricultural production may be retained (Clover, 2006).

64
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

We have investigated one such ‘smart’ solution by examining where aquatic biodiversity
lies within agricultural landscapes and how protection and mitigation measures may be
redistributed to provide more effective protection for key waterbody types whilst allowing the
continuation of agricultural production.

Smart solutions

New evidence shows that smaller waterbodies and in particular ponds support
comparatively very high levels of aquatic biodiversity across agricultural landscapes
(Williams et al., 2004; Davies, 2005; Davies et al., submitted). Although all waterbody types
contribute to aquatic biodiversity, it is likely that the protection and creation of a range of
smaller waterbodies may be a useful tool for securing aquatic biodiversity across a region.
This is made possible by the characteristically small catchment areas of these smaller
waterbodies (Davies et al., submitted) enabling resources to be concentrated into relatively
small pockets with extensive catchment deintensification, whilst retaining farming in adjacent
areas outside of these catchments.

References
Allan, J.D.; Erickson, D.L. & Fay, J. (1997) The influence of catchment land use on stream integrity
across multiple spatial scales. Freshwater Biology 37: 149-161.
Allan, J.D. (2004) Landscapes and riverscapes: the influence of land use on stream ecosystems.
Annual Review of Ecology, Evolution and Systematics 35: 257-284.
Brown, C.D.; Turner, N.L.; Hollis, J.M.; Bellamy, P.H.; Biggs, J.; Williams, P.; Arnold, D.J.;
Pepper, T. & Maund, S.J. (2006) Morphological and physico-chemical properties of British
aquatic habitats potentially exposed to pesticides. Agriculture, Ecosystems and Environment 113:
307-319.
Clover, C. (2006) Farmland ‘must go back to the Middle Ages’. Retrieved on May 2nd 2006, from:
www.telegraph.co.uk/news/main.jhtml.
Cresser, M.S.; Smart, R.; Billet, M.F.; Soulsby, C.; Neal, C.; Wade, A.; Langan, S. & Edwards,
A.C. (2000) Modelling water chemistry for a major Scottish river from catchment attributes.
Journal of Applied Ecology 37: 171-184.
Davies, B.R. (2005) Developing a Strategic Approach to the Protection of Aquatic Biodiversity.
Unpublished PhD thesis; Oxford Brookes University.
Davies, B.R.; Biggs, J.; Williams, P.J.; Lee, J.T. & Thompson, S. (submitted) A comparison of the
catchment sizes of rivers, streams, ponds, ditches and lakes: implications for protecting aquatic
biodiversity in an agricultural landscape. Hydrobiologia.
Defra (2002) Second Consultation Paper on the Implementation of EC Water Framework Directive
(2000/60/EC). Defra; London.
Donohue, I.; McGarrigle, M.L. & Mills, P. (2006) Linking catchment characteristics and water
chemistry with the ecological status of Irish rivers. Water Research 40: 91-98.
Fitzpatrick, F.A.; Scudder, B.C.; Lenz, B.N. & Sullivan, D.J. (2001) Effects of multi-scale
environmental characteristics on agricultural stream biota in eastern Wisconsin. Journal of the
American Water Resources Association 37: 1289-1507.
Wang, L.Z.; Lyons, J.; Kanehl, P. & Gatti, R. (1997) Influences of watershed land use on habitat
quality and biotic integrity in Wisconsin streams. Fisheries 22: 6-12.
Whitfield, J. (2006) How green was my subsidy. Nature 439: 908-909.
Williams, P.; Whitfield, M.; Biggs, J.; Bray, S.; Fox, G.; Nicolet, P. & Sear, D. (2004) Comparative
biodiversity of rivers, streams, ditches and ponds in an agricultural landscape in Southern
England. Biological Conservation 115: 329-341.
Woltemade, C.J. (2000) Ability of restored wetlands to reduce nitrogen and phosphorus
concentrations in agricultural drainage water. Journal of Soil and Water Conservation 55: 303-309.

65
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

Landscape ecology at the service of the policy maker: evaluating the Swiss agri-
environmental measures

F. Herzog, W. Richner , T. Walter

Agroscope Reckenholz-Tänikon Research Station ART, Zurich, Switzerland

e-mail: [email protected]

Introduction
“IALE encourages landscape ecologists to transcend boundaries and to work together
building theory and developing knowledge of landscape pattern and process, developing
integrative tools, and making them applicable to real landscape situations and applying them
to solve problems” (www.IALE.com). Evaluating the effects of policy measures is a
formidable challenge for landscape ecologists. Their broad interdisciplinary knowledge is
required to assess the effectiveness of policy measures and to distinguish their effects from
other factors affecting environmental parameters. The communication of results and the
involvement of stakeholders goes beyond mere scientific requirements (Figure 1).
In the early 1990’s a comprehensive agri-environmental programme was launched in
Switzerland, which was later (1999) transformed into a cross-compliance mechanism (“Proof
of Ecological Performance” – PEP). In order to quality for direct payments, whole-farm
nutrient budgets have to be balanced, measures for soil conservation, integrated crop
protection and animal welfare have to be implemented. Moreover at least 7 % of a farm's
land have to be managed as Ecological Compensation Areas (ECA).
Environmental objectives were formulated by the government and we were given the task
to assess goal attainment and the actual effectiveness of policy measures.

Figure 1. Nightmare of the landscape

ecologist evaluating agri-environmental
programmes. Reality proved to be less cruel
because the stakeholders agreed on a joint
interpretation of results.

Materials and methods

The methods comprised (i) monitoring of
ground- and surface water quality; (ii) long-
term monitoring of agricultural practices and
of their effects on nitrate leaching, soil
erosion and phosphorous run-off from grassland; (iii) scenario studies using nitrogen and
phosphorous simulation models; and (iv) large-scale surveys of vascular plants, arthropods
and birds on ECA. The research period covered 1995 – 2005.

Scientific findings
The main difficulty in evaluating the PEP’s environmental effects was that many factors
(e. g. climate, technical progress, socio-economic conditions of farmers) act both on farming
practices and on the environmental parameters under investigation. Only long-term
observations in conjunction modelling studies can single out the effect of policy measures.
Nitrogen and phosphorous emissions and immissions were substantially reduced in
comparison to the reference period 1990-92 (Herzog and Richner, 2005; Prasuhn and
Sieber, 2005). However, not all objectives were reached and there is still a considerable
amount of excess nitrogen and phosphorous leading to pollution problems.
Biodiversity was increased on ECA as compared to intensively managed fields (Herzog
and Walter, 2005; Aviron et al., 2007). In the lowlands, PEP rules protect these sites from
intensification whilst in mountain areas, ECA prevent them from abandonment. ECA largely
failed to promote threatened species.

66
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

Communication and stakeholders

In addition to scientific publications, results were communicated to the authorities and to
stakeholders. They contributed to the formulation of the new agricultural policy project for the
period 2008-011. At this stage, the scientific findings were evaluated also from a political
point of view. Farmers’ representatives tend to secure their income, environmental NGOs
emphasise the protection of natural resources, and regional representatives defend particular
regional interests. The administration tries to balance the different stakeholder interests
according to their political weight and ability to defend their stake.
What did we as scientists and landscape ecologists learn from this exercise?
1. Stakeholders may try to interfere already in the course of the project in order to
influence the results (Figure 1). However, they finally agreed that objective,
scientifically well founded results are most useful for everyone.
2. The requirements of science (e.g. sufficient sample size, sufficient time for data
acquisition and analysis) need to be defended against ‘’just-in-time’’ requirements for
results by decision makers.
3. Whereas scientific findings are unambiguous, they need to be interpreted. We found it
useful to integrate stakeholders in this process. They tried to steer the interpretation
in order to strengthen their position. But once a common interpretation was reached,
the results were generally accepted.
4. The communication of the results should be neutral and science oriented. Still it was
not always possible to prevent misuse by the press and by specific pressure groups.
In conclusion, we found this a rewarding exercise - both scientifically and in terms of
impact. We addressed four of six key issues (Wu and Hobbs, 2002) of landscape ecological
research: interdisciplinarity, integration with applications, international collaboration, outreach
and communication with the public and decision makers.

References
Aviron S., Jeanneret P., Schüpbach B. Herzog F. (2007) Effects of agri-environmental measures,
site and landscape conditions on butterfly diversity of Swiss grasslands. Agriculture, Ecosystems
and Environment (in press).
Herzog F. & Richner W. (eds.) (2005) Evaluation der Ökomassnahmen: Bereich Stickstoff und
Phosphor Zurich, Agroscope FAL Reckenholz, Schriftenreihe der FAL 57, 132 pp.
Herzog F. & Walter T. (eds.) (2005) Évaluation des mesures écologiques : Domaine biodiversité.
Zurich, Agroscope FAL Reckenholz, Les cahiers de la FAL 56. 208 pp.
Prasuhn V. & Sieber U. (2005) Changes in diffuse phosphorus and nitrogen inputs into surface
waters in the Rhine watershed in Switzerland. Aquatic Sciences 67, 363–371.
Wu J. & Hobbs R. (2002) Key issues and research priorities in landscape ecology: An idiosyncratic
synthesis. Landscape Ecology 17(4), 355–365.

67
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

The role of habitat networks in guiding integrated land-use planning in

agricultural landscapes

M. Smith, J.W. Humphrey

Ecology Division, Forest Research Northern Research Station, Roslin, Midlothian UK.
e-mail: [email protected]

Introduction

The development of habitat networks is seen as an important mechanism for reversing

the effects of fragmentation on biodiversity while delivering a range of other environmental
benefits such as creating more opportunities for public access and recreational enjoyment of
the countryside. In the Scottish lowlands, there is a desire to develop a more integrated
approach to planning land-use change, which takes account of conservation objectives for
the full suite of habitats and species associated with different types of land use.

Focal species modelling offers a potentially useful tool for developing different types of
habitat networks and informing the targeting of agri-environment and forestry incentives
Humphrey et al., 2005). Studies to date in Scotland have focused on the use of focal
species modelling in developing forest habitat networks (Ray et al., 2004), but there is now a
need to apply the approach to multi-network planning in more complex landscapes where
there are a mix of different land-uses. In this paper we report on a test of the focal species
modelling approach on the Isle of Tiree one of four case study areas in lowland Scotland with
contrasting patterns of land use and conservation objectives.

Study area description

The Isle of Tiree, covering approximately 80 km2 is situated some 50 km from the
Western Scottish mainland. Topographically the island is low-lying with the highest point
140 m above sea level. Tiree has a unique climate; although benefiting from the Gulf
Stream, precipitation at 1100mm/annum is much lower and temperatures are generally
higher and more equable through the year than on the mainland. Extensive agriculture using
the traditional crofting system is the main form of land-use on the island. This is based round
small Townships with groups of smallholdings sharing common grazings. Here, through
stakeholder groups, cattle grazing has been identified as the single most important
conservation management tool and issue for habitats of conservation concern on the island.
These are the Machair, unimproved coastal grasslands, and the Carex nigra dominated wet
heaths that are found on thin peat further inland. The enclosed land historically was made up
of small fields which were often further sub-divided by the crops grown on them, barley oats
and potatoes being the most common but also with patches of hay meadow intermingled
between the crops.

Land use change

Over that last 40 years land use has changed the balance on the enclosed in-bye land
notably through the introduction of black bag silage and increased use of non-organic
fertilisers. Tiree is an important area for breeding Corncrake (Cerex cerex) in Scotland and
agri environmental schemes for the protection of this species has contributed to this land use
change with close to 10% of agricultural land being under such schemes in 2005. This
involves the late cutting of silage to allow for a second brood and the maintenance of Iris
pseudacorus dominated mires, used as early cover for the birds. Over this same period
cropping, growing of barley and oats mainly and hay meadows have declined significantly

68
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

with only 20 ha under cropping in the whole island and hay cutting abandoned in all but the
driest of summers.

There has been a resultant decline in species associated with cropping and hay meadows
such as arable weeds and graniverous birds most notably the corn bunting (Miliaria calandra)
which became extinct on the island in 2000 though is still present, but declining, on nearby
islands. Winter cattle feed has now to be imported to the island from the mainland
threatening the sustainabiltity cattle production and therefore the use of grazing as a
conservation tool on the Machair and Carex nigra heaths. The Township of Barrapoll in the
south west of the island has an area of approximately 750ha and changes in land use here
reflect those over the whole island.

Over 50% of the 342Ha of present day permanent grassland in Barrapoll is in agri
environmental schemes 85% of which is late cute silage the rest as early cover for birds. The
42 ha of former cropping fields has resulted in 10 of these 14 being converted to late cut
silage under agri-environmental schemes for Corncrake, the remainder now being other
forms of permanent grassland. Since 1962 20% of the 447ha common grazings within this
township have been apportioned to individual crofts. This practice has gained momentum
over the last 25 years where crofters enclosed the area of common land they had the grazing
rights to. These apportioned areas have remained as Machair but the change in grazing
regime results in less extensive grazing over smaller areas and a subsequent decline in
habitat quality.

Getting the balance right

Corncrakes and corn bunting are both declining across much of Europe and where there are
populations or potential populations the habitats that support them should be safeguarded.
By using these as focal species as part of the development of integrated habitat networks a
balance can be struck that can maintain the corncrake population but also allow for the return
of the corn bunting. This would not only benefit the other species that utilise these habitats
but allow for the sustainability of cattle grazing. Winter feed would again be produced on
Tiree and so the quality of the Machair maintained for species such as the bee Colletes
floralis (another focal species used in this case study) which is an indicator of Machair habitat
quality

Acknowledgement

This work benefited greatly from the stakeholder involvement of Ross Lilley, David and Janet
Bowler and their knowledge of the ecology and land-use history of the island.

References
Humphrey, J.W.; Watts, K., McCracken, D., Shepherd, N., Sing, L., Poulsom, E.G. and Ray, D.,
2005. A review of approaches to developing lowland habitat networks in Scotland. ROAME No.
FO2AA102/2. Commissioned Report. 104. Scottish Natural Heritage, Edinburgh.
Ray, D., Watts, K., Hope, J.C.E. and Humphrey, J.W., 2004. Developing forest habitat networks in
southern Scotland. In: R. Smithers (Eds.), Landscape Ecology of Trees and Forests. Proceedings
of the twelfth annual IALE (UK) conference, held at the Royal Agricultural College, Cirencester,
21-24 June 2004. IALE UK, Stoke on Trent, pp 216-223.

69
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

An up-to-date cost benefit analysis of English agri-environment schemes: their

impact at the landscape scale and the cost of adequate monitoring

P.D. Carey, R.F. Pywell

Centre for Ecology and Hydrology, Monks Wood, Abbots Ripton, Huntingdon, UK.
e-mail [email protected]

Introduction

Four years ago in Darwin a summary of English agri-environment schemes and their
effectiveness was presented to IALE (Carey et al 2004). Since then there has been a well
mannered debate on the efficacy of schemes across Europe that in the end depended on
whether you think the ‘glass is half empty or half full’.
The new Environmental Stewardship scheme in England has the slogan “Look after your
land and be rewarded”. The scheme is split into two main parts in the ‘Entry Level Scheme’
(ELS) and the ‘Higher Level Scheme’ (HLS). The ELS is designed to attract 85% of all
farmland. By December 2006, there were approximately 25000 agreements covering 3.5Mha
(34%). The farmers are asked to reach a points target (30) for each hectare of land where
the points are gained for carrying out different and basic management prescriptions for things
such as the creation of wildflower strips in arable land, hedgerow management, and over-
wintering stubbles for birds. The Higher Level Scheme is more like the old schemes where
high quality land is targeted and the aim is to protect it and enhance if possible.
At Monks Wood we have been gathering data by experimentation on the efficacy of different
management prescriptions that will be given to the farmers. By applying these results to the
uptake figures for the new schemes we aim to predict the impact of the new schemes across
the countryside. As this abstract is written the data on uptake for different prescriptions are
not yet available.
Monitoring any project or programme against its objectives makes perfect sense. However,
in the world of conservation and government policies it is seldom done adequately. Why is
that? The simple answer is that it has not been budgeted for adequately. Recently the EU
fifth framework project EASY devised and tested a monitoring programme for agri-
environment schemes. We present how much the EASY methodology would cost to
adequately assess change.

The Cost of Monitoring – a fictional example

The region of Allsplat (total area 20,000 km2) follows EU directives by designing a scheme to
protect the treasured landscape and wildlife. Most of the wildlife exists in small areas of semi-
natural habitat in amongst the intensive agriculture. Like England we assume there are two
levels, one for ordinary land and one for rare species and habitats. The European Union
requires Allsplat to monitor the success of the scheme but does not say how, or give any
money to help do it (Carey 2001). The scheme paid out an average €80 per hectare of
managed land. The simplest and cheapest objective to monitor was uptake of the scheme
and to know how the farms were geographically located so a GIS was designed. It was found
that 50% of all farms were in the scheme and that overall 10% of all land was under scheme
prescriptions, at an annual cost of €16M. The uptake analysis and GIS cost €150K per year.

To monitor the effectiveness of the scheme requires: a baseline survey; and a monitoring
scheme of farms in the scheme and outside of it that will detect change in landscape and
wildlife that can be attributed to each of the prescriptions. A random sample of all
agreements will not detect rare species (as shown by Kleijn et al 2006) because they are
rare. We will show that for a rare weed that occurs on 5 farms the probability of detecting it is
K 8 x 10-5. This is obviously not an adequate probability to say anything about the efficacy of

70
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

a scheme. Targeted monitoring of various habitats in UK has shown that 200 plots can detect
a suite of rare plant species (Walker et al 2006).
If the aim is to show that habitats are getting closer to the pristine condition of Priority
Habitats power analysis is required to work out what sample size is required to achieve the
desired level of change detection. Critchley et al 2002 showed a sample of 1000 sites was
required. To this number of sites a control of 1000 should be added in fields not in the
scheme following the protocol suggested by Kleijn et al (2006)
To detect rare species and detect whether habitats are moving towards a pristine state would
require a sample of 2200. If each plot is visited twice in five years (the bare minimum to
detect change) and the cost of each visit is €300 then the cost of just collecting the data will
be €1.3M. Add to this the cost of project management, analysis and report writing, €0.5M and
5 years of uptake analysis and GIS, €0.75M then the total monitoring cost for the wildlife
objectives would be K €2.55M. Socio-economic and landscape monitoring would also be
required e.g. as carried out by Carey et al 2003 cost €4K per farm. A sample of 500 would
therefore cost €2M. Total payments for the scheme over 5 years in Allsplat would be 5x€16M
= €80M. The monitoring budget is therefore 5.6% of the total payments to farmers,
reasonable but expressed as a cost would annoy farmers and is probably outside the reach
of most regional governments who have hospitals, schools etc. to pay for.
There are 100 or so real regions in Europe the size of the fictional one and so the cost of an
adequate monitoring programme would be €455M over 5 years. Is there anybody out there
who has ever heard of an ecological monitoring programme getting anything close to that
from governments?

References
Carey, P.D. (2001). Schemes are monitored and effective in the UK. Nature 414: 687.
Carey, P.D.; Short, S.; Morris, C.; Hunt, J.; Priscott, A.; Davis, M.; Finch, C.; Curry, N.; Little; W.,
Winter, M.; Parkin, A. & Firbank, L.G. (2003). The multi-disciplinary evaluation of a national agri-
environment scheme. Journal of Environmental management 69:71-91.
Carey, P.D.; Manchester, S.J. & Firbank, L.G. (2005) Performance of two agri-environment schemes
in England: A Comparison of ecological and multi-discipliary evaluations. Agriculture Ecosystems
and Environment 108: 178-188.
Critchley, C.N.R.; Maskell, L.C.; Mitchley, J.; Adamson, H.F.; Burch,F.M.; Carey,P.D.; Firbank,
L.G.; Fowbert, J.A.; Parkin, A.B.; Smart, S.M. & Sparks, T.H. (2002). Review and
recommendations of methodologies to be used for botanical monitoring of agri-environment
schemes in England. Final contract report to DEFRA, pp. 224.
Kleijn, D.;Baquero,, R.A.; Clough, Y.;Díaz,M.; De Esteban,J.; Fernández, Gabriel, Herzog, A.
Holzschuh, R. Jöhl, E. Knop, A. Kruess, E. J. P. Marshall, I. Steffan-Dewenter,F.;
Tscharntke, T.;Verhulst, J.; West, T.M.; Yela, J.L. (2006) Mixed biodiversity benefits of agri-
environment schemes in five European countries. Ecology Letters 9: 243–254.
Walker, K.J.; Critchley, C.N.R.; Sherwood, A.J.; Large, R.; Nuttall, P.; Hulmes, S.; Rose, R.&
Mountford, J.O. (2006). An assessment of new agri-environment scheme options in England, UK.
Biological Conservation in press.

71
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

Spatial Ecology Models and Conservation Targeting in the UK

R.D.J. Catchpole

Natural England, Science and Evidence Team, Bull Ring House, Northgate, Wakefield, West
Yorkshire.
e-mail: [email protected]

Introduction

Mechanisms for targeting conservation effort rely heavily on existing definitions of

protected areas and species. Yet biodiversity has continued with observed declines in
woodland birds (Eaton et al., 2005), bumblebees (Kells and Goulson, 2003; Goulson et al.,
2006), vascular plants (Cheffings et al., 2005), woodland plants (Kirby et al., 2005), infertile
grassland plants (Bunce et al., 1999) and butterflies (Bergman, 2001; Fox et al. 2001; Swaay
et al. 2006). As conservation moves into the 21st century, never has the need been greater
to think about where future conservation resources might be spent. Decisions that may have
previously been made on either an ad hoc or opportunistic basis need to be placed within a
more clearly defined framework which supports judgements with an appropriate ecological
evidence base. This is especially the case beyond protected areas where land use planning
has generally failed to safeguard critical natural capital.

The application of landscape ecology thinking could offer significant benefits and help
develop a new paradigm for conservation in which the ecological flows between protected
areas and their wider landscape context are more clearly recognised and valued. This is not
a choice between the conservation of sites vs. whole landscapes but rather the development
of a more informed interpretation that goes beyond the urge to simply make patches bigger
or increase physical linkage. The use of ecological models, that are able to utilise currently
available information, offer some hope however. This paper will contrast the application of
species-based and habitat-based models. It will highlight key differences, explore the
practical value of the outputs and consider how this might change current patterns of
targeting, using England as an example.

Comparisons

One example of a habitat-based approach can be found in least-cost distance modelling.

This calculates the path of the lowest cost to movement between a series of focal patches.
These costs can be calculated from a modified land cover layer (e.g. Land Cover Map 2000,
ITE). The modification requires the attachment of estimates for the relative cost to
movement across different land cover types. These can be derived from a series of expert
judgements for a ‘generic focal species’ that might be associated with focal patches.
Although the approach can be applied to empirical data for individual species movement, its
value, as a conservation tool, lies in more general application. This can be done through the
use of the Spatial Analyst toolkit in ArcMap 9.0 GIS (ESRI, 2004). The algorithm generates
least-cost buffers around existing habitat patches which indicates where more permeable
areas of landscape might still occur.

One example of a species-based approach can be found in the ZONATION reserve

selection model (Moilanen and Kujala, 2006). This is a grid-based approach that supports
the simultaneous comparison of distribution data from a range of different species. It
sequentially removes the lowest scoring cells in such a way that the loss of biodiversity and
connectivity is minimised. Both masked and unmasked outputs can be generated for
comparison with other methods and priorities. Such masking allows ZONATION model
outputs to be constrained by where conservation action is currently prioritised, e.g. statutory

72
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

sites. Even though connectivity is evaluated through cell adjacency, it provides a structured
way to evaluate species distribution data. Currently available UK datasets for invertebrates,
lower plants and higher plants could be used within such an analysis.

Conclusions

The use of systematic and repeatable ecological analyses, that utilise currently available
information, are critical in prioritising spending to deliver the greatest benefits for wildlife.
There is a clear need to define what actions are appropriate in a spatially explicit manner so
that land managers and policy makers can be clear about what is expected beyond protected
areas. Although actions would need to be kept under review, some significant benefits could
be delivered through the use of such approaches. Questions not only need to be raised
about the assumed ecological benefits where current targeting fails to correspond to such
areas but also about the performance of newly targeted areas in the future. This is the only
option in a changing environment.

References
Bergman, K-O. (2001). Population dynamics and the importance of habitat management for
conservation of the butterfly Lopinga achine. Journal of Applied Ecology 38: 1303-1313.
Bunce, R.G.H., Smart, S.M., van der Poll, H.M., Watkins, J.W., Scott, W.A. (1999). Measuring
change in British vegetation. Ecofact Volume 2. ITE, Merlewood.
Cheffings, C.M., Farrell, L. , Dines, T.D., Jones, R.A., Leach, S.J., McKean, D.R., Pearman, D.A.,
Preston, C.D., Rumsey, F.J., Taylor, I. (2005). The Vascular Plant Red Data List for Great
Britain. Species Status 7: 1-116. Joint Nature Conservation Committee, Peterborough.
Eaton, M.A., Noble, D.G., Hearn, R.D., Grice, P.V., Gregory, R.D., Wotton, S., Ratcliffe, N., Hilton,
G.M., Rehfisch, M.M., Crick, H.Q.P., Hughes, J. (2005). The state of the UK’s birds 2004. BTO,
Thetford.
ESRI (2004). ArcMap 9.0 Geographical Information System. ESRI Inc.
Fox, R., Warren, M.S., Harding, P.T., McLean, I.F.G., Asher, J., Roy, D., Brereton, T. (2001). The
State of Britain’s Butterflies. Butterfly Conservation, CEH and JNCC. Wareham, Dorset.
Goulson, D., Hanley, M., Darvill, B., Ellis, J. (2006). Biotope associations and the decline of
bumblebees (Bombus spp.). Journal of Insect Conservation 10: 95-103.
ITE (2000). Land Cover 2000, Institute of Terrestrial Ecology, Merlewood, United Kingdom.
Kells, A.R., Goulson, D. (2003). Preferred nesting sites of bumblebee queens (Hymenoptera: Apidae)
in agro-ecosystems in the UK. Biological Conservation 109: 165–174.
Kirby, K.J., Smart, S.M., Black, H.I.J., Bunce, R.G.H., Corney, P.M., Smithers, R.J. (2005). Long-
term ecological change in British woodland (1971-2001). English Nature Research Report 653.
English Nature, Peterborough.
Moilanen, A., Kujala, H. (2006). ZONATION: Spatial Conservation Planning Framework Software v.
1.0 User Manual. Metapopulation Research Group, Helsinki, Finland.
Swaay, C., Warren, M., Loïs, G. (2006). Biotope use and trends of European butterflies. Journal of
Insect Conservation 10: 189-209.

73
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

Can agri-environmental schemes reduce habitat isolation and enhance

biodiversity? The example of butterflies in Switzerland

S. Aviron, P. Jeanneret, B. Schüpbach, F. Herzog

Agroscope Reckenholz-Tänikon Research Station ART – Reckenholzstr. 191, CH-8046

Zurich, Switzerland.
e-mail: [email protected]

Introduction

Agri-environmental schemes (AES) aim at integrating environmental concerns into

agricultural policy and at protecting biodiversity. AES can be considered as a real landscape
restoration experiment at a large scale (Herzog, 2005) and they allow testing, whether
restoration measures can reduce habitat isolation and enhance landscape diversity for
biodiversity. This study investigated the effect of AES measures at the field and landscape
scales on butterfly diversity in Switzerland. The Swiss AES consists in the conversion of 7%
of farmland into low-input habitats, the ecological compensation areas (ECA). We focused on
the two main ECA types in the lowlands, i.e. ECA grasslands (90’000 ha) and ECA orchards
(2.6 millions trees). We tested whether (i) butterfly diversity is higher on ECA than on
conventional grasslands, (ii) butterfly diversity is enhanced by the network of ECA and by
landscape diversity and (iii) the effects of ECA vary according to their landscape context.

Materials and methods

The study was conducted in two landscape units located on the Swiss ‘Plateau’.
Landscape unit 1 is characterised by mixed grassland-arable farming, whilst landscape unit 2
is characterised by intensive grassland farming. Within each landscape unit, butterflies were
recorded in ECA grasslands (N = 41), ECA orchards (N = 15) and conventional grasslands
(N = 34) in 2004. The selected grasslands were rotational grasslands initially sown with a
mixture of Italian ryegrass and clover. Explanatory variables were grouped into four sets of
variables: (i) grassland type (ECA grassland, ECA orchard, conventional grassland), (ii)
grassland quality (plant richness), (iii) landscape context of grasslands (amount of ECA,
semi-natural habitats, conventional grasslands and crop fields in a 200 m circle radius;
distance to the nearest ECA, semi-natural habitats, conventional grasslands and crop fields;
habitat diversity in a 200 m radius) and (iv) space (geographical coordinates of sampled
grasslands). For each landscape unit, the explanatory variables were tested on synthetic
indices of butterfly diversity (species richness, abundance, Shannon diversity) using general
linear models (GLM) and on butterfly assemblages by the means of redundancy analysis
(RDA).

Results – discussion

Butterfly diversity indices

Amongst the four sets of explanatory variables, only the type of grassland had a
significant effect on diversity indices. The species richness of butterflies in landscape unit 1
was in average higher in ECA grasslands and ECA orchards than in conventional
grasslands, whilst butterfly abundance was higher in ECA grasslands than in conventional
grasslands in both landscape units. This suggests a positive effect of extensification
measures at the field scale on butterfly communities, at least in one study area. The
Shannon index was not influenced by any of the tested variables.

Species assemblages
In both landscape units, ECA grasslands had different sets of butterfly species as
compared to ECA orchards and conventional grasslands (RDA, P < 0.05) (Figure 1). At the

74
Theme 1. Landscape, stakeholders, land use and policy
1.3 Open Session 4: Biodiversity conservation and agriculture

landscape scale, butterfly assemblages were influenced by the distance of grasslands to

other ECA grasslands (landscape unit 2) and/or to natural herbaceous and woody habitats
(both landscape units) (RDA, P<0.05). This indicates that ECA, together with existing natural
habitats, contributed to reduce the spatial isolation of butterfly habitats. Butterfly
assemblages were not affected by habitat diversity in the surrounding landscape (RDA,
P>0.05).

ECA
1.0

grassland
ECA
orchard
Conventional
Distance to herb. elements grassland
Axis 2

P. brassicae
P. rapae

% woodyP. napi
elements (200 m) O. venatus
T. lineola
C. hyale
V. atalanta
Plant A. cardamines % conventional
richness A. hyperantus
grasslands (200 m)
P. aegeria P. icarus
M. jurtina % crops (200m)
C. semiargus
A. paphia

% herbaceous
-0.6

elements (200 m)
-0.6 0.8
Axis 1

Figure 1. Ordination diagram from redundancy analysis of butterfly assemblages

constrained by grassland type (samples: polygons around each type), plant
species richness (arrow) and landscape context (arrows) in landscape unit 1.

Conclusions
Although AES focus on the implementation of good farming practices at the field scale, our
study suggests that they might bring some environmental benefits by restoring the
agricultural matrix (Aviron et al., in press; Donald and Evans, 2006). For butterflies, habitat
isolation can be reduced by connecting AES grasslands and existing natural habitats.

References
Aviron, S.; Jeanneret, P.; Schüpbach, B. & Herzog, F. (in press) Effects of agri-environmental
measures, site and landscape conditions on butterfly diversity of Swiss grassland. Agriculture,
Ecosystems and Environment
Donald, P.F. & Evans, A.D. (2006) Habitat connectivity and matrix restoration: the wider implication of
agri-environment schemes. Journal of Applied Ecology 43: 209-218.
Herzog, F. (2005) Agri-environmental schemes as landscape experiments. Agriculture, Ecosystems
and Environment 108: 175-177.

75
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

1.4 Open Session 5: Sustainability and agriculture

Silvoarable Agroforestry for Europe: environmental and economic performance

J.H.N. Palma1, A.R. Graves2, P.J. Burgess2, Y. Reisner3, F. Herzog1

1
Ecological Controlling Research Group, Agroscope FAL Reckenholz, Zurich, Switzerland,
e-mail: [email protected]
2
Institute of Water and Environment, Cranfield University, Silsoe, UK,
3
Swiss College of Agriculture, Zollikofen, Switzerland

Introduction

Many European landscapes suffer from intensive agricultural management which affects
environmental quality (e.g. water quality and landscape biodiversity). Silvoarable
Agroforestry (SAF) integrates use of trees and arable crops in the same field. It potentially
offers a range of environmental and economic benefits in comparison with conventional
arable cropping. A modeling approach was used to compare the environmental and
economic benefits of SAF with arable and forestry systems at a landscape and continental
scale.

Methods

At the landscape scale, biophysical and economic data were collected for a stratified
random sample of 19 Landscape Test Sites (LTS) in the Mediterranean and Atlantic regions
of Europe. Five tree species, holm oak (Quercus ilex subsp. ilex L.), stone pine (Pinus pinea
L.), hybrid walnut (Juglans sp), wild cherry (Prunus avium L.) and poplar (Populus spp) were
modeled with combinations of up to five crops; wheat (Triticum durum Desf. ), sunflower
(Helianthus annuus L.), oilseed rape (Brassica napus L.), grain maize and silage maize (Zea
mays L.). The environmental assessment focused on soil erosion, nitrogen leaching, carbon
sequestration and landscape diversity following the methodology suggested by Palma et al.
(2006a). The economic assessment focused on the farm’s infinite net present value over the
tree rotation period (Graves et al., 2006). At each LTS, different agroforestry scenarios were
modeled and compared to status quo arable production.
At the continental scale, data on soil, climate, topography, land cover and tree growth
were used to identify target regions for SAF, with the aim of finding areas where SAF could
reduce the risk of soil erosion, contribute to groundwater protection, and increase landscape
diversity.

Results and Conclusions

At landscape scale SAF had a positive impact on the four environmental indicators in
comparison with the status quo, but economic benefits varied according to tree species and
region. The extent of the environmental impact depends on the severity of the problems and
the SAF management options for each location. Benefits were predicted to be highest when
SAF was implemented on large areas (i.e. 50% of the farm) and on high quality land, where
current agricultural practices were intensive and associated with high levels of soil erosion
and nitrogen leaching (see details in Palma et al., 2006b). Economic predictions for the post-
2005 Common Agricultural Policy payments suggested that SAF with walnut and poplar in
France could provide a profitable alternative to arable systems. In Spain, it appeared that
holm oak and stone pine could be integrated into arable systems without substantially
reducing arable production for many years. Since these trees are of ecological and
landscape importance, rather than productive importance, additional support in the form of
an agri-environment payment could be justified (see details in Graves et al., 2006).

76
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

At a continental scale target regions

were found to make up about 40% of the
total arable area of the European Union
(Figure 1). Of these, 7% were found to
be at risk from soil erosion (erosion rate
> 5 t ha-1 a-1), 34% were in nitrate
vulnerable zones, and 59% were low
diversity arable landscapes (see details
in Reisner et al., 2006).

Figure 1. Target regions for

silvoarable agroforestry in Europe for
Juglans spp., Prunus avium, Populus
spp., Pinus pinea, and Quercus ilex.
Possible areas for agroforestry on arable
land in Europe, where at least one of the
five selected tree species can potentially
grow, and where at least one
environmental problem exists (soil
erosion, nitrate leaching, landscape
diversity). Target Areas I: well defined
areas; Target Areas II: scattered areas.
Based on Reisner et al (2006).

References
Graves, A., P. Burgess, J. Palma, F. Herzog, G. Moreno, M. Bertomeu, C. Dupraz, F. Liagre, K.
Keesman, & van der Werf, W. (2006) The development and application of bio-economic
modelling for silvoarable systems in Europe. Ecological Engineering in press.
Palma, J., A. Graves, P.J. Burgess, K. Keesman, H. van Keulen, M. Mayus, Y. Reisner, & Herzog,
F. (2006a) Methodological approach for the assessment of environmental effects of agroforestry at
the landscape scale. Ecological Engineering in press.
Palma, J., A. Graves, R. Bunce, P. Burgess, R. De Filippi, K. Keesman, H. van Keulen, M. Mayus,
Y. Reisner, F. Liagre, G. Moreno, & Herzog, F. (2006b) Modelling environmental benefits of
silvoarable agroforestry in Europe. Agriculture Ecosystems & Environment in press:
doi:10.1016/j.agee.2006.07.021.
Reisner, Y., R. De Filippi, J. Palma & Herzog, F. (2006) Target regions for silvoarable agroforestry in
Europe. Ecological Engineering in press.

77
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

Shifting cultivation or agroforestry? The dilemma of selecting a livelihood model

for the forest dwellers and preserving landscape ecology in north-east Himalayas

M. Choudhury

Centre for Himalayan Studies, North Bengal University, P.O. North Bengal University,
Siliguri, Pin Code 734013, District Darjeeling, West Bengal, India
e-mail: [email protected]

Introduction

There is an ever-growing debate between the advocates of shifting cultivation and the
proponents of agroforestry. According to some, shifting cultivation is agriculture on
forestland, and according to some others it is forestry on agricultural land (Kerkhoff, 2006). A
case of such dilemma has been observed in India’s northeast. Shifting cultivation, in spite of
being a time-tested age-old practice in the region, is now under the scanner of environmental
scientists. A section of bio-scientists working in the region are in favour of suitable
agroforestry model as an alternative to shifting cultivation by traditional slash-and-burn
method. However, this is often argued that “shifting cultivation is not to be confused with
slash-and-burn, since slash-and-burn is a mere land clearing method, whereas shifting
cultivation involves complete landscape management by customary institutions including the
management of forests as fallows and village forests” (International Centre for Integrated
Mountain Development, ICIMOD, 2006). The advocates of shifting cultivation are also of the
opinion that shifting cultivators help biodiversity conservation since they use a large variety of
plants and animal products in their everyday life and thus maintain these species, contrary to
the belief that they destroy biodiversity. The practice of shifting cultivation is believed to be
beneficial for certain species of wild biodiversity particularly the elephants that require open
space. The practice of land rotation creates patches and corridor for the easy movement of
elephant herds and other larger species of wild animals in northeast India, thereby
maintaining wild biodiversity. According to the ICIMOD, in a way shifting cultivation itself is a
traditional agroforestry practice that can be improved and adapted to modern needs, rather
than replacing it with alternatives.

Shifting cultivation or agroforestry?

India’s north-east, particularly the state of Arunachal Pradesh in the Eastern Himalayas is
densely forested. On an average, at any given point of time, more than 60 percent of the
geographical area of the state remains under forest cover though the percentage varies from
place to place due to shifting cultivation. Shifting cultivation is a prevalent agricultural system
in the state and affects an area of about 16,000 sq. km out of a total geographical area of
83,743 sq. km (Choudhury & Haridasan, 1993). The state is home to more than a hundred
tribes, the majority of who are forest dwellers and practice shifting cultivation for centuries.
They strongly rely on indigenous knowledge and oral tradition to maintain a balance between
their livelihood practices and environment (Borang & Borkotoki, 2002). Till very recently, the
abundant land and forests were thought to be sufficient to sustain the indigenous inhabitants
of the state. However, of late, some telltale signs of environment degradation are appearing
in certain areas that have put the practice of shifting cultivation in question. The increase in
population, though not alarming, has temped the forest dwellers to reduce the gap between
successive cycles of cultivation. The advocates of shifting cultivation are of opinion that it is
not so much the practice of shifting cultivation, which is in question; rather it is the length of
cycle that matters. Violating the indigenous norms of maintaining a gap of at least ten years
between two consecutive periods of cultivation, many tribal groups are returning back to the
forest clearings within three to six years (Sood, et al, 2002) that leave hardly any time to
regenerate forest and regain the soil nutrients lost during the period of cultivation. Whatever
the case may be, the gradual depletion of forest cover of the state is worrying the

78
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

environment scientists in the country. With a density of population of only 16 persons per sq.
km. (2001), the situation is hopefully not beyond repair. To preserve the landscape ecology
in this thinly populated state, a section of environment scientists are of opinion that, the
indigenous tribes should be trained to start agroforestry in place of shifting cultivation.
Considering the fact that there are several models of agroforestry, it is important to select a
model best suited for local needs, which may again vary according to the terrain condition,
soil cover, altitude and climate.

Agroforestry models

The International Centre for Research in Agroforestry (ICRAF) has developed

methodology for designing agroforestry models on the basis of regional productivity,
sustainability and adoptability. Following ICRAF guidelines several models have been
developed for the region by local as well as international agencies. The most talked about
models of agroforestry in the region are the multi-tier Agro-silvi-horti-pastoral model
developed by the Indian Council of Agricultural Research (ICAR) and the Sloping Agricultural
Land Technolology (SALT) model developed by the Mindanao Baptist Rural Life Centre in
the Philippines (Upadhyaya, et al, 2002). The present paper makes an attempt to assess the
dimensions of dilemma between two schools in favour and against shifting cultivation, and
viability as well as acceptability aspects of the agroforestry models prescribed for the area.

References
Borang, A. & Borkotoki, A. (2002) Indigenous people’s knowledge and alien culture in Arunachal
Pradesh, in R.C. Sundriyal, T. Singh & G.N. Sinha (Eds). Arunachal Pradesh: Environmental
Planning and Sustainable Development, G.B. Pant Institute of Himalayan Environment and
Development, Almora, India, pp. 537-548.
Choudhury, S.S. & Haridasan, K. (1993) Water conservation vis-à-vis forestry in Arunachal Pradesh,
Arunachal Forest News, Itanagar, India, 11(1): pp. 4-17.
Kerkhoff, E. (2006) A Dialogue on What Policy Makers Can Learn from Farmers, International Centre
for Integrated Mountain Development, (ICIMOD), Kathmandu, Nepal.
Sood, K. K; Singh, B. & Rethy, P. (2002) Shifting cultivation in Arunachal Pradesh – causes and
control, in R.C. Sundriyal, T. Singh & G.N. Sinha (Eds). Arunachal Pradesh: Environmental
Planning and Sustainable Development, G.B. Pant Institute of Himalayan Environment and
Development, Almora, India, pp. 161-165.
Upadhyaya, K; Arunachalam, A. & Khan, M.L. (2002) Designing proper agroforestry models for
Arunachal Pradesh: scope and challenges, in R.C. Sundriyal, T. Singh & G.N. Sinha (Eds).
Arunachal Pradesh: Environmental Planning and Sustainable Development, G.B. Pant Institute of
Himalayan Environment and Development, Almora, India, pp. 349-353.

79
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

Large-scale biomass production and potential effects on farmland habitats and

related biodiversity

B.S. Elbersen1, P.D. Carey2

1
Alterra, P.O.Box 47, 6700 AA Wageningen, The Netherlands,
e-mail: [email protected]
2
Centre for Ecology and Hydrology, Monks Wood, Abbots Ripton, Huntingdon,
Cambridgeshire PE262XH UK,

Introduction

In this paper the pressures from land use changes resulting from a politically induced
increased demand for biomass for bioenergy generation are assessed and translated into
potential effects on biodiversity.
The decision of the European Council in March 2007 set a binding overall goal for 20%
renewable energy sources and a minimum binding target share of 10% biofuels in the total
transport fuel consumption by 2020 Currently, only around 4 % (69 MtOE) of the EU’s total
primary energy consumption is met from biomass.
The study conducted for the EEA1 still based on the less ambitious targets of 5.75%
share of biofuels in total fuel consumption and 21% share of renewable energy in total
energy consumption by 2020 show that this would require an approximate 10% share of the
present Utilised agricultural area (UAA) in the EU-25 for targeted biomass crop production. It
is clear that the most recent more ambitious targets of 10% biofuels will even increase the
pressure on agricultural land further to a possible share of up to 18% of the UAA in EU-25.
In order to assess the potential effect on farmland biodiversity in the EU it is important to
understand how much land is required, but also what land use changes are induced and
where these changes are mostly likely to take place.

Potential land use changes and effects on biodiversity

Trying to assess the effects of any change in land-use on biodiversity and the
environment as a whole is extremely difficult because of imperfect knowledge. It is unclear
how organisms are distributed in the landscape, how they function, and how management
practices on the land affect them. But the effects of intensive agriculture on biodiversity are
conclusive and there is little doubt that further intensification of agriculture will occur in the
coming decades especially in the new EU countries. Economics along with the productive
capability of the land will determine which changes in land-use will occur first to
accommodate the needs of biomass production. But overall it is clear that biomass
production is most likely to be taken up first in those regions where land is released from
agriculture and on set aside land. The impacts on biodiversity of changing some extensive
land-uses to intensive arable or biomass production would be catastrophic but from an
economic and technical point of view these are less likely to occur.
However, a distinction needs to be made between extensive farmland categories that are
suited for potentially suited for biomass production such as ancient extensive grasslands like
the Dehesas/Montados and species rich hay meadows of the mountains and the extensive
grasslands that are more recent and have occurred due to land abandonment. This second

1
This paper is based on a study conducted for the European Environment Agency on the potential effects of a
wider demand for biomass in agriculture on European farmland habitats and biodiversity. Elbersen. B.;
Andersen. E; R. Bakker. R. Bunce. P.Carey. W. Elbersen. M. van Eupen. A. Guldemond. A. Kool. B.Meuleman.
G.J. Noij & J. Roos Klein-Lankhorst (forthcoming). Large-scale biomass production and agricultural land use –
potential effects on farmland habitats and related biodiversity. Technical report. EEA study contract
EEA/EAS/03/004

80
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

group are also valuable in biodiversity terms and if converted to intensive arable would lose a
great deal of environmental quality.
It is estimated that for most of the EU15 countries the negative biodiversity/environmental
impacts will affect a rather small portion of the land. The negative impacts will be higher in
Italy, Spain and Portugal. In Italy’s case this is mainly due to the large conversion of
permanent crops to intensive arable agriculture and the large demand for bioenergy and
therefore biomass. In Spain there are predicted losses of extensive grasslands which have a
high biodiversity value and there is also a large demand. Portugal, stands out as being
particularly badly affected under the storylines. If no considerable efforts are being made to
protect the environment in Portugal when biomass cropping is introduced it will be
catastrophic for farmland biodiversity.
The effects of biomass production on the new EU countries are not quantifiable using the
methods employed for the old EU countries but it is almost certain that biomass production
will cause negative impacts. However, they might not be greater than the impacts that
general intensification of agriculture in those countries will have (autonomous
development?).
There is little scientific information on the impacts of biomass crops on biodiversity, but
what can be inferred is that both short rotation coppice (SRC) and perennial biomass
grasses will provide shelter for animals although little food for them. Growing SRC or the
energy grasses on arable land will be very beneficial to earthworms and soil health generally
because it is a limited or no till systems. However, if biomass crops replace permanent
grassland there will be no benefits and probably negative effects on the soil and loss of soil
carbon.
The areas where High Nature Value (HNV) farmland is most at risk of being converted to
biomass crops is mapped by combining suitability information and HNV farmland
concentrations. The maps show that HNV and high suitability for ligno-cellulose production
are negatively correlated. This would be encouraging except that it has been postulated that
areas of high suitability are also areas of greatest agricultural production and are likely to
remain in their current land-use and it is areas of medium suitability and the better end of the
low suitability class that are more likely to be converted. These maps clearly show that large
areas of Spain, Portugal, Italy and Greece in the south have in theory large areas of HNV
with medium suitability for ligno-cellulose crops. In practice much of the areas in Greece and
the mountainous areas will be unsuitable because of the terrain. However, HNV in parts of
Italy and Portugal (extensive arable, extensive grasslands, permanent crops) may be at
threat. The low suitability areas are less at risk because they would be more difficult to
convert to ligno-cellulose cropping and would produce low yields. In the north of Europe
western Ireland, Wales, North Western England and Scotland have patches of medium
suitability land with a high density of HNV farming land but most of it is protected.
The Baltic States, Poland, Romania and Bulgaria all have a large proportion of HNV
(equivalent) in medium and high suitability categories and in these countries designations are
not as stringent as in UK and therefore are more at risk of conversion to ligno-cellulose crops
with a large impact on biodiversity. The area required for biomass crops in these countries is
relatively small and the conversion rate will be low unless carbon credit trading is undertaken
between member states.

81
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

Greenhouse planning criteria based on landscape ecology: a case study in the

South of Spain

A. Matarán Ruiz, and L.M. Valenzuela Montes,

Department of Urban and Spatial Planning, University of Granada,

Edificio Politécnico, Laboratorio de Urbanística y Ordenación del Territorio, Campus
Universitario de Fuente Nueva, Granada. Spain.
e-mail: [email protected].

Introduction: planning problems concerning greenhouses

Greenhouses are producing important territorial impacts and environmental externalities

related to unplanned land occupation and to the production of pollutants and discharges
(Matarán, 2005). Nevertheless, the extraordinary spatial expansion of greenhouses along the
last thirty years in the south east of Spain is leading to the saturation of the agricultural
landscapes of coastal plains (Matarán, Aguilera, and Valenzuela, 2006).

From the point of view of the environment, planning documents are obsolete both in their
methods and in their focus (Priemus, Rodenburg, and Nijkamp 2004). Agriculture is
considered as a secondary activity in most of the documents. However, there are very few
planning references to this important economic activity. In the case of local planning, the
urban focus is predominant, so agriculture land is always the rest of the territory that has not
got any other classification such as urban, urbanizable and protected. Furthermore, in the
case of regional planning, we have to add the lack of coordination and out of date landscape
and environmental analyses.

However, in this particular case, diverse historical, political, economic and environmental
circumstances have led to a situation where natural resources are still considered to have
high quantitative and qualitative territorial values (Matarán and Valenzuela, 2006). So, we
are still in time to produce another scenario overcoming the current trends and producing
planning and management measures taking into account new criteria based on the
multifunctionality of agricultural landscapes, as well as focusing on ecological corridors and
other landscape functions such as source, sink or buffer. The final objective of the proposed
measures is to generate a more efficient agricultural landscape in a Mediterranean context.

Planning criteria
We are proposing here some planning criteria as a response to those problems detected in
the planning documents. The following criteria are based on landscape ecology and on the
multifunctionality of coastal landscapes. The presented cartography includes a spatial
representation of these criteria in an area that is saturated by greenhouses (the Rambla of
Gualchos in the Coast of Granada –South of Spain-):
Criterion 1. A re-interpretation of the Mediterranean mosaic. Including a proposal of tree
based landscapes in two different categories: Dry crops like almond trees (light coloured)
and olive trees, subtropicals and forest trees (dark coloured).
Criterion 2. A re-consideration of pasture land. Including the promotion of vegetation
succession in two different categories: Pasture land with a buffer function (dark coloured)
and residual pasture land (light coloured).

82
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

Criterion 3. A re-generation of
diffusion nets. Including the
promotion of connectivity
functions in two different
categories: Natural net, mainly
rivers in this scale (dark
coloured) the net of percolation
ways between man made
landscapes (light coloured).
Criterion 4. A re-organization of
greenhouse zoning. Including
the integration of other criteria
in this area.
Criterion 5. A re-alocation of
urban zoning. Including also
the integration of the other
criteria in the area and
generating four categories of
density and urban design.

Figure 1. Zoning proposal in

the Rambla of Gualchos.

References
Matarán Ruiz, A. (2005). La valoración ambiental-territorial de las agriculturas de regadío en el litoral
Mediterráneo: el caso de Granada. PhD Thesis. University of Granada.
Matarán, A. and Valenzuela, L.M. (2006). Regional planning in Granada, south-east Spain taking
account the network of natural values. In R.G.H. Bunce and R.H.G. Jongman (Eds) 2006.
Landscape Ecology in the Mediterranean: inside and outside approaches. Proceedings of the
European IALE Conference 29 March – 2 April 2005 Faro, Portugal. IALE Publication Series 3, pp.
95 – 109.
Matarán, A.; Aguilera, F. and Valenzuela, L.M. (2006). Modelling future landscapes: greenhouse
expansion in the Mediterranean coast. In Meyer, B.C. (Ed.), 2006. Sustainable Land Use in
Intensively Used Agricultural Regions. Landscape Europe. Alterra Report No.1338, Wageningen.
Priemus, H. Rodenburg, C.A. and Nijkamp, P. (2004). Multifunctional urban land use. Built
Environment, 30, 4.

83
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

Multifunctionality of landscapes – rural development, landscape functions and

their impact on biodiversity

N. Guiomar, J. P. Fernandes

CEEM, Universidade de Évora, Rua Romão Ramalho, Évora, Portugal

e-mail: [email protected]

Introduction

Rural Landscapes constitute a particular good example of the integration through space
and time of multiple natural, social and economical functions. This integration is materialized
in a set of land use systems and social structures adapted to the particular natural
constraints and resources in the frame of the available technologies. Each structure tries to
find the best combined solution to the feasibility equation, involving the balance of labour and
other investments with the different products and alternative sources in order to maximize the
desired landscape functions. Under functions, the use and the response of a landscape to
human needs is understood. Multifunctionality consists in the integration of different functions
in a given spatial and/or temporal unit at a given scale. All landscapes are multifunctional but
their degree of multifunctionality can vary strongly with different environmental potentials and
resources. This paper aims at analysing the way in which the present land users of this
landscape equate their land use options and adapt to the different landscape functions and
social constraints (e.g. tradition, subsidies). The viability of the resulting system compared
with development scenarios is evaluated, as well as alternative income sources such as the
payment for environmental services.

Development of a conceptual framework

Over the past decades many of the traditional multifunctional Mediterranean landscapes
with their typical complexes of agro-, silvo- and pastoral components have changed
completely. Nowadays only relativelly few still exist. Their complex farming systems provide
at the same time a multitude of other functions than just agricultural production, such as
support for recreation, amenity, cultural identity, preservation of natural resources and
environmental quality (Pinto-Correia & Vos, 2004). De Groot (1992) classifies landscape
functions related to land use as production functions, which are related to the ability of a
landscape to ensure a functional balance as regulation functions and the functions related to
the ability to produce non material goods as information functions. These different functions
and the corresponding land use demands determine, at each moment and site, a particular
solution of land use intensity and typology and degree of landscape and resources stress.
There is a relationship between cultural-historical and biodiversity aspects of landscape
structures. Both intensive and extensive land use are expressed in the landscape: the
structure of the land, the size of the parcels and their area and the diversity of natural and
semi natural vegetation that is present. The spatial arrangement of land use at each moment
and in each location is the local integration of multiscalar decision and constraining factors of
economical, social and biophysical nature (Jongman, 2004). Land use intensification and the
resulting pressure on resources and regulation functions have determined the spatial
segregation of land use types, the functional specialisation and the degradation of the quality
and functionality of many landscape, process that is particularly perceptible in marginal rural
areas. This process is closely associated with the regression of traditional land use systems,
extensification and land abandonment due to human migration and the generalisation of land
use types associated to very low management efforts.
This situation is well illustrated by the particular situation of the depressed Portuguese
Mediterranean rural areas, where in the last 70 years land use has evolved from an intensive
use of the resources, either through agriculture or forestry and many times in a close
integration between agriculture and forest, to a situation of concentration of intensive

84
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

agricultural practice on good land, contrasted by an extensification and abandonment of

poorer or remote areas. The resulting increase in the forestry sector, both in Mountain areas
and on the plains, takes advantage of supporting policies. Many these plantations are poorly
or unmanaged forests which are prone to disease or fire.
In the frame of the European project “LACOPE-Landscape Development, Biodiversity and
co-operative Livestock Systems in Europe” the relations between extensive grazing and
biodiversity has been analysed as well
as the contribute of given forms of
grazing to the promotion of biological
quality and other landscape functions
(Kaule & Fernandes, 2006). The
particular situation of Portuguese
mediterranean agricultural systems,
where land use systems show a very
high liability and a high susceptibility to
disturbance, determine an urgent need
for integrated studies to assess the
contribution of each land use system to
the different landscape functions, as
well as assessing their economical and
social viability (Fernandes & Guiomar,
2006).

Figure 1. Multifunctional Landscape Management Mosaic

References
de Groot, R. S. (1992) Functions of Nature: evaluation of nature in environmental planning,
management and decision-making. Wolters Noordhoff BV, Groningen, The Neth.
Fernandes, J. P.; & Guiomar, N. (2006) Scenario approach – unpublished WP 12 final report Project
LACOPE-Landscape Development, Biodiversity and co-operative Livestock Systems in Europe.
Contract EVK2-CT-2002-00150, Stuttgart.
Jongman, R. (2004) Landscape linkages and biodiversity in European landscapes. R. Jongman R.
(Eds). The New Dimensions of the European Landscape, Wageningen EU Frontis Series,
Springer.
Kaule, G.; & Fernandes, J. P. (2006) Atlantic heath land rough grazing systems- Portuguese section
– unpublished report Project LACOPE-Landscape Development, Biodiversity and co-operative
Livestock Systems in Europe. Contract EVK2-CT-2002-00150, Stuttgart.
Pinto-Correia, T.; & Vos, W. (2004) Multifunctionality in Mediterranean landscapes – past and future.
R. Jongman R. (Eds). The New Dimensions of the European Landscape, Wageningen EU Frontis
Series, Springer.
Kaule, G.; & Fernandes, J. P. (2006) Atlantic heath land rough grazing systems- Portuguese section
– unpublished report Project LACOPE-Landscape Development, Biodiversity and co-operative
Livestock Systems in Europe. Contract EVK2-CT-2002-00150, Stuttgart.
Pinto-Correia, T.; & Vos, W. (2004) Multifunctionality in Mediterranean landscapes – past and future.
R. Jongman R. (Eds). The New Dimensions of the European Landscape, Wageningen EU Frontis
Series, Springer.

85
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

Landscape configuration, vegetation condition and ecosystem services in cotton

agro-ecosystems in southern Queensland, Australia
A.P.N. House1, N.A. Schellhorn2, S.D. Brown1, F.J.J.A. Bianchi2
1
CSIRO Sustainable Ecosystems, 306 Carmody Road, St Lucia, Qld, Australia.
email: [email protected]
2
CSIRO Entomology, 120 Meiers Road, Indooroopilly, Qld, Australia

Introduction

Cotton (both dryland and irrigated) is one of Australia’s most intensive agricultural crops.
One of the challenges for sustainable cotton enterprises is to optimise the production of
goods and services from all parts of the landscape. Most cotton farms support a wide range
of habitats suitable for native wildlife through the patches of native vegetation (woodlands
and grasslands) and other farming systems (cropping and grazing) that may benefit the
environment and sustainability of cotton production. The maintenance of habitat
heterogeneityin agricultural landscapes is the key to sustaining biodiversity (Benton et al.
2003), and there are links between levels of biodiversity and ecosystem function (Díaz &
Cabido 2001). However the means to manage for the maintenance of biodiversity and the
ecosystem services it provides on cotton properties and at larger catchment scales is poorly
understood.

Project components

Vegetation condition and ecosystem services

The ecological condition of vegetation may be as important to landscape function as
spatial arrangement (Debuse et al. 2006). Our project will assesses vegetation condition by
looking at surrogates of function – evidence of stand replacement, levels of ground cover,
structural complexity, diversity and abundance of key arthropod groups (ants, beetles,
spiders) – and analyse these in respect of landscape context (i.e. patch size, isolation,
shape, connectance). These spatial analyses will inform subsequent analyses of predator-
prey/parasitoids-host complexes and vertebrate habitat use.

Source:sink dynamics and bio-control

Colonisation of cotton and grain crops by predator-prey/parasitoid-host complexes is
influenced by the surrounding landscapes (Tscharntke & Brandl 2004). Crops and non-crop
habitat (i.e. native vegetation and weeds, and the condition of the vegetation) can be both
sources and sinks for pests and natural enemies (Tscharntke et al. 2005). However, little is
known of how the degree of synchrony between sources and sinks, and spatial
characteristics of these habitats influences (i) the timing and numerical response of
colonization by pests and natural enemies, (ii), species accumulation over time and (iii)
trophic interactions. This project addresses these issues using empirical and modelling
approaches. Experiments will quantify insect colonisation in crops and non-crop habitats,
identify habitats that function as sources and sinks of pests and natural enemies, and
monitor the population dynamics of pests and natural enemies in crop and non-crop habitats.
Modelling will focus on how the spatial arrangement and synchrony of sinks and sources of
pests and natural enemies impacts insect dispersal, trophic interactions, and population
dynamics.

Birds in cotton landscapes

Birds can play important roles in pest control in agricultural landscapes (Jones et al.
2005). We know that birds have been impacted by agricultural development through loss of
habitat, habitat fragmentation and habitat simplification (Radford et al. 2005). Our interest is
in the residual habitat value for birds of native and non-native vegetation elements in cotton-
dominated landscapes, and the potential pest control services that birds provide.

86
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

Contrasting landscapes
We will test the influence of landscape configuration on ecological and pest control
services in two contrasting landscapes. Cotton is a component of both and occupies similar
proportions of the landscape, but there are major differences in the proportions of irrigated
cropping, non-irrigated cropping and native vegetation, and in the spatial arrangement of
native vegetation patches (Table 1).

Table 1. Proportions of land use types and key landscape metrics in 2 contrasting cotton
landscapes in southern Queensland, Australia. Measures based on 5 km radii; landscape
metrics calculated using FRAGSTATS (McGarigal et al. 2002).

landscape 1 landscape 2
irrigated cropping 26.1 30.4
non-irrigated cropping 55.5 27.4
native vegetation 15.4 39.3

for native vegetation:

largest patch index 7.8 39.1
no. patches 16 4
splitting index 148.2 6.6

By examining the relationships between spatial pattern, ecological condition of vegetation,

and the provision of ecosystem services to both production and conservation, we will provide
scientific knowledge to inform regional resource management plans and support the
refinement of best management practice in cotton systems.

References
Benton, T.G., Vickery, J.A. and Wilson, J.D. (2003) Farmland biodiversity: is habitat heterogeneity
the key? Trends in Ecology and Evolution 18: 182-188.
Debuse, V.J., King, J. and House, A.P.N. (2006) Effect of fragmentation, habitat loss and within–
patch habitat characteristics on ant assemblages in semi–arid woodlands of eastern Australia.
Landscape Ecology.
Díaz, S. and Cabido, M. (2001) Vive la différence: plant functional diversity matters to ecosystem
processes. Trends in Ecology and Evolution 16: 646-655.
Jones, G.A., Sieving, K.E. and Jacobson, S.K. (2005) Avian Diversity and Functional Insectivory on
North-Central Florida Farmlands. Conservation Biology 19: 1234-1245.
McGarigal, K., Cushman, S.A., Neel, M.C. and Ene, E. (2002) FRAGSTATS: Spatial Pattern
Analysis Program for Categorical Maps,
<www.umass.edu/landeco/research/fragstats/fragstats.html >.
Radford, J.Q., Bennett, A.F. and Cheers, G.J. (2005) Landscape-level thresholds of habitat cover for
woodland-dependent birds. Biological Conservation 124: 317-337.
Tscharntke, T. and Brandl, R. (2004) Plant-insect interactions in fragmented landscapes. Annual
Review of Entomology 49: 405-430.
Tscharntke, T., Rand, T.A. and Bianchi, F.J.J.A. (2005) The landscape context of trophic
interactions: insect spillover across the crop-noncrop interface. Annales Zoologici Fennici 42: 421-
432.

87
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

Provision of non-commodities in agricultural production at farm and landscape

level in the MEA-Scope case study area Mugello (Italy)

S. Uthes1, C. Sattler1, A. Ciancaglini2, G. Piani2 and G. Lombardi2

1
Leibniz-Centre for Agricultural Landscape Research, ZALF, Institute for Socio-Economics,
D-Müncheberg, Germany.
e-mail: [email protected]
2
Università di Firenze, Dipartimento di Economia Agraria e delle Risorse Territoriali, DEART,
I-Firenze, Italy.

Within the framework of the FFP 6 EU project MEA-Scope, we investigate ex-ante the effects
of different policy options on the economic, ecological and social functions of agriculture in
seven European case study regions by using and combining micro-level bio-economic and
biophysical modelling techniques (Piorr et al., 2006).
In this contribution, we apply the bio-economic linear programming model MODAM to the
Mugello territory in Northern Tuscany, Italy (1127 km2) to model the provision of agricultural
land-use related non-commodities. The Mugello territory is characterised by small, in terms
of economic size, mixed crop-livestock farms mostly engaged in the total cow-calf line mixed
farming. The beef sector is made of traditional farms with forage crops or grassland for
grazing. Mountain pastures and permanent grasslands dominate the land-use; followed by
fodder crops such as alfalfa and forage sorghum. Important arable crops are grain maize,
barley and durum wheat. The actual multifunctional role of agriculture in Mugello
encompasses issues including the landscape and the hydro-ecological asset conservation
and the high quality local products, including the organic ones, as expressions of the specific
territory communicating values that are deeply-rooted in local history, custom and a
wholesome environment. In this paper, we will particularly focus on the environmental
functions in Mugello, namely water quality, soil conservation and biodiversity, which have
been specified by following indicators (cp. table 1)

Table 1. Environmental targets and indicators in the Mugello territory (MEA-Scope)

Function Indicator
Biodiversity Habitat potential for amphibians (Amph)
Biodiversity Habitat potential for wild flora species (Flora)
Biodiversity Habitat potential for field hares (Hare)
Soil conservation Risk of water erosion (WaEro)
Water quality Risk of nitrate entries into groundwater (NO3)
Water quality Risk of pesticide entries into ground- and surface waters (Pest)

To explore the relation between these indicators and agricultural land-use, we have carried
out a detailed computer-based survey to collect information on region- and site-specific
cropping practices and livestock systems. “Site-specific” means, that as the Mugello territory
is very heterogeneous with respect to geology, geomorphology, soils and elevation,
management and yield potential of the cropping practices were adapted to four main
landscape units: the Sieve River valley bottom, the hilly parts between the main axis valley
and the mountainous region both north and south of Sieve river, the mountain area south of
the river, between the Mugello valley and Florence valley, and the proper Apennines area in
the north/north-west of the territory. Altogether, two production systems, organic and
conventional, 192 cropping practices, and 12 livestock systems have been identified and
ecologically evaluated. The ecological evaluation in MODAM is based on a fuzzy-logic-
impact-assessment-method (Sattler et al., 2006) generating goal attainment values that
express the suitability of cropping practices with respect to the environmental indicators listed
in table 1.

88
Theme 1. Landscape, stakeholders, land use and policy
1.4 Open Session 5: Sustainability and agriculture

We combined the information on the agricultural management with FADN-information on

farm capacities (land, labour, etc.) and geo-referred cadastral data to create “virtual” that is,
model farm types that could be allocated in a raster grid of the region (Ungaro et al., 2006).
Input/output prices as well as information on different past and future policy framework
conditions, including 2nd pillar instruments were also added to the model. Possible future
policy options were identified with regional stakeholders, modellers, and representatives of
the European Commission leading to three main scenarios to be analysed: (i) a baseline as a
continuation of the status quo (BASELINE), (ii) an introduction of a single-area-payment to
replace the current single-farm-payment (SAP) and (iii) a free-market scenario (NOSUBS).
By optimizing the production plan of the virtual farm types in the three scenarios, we were
able to compare changes in land-use and economic performance of the farm types as well as
environmental impacts. For example, we could observe that the overall environmental threat
increases during the scenarios and varies considerably between conventional and organic
farming (figure 1).

NO3 NO3
100% 100%
80% 80%
60% 60%
Flora Pest Flora Pest
40% 40%
BASELlNE
20% 20%
SAP
0% 0%
NOSUBS

Hare WaEro Hare WaEro

Amph Amph

Conventional farming Organic farming

Figure 1. Average index of goal attainment in % for six environmental indicators in Mugello,
comparison between conventional and organic farming (MEA-Scope); 100 % indicating a
good and 0 % a poor ecological performance

We will present our results in the form of graphs and GIS maps, such as the average
environmental risk in each scenario, differences in environmental risk between production
systems (as shown in figure 1) or farm types, trade-off functions between economic
performance and environmental impacts etc.

References

Piorr, A.; Müller, K.; Happe, K.; Uthes, S. & Sattler, C. (2006) Agricultural management issues of
implementing multifunctionality: commodity and non-commodity production in the approach of the
MEA-Scope project. Mander et al.(Eds). Multifunctional land use meeting future demands for
landscape goods and services. New York, pp. 167-182.
Sattler, C.; Schuler, J. & Zander, P. (2006) Determination of trade-off-functions to analyse the
provision of agricultural non-commodities. – International Journal of Agricultural Resources,
Governance and Ecology 5: 309-325.
Ungaro, F.; Venuti, L. & Ciancaglini, A. (2006) Up-scaling approach for the Mugello Area I: soil
quality. MEA-Scope internal report, unpublished.

89
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

1.5 Open session 6 Biodiversity, management, policy and stakeholders

Identification of Conservation Opportunities: a study case of management and

conservation of biodiversity in rural landscapes based on field research in the
Central Andes of Colombia

F. H. Lozano-Zambrano, J. E. Mendoza , E. Jiménez, P. Caycedo-Rosales, W.

Vargas

Alexander von Humboldt Institute, Cra. 7 #35-20 Bogotá Colombia.

e-mail: [email protected]

The mountain forests are one of the most endangered ecosystems in the World due mainly
to the high degradation rates and habitat loss. This kind of forests has different kinds of
associated values such as, harboring high biological diversity, high proportion of endemism,
water sources, and biotic relationships (e.g. altitudinal migrations of species looking for
shelter and food), and resources for human communities, for instance, timber, charcoal,
medicinal plants, water sources for urban places, among others. In Colombia, the mountain
forests have around of 33.288.000 ha, 4.770.000 ha out of this amount of forest, is protected
areas and the rest are rural landscapes for livestock or agriculture. In Colombia, the
spreading productive system is livestock with more than 12.000.000 ha. In this way, is pretty
important go further and increasing the knowledge about structure and composition of biotic
communities in this kind of rural landscapes. We addressed four questions as the basis to
establish potential conservation strategies at local scale: What is the species composition in
fragmented landscapes? What is the turnover among different landscape elements? Are
rural landscapes harboring interesting species for conservation purposes such as endemic or
endangered species?

We worked with three target groups using rapid characterizations of birds, ants and plants
(trees and shrubs). This study was conducted in sub-Andean forests (1700 – 2100 m) in the
western slope of the Central mountain range in the Colombian Andes, specifically in the
Barbas river watershed, Municipalities of Filandia and Pereira, Colombia. We selected a
landscape window of 2500 ha with grasslands matrix, and recorded samples of our target
groups in six kinds of landscape elements, each one with different number of replicates.
Thus, eight big forest patches; five small forest fragments (3 – 6 ha), eight streams with
forest cover, five patches of forestry plantations (exotic species), and eight replicates in the
landscape matrix of grasslands. We made alpha, beta and gamma analysis for all three
target groups, and we are proposing a “conservation Index” to establish priority in landscape
elements which could represent higher conservation opportunities for all three biological
groups.

We found 94 ant species, 409 plant species (trees and shrubs, 74 spp. under threaten
category), and 156 bird species (4 globally endangered and 37 negatively affected by habitat
fragmentation). Alpha diversity is high for all three groups, especially in forest patches and
streams. For all three groups the species richness decreases within the gradient from forest
to productive systems. Regarding turnover (beta diversity), it was high for plants, while for
birds and ants has middle values. For birds we found 16 species were present in one of the
bigger forest fragments (Barbas river Canyon) which were not present in the other one
(Bremen forest preserve), and vice versa 20 bird species which were present in Bremen
forest preserve were not in Barbas river canyon forest patch. We also found differences in
the species abundances between these forest fragments.

Finally, we used three criterions (species richness, number of endangered and endemic
species) to build a “conservation index”. This index is based on the creation of three ranges

90
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

using percentiles (high, middle and low). The results for each biological group and for all
groups together are mapping. As we expecting, our “conservation Index” showed higher
values for forest patches. We suggest that activities such as maintenance and restoration of
this forest patches (through natural succession or ecological restoration) to increase size,
reintroduction of native species, and increase structural connectivity among patches using
linear landscape elements for instance hedgerows and biological corridors, for the elements
with higher values of the conservation index will ensure species conservation in our studied
landscapes.

This information is the first step in a conservation strategy for this landscape having in
consideration not only the necessities for biological conservation, but also social,
economical, and cultural interests for the inhabitants.

91
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

Investigation of the impact of herbivores on a Mongolian steppe vegetation and

the implications for conservation

M. A. van Staalduinen, M. J. A. Werger

Plant ecology and biodiversity, Utrecht University, Utrecht, The Netherlands.

e-mail: [email protected]

The large and relatively intact steppes of Mongolia have been a pastureland for thousands of
years and support large populations of wild ungulates and livestock. These steppes are of
great global importance.
Fig.1 Vegetation zones of Mongolia
(after WallisdeVries, 1996) with the
location of Hustai National Park
and the capital Ulaan Baatar.

In Mongolia the vegetation zonation

(Fig.1) corresponds rather well with
a gradient of decreasing
precipitation from the north to the
south (Van Staalduinen, 2005). The
forest steppe zone lays in the zone
with an annual precipitation
between 300 and 400 mm, and
consists of a mosaic of patches of forest and grassland. Further south, in the zone with a
precipitation between 200-300 mm, trees become rare and grasses dominate the vegetation.
This is the region of the typical steppe, or grass steppe, with vast areas of grasslands. The
Stipa steppe (Cymbario-Stipetum krylovii, Hilbig, 1995) is a characteristic community of this
region.
Nowadays overgrazing is a serious threat to the sustainability of the grasslands in Mongolia,
of which a significant proportion already is degraded. The high pressure on the steppe is
ascribed to the dramatic increase in the population and livestock over the past century. From
1918-1990 the population of the country increased almost three-fold (Kharin et al., 1999),
and the number of livestock, consisting of sheep, horses, cows or yaks, goats and camels,
increased 2.3 times (Kharin et al., 1999). These animals all had to live from the same plant
resources as before, since the country did not, or hardly, import any fodder, while the
biomass productivity of the steppe is very low (about 125 g m-2 y-1, Van Staalduinen, 2005).
The overgrazing resulted in a decrease of the vegetation cover, productivity and plant
species diversity in many degraded locations (Kharin et al., 1999). The original dominant
Stipa grasses were replaced by other grasses like Leymus chinensis, Cleistogenes
squarrosa and ruderal plant species (Hilbig 1995), or by the sedge Carex duriuscula. In
heavy degraded areas the grasses have disappeared altogether and are replaced by
Artemisia species. Sometimes the vegetation cover has completely disappeared and the
bare soil is left susceptible for wind erosion, leading to desertification.
In order to be able to develop a sustainable pasture management, a better
understanding is needed of the role of herbivory in the sustainability of the ecosystem. To
this end we conducted a field study in central Mongolia in the forest steppe region, on a
research site in Hustain National Park (Fig. 1), a national park located at about 100 km from
Ulaan Baatar. The takh (Equus przewalskii ) which is the last surviving ancestor of the
modern domestic horse, has been successfully reintroduced in the park. In 2005 a population
of 158 takh has established and lives freely in the Park.
We investigated the impact of different types of herbivores, a large and a small, on
the steppe vegetation (Van Staalduinen & Werger 2007; Van Staalduinen at al 2007). An
exclosure experiment was set up on the Stipa steppe, in which we examined the effects of 3
years of exclusion of takh and Siberian marmots on plant species abundance, plant biomass

92
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

and plant N-concentration (Van Staalduinen et. al 2007). We investigated whether shifts in
plant species dominance occur with increased grazing, and if this can be explained by
differences in growth responses to clipping. We tried to establish a link between the plant
species dynamics in the Mongolian steppe and physiologically based plant characteristics. In
a greenhouse experiment we compared the potential for compensatory growth of Leymus
chinensis and Stipa krylovii, two co-occurring grass species of the steppe, using a recently
developed technique of growth analysis. Compensatory growth was much stronger for the
rhizomatous Leymus chinensis than for the caespitose Stipa krylovii, and Leymus showed a
significant increase in its relative growth rate (RGR) after clipping, while for Stipa RGR was
negatively affected (Van Staalduinen and Anten, 2005). Apparently Leymus is more tolerant
to clipping (and by extrapolation to grazing) than Stipa and this could explain the shift in
dominance from Stipa to Leymus when grazing pressure increases in the Mongolian steppe.
The carrying capacity of the steppe areas, i.e. the maximum number of herbivores
which can be supported in an area, is an important issue for pasture management. When
estimating the carrying capacity the effect of compensatory growth should be accounted for.
Our results showed that compensatory mechanisms contribute enormously to the
productivity of plants after clipping.
Knowledge about the carrying capacity and forage quality of the steppe vegetation
was important for the reintroduction of the takh into Hustain National Park. The takh is seen
as a flagship species for the conservation of the park, whose general objectives are: the
restoration and conservation of the biodiversity (Bouman, 1998), development of a research
and training centre, development of eco-tourism, and improvement of pasture management
in the buffer zone around the park.

References
Bouman, I. (1998) The reintroduction of Przewalski horses in the Hustain Nuruu mountain forest
steppe reserve in Mongolia. An integrated conservation development project. Netherlands
Commission for International Nature Protection. Mededelingen 32.
Hilbig, W. (1995) The vegetation of Mongolia. SPB Academic publishers, Amsterdam, The
Netherlands.
Kharin, N.; Tateishi, R. & Harahsheh, H. (1999) Degradation of the drylands of Asia. Center for
environmental remote sensing (CEReS), Chiba University, Japan.
Van Staalduinen, M. A. (2005) The impact of herbivores in a Mongolian forest steppe. PhD-thesis.
Utrecht University, Utrecht, the Netherlands.
Van Staalduinen, M.A. & Anten, N.P.R. (2005) Difference in the capacity for compensatory growth of
two co-occurring grass species in relation to water availability. Oecologia 146:190-199.
Van Staalduinen, M.A. & Werger, M.J.A. (2007) Marmot disturbances in a Mongolian steppe
vegetation. Journal of Arid Environments 69: 344-351.
Van Staalduinen, M.A.; During, H.J. & Werger, M.J.A. (2007) Impact of grazing regime on a
Mongolian forest steppe. Applied Vegetation Science (in press).
Wallisdevries, M.F.; Manibazar, N. & Dugerlham, S. (1996) The vegetation of the forest-steppe
region of Hustain Nuruu, Mongolia. Vegetatio 122:111-127.

93
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

Analysis of recent ecosystem transformation processes in two sectors of the

buffer zone of Colombia's Pisba National Natural Park (1955 - 2001)

N.A. Ciontescu Camargo

Pontificia Universidad Javeriana. Department of Environmental and Rural Studies

e-mail: [email protected]

Introduction

National Natural Parks also host human communities that perform certain production
activities that may be transforming their natural landscapes. This research was aimed at
analyzing how anthropic activities performed in the buffer zones of the Pisba PNN have
influenced the space/time variations in the patterns of the landscape of the ecosystems
present in the park.

Results and discussion

The research found that the shrubby scrubland unit (Figure 1) is the natural unit bearing
the densest relative coverage. Nevertheless, anthropically originated units such as
grasslands, crop grasslands and shrubby grasslands also bear very dense relative coverage
rates.
1200

1000

800

ha 600

400

200

0
SG CG DF OF SS S OS Sc
Coverage

1955 1979 1991 2001

Figure 1. Dynamics of coverage changes for all periods (SG: Shrubby grassland; CG: Crop
grassland; DF: Dense forest; OF: Open forest; SS: Shrubby Scrubland; S: Scrubland; OS:
Open shrub; Sc: Scrubland)

Figure 2 shows the transformations of the various coverage units between the first and
the last period covered by the study. Natural units of dense forest, open forest and shrubbery
appeared highly transformed, while anthropically-originated units of scrublands and crop
grasslands gained a broader area during the past 46 years.

94
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

Open Open
Forest Shrub
Unit

Shrubby
Scrublan
d

Crop Scrubland
Grasslan
d

Shrubby
Grasslan
d Shrubs

Dense
Forest

Low (0-10%) Medium (10-20%) High (20-100%)

Figure 2. Flow chart of the main coverage changes during 1955-2001

Transformation processes have led to the development of agricultural and livestock

activities carried out basically by the local peasant communities. Human activities linked to
the slashing and burning of natural vegetation in an effort to broaden the agricultural and
livestock frontier have been a determining factor in the transformation of these highland
plateau zones. The shrubby scrubland unit has suffered the highest transformation rate. The
systems used to grow potatoes, lima beans, wheat and peas, and intensive and extensive
livestock-related activities have been responsible for the prevalence and increase of crop
grassland agro-ecosystems.

The rotation of agricultural and livestock activities around various coverage units
generates an increase in the number of patches of various vegetation mosaics in varying
succession states. According to Hofstede (2000), this is one of the main characteristics of the
current highland plateaus intervened by man.

According to Forman (1995), the consequences of fragmenting vegetation coverage units

such as these ones may include: increased isolation of patches, an increased number of
generalist species, greater number of habitats, an increased number of borderline species,
an increased number of exotic species and a higher extinction rate.

Other studies suggest that the transformation processes have occurred since pre-
Colombian and colonial times. Nevertheless, the current uses to which the territory is subject
have led to a constant acceleration in the transformation of the vegetation layer, resulting in
some remnant patches of forest and highland plateau vegetation.

References
Forman, R.T.T. (1995). Land Mosaics: the ecology of landscapes and regions. Cambridge University.
Hofstede, R. (2002). Los paramos andinos: su diversidad, sus habitantes, sus problemas y
sus perspectivas. Un breve diagnostico regional del estado de conservación de los
páramos. Congreso Mundial de Páramos.

95
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

Conservation priorities for threatened Yatsu valley landscapes in central Japan:

Evaluation based on ecological value and vulnerability

K. Takahashi1, K. Hara2 and K. Short 2

1
Tokyo University of Information Sciences, 1200-2 Yato-cho, Wakaba-ku, ChibaJapan.
e-mail: [email protected]
2
Tokyo University of Information Sciences, 1200-2 Yato-cho, Wakaba-ku, Chiba, Japan.

Introduction

To successfully balance development with preservation of biodiversity, a full and accurate

understanding of local topography and land use patterns, as well as zoning plans based on
this understanding, is required. Landscape ecology can be an effective tool for simplifying
and improving accuracy in this process (Forman & Godron 1986). Prior studies (Blankson &
Green 1991, Bunce 1996) have classified the natural environment by landscape types to
develop indicators for evaluating environmental value. In this research, landscape ecology is
employed to evaluate conservation priorities for threatened ‘Yatsu’ valley traditional
agricultural landscapes (Fujihara et al. 2005) in the southern Kanto region of central Japan.
Evaluation was based on both ecological value and vulnerability to development.

Study Area

The research was implemented in the city of Sakura, located in northern Chiba
Prefecture, about 40 kilometres northeast of Tokyo. The city is served by the Keisei and JR
railway lines, and is thus well within commuting distance of Tokyo. As a result, development
of suburban housing and associated commercial infrastructure is proceeding at a rapid pace,
thereby threatening traditional agricultural landscapes (Hara, 2000).

Methods

Landsat TM data and DEM were employed to classify landscape types, and five basic
types were identified and mapped:
• Marsh Landscape: Low-lying (less then 1 meter above sea level) marshland, comprising
Inba Marsh and its associated canals.
• Urban Landscape: Areas currently used as or designated for residential, commercial or
industrial development.
• Upland Agricultural Landscape: Farmland on top of relatively level uplands (20-30 meters
above sea level), used for non-irrigated vegetable fields, and fruit and chestnut orchards.
• Floodplain Agricultural Landscape: Low-lying areas comprising the original floodplain and
marshland that were reclaimed historically. Mostly planted in irrigated rice paddies.
• Yatsu Valley Agricultural Landscape: Narrow, branching valleys that cut deep into the
surrounding slopes. Valley bottoms traditionally planted in irrigated paddy.

In addition, 13 ecotopes, the most basic unit of which landscapes are comprised, were
identified and mapped against the distribution of the five landscape types. To evaluate the
ecological importance of each landscape type, the distributions of natural water seeps, as
well as that of two birds of prey, common buzzard (Buteo buteo) and gray-faced buzzard-
eagle (Butastur indicus), and five frog species were employed.

Results and Discussion

The results clearly showed that all these distributions concentrate in the Yatsu valley
landscape. Figure 1 evaluate individual Yatsu valleys, 17 valleys were identified and
96
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

assigned an overall ecological rating. In addition, vulnerability to future development was

estimated for each valley as a function of the distance from the nearest train station (Figure
2). The results of the research were used to prioritise valleys for conservation. The Azeta
Yatsu valley, for example, was ranked as the highest priority. Hopefully, these results will be
of use to government agencies involved in development and conservation planning in Sakura
City and in other structurally similar areas.

Figure 1. Evaluation of overall ecological Figure 2. Distance from nearest train

value for 17 Yatsu valley station (0.5-2.0 Km)
landscapes in Sakura City

References
Blankson, E. J. & B.H. Green. (1991) Use of landscape classification as an essential prerequisite to
landscape evaluation. Landscape and Urban Planning Volume: 21: 149-162.
Bunce R. G. H. (1996) ITE Merlewood Land classification of Great Britain. Journal of Biogeography
Volume: 23: 625-634.
Forman, R. T. T. & M. Godron. (1986) Landscape Ecology. John Wiley & Sons Inc, New York. pp.
619.
Fujihara, M., K. Hara, & K. Short. (2005) Changes in landscape structure of "yatsu" valleys: a typical
Japanese urban fringe landscape. Landscape and Urban Planning. 70: 261-270.
Hara, K. (2000) Landscape of Sakura-City. Natural Environment of Sakura City (ed Sakura Natural
Environment Survey Group). Shinzansha, Tokyo, pp. +7pp Sakura City. (In Japanese)

97
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

The rural policy as a tool for the natural resource management

D. Franco

Ministero dell'Ambiente e della Tutela del Territorio e del Mare, C. Colombo 44, Roma, Italy.
e-mail: [email protected]

Introduction

The European rural regions (OECD criteria) represent the 92% of EU25 landscape,
producing the 45% of the Gross Value Added and providing of the 53% of the employment.
Agricultural and forestry sectors account for the 8.3% of the employment and the 4.4% of the
GDP, covering the 77% of the land use, which is for the 12-13% designated as Natura 2000
and for the 10-30% designated as High Nature Value Farming Systems. Rural regions had
undergone a critical period in the last decades. The underlying approach adopted for rural
areas is the sustainable development, that is the engine of the EU policies on the basis of the
Göteborg - Lisbon strategies for a "European model" of development, where the employment
and development growth capacity has to be triggered by a best knowledge and sustainable
use of natural resources. The application of these strategies throughout multi-annual policies
undergoes to a monitoring and evaluation process that should allow an on-going comparison
between policies expectations, results and innovative knowledge.

The adaptive process to fit the expectations and the results: does it work?

In general the rural development policy (related to the compulsory integration and
subsidiarity with other EU, National and regional policies) appears to be a good strategic
approach in achieving a sustainable development. Policy programming tools seem actually to
cope with competitiveness, employment and natural resources sustainable management,
giving to the "externalities" a new marketable perspective and transforming the environment
in a competitive boost. The related economic instruments (schemes of measures) seem to
be progressively able to ensure the strategies aims, mostly coupled with a (annually!) review
system that should optimise the local fit to the policy. All these from the economic,
administrative and financial point of view; but two elements appear to limit this effectiveness.
The first is the bureaucratic viscosity in the strategies sharing and application at the
management level (in between EU and rural communities’ awareness); the LEADER axis in
the 2007-13 Rural Dev. Fund should contributes to correct this problem encouraging the
bottom up approach.
The second, and more general, is the lack of information feedback of the best knowledge
to the policy input, a central principle of the "European model" of sustainability that waken the
policy adaptive process implementation. This is related to the difficulties of science to inform
policy and management, to interact each other to jointly inform the policy and management
decision, and to the delay in the upgrade of policy and management output. These
information feedback difficulties, (very human) are coupled to: (i) the complexity of the rural
landscape, that intrinsically brings uncertainty that has to be communicated to the policy
makers and managers, to allow them to progressively adjust the solutions adopted; (ii) the
fact that disturbance (human and not), openness and heterogeneity are intrinsic features of
rural landscape, and that composition, structure and functions of a single rural ecosystem are
contingent on its history and spatial context. Then the complexity and uncertainty of this
system is coupled with the complexity of the policy measures used to implement the rural
sustainable development.

Research and feedback needs

Monitoring the environmental effect of policies and management solutions is essential to

cope with the high variability of rural systems, and on ground data are necessary to verify

98
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

expectation and to cope with uncertainty: this awareness is actually unclear in the ongoing
monitoring schemes. The actual lack of science-policy-management feedback and its
consequence has been already pointed out in the assessment of agro-environmental
measures effects on the environmental policy objectives: uptake figures do not give factual
information about the environmental results of their implementation, and do not give efficient
information to review the schemes to cope with the policy objectives. On the landscape
preservation and structural transformation side, similar consideration may be drawn: natural
resources' policy strategies and programs decoupled with spatial planning are not
necessarily correspondent to the policy objectives (e.g. Madsen, 2002). Considering that
uncertainty is a core concept of the nowadays non-equilibrium ecology, and most ecological
knowledge comes from managed systems far from a human free equilibrium climax, a first
solution could be to enforce the information feedback between theory and application by
means of a direct engagement of the scientific world with society, to promote upgraded
awareness in the policy makers to correctly drive the bureaucratic engine. Examples exist to
feed this exchange, which account for the intrinsic characteristic of the system and/or the
embedded social values, by means of participatory processes or considering the valuation of
shared societal values (e.g. Nassauer & Corry, 2004; Hughes, 2005; Bastian et al. 2006).

Conclusions

Some key points can be considered in the future landscape ecology contribution to the
process: 1) scientific bodies should encourage the participative approach with the local
actors and stakeholders before and during the programmes implementation; 2) the best
knowledge available should contribute to a clearer definition of environmental objectives at
the landscape scale pursued by single and mix of measures in each program; 3) a long term
scientific on the ground evaluation of environmental measures impacts has urgently to be
embedded in the programming structure, and a better evaluation at the landscape scale
would be possible by geo-referencing the measures application, allowing synergies with risk
assessment and natural resources management and planning; 4) environmental services
can represent a new market for rural enterprises’ income, but local research it is urgent to bid
them inside the schemes as shared public benefits, linking them to other emerging markets
such as bioenergies for their implication on climate change carbon market and on renewable
energy policies.

References
Bastian O; Krönert R. & Lipski Z. 2006. Landscape diagnosis on different space and time scales – a
challenge for landscape planning. Landscape Ecology. 21: 359-374
Hughes, F. M. R; Colston A; Owen Mountford. J. 2005. Restoring riparian ecosystems: the
challenge of accommodating variability and designing restoration trajectories. Ecology and Society
10: 12. [online] URL: http://www.ecologyandsociety.org/vol10/iss1/art12
Madsen L.M, 2002. The Danish afforestation programme and spatial planning: new challenges.
Landscape and Urban Planning. 58: 241-254.
Nassauer J.I; Corry R.C. 2004. Using normative scenarios in landscape ecology. Landscape
Ecology. 19: 343-356.

99
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

A computer aided, modular, spatial support tool for the systemic land-use
management toward environment conservation and socio-economic development

M.F. Falcetta1, F. Attorre2, C. Corugati 3

1
PROGES - Planning and Development Consulting, Via Appennini, 46 Roma, Italy,
e-mail: [email protected]
2
Department of Plant Biology – Sapienza University of Rome, P.le A. Moro 5, Roma, Italy.
3
Platypus S.r.l.Via Pedroni, 13 - 20161 Milano, Italy

Introduction

These days there is an internationally generalised concurrence on the merits of tackling

planning and management of environmental conservation and development in an
“Integrated”, “Holistic”, “Comprehensive”, etc way. Still, such a convergence in principle is far
from being effectively paralleled at the actual implementation level. The process toward a
consistently systemic action in ecological sustainability and human well being pursuits is still
very immature and cannot yet rely on standards and internationally accepted systemic
procedural/methodological mechanisms. During the last decade, the Italian Development
Cooperation has being promoting and supporting truly systemic objective planning and
management decision making. In pursuing such a challenge, a general-purpose DSS
software shell has developed which can be customised to implement Decision Support
System (DSS) systemic applications for Specific territorial and thematic contexts such as
Lake Turkana area -Kenia, Madre de Dios District – Peruvian Amazon, Bay of Guanabara –
Brazil and Galapagos Archipelago - Ecuador. DSS has been developed to help the
understanding and assessment of the relationships between the ecosystem components and
their related management issues; this in order to enable policy maker to take informed
decisions and monitor effectively their performances.

DSS description

Why use a DSS

• Facilitates the organization and effectiveness of the data gathering process.
• Stores all relevant data and information in coherent databases (GIS and tabular).
• Provides a conceptual framework for the analysis of information for both monitoring
and management.
• Allows specific management actions to be selected from a list of options, or to be
identified as “new options” to be added to such list.
• Allows to set target values for selected indicators for the prescribed management
actions.
• Facilitates efficient monitoring of the prescribed management actions .
• Provides efficient presentation tools to its users.
•
DSS core modelling framework: the Ecosystem Management Model (EMM)
• Is developed through participatory and iterative processes involving key stakeholders.
• Defines basic structural components and key functions of the involved ecosystems.
• Identifies key management issues and their cause-effect relationships in the
framework of relevant ecosystem components and functions.
• Defines a set of specific and quantitative indicators within Drivers-Pressures-Impacts-
States-Responses framework. Allows the integration of the identified management
issues and indicators into the relevant areas.
• Assists in monitoring the performances of management actions and/or policies by
evaluating their effects against a number of selected references, baselines and
benchmarks.
•

100
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

Base attributes for each index and indicators (benchmarks)

• Baseline - is the value at the time the SSDSS is put to use by the decision makers.
• Reference - is the ideal value which corresponds to the optimal or equilibrium state .
• Thresholds - are values (upper and lower with respect to the baseline) marking the
level above/beyond which an action must be taken.
• Targets - are goals of a prescribed management action.
•
Architectural framework of the DSS

Knowledge data base Core modelling framework: Ecosystem User interface for:
for: Management Model (EMM) developed
through participatory and iterative - Data entry and retrieval
- Problem diagram processes involving key stakeholders
for: - Upgrading and updating
- Indexes and knowledge data base
Indicators - Integrates sector knowledge into a
- Problem and solution
- Models collection systemic framework (builds EMM)
- Reference for monitoring (definition
analyses
- Definition and selection of
(computational modules of benchmarks and thresholds for
for sector simulation management options
indexes and indicators)
and scenario forecast) - Provision of alternative Integrated
scenarios
- Monitoring
- On screen and printout
presentations.

Tabular and spatial (GIS) Monitoring data base for:

- Values for indexes and indicators

- Thematic mapping layers (spatial reference for indexes and indicators)

DSS functionalities

- Socio-economic and
environmental
monitoring:
analysis of current
status and trends of the
selected indicators
within the DPSIR model.
- Management actions:
management actions
can either be selected
from a list of options
prompted by the DSS,
or identified by the user
and added as “new
options” to such list.
-Monitoring of
t ti

101
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

The effects of the attitude concept, public education campaigns, social discourse
and socio-demographic factors on water conservation behaviors in residential
landscapes

B.J. Andersen

Environmental Science Program, University of Idaho, Moscow, Idaho, USA.

e-mail: [email protected]

Since the early decades of the 20th century, the study of attitudes has been a predominant
theme in contemporary social psychology (Eagly and Chaiken, 2005). Attitudes consist of
psychological tendencies to evaluate a specific object, whether that object is a person, thing,
or concept and may be either enduring or temporary; either implicit or explicit; and simple
and unitary, dual or multiple in their nature. Therefore, the attitude concept is a common tool
utilized in attempting to understand human behavior. My literature review examines this
concept in relation to what we know about pro-environmental behaviors in residential
landscapes. One interesting gap in existing research, building on the Theory of Reasoned
Action and Theory of Planned Behavior work of Ajzen and Fishbein (2005) and others,
involves the lack of examination of the inconsistencies that occur when people fail to carry
out their intentions.

Recent evaluations of energy conservation, water conservation and recycling

behaviors show that the overall theory of attitudes influencing actions is useful. However,
using evidence from studies of residential landscape water consumption behavior and
energy conservation behavior (Stets and Biga, 2003), I show that local environmental
education campaigns, discussions with friends and family, and socio-demographic variables
also shape environmental behaviors (Campbell et al., 2004). For example, the Eco Team
Program approach studied in several Dutch cities by Staats et al. (2004), found evidence for
the informal advice of neighbors, friends, and family having a positive influence on the
adoption of proenvironmental behaviors. However, overcoming habitual behaviors, i.e.,
those not preceeded by overt, conscious intention, requires the aim to act differently and a
commitment to try and sustain a new action. Campbell et al. (2004) studied individual
household water consumption in Phoenix, Arizona (USA), assessing the effectiveness
(through actual individual water consumption records, rather than self-reports) of different
policy intervention techniques: devices/engineering solutions (low flow showerheads), people
(having seniors go to senior households to help install water-saving devices), prices, and
rules. Prices and rules were more certain in reducing water consumption than were people
or devices.

Municipal officials trying to convince residents to implement water conservation

practices in their domestic landscapes should concentrate on policies that will provide them
with opportunities to implement public education, promote social discourse, and focus on
people typically not practicing pro-environmental behaviors as well as addressing the pro-
environmental attitudes and values of people already practicing water conservation (Schultz,
2000). Values appear to be a fruitful avenue to explore because of data supporting them as
a basis for environmental attitudes (Schultz and Zelezny, 1999). Connected to this, Schultz
and Zelezny (1999) found self-enhancement to be associated with self-benefit and self-
transcence to be related with a broader cognitive representation of self, rather than
proenvironmental attitudes and behaviors being linked with altruism, egoism, and prosocial
behaviors. Analysis of this aspect of the social/human dimension of landscape challenges
landscape ecologists to integrate social science with biophysical science for a broader
landscape perspective.

102
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

References
Ajzen, Izak and Fishbein, Martin. 2005. The Influence of Attitudes on Behavior. Pp.173-221 in
Albarracín, Dolores, Johnson, Blair T., and Zanna, Mark P. (eds.). The Handbook of Attitudes.
Mahwah, NJ: Lawrence Erlbaun Associates.
Campbell, Heather E., Johnson, Ryan M., and Larson, Elizabeth Hunt. 2004. Prices, Devices,
People, or Rules: The Relative Effectiveness of Policy Instruments in Water Conservation. Review
of Policy Research 21(5):637-662.
Eagly, Alice H. and Chaiken, Shelly. 2005. Attitude Research in the 21st century: The Current State
of Knowledge. Pp.743-767 in Albarracín, Dolores, Johnson, Blair T., and Zanna, Mark P. (eds.).
The Handbook of Attitudes. Mahwah, NJ: Lawrence Erlbaun Associates.
Schultz, P. Wesley and Zelezny, Lynn. 1999. Values as Predictors of Environmental Attitudes:
Evidence for Consistency Across 14 Countries. Journal of Environmental Psychology 19:255-265.
Schultz, P. Wesley. 2000. Empathizing with Nature: The Effects of Perspective Taking on Concern for
Environmental Issues. Journal of Social Issues 56(3):391-406.
Staats, Henk, Harland, Paul, and Wilke, Henk A. 2004. Effecting Durable Change: A Team
Approach to Improve Environmental Behavior in the Household. Environment and Behavior
36(3):341-367.
Stets, Jan and Biga, Chris F. 2003. Bringing Identity Theory into Environmental Sociology.
Sociological Theory 21(4):398-423.

103
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

Catchment management in lowland England: an integrated, multifunctional

approach

C. Stoate

The Game Conservancy Trust, Allerton Project, Loddington, Leics. UK

e-mail: [email protected]

Introduction

The requirements of the Water Framework Directive (WFD) for improved chemical and
ecological status of watercourses are an important driver for land management and research
at the catchment scale. Associated wildlife and landscape, including their use by local
people are additional social and economic drivers for a new approach to catchment
management. This paper is based on the 67,000 hectare Eye Brook catchment, a tributary
of the River Welland in the English East Midlands. The catchment comprises arable and
livestock farming on clay soils, ancient semi-natural woodland, and Eyebrook Reservoir (a
trout fishery). Since 1992, the Allerton Project research and demonstration farm business in
the centre of the catchment has been the focus for research into farmland ecology, the
relationship between soil management and water quality and ecology, and practical
management measures for environmental improvement, including EU and UK government
and industry funded projects in collaboration with numerous research partners. Loss of soil
and nutrients, especially phosphorus (P), from land to water is a major concern because of
costs of water treatment and environmental consequences such as eutrophication. This
paper provides a summary of this research in the context of the changing agricultural and
economic climate.

Stream ecology

An electro-fishing survey of brown trout in the Eye Brook revealed low numbers at all ten
sites, and very low numbers of young fish at all but one site. Breeding success is low
because of sedimentation of gravel spawning habitat resulting from soil erosion associated
with surface runoff from arable land. The physical, chemical and biological processes
associated with transport of soil and nutrients from arable land to the stream are investigated
by the Phosphorus from Agriculture: Riverine Impacts Study (PARIS) which is based mainly
on a low input grass sub-catchment and an arable sub-catchment at Loddington. Median
particulate P concentrations are ten times higher in the arable catchment than in the low
input grass catchment, and in both, highest concentrations are associated with storm events.
Allerton Project monitoring revealed that field drains also contribute P to water, exceeding
concentrations likely to cause eutrophication on 60 – 100% of sampling occasions. Septic
tank discharges result in further elevation of soluble reactive P in some headwater streams.
The effect of this is diluted in winter but can be substantial under low summer flows. Diatom
and macro-invertebrate communities show a response to increasing P.

In-field mitigation

Two erosion plot based studies at Loddington assess the loss of sediment and nutrients
from fields via surface runoff. The Soil and Water Protection (SOWAP) project compares
minimum tillage with ploughing in Belgium and Hungary as well as at the Loddington site.
Loss of soil and nutrients to water are generally lower under minimum tillage than under
plough and soil organic matter, earthworm density and soil microbial biomass are also higher
under minimum tillage. Minimum tillage could have multiple environmental benefits, as well
as mitigating against diffuse water pollution. Crop establishment costs can be lower under
minimum tillage, with potential benefits for the farmer, but this is sometimes offset by
increased herbicide costs. The Mitigation Of Soil and Phosphorus loss (MOPS) project

104
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

compares the mitigation potential of cultivation direction and in-field barriers (‘beetle banks’),
as well cultivation type. Cultivation direction has the overriding influence on runoff and
nutrient loss.

Field edge mitigation

UK farmers prefer to avoid in-field measures to meet environmental objectives and the
Allerton Project has a long history of developing field edge measures for environmental
improvement. Perennial grass buffer strips are adopted at Loddington to provide wildlife
habitat, over-wintering habitat for beneficial predatory invertebrates, and to reduce surface
runoff entering watercourses. In one case, ditches and field drains are diverted into series of
pools within a wide buffer strip in a small food plain. P concentrations are 40 – 50% lower at
the outlet than at the inlet of these pool sequences. Although small and eutrophic, these
pools provide a valuable habitat for several aquatic invertebrates, including fifteen Odonata
species and seven nationally scarce beetle species. The associated vegetation also
supports five species of Orthoptera and breeding bird species such as whitethroat (Sylvia
communis) and reed bunting (Emberiza schoeniclus), and wintering snipe (Gallinago
gallinago), jack snipe (Lymnocryptes minimus) and water rail (Rallus aquaticus). Continuing
research is investigating the potential of smaller scale field edge features for nutrient
management and wildlife conservation, including nutrient behaviour, macro-invertebrate
communities, and birds in ephemeral ditches.

The social dimension

The project also introduces a social dimension. While regulation and economics are
important influences on catchment management, cultural influences also play a role. Values
attached to landscape provide an example. Knowledge of local history increases local
people’s cultural identity, sense of place, and ‘ownership’ of environmental problems and
opportunities. This is addressed through a four year project which engages local people in
exploration of the cultural and natural heritage at the catchment scale.

Integration and multifunctionality

This paper describes how hard science, as applied locally by national and international
institutions, can combine with cultural values of local people to build a sustainable approach
to catchment management. The integration of a number of research themes is essential to a
comprehensive understanding of catchment issues. The examples of grass field margins,
minimum tillage and buffer strip pools illustrate how this approach can bring multifunctional
benefits which can be cost effective and attractive to stakeholders. Meeting multiple
objectives simultaneously can therefore involve cost sharing. Such an approach is most
likely to succeed in meeting policy objectives such as those of the WFD, as well as those of
local people.

Acknowledgements
The research described here is in collaboration with the universities of Cranfield, Harper
Adams, Lancaster and Leicester, and ADAS, Centre for Ecology and Hydrology, the Ponds
Conservation Trust and RSPB, with funding from Defra, the EU, Environment Agency, HART
and the Lottery Heritage Fund.

105
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

Integrated river basin management plan: public involvement and stakeholder

participation

Z. Gulbinas

Nature Heritage Fund/Vilnius Pedagogical University, A. Juozapaviciaus 9, Vilnius, Lithuania.

e-mail: [email protected]

Introduction

In year 2003 PIN-Matra project “Management and Restoration of Natura 2000 sites
through an Integrated River Basin Management Plan of the Dovine River”, financed by the
Netherlands Ministry of Agriculture, Nature and Food Quality, was started in Lithuania. The
overall purpose of the project was to produce a Management and Restoration Plan for the
Dovine River Basin (total area 588.7 km2).

Approach

In 2002 the Zuvintas Lake and adjacent areas (Almalvas wetlands, fen meadows, drained
peatlands and buffer zones) were designated as a biosphere reserve. As these territories are
on the list of SCI, Lithuania is obliged to maintain its favourable conservation status. This
implies that the eutrophication has to stop and relevant management and restoration
activities have to be carried out.
Both the obligation to maintain the favourable conservation status of the Zuvintas Lake,
adjacent wet grasslands and bogs (based on the Habitats Directive) and the obligation to
restore the reference situation of the Dovine river (according to the WFD) make the area an
excellent pilot for integrating of the protection and restoration of nature values of European
significance into integrated river basin management. The project provided the bases for the
long term protection of the Zuvintas Lake, the re-naturalisation of the Dovine River and the
restoration of drained wetlands.
The project helped to improve the capacity and organisational build up of involved
Lithuanian governmental organisations in the implementation of Natura 2000 and the WFD.
Training was provided for local and central staff on the implementation of the WFD and the
Natura 2000. Dissemination of the lessons learned to other regions in Lithuania took place.
Involvement of local stakeholders in decision making and the awareness and knowledge of
the local population on biodiversity protection and sustainable water management and the
role of relevant EU legislation has been increased (Zingstra et. al., 2006).

Activities and results

Seminars and trainings

The project carried out a series of workshops and trainings:
1. Seminar on introduction of the project activities and EU requirements.
2. Seminar on the status of the Natura 2000 habitats and species in the Dovine River Basin.
3. Seminar on defining and assessing Favourable Conservation Status of Natura 2000
species and habitats.
4. Seminar on monitoring of biological quality elements for assessing WFD typology and
ecological status of inland surface water bodies.
5. Peatland Restoration Workshop with focus on the Amalvas and Žuvintas fen complexes.

Dissemination of project activities and results

An information leaflet introducing the project idea and tasks, presenting implementation
organizations, presenting information about the Dovine river basin was prepared and

106
Theme 1. Landscape, stakeholders, land use and policy
1.5 Open Session 6: Biodiversity, management, policy and stakeholders

published, and distributed for the national, regional and local stakeholders. At the end of the
project a “Door-to-door” newsletter was prepared and published where main results and
outcomes of the project with recommendations were presented. This publication was
oriented mainly to regional and local stakeholders and to the people living in Dovine basin.
The web site for Žuvintas Biosphere Reserve was created and domain www.zuvintas.lt
was registered. Project home page and Extranet were developed inside the Žuvintas
biosphere reserve web site.
Information about the project activities and project results were published in local and
national newspapers. In May 2006 the Lithuanian National TV broadcasted a documentary
about the Žuvintas Lake and the PIN-Matra Project.

Activities in the region

Municipal schools were visited to educate and teach school teachers and students in the
Dovine Basin in knowledge about the importance of biological diversity and nature protection.
8 meetings in three municipalities were organized and about 630 pupils and 90 teachers took
part in the meetings. For different levels of the pupils lectures were presented on subjects
like: “Why are wetlands useful for us?”, “What are the watersheds? Survey the Dovine
watershed”, “What are the habitats of Natura 2000? Why is it necessary to safe and
restore?”, “Wetlands habitats and their state in the Dovine watershed”, “Forests, meadows
and water habitats and their variety in the Dovine watershed”. For the primary classes’ pupil
after a lesson about wetlands practical period “I draw a wetland” was organized. For the
teachers special discussion at the round table “Natura 2000 habitats, their protection and
restoration: ecological education and local community view” was arranged.
The main results and outcomes of the project were presented during the regional
conference organized for regional and local stakeholders. Representatives and responsible
specialists from different institutions such as counties, municipalities, Environmental
Protection Departments, forest enterprises, localities and protected areas administrations
acting within the basin took part in this conference.

Final conference
Project implementation process, main results and outcomes, also project evaluation and
recommendations prepared by project were presented during the Final conference. Around
60 participants took part in the conference: members of the project Steering Committee,
responsible officers from the Ministry of Environment, officers from Environmental Protection
Agency and State Service for Protected Areas, foreign and local experts of the project,
guests discussed the main project issues.

Acknowledgments: Support from the Dutch Government through PIN-Matra programme (project No.
2003/040) is greatly acknowledged.

References

Zingstra, H (final edit); Gulbinas, Z; Kitnaes, K. & Querner, E. (2006) Integrated Water and
Biodiversity Management in the Dovine River Basin, Lithuania. Wageningen International, the
Netherlands.

107
Theme 1. Landscape, stakeholders, land use and policy
1.6 Open Session 8: Landscape and nature perception

1.6 Open Session 8: Landscape and Nature Perception

From re-discovering to dealing with nature. The Netherlands as an example.

S. Kost

University of Kassel, Department of Architecture, Urban Planning, Landscape Planning,

Institute for Empirical Research of Planning Issues, Georg-Forster-Str. 7, Kassel, Germany,
e-mail: [email protected]

Introduction
In the 1980s there started in The Netherlands a kind of rediscovering of nature connected
with a change in how nature is seen and evaluated by experts and people. In regions where
so-called new nature was constructed new forms of economy and cooperative planning have
start to develop. More and more people were involved in the discussion as how to preserve
and develop nature. New nature became a symbol for the change in politics and planning in
general.

New ways to handle landscape

Transformation processes either in the meaning of and handling of nature depends on the
accumulation of problems and events. A beginning is seen to be visible in different fields,
such as water management, spatial planning or ecology. In these fields experimental forms
of planning and projects have been developed, which combine different aspects, for example
ecology and economy, ecology and aesthetics, housing and landscape conservation etc.
Step by step a process of invention and innovation in handling nature is taking place, with the
elements of a new paradigm are becoming visible.

Making Landscape

Two key issues are important in the development of landscapes in The Netherlands:

The shortage of land leads to the production of new land for agriculture and settlement. The
knowledge of land engineering and land-management corresponds with an „everything goes
mentality“- everything is thinkable anywhere. This has been true up to now: one can produce
any kind of landscape as an economic resource anywhere.

The renaissance of nature in the middle of industrialized and technical designed territories is
made clear: the nature, which was believed to be lost, develops spontaneously and in an
unplanned way. The concept „new nature“ refers to this process of self-creating landscape
and renewed discourse on nature as a potential of the economic development of regions.
New nature has become a marketing strategy for several economic branches and political
institutions. Thus nature has become an integral part of regional socio-economic
development and planning.

Dealing with nature

If one concentrates on the active participants of this process one sees how the status of
power of different groups is changing and a new pattern of strategies has been developed:
The issue of new nature opens new forms of cooperation with former conflicting parts:
economic and ecologic interests, business and nature conservation are finding new
coalitions in an instrumental sense.

108
Theme 1. Landscape, stakeholders, land use and policy
1.6 Open Session 8: Landscape and nature perception

Companies are becoming sponsors for nature-protecting projects and demonstrating an

image which conforms to natural processes; in sponsoring projects of new nature they
convince the local government that their business will not destroy the natural environment.
One outstanding example is the re-naturalization of the Maas, where the river will be more
natural in the future by changing its course. The exploitation of the gravel is the basis of
economic surplus. Another example is the planning of new nature means of making the
planning of new houses more attractive and developing new recreational areas. New nature
is becoming a symbol of any kind of innovative action.

In general, in different economic fields and geographic regions, one can show that the idea of
new nature has already been helpful in constituting new alliances of active participants,
innovative coalitions of development and a creative correlation of economic interests and
nature conservation. The presentation will show this on the basis of a number of case studies
and interviews with experts and citizens in different parts of the Netherlands.

109
Theme 1. Landscape, stakeholders, land use and policy
1.6 Open Session 8: Landscape and nature perception

Do people experience landscapes as they state they do? Stated and observed
landscape preference assessment on site.

M. Sevenant, M. Antrop

Ghent University, Department of Geography - Krijgslaan 281, S8, Ghent, Belgium.

e-mail: [email protected]

Introduction

Research on landscape preference should consider a series of preferences for

landscapes as elicited (1) in general and verbal terms, (2) in most landscape preference
studies, i.e. using visualisation, and (3) in concrete behaviour in daily life (de Groot and van
den Born, 2003). The present study attempted to compare stated and observed landscape
preferences. The research questions concern the agreement between what respondents
express as their landscape preferences and the observations and judgements they make in
the specific landscapes, and how these relate to their environmental beliefs.

Research approach and method

Questionnaire
A questionnaire on site was carried out during field excursions in France and Belgium, on
different days with similar weather conditions during April-May 2006. Respondents (n=108)
were recruited within different students’ groups who followed a similar course on landscape
science. In the first part of the questionnaire, they were asked about their general landscape
preference in an open-ended question. Their environmental beliefs were assessed using the
original 12-item version of the New Environmental Paradigm scale on 5-point scales, with
high codes corresponding to high ecocentrism (Dunlap and Van Liere, 1978). The second
part dealt with valuation of 11 concretely experienced landscape vistas in situ. Respondents
were asked to describe, in decreasing order of importance, three aspects why they did (not)
like the landscape (e.g. Galindo and Rodríguez, 2000) and to rate the vista on ‘beauty’,
giving a score between 0 and 10 (e.g. Kaplan and Kaplan, 1989; van den Berg et al.,1998).

Analysis
The 4 open-ended questions on stated and observed landscape preference were first
text-analyzed using WinMAX software. Per question, all text segments were assigned a
(sub)code in a hierarchical code system, covering the aspects of landscape evaluation as
summed up by the respondents. The 19 most frequent (sub)codes in all 4 questions (see
table 1) were then transformed into dichotomous variables, with the values depending on
whether the respective aspect being mentioned or not in the answer per question. This
yielded 19 dichotomous variables for each of the 4 open-ended questions, ready for further
quantitative analysis.

Results

Nine of 19 most frequent aspects that were mentioned in stated general landscape
preference also show up in landscape observations on site (see table 1, italic font style). The
average score on ‘beauty’ differ with the occurrence or otherwise of only 3 of the 19 aspects
of stated landscape preference (t-tests or Mann-Whitney U tests, p<0.05). These are:
diversity, relief, and human influence s.l. The average score given by the respondents on
‘beauty’ of the landscape on site is not significantly related to the average score on New
Environmental Paradigm (Spearman’s rho correlations).

110
Theme 1. Landscape, stakeholders, land use and policy
1.6 Open Session 8: Landscape and nature perception

Table 1. Agreement of most frequently mentioned aspects between stated and observed
landscape preference (italic font style: significant agreement; (**): p<0.001; (*): p<0.05).

Stated general landscape Observations of landscape preference in situ

preference First aspect Second aspect Third aspect
(n= frequency of mentioned Mc Kappa Mc Kappa Mc Kappa
aspect) Nemar Nemar Nemar
Agriculture (n=1) .000 -.018 .000 -.018 .000 .060
Buildings (n=17) .000 .037 .000 .104 .002 .160
Closed landscape (n=0) (not calculated as n=0)
Coherence (n=20) .011 .058 .115 .274 (*) .012 .201
Colours (n=16) .503 .335 (**) .041 .417 1.000 .120
Contrast (n=1) .004 .168 .219 -.016 .375 -.015
Diversity (n=25) .065 .198 (*) .851 .291 (*) .324 .123
Human influence s.l. (n=10) .041 .239 .008 .045 .629 .174
Infrastructures (n=4) .035 .065 .000 .039 .000 .112
Landscape structure (n=16) .000 .191 .000 .106 .003 .089
Nature s.l. (n=38) .001 .262 .000 .182 .000 .077
Open landscape (n=14) .267 .371 (**) .210 .197 (*) .057 .241 (*)
Quietness (n=57) .000 -.043 .000 .173 .000 -.003
Relief (n=2) .000 .000 .000 .012 .000 .029
Vast (n=25) .024 .260 1.00 .050 .020 -.035
Vegetation (n=27) .004 .180 .012 .163 .081 .140
View (n=18) .014 .175 .080 .116 .678 .281 (*)
Water (n=15) .362 .006 .265 .061 .824 .267 (*)
Weather (n=26) .000 .038 .000 .216 .000 .056

Conclusion

From this study it appears that people do not always pay attention to the same aspects in
the landscape as the elements they state to be important for evaluation when not being
exposed to a specific landscape. Similarly, their environmental beliefs are not significantly
related to the landscape judgements on ‘beauty’ they make on site.

References
de Groot, W.T. & van den Born, R.J.G. (2002) Visions of nature and landscape type preferences: an
exploration in The Netherlands. Landscape and Urban Planning. 3: 1-19.
Dunlap, R.E. & Van Liere, K.D. The 'New Environmental Paradigm': a proposed measuring
instrument and preliminary results. Journal of Environmental Education. 9: 10-19.
Galindo Galindo, M.P. & Corraliza Rodríguez, J.A. (2000) Environmental aesthetics and
psychological wellbeing: relationships between preference judgements for urban landscapes and
other relevant affective responses. Psychology in Spain. 4: 13-27.
Kaplan, R. & Kaplan S. (1989) The experience of nature: a psychological perspective, Cambridge
University Press, Cambridge .
van den Berg, A.E.; Vlek, C.A.J. & Coeterier, J.F. (1998) Group differences in the aesthetic
evaluation of nature development plans: a multilevel approach. Journal of Environmental
Psychology. 18: 141-157.

111
Theme 1. Landscape, stakeholders, land use and policy
1.6 Open Session 8: Landscape and nature perception

Integrating geographic information into scenic assessment of a catchment basin

S. Yamashita 1and N. Nakamura 2

1
Department of Civil and Urban Design Engineering, Kyushu Sangyo University, 2-3-1
Matsukadai, Higashi-ku, Fukuoka, Japan.
e-mail: [email protected]
2
Center for Landscape Research, Kyushu Sangyo University, 2-3-1 Matsukadai, Higashi-ku,
Fukuoka, Japan.

Introduction

Suburban areas, or middle landscapes, can serve as ideal human habitats and
demonstrate how human-environment synergy can be sustained (Marx, 1964). Middle
landscapes tend to vary in type and are susceptible to either urban or natural encroachment
(Tuan, 1998), which makes assessment and management of aesthetic quality quite difficult.
There are two different types of approaches to landscape quality assessment: the
ecologist’s approach and the engineer-architect’s approach. The former, landscape ecology,
emphasizes exploring causes and effects of the spatial heterogeneity of a region whereas
the latter, landscape architecture/engineering, focuses on investigating perceptual features of
the region for design and its management. Landscape ecology often uses and analyzes
aerial photographs and satellite images. Landscape architecture/engineering establishes a
designated viewpoint on the ground representing ordinary, “down-to-earth” experiences of
landscapes. Landscape ecology addresses a wide-ranging area of the region whereas
landscape architecture/engineering emphasizes a relatively small locale encompassing the
viewpoints.
Combining these contrast approaches enables environmental planners and managers
to address the geographic information relevant to ordinary perceptions and aesthetic
assessments of landscapes. It also allows us to appreciate the aesthetic attributes of the
region and examine places beyond the vicinity of the designated viewpoint. The combination
of these two approaches may thus contribute to identifying elusive middle landscapes and
assessing them aesthetically in the geographic context of the region.
This study will identify the middle landscapes in a catchment basin of a Japanese river
and obtain information relevant to the assessment and management of the basin by
integrating GIS information of the basin into an aesthetic assessment of the river landscape.
The study explores the roles the middle landscape plays in the perception and evaluation of
landscapes.

Procedure

The selected catchment basin is of the Mikasa River, which runs from upstream,
natural areas to downstream, intensely-urbanized/densely-populated areas.
To integrate geographic information into perceptual/aesthetic data, a digital surface
model of the catchment area was created using aerial photographs which were classified
according to land use. Panoramic video images were taken both upstream and downstream
from all the 66 bridges along the mainstream of the river. View-shed areas within the frame
of each of the panorama images were generated for the 3-D individual land use areas. The
composition of each panorama image was analyzed in terms of water surface, vegetation,
man-made structure and sky. Forty-one college students evaluated and rated the images
using the Scenic Beauty Estimation procedure to standardize the perceptual assessment
data thus obtained.

112
Theme 1. Landscape, stakeholders, land use and policy
1.6 Open Session 8: Landscape and nature perception

Results

Three types of middle landscape were identified by factor analysis and cluster analysis:
1) a perspective of distant mountains and forests with anthropogenic structures in the
foreground (mountain-forest type); 2) view consisting of rice paddy fields, farmlands, natural
areas and/or vacant land (country-nature type); and 3) view of water surface with commercial
facilities (developed waterfront type). A “naturalness” factor, which is relevant to vegetation
and man-made structure and is independent of the three types, was also identified; views
were characterized by this factor as well.
The mountain-forest type was observed upstream from the viewpoints at the bridges
over the middle reaches of the river. The viewpoints of the country-forest type were located
within the upstream areas, and its directions of view were confined to downstream. The
developed waterfront type was located in the mid-to-lower reaches and viewed in the
directions of both up- and downstream. The views in the lower reaches tended to be
characterized by the “naturalness” factor.
The mountain-forest type middle landscape and the country-nature type tended to be
rated high in preference whereas the developed waterfront type tended to be less preferred.
The factor score for the mountain-forest type had a significant correlation with the evaluation
of the river landscapes (Table 1). The “naturalness” factor score was slightly correlated with
the landscape evaluation (Table 1); the projected area of man-made structure in an image
was significantly correlated with preference (r=-.32, p<.01); however, the relationship
between the area of vegetation and preference was not significant due to vegetation’s
complex effects of obstructing visual penetration and covering man-made structure.

Conclusions

The findings indicated that the visibility of middle landscapes, particularly the
perspective of distant mountains and forests, has a positive impact on the view from the
river, even if they have conspicuous anthropogenic structures in the foreground. Conversely,
the combination of commercial facilities with water tends to be rated relatively low.
It is clear that the perspective in the upstream direction from the mid-stream areas of
the river is most important. Keeping buildings low in this direction is thus essential in the
planning and management of the catchment area. From middle- to downstream areas, the
combination of man-made structure, vegetation and water surface needs to be improved by
taking the ambivalence toward vegetation into account.

Table 1. Correlation coefficients between factor scores and evaluation (SBE).

Mountain-forest Country-nature Developed waterfront Naturalness

0.408** 0.087 0.086 0.186+

**: p<.01, +: p<0.1

References
Marx, L. (1964) The Machine in the Garden: Technology and the Pastoral Ideal in America. Oxford
University Press, New York.
Tuan, Y.-F. (1998) Escapism. Johns Hopkins University Press, Baltimore and London.

113
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

1.7 Posters

Improved Decision-Making for Food Security, Income and Environmental Services

through Modeling Cropping Systems and Optimizing Land Use Options: A Case
Study of Taita Hills, Kenya

D. W. Odede-Oremo, MSc

e-mail: [email protected]

Regional GIS Office, Juba Southern Sudan, United Nations Mission in Sudan

Climate changes and rapid population growth cause increasing pressure on the East African
highlands. The results of the pressure are manifold: intensified agriculture, decreasing
amount of forestland, loss of biodiversity, intensified land degradation and soil erosion.
These consequences introduce high demands on land use and land use planning in these
areas. The Taita hills are one of the highland areas in East Africa which has gone through
significant changes in land use over the past decade. Being part of the Eastern Arc, Taita
Hills is very valuable and rich in biodiversity and has many endemic mammal, bird, and
butterfly species. The hills were once forested with cloud forest, but today, only few larger
patches of indigenous forests are left. These forest patches are not only valuable for the
endemic animal and vegetation species, but also for local people because of their
importance as sacred forests. Parts of these forest remnants are sacred or traditionally
protected. The forests have suffered substantial loss and degradation since the early 1960s.

The current land use practices in Taita Hills exhibit inefficient patterns that are of a major
concern for sustainable rural development. The biotopes in Taita Hills have suffered from
quantitative as well as qualitative decline primarily as a result expansion of land under
agricultural use. Currently, there is growing need to implement on-farm biodiversity and soil
conservation strategies and re-forest the biotope. Faced with decreasing farm sizes due to
constant land sub-division, the local farmers’ constantly make adaptations to their farming
practices to meet the basic house-hold food economy demand. While it is prudent that
farmers’ minimize environmental impacts of their decisions, their decisions on choice of
enterprises is constrained by poverty, limited access to technologies, lack of stable markets
and weak institutional capacity to respond to environmental and market shocks.

For each parcel of land, the farmers choose a type of use from which they expect to derive
the most benefit considering a variety of issues: (i) their personal objectives and constraints,
(ii) the given set of biophysical parameters, and (iii) the institutional, cultural, and legal
attributes of the land parcel. As these issues vary in space and time, so does land use
choices, resulting in a spatio-temporal mosaic in land use types. Little is known about the
long-term ecological and economic consequence of such changes in land use hence, making
sustainable rural land-use policy formulation a daunting task. For instance, changing the plot
size of one crop and re-allocating to another has practical implication on land management
and soil erosion as the change in crop type may affect soil water infiltration, run-off, evapo-
transpiration and plant-water budget.

The objectives of the proposed doctoral study are to: (i) develop cropping strategies that can
improve cash income and nutritional quality using optimizing models, (ii) evaluate the
implications of the change in crop enterprise choice on soil erosion and biomass production,
(iii) assess the costs of agricultural land use changes and compare them with the ecological
benefits from expected vegetation changes and reduced fragmentation of (semi-) natural
terrestrial biotopes, and (iv) develop a land use conflict resolution process that uses a
Geographic Information System (GIS) based decision support system to optimize land use
allocation in Taita Hills.

114
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

The multifunctional character of the landscape in areas of intensive agriculture:

towards a sustainable planning strategy

J. Ruiz, G. Domon

Department of Landscape and Environmental Design, University of Montréal, C.P. 6128,

succ. Centre-ville, Montréal, Québec, CANADA.
e-mail: [email protected]

Introduction

Many functions of landscapes were greatly altered over the last decades following the
transformations occurring in agriculture. In areas suitable for agriculture, trends towards the
homogenization of landscapes were coupled with a decrease in the number of farms,
creating a risk of devitalization for rural zones. Thus, the restoration of various landscape
functions, particularly the environmental functions, is urgent. In addition, issues related to
quality of life, because of the growing demands of urbanites towards landscapes, will have to
be considered. In that context, an interdisciplinary research project was conducted for three
years within a southern Québec agricultural watershed (1200 km²). The main objective of the
project was to identify landscape structures that would allow for both the reintroduction of
past functions, as well as being responsive to new functions (residential, amenities, etc.).
This project posited that the multifunctional character of landscapes could be maintained or
reintroduced solely if farmers, non-farmers and communities would receive benefits.

Methodological strategy

First, an ecological classification of the study area allowed us to select three types of
landscape patterns representative of areas of intensive agriculture in Quebec. Within these
three landscape contexts, the research project was developed in three parts. The objective of
the first part was to find landscape structures and patterns likely to promote the presence of
insects beneficial to agriculture, in order to decrease the use of insecticides by farmers.
Aphid populations present in corn fields were sampled, and using multivariate statistics
(redundancy analysis), these insect data were analyzed jointly with landscape
characteristics. The second part was to focus on the landscape structures valued by farmers
and non-farmers. A total of 46 semi-directed interviews were thus conducted among resident
populations of the three different landscape contexts. These data were analyzed
qualitatively. The third part was to measure the influence of agricultural patterns on water
quality during the last five years. Finally, the development of landscape scenarios (presented
visually and using maps) integrated the results of the three parts of the project.

Results and discussion

The results suggest the existence of landscape structures that are beneficial to both
individuals and communities, and that are able to support multiple functions (environmental,
aesthetics, amenities, etc.). The importance of forested areas, the size of agricultural fields,
and the types of crops were among such landscape structures and patterns. From the point
of view of landscape design in areas of intensive agriculture, these results also suggest the
necessity of developing new practices, no longer solely oriented towards protection
measures, but also capable of allowing the coexistence of the different functions of
landscapes, including those associated with aspects of the quality of life of residing
populations.

115
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Understanding graziers’ decision-making in the Great Barrier Reef catchment to

improve environmental, social & economic objectives

I. Bohnet

CSIRO Sustainable Ecosystems – Tropical Forest Research Centre, PO Box 780, Atherton
4883, Queensland, Australia. e-mail: [email protected]

Grazing lands cover over 75% of the Great Barrier Reef (GBR) catchment. The major
livestock production—beef—contributes over $3,000M to the national economy annually and
employs approximately 9000 people (Productivity Commission, 2003). These grazing lands
also contribute a substantial proportion of the sediments and nutrients that drain into the
GBR lagoon (Brodie et al., 2003). Recent research shows that the quantity of sediments and
nutrients lost from these grazing lands is strongly dependent upon grazing management
practices (Roth et al., 2003). This has led to growing public concern about the environmental
performance of the beef industry and increasing pressures on graziers to change their
management practices to decrease off-property impacts.
It is believed that improvements in the quality of water draining into the GBR lagoon may
be more easily achieved if science-based improved grazing management practices can
demonstrate improved economic and environmental performance for graziers, leading to
“win-win” outcomes for all concerned (Gordon and Nelson, 2007). However, graziers’
decision-making is largely based on experiences and the perception of potential outcomes
from changes in management actions. He or she will select the management actions that
provide, in his/her opinion, the highest chance of attaining their personal objectives. The
various combinations of management actions from which a grazier might choose are almost
endless. Therefore, this research suggests that it is important to understand graziers’ mental
models (Abel et al., 1998) to develop schemes, e.g. educational programs, demonstration
sites, financial incentives, tax rebates, that can be taken up by graziers and assist them to
achieve a multitude of objectives.
Results from nine in-depth interviews with graziers and land assessments of their
properties in the Bowen-Broken, a small sub-catchment of the GBR catchment, support the
importance of identifying graziers’ mental models. The interviews provide insights into a
range of decision-drivers that have led to degraded grazing lands, indicating that the number
of schemes currently available to graziers only attracts a small number of graziers, and
therefore, additional schemes, that are tailored to address the constraints graziers face, and
regulatory measures, may be required to achieve the desired outcomes.

References
Abel, N.; Ross, H. & Walker, P. (1998). Mental Models in Rangeland Research, Communication and
Management. Rangeland Journal 20(1), 77-91.
Brodie, J.; McKergow, L.A.; Prosser, I.P.; Furnas, M.; Hughes A.O. & Hunter H. (2003). Sources
of sediment and nutrient exports to the Great barrier Reef World Heritage Area. Australian Centre
for Marine and Tropical Freshwater.
Gordon, I. J. & Nelson, B. (2007). Reef Safe Beef: Environmentally sensitive livestock management
for the grazing lands of the Great Barrier Reef catchments.
Productivity Commission (2003) Industries, Land Use and Water Quality in the Great Barrier Reef
Catchment. Research Report, Canberra, Australia.
Roth, C.H.; I.P. Prosser; Post, D.A.; Gross, J.E. & Webb, M.J. (2003). Reducing Sediment Export
from the Burdekin Catchment. CSIRO Land & Water, NAP3.224, Burdekin Dry Tropics Board,
Townsville, Australia.

116
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Do woods help biological control of cereal aphids by hoverflies?

F. Arrignon 1, A. Ouin 2, B. Bouyjou2, M. Deconchat, 2, C. Monteil2, J.P. Sarthou2

1
UPRES EA 1293 ECODIV, Université de Rouen, Bât. IRESE A, place Emile Blondel, Mont Saint
Aignan.
e-mail: [email protected]
2
UMR DYNAFOR, INRA / INP-ENSAT, Avenue de l'Aagrobiopôle, BP 32607, F-31326 Castanet
Tolosan

Introduction
As an aphidophagous, Episyrphus balteatus (De Geer, 1776) is an auxiliary species for
agriculture, particularly at the end of winter-beginning of spring; its densities may be
influenced by the amount of natural habitat in landscape. Our hypotheses were that i) E.
balteatus overwintering success and spring abundances are higher in woody landscape
(providing shelter and feeding) than in a less wooded landscape.

Methods
Spring abundance and winter survival were studied in the same two landscapes
differing by wood density. The field studies of the spring abundances held in 2003, 2004 and
2005 in 14 wheat parcels spread in two landscapes of south western France. Aphids,
syrphids eggs and larvae were counted. A multi-agent model of winter survival was
developed, on CORMAS plateform to predicts the abundance of hoverfly at the end of the
winter, according to winter temperature, landscape composition and structure, and individual
behavior (Arrignon et al. 2007).

Results
Winter survival was higher in woody than in less woody landscapes, the importance of
flowered hedge bank or meadows was also pointed out. While in 2003 simulated winter
survival corresponded to high early abundance in wheat, in 2004 and 2005 early abundance
in wheat did not differed between landscapes (Figure 1).

Figure 1. Simulated survival at the end of

the winter (mean and SD) and hoverfly
(eggs and larvae) wheat field abundance
(mean and SD) in the following early spring
(4 first samplings), in 4 landscapes:
woody1, woody2, LessWoody1,
LessWoody2 (wood cover: 25%, 17%,
3.6% and 1.8%) in 2003, 2004, 2005.

References

Arrignon, F., Deconchat, M., Sarthou, J.P., Balent, G., & Monteil, C. (2007) Modelling the
overwintering strategy of a benficial insect in a heterogeneous landscape using a multi-agent
system. Ecological Modelling, accepted.

117
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Five centuries of landscape change in Colombia: a conservation perspective

A. Etter

Depto. Ecología y Territorio, Facultad Estudios Ambientales y Rurales, Universidad

Javeriana, Tr 4 Nr 42-00 Piso 8, Bogotá, Colombia.
e-mail: [email protected]

Introduction
The spatial extent and the patterns of ecosystem and landscape transformation we
observe today are the consequence of historic settlement patterns dating back hundreds or
thousands of years (Foley et al., 2005), and with varying cultural, socioeconomic and political
contexts (Wardell et al., 2003; Lunt and Spooner, 2005). Analyzing and reconstructing land
use change and the historical landscape patterns can help advance the understanding of the
dynamics and persistence of present-day landscapes and ecosystems, and therefore
contribute to their conservation. However, longer term historical analyzes are still very
scarce. This paper presents a spatially explicit historic approach to analyze the human
transformation of broad Colombian ecosystems at a national scale, identifying areas and
ecosystems with higher transformation impacts.

Methods
I applied a historic spatial approach to reconstruct the landscape change since 1500,
based on historic demographic, settlements, economic and land use data, biophysical data,
and present day and reconstructed ecosystem maps, applying GIS analysis. The analysis is
based on meaningful historic periods and their land use change drivers.

Results
Overall rates of landscape clearing are estimated to have increased from around 10,000
to 280,000 ha per year during the last 500 years. There are large contrasts in the rates,
spatial extent and patterns of transformation of different ecosystems. The largest impacts are
on Tropical Dry and Sub-humid forests, and the high-Andean forests. However, during the
last 100 years Humid Tropical Lowland and the sub-Andean forests show the highest rates of
change. Cattle grazing progressively became a major driver of land clearing and
transformation in Colombia. Due to historic events some areas show processes of land
abandonment, regrowth and eventual re-clearing decades or centuries later. Regions with
longer and more intense disturbance regimes should show stronger and larger impacts on
native ecosystems and species. The study provides an insight into the environmental history
of Colombia, inviting the formulation and testing of new hypotheses about the impacts of
human land use on ecosystems in a longer time frame.

References
Foley, J.A., DeFries, R., Asner, G.P., Barford, C., Bonan, G., Carpenter, S.R., Chapin, F.S., Coe,
M.T., Daily, G.C., Gibbs, H.K., Helkowski, J.H., Holloway, T., Howard, E.A., Kucharik, C.J.,
Monfreda, C., Patz, J.A., Prentice, I.C., Ramankutty, N. and Snyder, P.K., 2005. Global
consequences of land use. Science, 309 (5734): 570-574.
Lunt, I.D. and Spooner, P.G., 2005. Using historical ecology to understand patterns of biodiversity in
fragmented agricultural landscapes. Journal of Biogeography, 32 (11): 1859-1875.
Wardell, D.A., Reenberg, A. and Tottrup, C., 2003. Historical footprints in contemporary land use
systems: forest cover changes in savannah woodlands in the Sudano-Sahelian zone. Global
Environmental Change, 13 (4): 235-254.

118
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Environmental analysis and diagnosis of recreational zones in protected areas of

Madrid (Spain)

C. Rozas, B. Martín, N. Roblas, C. Muñoz & L. Jiménez

Centro de Investigaciones Ambientales de la Comunidad de Madrid “Fernando González

Bernáldez” C/ San Sebastián 71. Soto del Real MADRID.
e-mail: [email protected]

There has been a noteworthy increase in the recreational use of protected natural areas
over the last few years. With the aim in mind of minimising the impacts caused by the
numerous visitors, a series of recreational areas has been set up in order to concentrate
them therein.

On locating a recreational area it is important to evaluate the possible impacts by visitors

in order to avoid sensitive areas with high ecological or landscape values. Furthermore,
these areas must also have a certain degree of naturalness and of landscape and
environmental quality if they are to satisfy the expectations and news of visitors and they are
to be used as a resource for the development of environmental education activities.

The objectives of this study involve analysing the recreational areas of the Madrid
Regional Autonomy, evaluating their degree of naturalness, their state of conservation and
the possible impacts deriving from the recreational activity in the surrounding area. The study
focuses on areas situated along riverbanks or streams within a Protected Natural Area. The
methodology is based upon the analysis of the quality of the riparian vegetation and the state
of conservation thereof and involves the application of the QBR riverbank quality index
(Munné et al; 1998), and an estimation of the impacts caused by the recreational use in the
area itself (Gómez-Limón et al, 1996).

The results obtained show that the quality of the riparian ecosystems is significantly lower
in the recreational areas than in the control sites. According to the QBR index, 45% of the
sampling sites present a level from acceptable to optimum in the recreational areas,
compared with 90% in the control sites.
Over 70 % of the recreational areas studied are in a poor state of conservation, according
to the impact assessment index. The factors with most influence are the width of the main
pathway and the percentage of ground with no cover, both of these being indicators of a high
number of visitors.
In exploratory studies, no correlations have been detected between both indices, and it
would therefore be of interest to study the factors influencing the variation of the QBR inside
and outside the recreational areas.

119
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Easy evaluation of the abundance of carnivores through habitat use map, in a

Brazilian savanna region

M.C. Lyra-Jorge, M.C. Ribeiro, V.R. Pivello

Instituto de Biociências – Universidade de São Paulo.

e-mail:[email protected]

Introduction
São Paulo is the most urbanized state in Brazil; therefore a few fragments of natural
ecosystems remain. Land cover in the study area (UTM 200000–7620000; 240000–
7590000, SAD69 Brazil 23S) comprises patches of savanna and semideciduous forest
surrounded by a matrix of sugar-cane and Eucalyptus monocultures. Despite the intense
agricultural use, large mammals are still found in that region. To establish conservation
strategies focusing on these animals it is necessary to understand how they respond to land
cover variation and to landscape structural patterns. In this study, we intended to register the
carnivorous mammal species which are using the mosaic of native vegetation patches and
Eucalyptus plantations.

Methods
We registered species occurrences using 23 camera traps placed in fragments of typical
native savanna (CSS), dense savanna (CAO), semideciduous forest (NSF), and Eucalyptus
plantation (EP). We located sample points on a land cover map generated from a Landsat-
TM5 satellite image and calculated landscape indices – percentage of remaining habitat
(PLAND), edge density (ED), and habitat patch shape (SHAPE) – in areas defined by 250 m,
500 m, 1,000 m and 2,000 m of radii around each sample point. ANOVA was undertaken to
verify habitat effect on species occurrence but no significant habitat effect was found
(F=2.84; p>0.05). We created a carnivorous habitat use map through a model generated by
stepwise regression analyses and data on species richness, occurrence, habitat type, and
landscape indices.

Results
The best regression model for species occurrence considered the parameters:
radius=250m; 0.04219*PLAND(CAO) + 0.01174*PLAND(CSS) + 5.04857* SHP(EP) –
0.26911*ED(EP) (R2=0.88; p<0.01). The map showed that Eucalyptus plantations and
fragments of dense savanna are intensively used by mammals such as Puma concolor,
Chrysocyon brachyurus, Leopardus pardalis, Eira Barbara. The results indicate that some
non-native agrosystems - as, in this case, the eucalyptus plantation - may also have a high
conservation value for large and medium carnivores, as they perform as permeable matrices
and probably offer environmental resources to the animals, such as refuge or simply an
access route.
Figure 1: Habitat use in Mogi-Guaçú river basin according
to the occurence of carnivours mammals. (Darker areas
represent high intensity of use, as the bright areas
represent less intense use; dashed areas were not
analysed; dashed enclosed areas are uban zones. Each
point represents the camera trap and a 250m buffer.

Financial support: FAPESP, CNPq, NGC.

120
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Relationships between the naturalness of the landscape and the nature

conservation status of a particular habitat in it

E. Illyés, Z. Botta-Dukát, F. Horváth, Zs. Molnár, J. Bölöni, I. Pándi

Institute of Ecology and Botany of HAS – 2163, Vácrátót, Alkotmány 2-4., Hungary
e-mail: [email protected]

Introduction

Loss of area and fragmentation are the two main endangering factors of natural habitats
and landscapes (Millennium Ecosystem Assessment 2005). Conservation efforts often
attempt to maintain (semi)natural landscapes as a whole with the biggest possible
unfragmented area (Margules & Pressey 2000). This approach seems to be successful on
landscape level, but is it suitable for the conservation of particular habitats in the landscape?
We assessed the correlation between the overall naturalness of the landscape and the
nature conservation status of a particular habitat in it to answer this question.

Method

We worked with the database of the Landscape Ecological Habitat Mapping Program of
Hungary (MÉTA) for this study, which contains data on 86 (semi)natural habitats based on
actual field survey. Hexagons of 35 hectares covering whole Hungary are the basic units of
the database. List of (semi-natural) habitats and their area, naturalness based conservation
value and 16 other attributes are stored in the database for all hexagons. We choose forest
steppe meadow as our modell habitat, therefore we performed our analysis on the 3969
hexagons where forest steppe meadow occured.
We defined nature conservation status of the chosen habitat by the area of the habitat in
the hexagon and its naturalness based conservation value derived from the database. The
naturalness of the landscape was defined in two spatial levels, by the overall area of
(semi)natural habitats within the same hexagon and by the overall area of (semi)natural
habitats in landscape section containing app. 100 hexagons (we had 340 such sections).
We used Spearman’s rank correlation to explore the relationships in these two spatial
scales. We worked with the Statistica program package.

Results and conclusion

We found that there was a significantly positive relationship (p<0,01) between the overall
area of (semi-)natural vegetation and the area of the forest steppe meadows on both spatial
scales, although the relationship between them was low (r=0,433 and r=0,437). The
naturalness based conservation status of the habitat was significantly correlated (p<0,01)
with the overall area of natural and semi-natural vegetation on the finer scale, but the
relationship was even weaker (r=0,205). It seems that a more natural landscape somewhat
contributes to the higher conservation status of the habitats in it, however this effect is rather
weak and may not be driving factor in the future maintenance of the habitats.

References
Margules, C.R. & Pressey, R. (2000) Systematioc conservation planning. Nature 405: 243-253.
Millennium Ecosystem Assessment (2005) Ecosystems and Human Well-being: Biodiversity
Synthesis. World Resources Institute, Washington, DC.

121
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Adjacency analysis: a method to evaluate landscape conservation status

L. Zavattero1, D. Smiraglia2, M. Marchetti1 & C. Blasi2

1
Department S.T.A.T. – University of Molise, Contrada Fonte Lappone, Pesche (IS) Italy.
e-mail: [email protected]
2
Dept of Plant Biology – University of Rome “ La Sapienza”, P.le Aldo Moro 5, Rome – Italy.

The replacement of traditional agricultural, silvicultural and pastoral practices by more

intensive exploitation has strong impact: structural simplification, changes in spatial diversity
and boundaries, and modification of matter and energy flows between patches. To
characterize and quantify the landscape heterogeneity level in space and time the patch
boundary analysis could be more explanatory in a multitemporal study because it gives
information on how a landscape has changed and not only on how much. So, the spatial
arrangement of land use types across a given landscape can be explained by using the
nature of the different contact types (Acosta et al. 2003). Depending on the spatial
distribution of the patches and the shared perimeter between different land cover types is
possible to assess the landscape conservation status.
The aim of this study is to describe the dynamics of the landscape through patch
boundaries dynamics, taking into account the nature of the spatial contacts between adjacent
land cover types and using the spatial distribution of land units as reference model.
We considered recent historical changes (1954-2000) in the landscape of Lepini Range
(Central Italy) focusing on the boundaries between patches for the characterization of spatial
organization. A spatial delimitation in “land unit” of the study area is better suited for
assessing landscapes, thus it is possible to study ecological processes in an ecosystem
classification framework. In this paper we apply the method for classifying and mapping land
units at different scales developed by Blasi et al. (2000). The method integrates the basic
physical aspects of the landscape and is particularly useful for multipurpose applications. In
this case, the emphasis is focused on the boundaries analysis in environmental
homogeneous units (to Potential Natural Vegetation units). The term “boundary” refers to two
different neighbouring land use types having a common or shared perimeter. We take into
account 5 kinds of contacts between land cover types: 1) artificial surface/agricultural land, 2)
artificial surface/natural areas, 3) agricultural land/natural areas, 4) between natural areas
not dynamically related: that is between coenoses which do not represent successional
stages (forests/grassland), 5) between natural areas dynamically related: that is between
cenoses which represent successional stages (forest/shrubland or shrubland/grassland).
The results show a reduction in the number of contacts agricultural lands/natural areas
and an increase of contacts artificial surfaces/agricultural lands and artificial surfaces/natural
areas. Generally, this involves a retrieval of forest cover in abandoned mountains with a
decrease of agricultural lands and natural areas such as shrublands and grasslands and, to
the opposite, a more intensive exploitation of flat areas. In particular, in the analysed land
units, we noticed in the Mediterranean Region a decrease of adjacency grasslands
/shrublands and an increase shrublands/forests. In the Temperate Region land units instead
there is a decrease grasslands/shrublands and also shrublands/forests and, to the opposite,
an increase forests/grasslands.

References
Acosta, A; Blasi, C; Carranza, M.L. & Ricotta C. (2003) Quantifying ecological mosaic with a new
topoecological index. Phytocoenologia 33(3): 40-51.
Blasi, C; Carranza, M.L; Frondoni, R. & Rosati, L. (2000) Ecosystem classification and mapping: a
proposal for Italian Landscape. Applied Vegetation Science 3: 233-242.

122
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Estimating the effect of protected lands on the development and conservation of

their surroundings

R.I. McDonald1, C. Yuan-Farrell2, C. Fievet3, M. Moeller4, P. Kareiva2, D. Foster1,

T.Gragson3, A. Kinzig4, L. Kuby5, C. Redman5
1
Graduate School of Design, 48 Quincy Street, Harvard University, Cambridge, MA USA.
e-mail: [email protected].
2
The Nature Conservancy, Seattle, WA, USA 3Department of Anthropology, University of
Georgia, Athens, GA, USA 4School of Life Sciences, Arizona State University, Phoenix, AZ,
USA 5International Institute for Sustainability, Arizona State University, Phoenix, AZ, USA

The fate of private lands is widely seen as key to the fate of biodiversity in much of the
world. Organizations that work to protect biodiversity on private lands often hope that
conservation actions on one piece of land will leverage the actions of surrounding
landowners. However, few researchers have examined whether protected lands do in fact
encourage land conservation nearby or how protected lands affect development in the
surrounding landscape. Using spatiotemporal datasets on land cover and land protection for
three sites (western North Carolina, central Massachusetts, and central Arizona- see Figure
1), we examined whether the existence of a protected area correlates with an increased rate
of nearby land conservation or a decreased rate of nearby land development. At all sites,
newly protected conservation areas tended to cluster close to pre-existing protected areas.
This may imply that the geography of contemporary conservation actions is influenced by
past decisions on land protection, often made for reasons far removed from concerns about
biodiversity. On the other hand, we found no evidence that proximity to protected areas
correlates with a reduced rate of nearby land development. Indeed, on two of our three sites
the development rate was significantly greater in regions with more protected land. This
suggests that each conservation action should be justified and valued largely for what is
protected on the targeted land, without much hope of broader conservation leverage effects.

Figure 1. Protected area by ownership since 1900: Chart for each site, showing the
proportion of land protected by different groups: federal (black), fee-simple NGO (dark grey),
owned by state (grey), easement NGO (light grey), and municipal/county (white). Coweeta is
dominated by federal ownership, Quabbin by state ownership, and Phoenix by a mix of
federal, state, and municipal ownership.

123
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Identification of sites for mitigation banks using multiple ecological and social
criteria

N.Hope, M.C. Fessey, N. Bailey, S. Thompson

Spatial Ecology and Landuse Unit (SELU), School of Life Sciences, Oxford Brookes
University, Headington, Oxford UK.
e-mail: [email protected]

Introduction to mitigation banking

In order to address key landscape planning concerns such as; habitat loss and
fragmentation, accessible greenspace, and potential climate change demands, mitigation
banking is proposed. This approach to compensatory mitigation operates by developers
offsetting adverse effects and habitat loss from developments by the purchase of credits in a
land ‘bank’ with a pre-determined nature conservation value (Treweek 2000). This allows
larger scale and higher quality habitat creation and management to be achieved, than current
systems, through accumulation and targeting of funds (Wende et al. 2005). Mitigation banks
must become part of the green infrastructure resource. Therefore, their location and
constituent habitats need to be planned from a multifunctional perspective.

Multifunctional landscape assessment

Four linked assessment modules; ecological networks, enhancement of landscape character,
recreation and amenity provision and natural resource protection, will be used to address
landscape planning concerns. The results from two modules; ecological networks and
recreation and amenity provision are explored here using the Milton Keynes and South
Midlands (MKSM) Growth Area as a case study.

Ecological networks and recreation and amenity modules

Using generic focal species (Watts et al. 2005), ten scenarios were developed for woodland,
grassland and wetland. The ecological network model then used resistance surfaces to
measure functional landscape connectivity. Existing ecological networks were identified,
followed by determination of optimum locations for further habitat creation. Categorisation of
greenspace using a typology based on national planning guidance formed the recreation and
amenity baseline. Allocation of resistance values subsequently allowed people’s accessibility
zones to be mapped accurately. Using national accessibility standards identified zones were
analysed to determine current level of provision and potential expansion areas. Compilation
of the two modules allowed areas with potentially multiple functions to be identified.

Results
Comparison of habitat expansion areas with biodiversity targets (Bedfordshire and Luton
Wildlife Working Group 2000) reveal expansion zones to have the potential to exceed targets
for woodland by 44320 ha, grassland by 4240 ha and wetland by 840 ha. Existing recreation
and amenity greenspace does not currently meet either Local Nature Reserve or 20 ha site
standards. Recreation expansion zones have the potential to rectify this shortfall with 110
sites over 20ha (4770 ha) identified. Compilation of the two modules indicates habitat
creation areas can provide multiple functions with 5740ha woodland, 720ha grassland and
50ha wetland featuring in both modules.

References
Bedfordshire and Luton Wildlife Working Group (2000). Bedfordshire and Luton BAP. Luton.
Treweek, J. (2000). "Royal Commission on Environmental Pollution." (online).
Watts, K., M. Griffiths, et al. (2005). Towards a Woodland Habitat Network for Wales. Report 686,
Countryside Council for Wales and Forestry Commission.

124
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

A conceptual model for restoration site selection based on a review of reserve

selection procedures

J. Imanishi, A. Imanishi, Y. Natuhara, Y. Morimoto

Graduate School of Global Environmental Studies, Kyoto University, Yoshida-Honmachi

Sakyo-Ku, Kyoto, Japan.
e-mail: [email protected]

Introduction

The Law for the Promotion of Nature Restoration of Japan, which aims at recovering the
ecosystems and other natural environments that have been degraded or destroyed in the
past, was enforced in January 2003. However, procedures for nature restoration at national
and regional scales are still unclear. The purpose of this study is to propose a conceptual
model for restoration site selection based on a review of literature on procedures for
selecting nature reserves to conserve biodiversity in a nation and a region.
We, herein, define ecological restoration as the process of assisting the recovery of an
ecosystem that has been degraded, damaged, or destroyed (SERI 2004). Although activities
of ecological restoration partly overlap with conservation by reserves, it is difficult to find
restoration opportunity using usual reserve selection procedures for the ecosystems that
have already been altered. Thus, it is needed to integrate restoration site selection into
reserve selection procedures in conservation planning.

A conceptual model for national biodiversity conservation planning of Japan

The three attributes, i.e. composition, structure and function, seem well recognized as
elements of ecosystems in current conservation biology. A conceptual model, which
separately assesses the three elements, was proposed as one of the practicable procedures
based on the availability of necessary information.
The information on ecosystem composition for national biodiversity conservation planning
is available as distribution data of endangered species in coarse geographical units and as
maps of vegetation communities in 1:50,000. The concentrated areas of endangered species
are identified as primary reserves in a “reactive” manner. Then, abiotic environmental units
are analyzed in relation to the distribution of vegetation communities as a surrogate for
species diversity with a similar procedure of Imanishi et al. (2005) to find gaps in present
conservation networks and to find possible restoration sites in a “proactive” manner.
An aspect of ecosystem structure is analyzed in landscape pattern at regional to
watershed levels using medium resolution satellite imagery. Habitat models for species,
which rely large area, “area-demanding umbrella species,” and which use multiple types of
habitats, e.g. amphibians, are utilized for the assessment.
A feature of ecosystem function is assessed by distribution of artificial structures such as
dams and levees at watershed level. It is a prerequisite for the restoration of interface
ecosystems between land and water, which usually maintain high biodiversity.

References
Imanishi, J, Shimabayashi, Y & Morimoto, Y. (2005) A new analytical method for wildlife habitat
conservation planning on a city scale using the classification of physically homogeneous areas.
Landscape and Ecological Engineering 1(2): 157-168.
Society for Ecological Restoration International Science & Policy Working Group (2004) The
SER International Primer on Ecological Restoration. Society for Ecological Restoration
International, Tucson, USA. available online at www.ser.org.

125
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Preservation of sand scenery and native plants conservation in the coastal

dunes, the landmark of Tottori, Japan

D. Nagamatsu, S. Tominaga

Department of Regional Environment, Tottori University, Tottori, Japan.

e-mail: [email protected]

Introduction

The Tottori Sand Dunes are typical costal dunes in Japan. They are the largest dune park
in the Japanese National Parks and the most famous tourist attraction for this region. But the
dunes were regarded as a nuisance for local people for centuries. They planted pine trees
and tried to develop the dunes into farmland. After WWII, the dunes were therefore reduced
in area. The remaining dunes were converted into grassland (Shimizu & Shibata 1991). The
dunes then declined in their value as a miniature desert. Local government and citizens are
now focusing their efforts on trying to preserve sand scenery by weeding for the past 15
years. Vegetation in the dunes has changed gradually due to environmental change and
alien plant removal by human beings. But vegetation change in the dunes has not been yet
objectively examined. So the changes in plant communities and the effect of exotic removal
have been investigated.
We measured distribution of plant community using a GPS receiver and studied species
composition and abundance of the community in the dunes (c.a. 130 ha). To estimate past
vegetation, we used scientific reports (e.g. Shimizu & Shibata 1991) over 40 years. We made
a detailed vegetation map and analyzed community structure by GIS software.

Results and Discussion

Based on vegetation maps from 1967, 1979, 1991 and 2006, Carex kobomugi and
Ischaemum anthephoroides, the native plants of the dunes dominated over 40 years.
Although no alien plant communities were recorded until 1967, Digitaraia ciliaris and other
alien communities became distributed widely between 1979 and 1991. It is suggested that
plant communities of the dunes were once widespread in the 1960's to the 1990's and then
declined until the present. Alien weed communities were clearly decreased by recent
weeding but have not been eliminated. Now native plants occupy most of the dunes, the
balance of the sand scenery and native plant communities is the main subject for discussion
in the dunes.

Reference
Shimizu, H. & Shibata, M. (1991) Expansion of grassland in Tottori Sand Dunes. The research
committee of the Tottori Sand Dunes (Ed). Science Report of the Tottori Sand Dunes, San-in
Kaigan National Park, Tottori Prefecture, Tottori, pp. 14-24 (in Japanese).

126
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Landscape structural analysis at human scale in a Mediterranean nature park

L. Hadar1, Z. Henkin2, I. Noy-Meir3 and H. Muklada1

1
Ramat Hanadiv Nature Park, P.O.B. 325 Zichron Ya'acov, Israel.
e-mail: [email protected]
2
Beef Cattle section, Newe-Ya'ar Research Center, Department of Natural Resources,
Agricultural Research Organization, P.O. Box 1021, Ramat Yishay, Israel.
3
Institute of Plant Sciences, Faculty of Agricultural, Food and Environmental Quality
Sciences, the Hebrew University of Jerusalem, Rehovot, Israel.

Introduction

Vegetation description at the landscape level often focuses on the botanical elements,
such as communities, dominant species and species diversity. However, for management
and planning objectives, vegetation units must also be defined on the basis of their structural
qualities and with relation to the human scale. Structural criteria relate to basic human needs,
such as transparency, accessibility, shade and micro-climate, air and open space. When
combined with visual and aesthetic values, they define the extent to which different areas are
attractive for human activity (Jacobs, 1993).

Methods

An inter-disciplinary approach to landscape vegetation description was developed and

tested in northern Israeli rangelands (Henkin et al., 2007) and then applied to a well
managed nature park in the Mediterranean region of Israel. The landscape was described
using a set of 'architectural' criteria, by mapping the dimensions, basic shapes, and
distribution of gaps between individual plants, to distinguish between areas with different
management regimes. The shapes of the trees and the bushes were drawn in situ in order to
characterize the various structural profiles of the vegetation.

Application

The Ramat Hanadiv Nature Park is managed as an open space for nature, landscape,
public activity, research and education. The park's animal populations are an integral part of
the landscape and much effort is invested in their monitoring. An application of the proposed
method to the scale of the park will also produce a 'structural vegetation map' to which
animal data could be related in a more reliable way.
As the drawing of structural profiles is an intuitive, easy and a 'user friendly' technique, an
application of the above method to be used by teachers and high school students was tested
and is now widely used in the park.

References
Henkin, Z; Hadar, L & Noy-Meir, I. (2007) Human-scale structural heterogeneity induced by grazing
in a Mediterranean woodland landscape. Landscape Ecology (in press).
Jacobs, A. (1993) Great Streets. M.I.T. Press, Boston.

127
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Impact of agriculture on water erosion and new organization of landscape

I. Šimonides1, T. Hrnčiarová2

1
Horticulture and Landscape Engineering Faculty, Slovak University of Agriculture in Nitra,
Hospodárska 7, Nitra, Slovak Republic,
e-mail: [email protected]
2
Institute of Landscape Ecology of Slovak Academy of Sciences, Štefánikova 3, P. O. Box
245, Bratislava, Slovak Republic

Conservation of the soil production potential, involves the elimination of the negative
processes which can be achieved by several measures: a new structure of crops (distribution
of agricultural crops or crop rotation from the aspect of erosion control effects), a change in
plot borders, plantations of erosion control vegetation, reductions in slope length, erosion
control vegetation belts, the spatial arrangement of vegetation according to runoff relations in
micro catchments and stabilization of talwegs.
We studied the erosion processes in detail in the study area of the Smrečianka river
catchment in the Liptovská kotlina Basin. It is an intensively used agricultural area with
almost no non-forest woody vegetation. The area of agricultural plots varies from 20 to 50 ha;
the overall length of the slope reaches a maximum of 800 m. Erosion processes are
influenced mainly by slopes length and are only partially modified by changes of use. Relief
microforms are other factors causing the concentration of surface runoff, the consequence of
which is an increase of erosion processes. Major water runoff takes place during rainstorms
forming large erosion rills developing into gullies in the upper part of the slope. From the
viewpoint of landscape ecological use, it is necessary to proceed to land use zoning of plots
lying above each other, to specify the barrier in order to modify the surface runoff, e.g. to
complete also the elements of ecological network and to propose different anti-erosion crop
rotations dependent on the limits of erodibility.
This work was supported by the Ministry of Education of the Slovak Republic under the
contract No. 1/4404/07 Influence of erosive processes on the agricultural landscape
organizational change and No. 2/7027/27 Evaluation of landscape diversity changes and by
the Slovak Research and Development Agency under the contract No. APVV-51-035102
Creation of environmental limits for sustainable development (on example of model
territories).

128
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Using public inquiries in the validation of expert based landscape assessment: a

case study in Southern Portugal: Castelo de Vide

H. Menezes1, T. Pinto-Correia2, I. L. Ramos3

1
National Agrarian Station, Quinta do Marquês, Oeiras, Portugal.
e-mail: [email protected]
2
University of Évora, Largo dos Colegiais 2, Évora, Portugal.
3
CESUR – Technical University of Lisbon, Avenida Rovisco Pais, Lisbon, Portugal.

There is an increasing awareness at the European level (e.g. European Landscape

Convention, 2000) of the importance of classification of all landscapes. Public participation
can provide more information, to help identify areas of distinctive character and determine
key characteristics and special qualities (Bingham et al., 2002). As a result, the
understanding of people’s relationship with the landscape is required to complement
technical characterization.
The goal of this study is to use the public’s perception of landscape as the validation of an
expert based landscape classification. The case study is the municipality of Castelo de Vide,
in the centre of Portugal, near the border with Spain.
The methodological flow applied in this study is represented in the poster. The first phase
(Abreu et al., 2004) consists of the interactive combination of literature review, field work,
expert judgement and the incorporation of map data. The human experience of the
landscape and the opinions and rights of stakeholders, represent relevant topics within the
landscape character assessment for further definition of landscape units (Groom, 2005). A
photograph based questionnaire was developed, with representative photos of each
landscape unit, to understand if the assessment of landscape areas done by experts was
consistent with people’s perception. The sample (n=30) was stratified according to age,
gender, profession/activity and education level.
Three landscape areas have been identified, in which two are sub-divided (total of six
landscape areas). In general, the landscape of the municipality is not seen as much widely
differentiated. However, the majority of the interviewed people could identify the main visible
characteristics of each area. People connected to farming showed more knowledge in issues
regarding landscape management, with a finer perception regarding landscape issues. Land
abandonment was frequently raised by those interviewed, pointing out the increase of shrubs
as the main visual indicator, and the decrease of people connected to farming as the main
cause. It was possible to understand which elements were mostly appreciated and valued by
the interviewed.

References
Abreu C., Correia, T., Oliveira, R. (2004) Contributos para a identificação e caracterização da
paisagem em Portugal continental. DGOTDU, Colecção de Estudos 10, Lisboa.
Bingham L., Parfitt A., Swanwick C. (2002) Landscape Character Assessment Guidance for
England and Scotland - Topic Paper 3: Landscape Character Assessment – how stakeholders can
help. The Countryside Agency and Scottish Natural Heritage.
Groom, G. (2005) Methodological review of existing classifications. In: Wascher. D.M. (ed).
European Landscape Character Areas - Typologies, Cartography and Indicators for the
Assessment of Sustainable Landscapes. Final Project Report as deliverable from the EU`s
Accompanying Measure project European Landscape Character Assessment Initiative (ELCAI),
funded under the 5th Framework Programme on Energy, Environment and Sustainable
Development (4.2.2).

129
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Saxon cultural landscapes: what amphibians would like most? Getting baseline
data for the management of a woodpasture

K. Öllerer, T. Hartel

Institute of Biology – Romanian Academy, Splaiul Independentei 296, Bucharest, Romania,

e-mail: [email protected]

Introduction
The Breite woodpasture is on a 135 ha plateau, at about 500 m elevation, situated nearby
the town of Sighişoara (Transylvania, Romania). A 70 ha area of the plateau is legally
protected (IUCN IV) due to the large density of scattered several-century-old oaks (Quercus
robur and Q. petraea, most of the trees being hybrids). The area is currently threatened by
desiccation conditions that appeared after fifteen drainage ditches were built in the 1980s,
and which led to major decrease of the initial marsh system. The results of the
disappearance of former humid areas are: (i) the reduction of spatial heterogeneity, (ii)
degradation of the veteran oaks as key determinants of this landscape, and bearers of saxon
cultural heritage (iii) reduction of suitable habitat of several species of conservation interest.
Beside the oaks, several plant and animal species of conservation interest are negatively
affected by the progressive drainage. It was therefore decided to study amphibians, due to
their dependence on the aquatic habitats. Therefore, we’ve began an amphibian monitoring
project in 2003, with the aim to create the basis of a management plan for the whole plateau.

Methods and results

102 temporary water bodies (10 drainage ditches, 20 archaeological ditches and 72
ponds) were studied and eight amphibian species were identified. The water bodies were
grouped in three categories, based on their hydroperiod: ephemeral (that dried up in 2-3
weeks after filling up with rainwater), transient (those that dried up once or twice per season)
and constant (those that hold water throughout the year in characteristic climatic conditions)
(Hartel et al. 2005). Large variations were recorded regarding (i) the number of water bodies
with reproductive amphibians, (ii) number of water bodies used for reproduction and (iii) the
number of water bodies with successful reproduction. Reproductive success was estimated
based on the number of water bodies where metamorphosis occured. The number of species
breeding in the specific water bodies and the success of reproduction was mostly dependent
on the hydroperiod. The percentage of water bodies with successful metamorphosis varied
between 0 and 39%. Bufo bufo, Triturus cristatus, Pelobates fuscus and Hyla arborea had
complete reproductive failures in 2003 and 2004 whereas Bombina variegata was the most
successful species.

Discussion
Our results show that the most suitable habitats for amphibians are represented by the
drainage ditches. Based on the four-year observations we predict that an increased
hydroperiod of the existing aquatic habitats will positively influence the amphibian
community, both through increases in the reproductive success and in the number of species
present in specific water bodies. Therefore, the ditch enclosures would be the most suitable
management intervention, both for the amphibian community, and would also help regaining
the initial marshy characteristic of the plateau, key and necessary action for the maintenance
of the veteran oaks and species of conservation interest.

References
Hartel, T., Moga, C.I., Nemes, Sz. (2005) Use of temporary ponds by amphibians in a wooded
pasture, Romania. Biota 6: 21-28.

130
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Integrating livelihoods and multiple biodiversity values in landscape mosaics: a

new global initiative

P.K. Koponen 1, J-L. Pfund1, M. van Noordwijk2

1
CIFOR Jalan Cifor, Sindang barang, 16680 Bogor barat, indonesia . e-mail:
[email protected]
2
ICRAF Jalan Cifor, Sindang barang, 16680 Bogor barat, Indonesia.

Introduction
Conventional approaches to biodiversity conservation are still mainly based on the
segregation of protected areas and biodiversity corridors from the surrounding landscape.
The joint CIFOR-ICRAF Biodiversity Platform aims to promote flexible and broadly applicable
approaches for integrating livelihoods and biodiversity conservation (Pfund et al 2006). In a
new project, our objective is to find negotiated ways to combine local resource use and
global conservation objectives in forest landscape mosaics of five main sites (Cameroon,
Indonesia, Kenya, Laos, Madagascar).

Objectives and approach

The aim is to design desirable multifunctional landscape states in a participatory and
bottom up process, and to negotiate the needed changes of social, economic and ecological
conditions towards such states. Key elements of the project are 1) analyzing the driving
forces that have influenced conservation and development at the landscape level, 2)
assessing the current biodiversity and livelihood values of local and external key actors, local
rules, how local rules are incoporporated into customary and regulatory frameworks and
potential incentives for biodiversity conservation, 3) projecting and negotiating desirable
changes and 4) developing commitments to implement measures to change.

In order to have robust information to feed negotiations between local and external
stakeholders CIFOR and ICRAF have developed tools, such as Multidisciplinary Landscape
Assessment (MLA) and Rapid Agrobiodiversity Assessment (RABA) to assess what, where
and why certain land types, species and services are valued (multiple reasons are analyzed
in time and space) by the local people and potential environmental service buyers. Chosen
research sites are well known and accumulated information is our advantage. Improvements
of specific locally and externally identified biodiversity related issues will be studied from the
angle of 1) traditional knowledge on biodiversity products, services and related livelihoods 2)
incentives and reward mechanisms for biodiversity conservation 3) consideration of local
rules and effects of external regulation and 4) understanding and modeling landscape
patterns and processes. One of the sites in Jambi province of Sumatra, Indonesia, is
presented as an example of what the approach will focus on and how. In Jambi the
widespread and ongoing conversion of forest to oil palm and rubber monoculture is resulting
in great loss of biodiversity. At landscape level high biodiversity diversity is found in natural
forest that mainly remains in national parks and traditional diverse rubber agroforests.
Diverse reward mechanisms are sought for biodiversity conservation such as green
marketing or improved management and access rights. Sampling of local landscapes will be
conducted along a gradient from low to high land use intensity.

References
Pfund J-L; O´Connor T; Koponen P. & Boffa J-M. (2006) Transdisciplinary research to promote
biodiversity conservation and enhanced management of tropical landscape mosaics. Lafortezza
(Ed) Patterns and processes in forest landscapes, consequences of human management. IUFRO
Landscape ecology working party, Locorotondo, Bari, Italy

131
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Arasbaran forest ecological conditions and studying to establish forest park

(case study: Ilgineh Chay basin)

M. Akbarzadeh, S. Babaei Kafaki

Islamic Azad University, Science & Research Branch, Tehran, Iran.

e-mail: [email protected]

Introduction

Arasbaran is located in the north-eastern part of East Azerbaijan province and due to the
increased destruction along time the mountainous forests were converted into a fragile
ecosystem.

Material, methods and conclusions

Ilgineh chay basin

These forests have special ecological conditions, with a unique fauna and flora. There is
also much topographical and climatologically structure in this forest which is different from
any another forest structure in Iran. Ilghineh chay basin is the biggest basin of this district,
with an area about 58000 ha.

Ecological evaluations for establishing the park

In this study, satellite and field data were used from the study area, and matching these
data together, and with a Digital Evaluation Model (DEM), finally 4 points were determined as
suitable for the forest paradise park (Makhdum, 1999 & Darvish Sefat, 1990). The total area
of the 4-forest Paradise Park is 3000 ha and forest management model is the basis of the
study foundation. In this model, each district has segmentation to three main parts; one (part
alpha): the regions under protection programs (time of programs in each site is different with
another one). Two (part beta): the regions under reforestation and silvicultural growing
program, Three (part gamma): the regions under harvesting with secondary, forest
production. Time of protection = (Number of trees in each ha * species situation *
topographical condition).

References
Makhdum, M.F. (1999). Introducing a methodology for integration of ecological and socio economic
data in regional planning. The Karolinum Press.Prague: 390-393-414.
Darvish Sefat, A.A. (1990) Information analysis of a spatial database for ecological land
classification. Tehran University Press, Tehran, Iran

132
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Representing biodiversity: Spatial analysis of multi-taxon species composition in

the Netherlands

A. Barendregt, M. Schouten

Department of Environmental Sciences - Utrecht University Copernicus Institute for

Sustainable Development and Innovation, PO Box 80115, Utrecht – The Netherlands
e-mail: [email protected]

Landscape ecology combines the information of physical and biological components in

spatial distribution. The analysis of the spatial distribution of biodiversity in combination with
environmental variables is crucial to understand the biogeographical pattern and can be
applied as effective source of information for nature conservation.
Here, we used species occurrence data of a broad array of taxonomic groups (Syrphidea,
Orthoptera, Odonata, herpetofauna and bryophytes) to produce a quantitative classification
of the Netherlands. We focused hereby on regions that contain several characteristic species
of different taxonomic groups, so-called umbrella regions, as these are the most interesting
areas from a conservation perspective.
For each of the taxonomic groups we were able to define regions that are clearly distinct
in terms of species composition. However, not all regions comprised characteristic species.
We identified five regions that appeared in at least two taxonomic groups and are
characterized by a set of unique species (e.g., see fig. 1). Stepwise discriminant analysis
revealed significant environmental differences among these five regions.
These five regions each comprise a unique set of species from several taxonomic groups.
The combination of these regions covers the majority of the represented species of the
studied groups in the Netherlands; these regions represent the characteristic biogeography
within the country. Therefore the umbrella regions can play a leading role in future nature
conservation planning.

Nederland.shp Nederland.shp
Grid5.shp Grid5.shp
1 1
2 2
3
3 4

Figure 1. Distribution of two multi-species regions in The Netherlands. Numbers in grids

refer to the number of taxonomic groups for which a grid square is identified to the regions:
left = peaty wetlands; right = southern sandy soils.

133
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Indigenous landscape conception, management and conservation: lessons for

bird conservation from Mexico.

L. Cabrera
1
McGill University, Department of Geography, Montreal, Quebec, H3A 2K6, Canada.
e-mail: [email protected]

Introduction
Natural resource managers increasingly recognized the relation between biodiversity rich
regions and the traditional societies owning and managing them. Few studies have yet
investigated what the social-ecological mechanisms are subjacent of this amalgam,
especially on reference to bird conservation. This study addresses this gap by examining
how local communities are related to the endangered Sierra Madre sparrow’s core
distributional area, and how these are compatible with bird conservation goals.

Methods
The study employed a combination of ethnographic, participative and ecological
approaches to address human-land interactions and their impacts on the bird habitat. Three
types of data were generated for this study with local participation: historical (land occupation
and demarcation), traditional knowledge (landscape management and social institutions) and
ecological (breeding bird habitat suitability).

Results and discussion

Local inhabitants or Momoxcas own a particular ethnic history. They consider themselves
direct descendants of the Aztecs and inheritors and defenders of the land. People is
geographically and culturally embedded in this landscape. They recognized to own their
ethnic identity and survival to the monte comunal or communal mountain. For them, “the
monte is everything”. Their conception corresponds with the core fundament of landscape as
a holistic physical-social entity (Troll, 1938 cited by Bocco, 2007 and see Sauer, 1939).
Hence, for Momoxcas, the Sierra Madre Sparrow and its habitat are only part of the
landscape and they can not be abstracted for independent scientific conservation purposes.
The bird habitat is a tall bunchgrassland traditionally managed with rotation of fire and
grazing to achieve diverse purposes, including pasture renewal, species selection, and
prevention of dangerous fires. A rich traditional ecological knowledge system was
documented to exist behind these disturbance-based management practices. Through this
historical interaction of land dominance and management, the Sierra Madre Sparrow’s
habitat is renovated and guarded by legitimate land stewards.
Yet what is not clear is for how long this bird will find its last refuge in this land. External
conservation interests and perspectives, land use change pressures, and internal land
disputes are lessening the social-ecological resilience of this cultural landscape and the
conservation possibilities of this vulnerable bird species. What is clear is that collaborative
research with the Momoxca is needed to strengthen local organizational capacities and
integrate ecological knowledge to impulse adaptive and co-management resource
management projects.

References
Bocco, G. (2007). Carl Troll y la ecologia del paisaje. Instituto Nacional de Ecologia, Mexico City,
Online publications, retrieved on March 15, 2007 from:
http://www.ine.gob.mx/ueajei/publicaciones/gacetas/399/bocco.html#top].
Sauer, C. 1939. Man in nature: America before the days of the white men. A first book in Geography.
Scribners, New York.

134
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Soft-system approach to stakeholder participation in landscape planning &

management

L.F. Cassar, S. Morse, G.H. Griffiths

Department of Geography, School of Human and Environmental Sciences,
Faculty of Science, University of Reading, Whiteknights, Reading UK.
e-mail: [email protected]

Introduction
Although there is increasing emphasis on the need to involve stakeholders in landscape
planning and management, conventional social science methodologies are constraining
when addressing complex issues such as landscape. Simple questionnaires and interviews
can be limiting in the depth of data acquired, where a complexity of natural, cultural and
social elements are concerned. Direct participation and expression may also be difficult for
stakeholders. This study presents results derived from research on landscape pressures
which used a soft systems methodology (systemic sustainability analysis - SSA) for
assessing stakeholder perceptions of landscape threats.
Methodology
Varied groups of key respondents were brought together in three workshop-style fora on the
island of Gozo, part of the Maltese archipelago. The respondents were briefed on the SSA
methodology, which involves the drawing of ‘rich pictures’ from which issues are
subsequently prioritised (Bell & Morse, 2003). A further stage comprises the identification of
appropriate sustainability indicators. The results derived were subsequently used as the
basis for a ranking exercise with a wider group of stakeholders (230 individuals). Cards
containing photographs that were recognizable and simple to understand were presented to
stakeholders, and these were asked to rank the pressures in relation to their perceived
importance as agents in changing the Gozitan landscape.
Results
The key respondent groups identified eight key issues affecting landscapes in Gozo, namely
urbanisation, quarrying, pollution from agriculture, visitor pressure, hunting & trapping,
grazing, landfill, and reclamation, land abandonment & proliferation of alien species. The
ranking exercise highlighted urbanisation as the predominant pressure on landscapes, a
result common to many Mediterranean landscapes. Key differences between stakeholder
groups were also evident.
Conclusions
The methodology utilised was found to have numerous advantages over conventional
questionnaires and interviews, as there was enhanced stakeholder participation. The
process of drawing ‘rich pictures’ illustrating the landscape and perceived pressures was met
with enthusiasm, and vibrant discussion between stakeholders, and the various ‘rich pictures’
derived succeeded in presenting a significant depth of data in a simple, understandable
format. The subsequent ranking exercise also benefited from the simplicity of the method
used. Both methods in turn have much potential for adaptation and use as rapid participatory
methodologies for situations where more time-consuming methods may not be suitable.

References
Bell, S & Morse, S. (2003). Measuring sustainability. Learning from doing. Earthscan, London.

135
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

The Influence of Agricultural Roads and Water Channels on Farm Landscape

Structure in Taiwan

Y. C. Chen 1, C. P. Chang 2, Z. F. Wu 3
1
Chin-Yi Institute of Technology
2
Chung-Chou Institute of Technology,–6,Line2,Sec.3,Shan-Chiao Rd, Yuanlin
Changhwa.51003, Taiwan.
e-mail: [email protected]
3
Guangdong Institute of Ecology, Environment and Soil Science

Farmland consolidation has been successfully implemented over 391,000 hectares in

Taiwan. The consolidation work established standard blocks and installed agricultural water
channels so that each block could be directly connected to road, irrigation, and draining. For
this reason, the roads and channels are increasing quickly and that changed the farm
landscape structure greatly after farmland consolidation. This study used the network
connectivity and circuitry indices of landscape ecology theory to characterize the corridors
changing in Liu-Xiang farmland consolidation area, Ping-Dong County, the south of Taiwan.
The results revealed that connectivity and circuitry of farm road corridor increased after
consolidation, the connectivity from 0.405 to 0.504, and the circuitry from 0.102 to 0.267.
These results showed that farm roads brought conveniences and added the paths for people,
but disturbed the way other creatures pass through the landscape. The connectivity and
circuitry of waterway corridor also increased after consolidation, connectivity from 0.35 to 0.50,
and circuitry from 0.to 0.241. The results show that transmission’s capability of channel
corridors is better for creatures after farmland consolidation. If we used the ecological
engineering for water channel and planting extensively in farm roads and banks that will be
well for the farmland consolidation areas.

136
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Modelling the impact of agriculture policy changes on farmland birds

T.K. Gottschalk, V. Wolters

Justus Liebig University, IFZ, Department of Animal Ecology, Heinrich-Buff-Ring 26-32,

Giessen, Germany.
e-mail: [email protected]

Our objectives were to forecast and to analyse the effects of agricultural subsidies on faunal
richness by applying a spatially explicit biodiversity model to socio-economic scenarios. We
exemplarily used changes of the EU Common Agricultural Policy (CAP) from “volume of
production dependent subsidies” to "single farm payments".
We used a GIS-based model GEPARD (Geographically Explicit Prediction of Animal
Richness Distributions) to estimate the current pattern of bird diversity in different regions of
Hesse, Germany, and to predict future avian diversity patterns under land use scenarios
based on the CAP policy. The model combines (1) a spatial regression model, (2) maps of
spatial resource coverage, (3) quantifications of landscape pattern and environmental factors
within different scales, and (4) information from an agro-economic simulation model
predicting future spatial patterns of land use. For each study site, the surrounding landscape
was analysed within a radius varying from 200 to 1000 m to take into account species-
specific differences in the perception of the landscape matrix. Accordingly, simulations were
conducted using circular-moving window analysis for landscape metrics. To quantify the
impact of a changing agriculture policy, the model was used to predict areas with gains or
losses of farmland birds.
The results show that alterations of species richness are caused not only by gains and
losses of major land use types but also by overall changes in landscape configuration. The
model predicts the breeding population of the skylark Alauda arvensis to profit from
production dependent subsidies. Population density will even more increase under the CAP
reform, compared to the present situation. Total species richness of the avifauna, in contrast,
is predicted to decline in many areas under both policy scenarios. In general, however,
income support programs allowing farmers to improve environmental performance in specific
resource and management settings have a more positive effect on farmland bird diversity
than production based subsidies. The spatially explicit nature of the model’s output provides
a powerful tool for analysing the current and the future diversity and abundance of key bird
species. The model can thus support decision makers and nature conservationists to
evaluate trade-offs between economic values and values of nature conservation.

137
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

The West Weald Landscape Project, south east England

R.T. Howorth, T. Whitbread

West Weald Landscape Project, Sussex Wildlife Trust, Woods Mill, Henfield, W. Sussex,
UK.
e-mail: [email protected]

Overview
The West Weald Landscape Project (WWLP) is a partnership initiative working to promote
integrated management of a viable and enhanced landscape in the West Weald for people
and nature. It focuses on improving wildlife conservation, ‘ecosystem service’ provision,
environmental quality and public involvement in particular. The specific objectives are:
1. Conservation, enhancement and expansion of core areas of ecological interest.
2. Better-connected habitats and species populations across the whole landscape, through
improved linkages and less intensive land use and management.
3. Improved and more integrated delivery of conservation mechanisms to support targeted
land management to achieve greater environmental improvements.
4. Enhanced public enjoyment, understanding, access to and inspiration from the
landscape, with greater provision of local ‘sustainable’ natural goods and services.

This landscape-scale project covers an area of 240 km2 in the English counties of West
Sussex and Surrey, located at the western end of the clay-based Low Weald character area
in south east England. The West Weald Landscape (WWL) is internationally important for
nature conservation, a unique lowland environment dating back to the early-medieval period
that is one of the best wooded landscapes in the UK (woodland covers one-third of the area).
It is thus a rare example of a part-functioning lowland forest ecosystem in Western Europe,
as well as being an island of tranquillity in the built-up south east region of England.

Landscape ecology & land management

The ancient semi-natural woodland of the core forest areas in the WWL occur within a
surrounding matrix of mostly intensive agriculture, and thus are relatively isolated and
fragmented in the landscape. An analysis of functional ecological connectivity is being
conducted, based upon the distribution of semi-natural habitat patches and priority focal
species (listed under the UK Biodiversity Action Plan), to identify key areas and management
options for ecological enhancement. This contributes to a spatial targeting framework being
developed for proactive application of state-funded land management schemes, to secure
effective measures for biodiversity, the historic landscape and public access.

Research
A comprehensive baseline assessment has been carried out of the West Weald’s whole
natural environment and peoples’ use of it, as a foundation for implementation of the WWLP
and to enable monitoring of future change at this landscape scale. Detailed research has
been conducted additionally on the provision of ‘ecosystem services’ by more natural
landscapes at the scale of river catchments, establishing links between past land use change
and the water environment. This has highlighted impacts on hydrology and water quality from
intensive arable agriculture and the potential to reduce flood risk to human populations
downstream. Further project studies have assessed the development of pasture woodland
from ex-arable land reversion subject to a naturalistic grazing regime, and long-term
monitoring of forest dynamics in a minimum-intervention woodland. Project research reports
are available at www.westweald.org.uk.

138
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Resolving landuse conflicts through pedogeomorphic knowledge of farmers in

Hararghe highlands, Ethiopia

D. Kassahun

Environmental Researcher, Forum for Social Studies (FSS), P. O. Box 25864 Code 1000,
Addis Ababa, Ethiopia.
e-mail: [email protected]

In Ethiopia, the commonest land use conflicts were taking place between the arable and
pastoral farming farmers (Tesfaye, 2004). Since recently, however, a new type of conflict has
emerged in the Hararghe highlands. This is due to the rapid expansion of Chat (Catha
edulis), a mild narcotic crop, which is a highly competitive cash crop, affecting the food
security through replacing subsistent food crops. While farmers and businessmen want to
boost the cultivation of chat, government has been against it, leading to land use conflict.
This paper has attempted to trade-off the conflicting approaches through systematic collating
of farmers’ pedogeomorphic knowledge.
Farmers drawn from sample districts of East Hararghe were involved in the questionnaire-
based survey. Attempts were made not only to capture the information on the comparative
environmental, economical, and social advantages or disadvantages of chat vis-à-vis food
crops, but also to let farmers allot each topographic units with either chat, cereal crop, or the
combination there of.
Results of the study shows that chat fulfils the five pillars of sustainability but with a range
of suitability to different units of toposequence, called “cathasequence”. This is a pattern
where the density of chat is getting scanty down the slope while the reverse is true for cereal
crops. Such finding logically resonates with the biophysical gradient of pedo-
geomorphological models developed by Wrieght (1993), Wilson and Gallant (2000), and
Ruhe (1975).
Results of this study would help not only to unravel the emerging landuse conflict, but also
to lay the basis for low-tech and site-specific landuse planning in the region. Therefore,
experimental-based research is recommended for wider application.

References
Ruhe, R. (1975) Geomorphology. Boston: Houghton Mifflin.
Tesfaye, T. (2004) Natural resources scarcity and rural conflict, Environment, Poverty and Conflict.
FSS Studies on Poverty, Addis Ababa, No. 4, 1-14.
Wilson, J. & Gallant, J. (2000) Terrain Analysis: Principles and Applications. John Willey & Sons, Inc.
Wrieght, L. 1993. Environment systems and human impact. Cambridge University Press.

139
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Naturalistic grazing in modern British landscapes

K.J. Kirby

Natural England, Northminster House, Peterborough, UK,.

e-mail: [email protected]

Introduction

There is increasing interest in Britain in ‘re-wilding’ and naturalistic grazing, where animals
live largely free from human intervention (Hodder et al. 2005), inspired by the Dutch
experience at Oostvaardersplassen and ideas put forward by Vera (2000). Rather than
grazing being a tool to achieve specific habitat or species targets, under ‘re-wilding’
scenarios the grazing process itself becomes part of the conservation aim. Primitive breeds
of stock (typically cattle) are used as analogues of past natural grazers such as the aurochs.

Key elements of naturalistic grazing are that the animals are allowed to breed, form family
groups and may eventually die on site. (In practice welfare considerations mean that
animals are likely to be removed from the site before this happens or are humanely killed
when ‘beyond hope’ of recovery.) The populations are resource-limited, rather than kept at
prescribed densities and fluctuations from year-to-year lead to variations in grazing pressure.

Factors to consider under British conditions

The outcome of ‘naturalistic grazing’ is unpredictable: it may or may not produce the type
of mosaic of open and closed habitats suggested by Vera (2000); or as seen in the New
Forest (Hampshire) which is a managed grazing system. Habitats and species currently on
the site may be lost. Conflicts may arise with landscape and cultural heritage interests.

The reserves must be large enough to sustain all-year grazing of sufficient animals to
allow social groups to form. Cattle and other domestic stock, even if allowed to run wild are
still subject to health and welfare legislation, which means that leaving the animals
completely alone is not an option. Often the sites being considered for such trials are also
open to public access, such that human health and safety is also an issue.

Active management or rewilding?

Large herbivores can play a key role in large-scale conservation schemes; they are likely
to contribute to creating and maintaining habitat mosaics. By reducing the level of
intervention, i.e moving towards more naturalistic grazing systems, we may create systems
that are more resilient in the face of environmental change. However Britain is a
predominantly cultural landscape and over much of the countryside, including most sites
protected for nature conservation, active, targeted management is likely to be the normal
approach to meeting our conservation aims.

References
Hodder, K.H., Bullock, J.M., Buckland, P.C. & Kirby, K.J. (2005) Large herbivores in the wildwood
and modern naturalistic grazing systems. Peterborough: English Nature (Research Report 648).
Vera, F.W.M. (2000) Grazing ecology and forest history. CABI, Wallingford.

140
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Are traditional and new planted hedgerows part of the sustainable development
of agricultural landscapes?

A. Lotfi, F. Burel

Ecobio, CNRS-Université de Rennes 1, campus de Beaulieu, 35042 Rennes cedex, France,

e-mail : [email protected]

Introduction
Typical agricultural landscapes of western France called”bocage”. They are constituted of
hedgerows networks limiting pastures and corps parcels. In these landscapes hedgerows
have been used for a long time, they produce wood for timber and energy and provide many
services to farmers (Baudry et al., 2000, Burel and Baudry, 1995). In recent decades
agricultural intensification has led to a drastic decrease of these elements in most of
European agricultural landscapes (Baudry and Jouin, 2003).
We study the effects of these changes on fire wood production as a mean to assess
sustainability of the current landscapes. Our objectives are to understand the spatio-temporal
distribution of pruning, its recent changes and its relations with landscapes characteristics.

Methods and study area

The study area is situated in northern Brittany, France and is a long term ecological site. In
order to test differences among landscapes, we chose three sites (650 ha to 1800 ha) that
differ in hedgerow density and agricultural characteristics. For study the changes of
hedgerows (identification of pruning or coppicing of all the hedgerows and their date), the
aerial photos at scale of 1:10 000 acquired yearly on the study site during 10 years (from
1996 to 2005) were used.

Results
At the level of landscape, our results indicate that there is a relation between the presence
of the hedgerows, their management and the land use type of the adjacent parcels. Around
pastures there are more hedgerows than around the cultivated lands. On the other hand, the
hedgerows surrounding the cultivated fields are pruned more often than around the pastures.
Time between two pruning increase, as most of the hedgerows were not managed during the
surveyed period.
At the farm level, a study of 18 farms shows that there is no relation between the existence
of the hedges, the rate of pruning and the type of farming system. There is a significant
relation between the rate of the pruned hedges and the use of wood to heat house.

Conclusion
In spite of importance role of hedgerows in agricultural landscapes, they are the
threatened elements in these landscapes. Agricultural intensification, mechanization,
changes in management techniques and rhythms are a threat for these landscapes. new
planted hedgerows have to be integrated within this new socio-ecological context.

References
Burel, F. & Baudry, J.( 1995). Social, aesthetic and ecological aspects of hedgerows in rural
landscape as a framework for greenways. Landscape and urban plan. 33, 327-340.
Baudry, J; Bunce, R.G.H. & Burel, F.(2000) Hedgerow: an international perspective on their origin,
function, and management. J. Environ. Manage 60, 7–22.
Baudry, J.& Jouin, A.(2003). De la Haie Aux Bocages.Organisation, dynamique et gestion, INRA,
Paris

141
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

The spatial plan as instrument for protection of the underground water basin

S. Hadzi Pecova1, P. Penev2, K. Taleska3

1
Faculty of Agricultural Sciences and Food, University ”St. Cyril and Metodius” Skopje, Bul.
A. Makedonski bb, 1000 Skopje e-mail: [email protected]
2
Ministry for Local self government, ul. Dame Gruev 14, 1000 Skopje
3.
Faculty for Natural Sciences and Mathematics, University ”St. Cyril and Metodius” Skopje,
Gazi Baba bb, 1000 Skopje

Environmental protection and conservation of natural resources could be efficiently

managed through development of spatial plans. Ministry of Environment and Physical
Planning in the Republic of Macedonia initiated and conducted development of the Spatial
Plan for the protection zones of the spring Rasche, which is the main water supply source for
Skopje, the capital of Macedonia and above 500.000 inhabitants.

Spatial plan of the Rasche groundwater aquifer and adhering protection zones
subsequently defining four grades of protection regimes (2003) encompass several
municipalities. Using predominantly hydro-geological criteria (permeability of the subsoil as
well as geological features allowing for migration of pollutants), the Plan has distinguished
four areas of different vulnerability; accordingly, four protection regimes setting certain
development indicators and thresholds have been established.

Spatial development, environmental and natural resources management have been

recently delegated as a new duty to the local administration. Consequently, municipalities are
solely responsible for the implementation of the Spatial Plan. It would require local
governments to adopt the action plan, and to establish an institutional mechanism for inter-
municipal cooperation, originating from the regional character of the protected area. Local
Self Government is not yet prepared to cope with delegated responsibilities and with the
environmental management in particular. Dynamic changes in the national legislation
towards aligning with the EU requirements can not be followed by the current local
capacities. This point out the discrepancy between national policies for environmental
protection and relatively low capacities of the local self government to implement locally
concrete measures in line with broadly recognized national priorities. In the article the policy
for spatial development combining restrictive and stimulation measures towards ensuring
appropriate protection regimes for the groundwater aquifer of the spring Rasche are
presented.

References
Public Enterprise for Spatial and Urban Plans, (2003), Spatial Plan of the Protected Zones of
Rashche Source, Government of the Republic of Macedonia.
Public Enterprise for Spatial and Urban Plans, (2004), Spatial Plan of the Republic of Macedonia,
Government of the Republic of Macedonia.
Institute for Spatial Planning, (1986), Spatial Plan of the Polog Region - municipalities Tetovo,
Gostivar, Debar, Government of the Republic of Macedonia.

142
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

What is the role of riparian areas set aside from production in a large-scale
industrial tropical plantation landscape in Riau, Sumatra for vegetation and
primates?

P. Koponen1, R. Nasi 1,2, J.G. Poulsen1, M. Buitenzorgy3, W. Rusmantoro4

1
CIFOR, Bogor Barat 16680, Indonesia,
e-mail: [email protected].
2
CIRAD, Montpellier, France 3University of Wageningen, The Netherlands 4Lantana,
Indonesia

Introduction
At present more than 2 million hectares across Indonesia are in fast growing plantation
forests and 9 million hectares are targeted for plantation development (FWI/GFW 2002). In
Indonesia plantations are generally consist of large mono-specific blocks interspersed with
natural forest remnants. The extent of these forest remnants vary and legislation for
sustaining the ecological basis of plantation establishment is fragmented and/or left to the
interpretation of the plantation companies. In Riau, Central Sumatra, the Tesso Nilo natural
forest complex represents one of the large intact diverse rain forests remaining on the island
of Sumatra (Gillison 2001). Our study comprises off three large scale industrial Acacia
mangium plantation sectors bordering Tesso Nilo, with average of 18 % of their area being
set aside from production, mainly as natural riparian forests protecting the riparian areas.
Methods
We sampled natural vegetation diversity in the Baserah sector (systematic sampling of set
aside areas, involving a total of 347 plots measured giving a sampled area of 0.14 ha, 0.9 ha,
3.5 ha and 13.9 ha for seedlings, saplings, poles and trees respectively), Tesso Nilo National
Park (29 plots, area of 0.012 ha, 0.07 ha, 0.3 and 1.6 ha respectively) and examined patterns
of primate species richness and abundance using transects in set aside and in planted areas
in three plantation sectors. Results were analyzed in relation to spatial arrangement and
dimensions. We tested the hypothesis that patterns of primates are independent of
connectivity, width, distance to roads, crown closure, age of neighbouring plantation stands
and height, respectively, riparian forests set aside from plantation.

Results and discussion

Natural forest remnants inside plantations may, if appropriately designed, be used to
mitigate the negative impact of large scale industrial plantations on biodiversity by providing
some degree of connectivity with and between remaining natural forest patches (MacDonald
2003). With eight species of primates and more than 250 tree species recorded from riparian
forests, set aside areas inside plantation are comparable with primate diversity to Tesso Nilo,
but have lower tree diversity. Primates were not abundant in any of the A. mangium
plantation stands, irrespective of stand age, size and location. They were only observed in
riparian forests that were connected to either a larger conservation area inside the
concession or other natural forest areas. A well-connected network of natural forest
corridors is of critical importance to maintain primates in such landscapes. This study
provides a rigorous baseline data but further monitoring is needed
References
FWI/GFW (2002) The state of the forest: Indonesia. Forest Watch Indonesia and Washington DC:
Global Forest Watch, Bogor.
Gillison, A.N. (2001) Vegetation survey and habitat assessment of the Tesso Nilo forest complex.
WWF-US, Pekanbaru.
MacDonald, M.A. (2003) The role of corridors in biodiversity conservation in production forest
landscapes: a literature review. Tasforests 14:41─50.

143
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

One step back, one step forth: agricultural expansion and landscape change
trajectories in shifting cultivation system of Southern Cameroon

V. Robiglio, F. L. Sinclair

University of Wales, Bangor, Gwynedd, UK.

e-mail: [email protected]

Forests in the Humid Forest Zone (HFZ) of Cameroon are encroached by logging and small-
scale farming. Local land users combine shifting cultivation with cocoa agro-forestry,
gathering of NTFPs, hunting, fishing, and small-scale logging for domestic use. The
proportion of land annually cleared for cultivation, its spatial distribution, together with fallow
sequence and length affect shifting cultivation landscape dynamics. Here we studied on a
fine scale the land cover dynamics in Southern Cameroon. We quantified the impact of
agricultural expansion on the landscape pattern of two contrasting villages: the magnitude,
spatial configuration and distribution of land-cover transitions and trajectories. We focused on
secondary forest regrowth and the patterns of shortening fallows.

Fine-resolution GIS-based models captured Land Use/Land Cover trajectories. Modelling

was based on analysis of land-cover patch dynamics over 50 years. An interdependent on-
screen interpretation, based on a LCCS-derived hierarchical legend, was applied to historical
aerial photographs (1951, 1984) and very high resolution satellite images (Ikonos-2001).
Transitions were classified in modification or conversions depending on the distance
between classes in the legend frame. Patch trajectories were defined from the sequence of
classifications as either agriculture or forest for each observation date. The relation of the
spatial pattern of trajectories with locational factors was assessed (distance from the
residential unit-accessibility, distance from watercourses, distance from already farmed
fields).
Quantitative analysis of transitions highlighted an intense clearing/secondary regrowth
dynamic, and a significant shortening and intensifying of fallow rotations. Forest loss rates
increases but not due to an increase in primary forest clearing. Spatial analyses of
transitions and trajectories highlighted the fine-grained fragmentation of land cover dynamics
in our study. Similar contiguous processes aggregate, originating wider patches. These are
mosaics of farmed fields or fallows in the same cohort.
In Akok, the southernmost village, a balance existed with intensification and shortening fallow
cycles on already cleared areas, yet long cycles of secondary vegetation regrowth in fallow
areas and little change in primary forest. In Awae, the more accessible village, continuous
clearing has almost exhausted all natural old growth forest with little secondary forest
regrowth. Land Use / Land cover trajectories present different spatial patterns. For example
stable agriculture patches are rarer with increasing distance from the village but average
patch size is constant. Forest regrowth patches are larger with increasing distance from the
village.
These aspects are relevant for integrating macro and meso-level assessments of
deforestation in Central African forests and for evaluating the potential of shifting cultivation
landscapes in providing ecological and environmental benefits for local communities and of
global environmental services.

Future research
Influence of fragmented landscape structure on secondary vegetation processes will be
assessed through patch level vegetation analysis in two1 Km2 quadrates per village (45
sampling plots per quadrate along ideal transects crossing bordering vegetation units).
Information on local ecological knowledge, on locational properties of land cover units that
influence farmers’ decision-making, on land use practices (e.g. fallow management) as well
as on tenure and demographic dynamics over time will be integrated into a panel data set for
the analysis of dynamics at the plot level.

144
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Rationalising biodiversity conservation in dynamic ecosystems (RUBICODE

project)

P.A. Harrison1, R. Bugter 2, RUBICODE partners 3

1
Environmental Change Institute, Oxford University Centre for the Environment, South Parks
Road, Oxford OX1 3QY, UK.
e-mail: [email protected]
2
Alterra, P.O.Box 47, 6700 AA Wageningen, The Netherlands.
3
see www.rubicode.net

Rationalising Biodiversity Conservation in Dynamic Ecosystems (RUBICODE) will review

and develop concepts of dynamic ecosystems and the services they provide. Nature is
fundamentally dynamic, as are the pressures of human activities on biodiversity, yet most
conservation strategies focus on protected areas and are still developed around a static view
of nature. For the realisation of future conservation objectives it is critical that new strategies
and policies incorporate these dynamics. RUBICODE will address this by developing
innovative concepts for conservation strategies that concentrate on managing dynamic
ecosystems for maintaining their capacity to undergo disturbance, while retaining their
functions, services and control mechanisms (ecological resilience; Gunderson (2000)).

Our approach in developing concepts of dynamic ecosystems and the services they
provide is based on a new category of population, the Service Providing Unit (SPU), which
provides a recognised service at some temporal or spatial scale (Luck et al., 2003). For
example, if the service is conservation of a rare insect, the SPU would be the population of
its host plant present in suitable habitat in the area over which special conservation
measures are required. SPUs often comprise more than one species and any given species
may contribute more or less than another species to a given service. They may also
contribute to more than one service or be antagonistic to another service (termed Service
Antagonizing Units (SAUs)). For example, wild flowers may support crop pollinators and
biocontrol agents, but also harbour pests or compete directly with crops. Quantifying the
potential positive effects of biodiversity on a service should involve the subtraction of the
effects of SAUs. Additionally, one person’s SPU may be another’s SAU and thus,
stakeholder involvement in the process is required from an early stage.

The SPU/SAU concept provides a framework for linking changes in key characteristics of
populations with implications for service provision. Indeed, identifying quantitative links
between components of ecosystems and service provision is crucial to providing specific
rather than vague management guidelines for policy makers and land managers.
Relationships between SPUs and socio-economic and environmental drivers of biodiversity
change are being evaluated in case studies covering the main ecosystem service categories
of the Millennium Ecosystem Assessment across multiple scales. Methods for linking
biodiversity traits to service provision and for improving and testing indicators are being
developed and used to explore effective management strategies for the SPUs and to inform
priorities for biodiversity conservation policy. RUBICODE will also identify current gaps in
knowledge and propose a roadmap for future research.

References
Gunderson, L. (2000) Ecological resilience - in theory and application. Annual Review of Ecological
Systems 31: 425-439
Luck, G.W.; Daily, G.C. & Ehrlich, P.R. (2003) Population diversity and ecosystem services. Trends
in Ecology and Evolution 18: 331-336.

145
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

EucaLand European Culture expressed in Agricultural Landscapes

G. Pungetti1, A. Kruse2
1
CCLP – Cambridge Centre for Landscape and People, 9 Selwyn Road,
Cambridge CB3 9EA, UK,
e-mail: [email protected]
2
BfLS – Bureau for Landscape & Services, 162 Av. de Paris, 92320 Châtillon, France,
e-mail : [email protected]

Land use is interlinked with the culture and history of its society, and the Agricultural
Landscape (AL) is the one of most evident expressions of it. Valuable and typical ALs are
part of European heritage and character. Explaining this character makes people aware of
their landscapes and helps in turn to foster a European identity. Without this, future
generations will perceive ALs from a mere industrial land use point of view and will abandon
them losing identity.
Agricultural landscapes are indeed expression of our identity and heritage, and yet retain
ecological values and play an important role in sustainable development. However they are
currently mostly considered as places for production. To overcome this limit, EucaLand has
been set up with the following goals:
• To consider the European Agricultural Landscapes (EALs) as part of our cultural
heritage, including their values and meaning for the Europeans;
• To describe past and present EALs for a common classification;
• To produce recommendations on alternatives to deal with future EAL, culturally and
sustainably oriented, addressed to scientists and planners, but also to policy makers and
especially to the European public.

EucaLand unites a large network of research departments and institutes with interdisciplinary
and intercultural vision: more than 40 partners from over 30 countries have already started to
work together towards a Pan-European description of EALs history and types, raising
awareness on the cultural values of EALs and constructing a website to disseminate the
finings. Links with the CoE European Landscape Convention, UNESCO World Heritage
Centre, ECOVAST, FAO, WWF and IUCN are established.

The EucaLand network aims to disseminate knowledge on EALs, and in particular to:
• Scientific experts and cultural operators, which will gain know-how in disciplines
not directly linked to heritage preservation but essential for it;
• The public: people living and working in the countryside, which can become familiar
and aware of EAL heritage;
• Authorities: institutions at different levels dealing with agricultural land and its
development, which will benefit of cross border connection of regions of same AL
type;
• Economic sectors: people in disadvantaged areas, land owners and land
developers from public and private sectors, and local tourism in the agricultural
countryside, which will be able to use the EucaLand toolkits and recommendations;
• CAP: the Common Agricultural Policy will benefit of an elaborated concept on AL,
including guidelines to address their specific values and types. The EucaLand
findings can assist the development of a more rationalised agricultural policy,
regarding AL as part of our common and valuable cultural heritage. Therefore a
strong link of each partner to its agricultural ministry is envisaged.

146
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Landscape context affects the abundance & diversity of bees on annual crops in
Europe

G. Carré1*, B.E. Vaissière1, R. Chifflet1, N. Morison1, R. Bommarco2, S.G. Potts3,

S.P.M. Roberts3, G. Rodet1, J. Settele4, I. Steffan-Dewenter5, H. Szentgyörgyi6, C.
Westphal5, M. Woyciechowski6 and P. Roche7
1
INRA, UMR Ecologie des Invertébrés, Avignon, France. e-mail:
[email protected]
2
Department of Entomology, Swedish University of Agricultural Sciences, Uppsala, Sweden.
3
Centre for Agri-Environmental Research, University of Reading, UK.
4
Umweltforschungszentrum–Centre for Environmental Research, Halle, Germany.
5
Department of Animal Ecology, University of Bayreuth, Bayreuth, Germany.
6
Institute of Environmental Sciences, Jagiellonian University, Krakow, Poland.
7
CEMAGREF, Aix en Provence, France.

Introduction

Bee populations are declining worldwide (Biesmeijer et al. 2006), and this loss may alter
the stability of the pollination service bees provide (e.g. Chapin et al. 2001). It is therefore
urgent to better understand the relationship between bee populations and biogeographic
conditions. We assessed the impact of landscape contexts on the abundance and diversity of
bees. The same protocol was used in 5 countries over Europe (France, Germany, Poland,
Sweden & the United-Kingdom) in order to be able to generalize results on a larger scale.

Methods

Bees in 5 crops (one per country) were sampled over a total of 45 sites x 4 dates in 2005
and a detailed land-use classification was used in a 3 km radius around each study site
((CORINE Land cover database). We extracted landscape parameters (nature & structure)
using GIS & specific software (GRASS & FRAGSTAT). Then we used multivariate methods
(correspondence, clustering, and partial least-square regression analyses) to quantify the
impact of landscape features and geography on the abundance and diversity of bees in
Europe at the sub-generic taxonomic level.

Results

Three groups of bees emerged based upon their diversity and abundance patterns (1-
Poland, Germany & Sweden; 2-United-Kingdom; 3-France). Overall abundance and diversity
were positively affected by some natural habitat, such as transitional woodland-shrub, but
also by some urban habitat, such as sport and leisure facilities and crops, such as pastures.
But there were differences among the three groups due to the specific habitat response from
the bees that dominated each group. We characterized bee groups that were positively
affected either (i) only by natural habitat, (ii) only by urban habitat, (iii) by some natural, urban
and crops habitats. Despite different pollinator guilds among the countries and crops, the
landscape context still had a significant effect on the abundance and diversity of bees.

References
Biesmeijer, JC et al. (2006) Parallel declines in pollinators and insect-pollinated plants in Britain and
the Netherlands. Science, 313, 351-354.
Chapin, FS et al. (2000) Consequences of changing biodiversity. Nature, 405, 234-242.

147
Theme 1: Landscape, stakeholders, land use and policy
1.7 Posters

Landscape organization and plant biodiversity in an intensively used agricultural

region of central France

F. Di Pietro

Université de Tours, Equipe Dynamiques Environnementales et Paysagères, UMR 6173,

Parc de Grandmont, F-37200 Tours;
e-mail: [email protected].

European agricultural landscapes were shaped by agroecological units, now changing into
new landscape units more linked to farms’ spatial patterns. Therefore farms, which are main
management units, are nowadays also main landscape units and play an essential role in
enhancing biodiversity, through agricultural practices of field management and field margins'
care. Field margins are a crucial landscape element for supporting biodiversity (Marshall,
2002; Le Coeur et al., 2002). Especially in intensively used agricultural regions, because of
the scarcity of grasslands, field margins play a key role in allowing species dispersion and
thus in enhancing biodiversity (Petit & Burel, 1998).
In this poster our aim is to outline the results of a several-years research on landscape
organization and field margins structure and botanical composition in an agricultural
landscape of central France (Gâtine lochoise) (Di Pietro, 2006).
Our objectives are: (i) to study the relationships between agriculture and landscape units,
and to understand at which extent farms can represent landscape units, in addition to being
management units; (ii) to assess the respective contribution of landscape factors Vs
agricultural practices to plant biodiversity of field margins.

Methodology and results

In order to deal with this issue, several units have been studied: farms, fields, field
margins. The contribution of many variables to field margins’ botanical composition of about
400 margins has been analysed using Canonical Correspondence Analysis (CCA), and has
led to a hierarchy of relevant variables.
Among these variables we will focus particularly on land use characteristics including crop
rotations and agricultural practices of management of field margins. Their impact over life
traits of field margins plants, particularly perennials versus annual weeds, will be examined.
We emphasize the role played by some agricultural variables, especially spatial pattern of
farms, farm size, and crop rotations involving grasslands on the one hand, and by some
structural parameters such as forest edges on the other hand. The role of landscape factors,
such as density and size of woodlots, is suggested. We particularly underline the major effect
of farm size over the plant composition of field margins. The increase of field size is a major
trend of modern agriculture; its effects over biodiversity, by related loss in habitats and
corridors, are known. We suggest that also the increase of farms size has a dramatically
harmful impact on biodiversity, because of the more drastic management of field margins
that it entails.

References

Di Pietro, F. (2006) Agriculture and biodiversity: assessing the contribution of agricultural and
structural parameters to field margins plant diversity. A case study in a crop field region (Centre
region, France), In: Meyer BC (Ed). Sustainable Land Use in Intensively Used Agricultural Regions,
Landscape Europe, Wageningen, pp. 140-151.
Le Coeur, D.; Baudry, J.; Burel, F. & Thenail, C. (2002) Why and how we should study field
boundary biodiversity in an agrarian landscape context. Landscape and Urban Planning 89:23-40.
Marshall, E. J. P. (2002) Introducing field margin ecology in Europe. Agriculture, Ecosystems &
Environment 89:1-4.
Petit, S. & Burel, F. (1998) Quelle biodiversité en zone de grande culture ?, Ministère de
l'Environnement et CNRS.

148
Theme 2: Urban environment and transport

Theme 2: Urban Environment and Transport

149
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

150
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

2.1 Symposium 2: Efects of roads and traffic on wildlife populations and

landscape functions

Effects of wildlife passages on the viability of badger populations in the

Netherlands

E.A. van der Grift, J. Verboom

Alterra, Wageningen University and Research Centre – P.O. Box 47, Wageningen, The
Netherlands,
e-mail: [email protected]

Introduction

In the Netherlands badger (Meles meles) populations declined rapidly in the 1970s and
1980s. Besides habitat loss and habitat fragmentation, mortality due to collisions with traffic
at roads and railroads was found to be one of the main reasons for this decline (Van der Grift
et al. 2001). To reduce road-kill and counteract the barrier effect of transport corridors
numerous wildlife crossing structures, such as badger tunnels, have been built for badgers
since the 1970s in the Netherlands. The use of these structures by badgers has been
repeatedly documented. Whether the restored habitat connectivity resulted in an increase in
badger population viability is, however, a question that is harder to answer than whether the
planned mitigation measures will be sufficient to ensure the survival of (local) populations.
The planned extension of highway 73 will intersect core badger habitat. To prevent habitat
fragmentation and an increase in the number of road-kill, a plan has been worked out for
mitigating the negative impacts of the new road, including the construction of wildlife
passages and fences to keep the animals from entering the road. Furthermore, the plan
contained proposals for habitat restoration and the development of small-scale linear
landscape elements in order to improve connectivity between foraging areas and sett sites,
and to guide the animals to the wildlife passages. Our objective was to assess whether the
proposed set of mitigating and compensating measures for the planned extension of highway
73 would be sufficient to ensure the survival of badgers in the area.

Methods

We used the model DASSIM to assess badger population viability after the proposed
construction of the southern stretch of highway 73 in the Netherlands. DASSIM is a spatially
explicit, dynamic (meta)population model (Lankester et al. 1991). The modelhas the ability to
distinguish individual badgers, badger clans, local populations, i.e. clusters of badger clans,
in which random mating occurs, and metapopulations, i.e. clusters of local populations,
connected by dispersal. Local populations are spatial explicit, which is expressed in
differences in the exchange probability between local populations. In addition, differences in
mortality probability between local populations due to high or low road densities or
differences in traffic volume can be applied. Other characteristics of DASSIM are the
inclusion of demographic stochasticity, the ability to distinguish male and female badgers in
two age classes (juvenile and adult), and the ability to include knowledge about the social
structure of badger populations (Verboom 1996, Van Apeldoorn et al. 1998).
We analysed three scenarios for highway construction, which differ in the amount of
mitigation and compensation measures taken, and the expected effectiveness of these
measures (100% versus 50% effectiveness, i.e. half the wildlife tunnels cannot be used and
half the fences are defective). These scenarios were compared with the situation that no
highway is constructed.

Results

151
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

The study showed that highway construction is not necessarily bad for the viability of badger
populations in the region. That is, if all proposed wildlife passages and fences are
constructed and all these measures remain functional, i.e. all wildlife tunnels can be used
year-round and fences show no failures. Proper management of measures appears to be of
decisive importance to the survival of the species in the region: if only half of the measures is
effective the badger population is likely to disappear (fig. 1). If no construction of the highway
takes place, and consequently no mitigation measures are installed, expected autonomous
increase in traffic volume at local and regional roads would result in high badger mortality.
That is why badger population viability is expected to be lower for this scenario than in case
the highway, including all wildlife measures, is constructed.

160
140
Mitigation
120
measures 100%
of badgers
dassen

100 effective
au ton oom m et, a
m itig atie
80
Mitigation
au ton oom m et, 5
Numberaantal

60 measures
effectiviteit 50%
40 effective
20
0

Figure 1. Simulated trends in badger numbers after highway construction in case all
mitigation measures (1) or only half of the mitigation measures (2) function well.

Conclusions

Road planning more and more requires the assessment of impacts on nature and the
environment. Although impacts on individual animals have to be addressed, e.g. expected
road kill rates, more emphasis should be put on the impacts of road construction and road
use on the viability of populations. Models and expert systems may be helpful tools to assess
population viability. These tools give the possibility to predict changes in viability, or even
threats to the (local) survival of a species, before road construction is started and thus may
play a key role in comparing scenarios and in decision making.

References
Lankester, K.; van Apeldoorn, R.C.; Meelis, E. & Verboom, J. (1991) Management perspectives for
populations of the Eurasian badger (Meles meles) in a fragmented landscape. Journal of Applied
Ecology 28: 561-573.
Van Apeldoorn, R.C.; Knaapen, J.P.; Schippers, P.; Verboom, J.; Van Engen, H. & Meeuwsen, H.
(1998) Applying ecological knowledge in landscape planning: a simulation model as a tool to
evaluate scenarios for the badger in the Netherlands. Landscape and Urban Planning 41: 57-69.
Van der Grift, E.A.; Hoogeveen, Y.R.; Kamphorst, D.A.; Jaarsma, C.F.; Kleijberg, R.J.M.;
Piepers, A.A.G.; Snep, R.P.H.; Soesbergen, M.; Veenbaas, G. & de Vries, J.G. (2001)
Fragmentation by existing infrastructure. A.A.G. Piepers (ed). Infrastructure and nature;
fragmentation and defragmentation. Road and Hydraulic Engineering Division, Delft, The
Netherlands, pp 47-72.
Verboom, J. (1996) Modelling fragmented populations: between theory and application in landscape
planning. Institute for Forest and Nature Research, Wageningen, The Netherlands.

152
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Species level differences in the road effect zone of a major motorway for anuran
populations in Ontario, Canada

F.Eigenbrod1, S.Hecnar2, and L. Fahrig1

1
Geomatics and Landscape Ecology Research Laboratory (GLEL), Department of Biology,
Carleton University, Ottawa, Ontario, Canada K1S 5B6
e-mail: [email protected]
2
Department of Biology, Lakehead University, Thunder Bay, Ontario, Canada P7B 5E1

Introduction
The term ‘road effect zone’ (Forman & Alexander, 1998) describes the extent of
significant ecological effects from the edge of a road. Anurans (frogs and toads) are known
to be negatively affected by roads at the landscape scale (Vos & Chardon, 1998; Houlahan &
Findlay, 2003), with the major negative effect of roads on anurans thought to be direct
mortality from traffic (Fahrig et al., 1995; Hels & Buchwald, 2001). However, the extent of the
road effect zone of a single high-traffic road on anurans has never been quantified.

Methods
We measured the population abundance of seven common species of anurans at 36
ponds located between 50 m and 3600 m from a 60 km section of a major motorway
(highway 401) in a rural portion of eastern Ontario, Canada in the spring and summer of
2006. We measured relative abundance as the sum of call intensity indices (scale of 0-3) for
each species over 8 nocturnal surveys. For two species (leopard frog Rana pipiens and
green frog R. clamitans), we also included the number of adult frogs seen at each pond as
an additional measure of relative abundance. We regressed relative abundance against
distance to the motorway using generalized linear models.

Preliminary Results
We found a significant positive relationship between distance from the motorway and
relative abundance for three of seven species. All three of these species – wood frog (R.
sylvatica), spring peeper (Pseudacris crucifer), and gray treefrog (Hyla versicolor) - showed a
threshold-type response within 300 m of the motorway (Figure 1). We also found a borderline
significant effect of distance to the motorway for two other species (leopard frog and green
frog), but for these species there was gradual decrease in abundance with decreasing
distance, rather than the clearly defined road effect zone exhibited by the previous three
species. There was no significant relationship between the distance to the motorway and the
relative abundance of two other species – American toad Bufo americanus and western
chorus frog P. triseriata. Re-sampling of a subset of the study ponds is planned for the spring
of 2007 to confirm that the 2006 results are not simply an artefact of the year of sampling.

Discussion
Our results indicate that the motorway in this study has a severe road effect zone of 300
m for at least three species of anurans. Interestingly, the three species which showed the
strongest effect of distance to the motorway – wood frog, spring peeper and gray treefrog –
have been previously shown to be relatively unaffected by traffic density in the landscape
(Felix Eigenbrod, unpublished data). However, the motorway in this study (highway 401) is
unusual in its very high levels of night-time heavy truck traffic. We suggest that the 300 m
road effect zone we found for these species may be due to their avoiding ponds near the
motorway as the very high noise levels found at these sites may interfere with their nocturnal
mating calls. The avoidance of roads with high night-time traffic volumes may thus be
another factor besides the well-documented negative effect of traffic mortality on anuran
populations in human-dominated landscapes.

153
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Figure 1: The relationship between the distance of breeding ponds to a motorway and the
relative abundance (summed call intensity indices) of three species of anurans.

References
Fahrig, L.; Pedlar, J.H.; Pope, S.E.; Taylor, P.D. & Wegner, J.F. (1995) Effect of road traffic on
amphibian density. Biological Conservation 73:177-182.
Forman, R. T. & Alexander, L.E. (1998) Roads and their major ecological effects. Annual Review of
Ecology and Systematics, 29: 207-231.
Hels, T. & Buchwald, E. (2001) The effect of road kills on amphibian populations. Biological
Conservation 99: 331-340.
Houlahan, J. E., & Findlay, C. S. (2003) The effects of adjacent land use on wetland amphibian
species richness and community composition. Canadian Journal of Fisheries and Aquatic
Sciences 60:1078-1094.
Vos, C. C., & Chardon, J. P. (1998) Effects of habitat fragmentation and road density on the
distribution pattern of the moor frog Rana arvalis. Journal of Applied Ecology 35:44-56.

154
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Effects of traffic infrastructure on patch dynamics

H. Reck, D. Lorenzen, J. Schrautzer

University of Kiel, Ecology Centre, Department of Landscape Ecology; Olshausenstrasse 75;

D- 24118 Kiel (Germany);
e-mail: [email protected]

Introduction
Spatiotemporal interactions between habitats and populations are complex and difficult to
communicate especially in planning procedures concerning road traffic. For the analysis of
population vulnerability due to traffic infrastructure or to forecast impacts of barriers and
compensation measures on landscape level we should combine ideas from the theories
about: 1. patch dynamics, 2. intermediate disturbance (both e.g. triggered by migrating
megaherbivores) and 3. metapopulations (Fig.1). Regarding those theories, potentials for
migration of species strongly determine the functioning of changing landscapes for
safeguarding biological diversity. But until today local metapopulations are within the focus of
impact assessment at best. Local habitat dynamics are mostly neglected and dynamics of
species or habitats on regional level are ignored.

Fig. 1: Conceptual
framework; main factors for
the spatiotemporal viability
of populations

Experiments
To find out how patch
dynamics, habitat
dependant mobility and
barriers could affect each
other we use several
monitoring projects. Our
most integrative
investigation started when
in 1998 a new habitat
corridor of 0,5 km x 8 km, the “Eidertal Valley” (ETV) near Kiel, was implemented as part of
the Schleswig-Holstein (SH) ecological network which is thought to compensate large scale
landscape fragmentation. SH, the 2nd smallest state of Germany is 15.761 km2 in size and
strongly dissected by traffic infrastructure. The current landscape analysis of Lorenzen
shows that 50 % of the SH area is already divided by canals or by roads (with a traffic flow of
more than 1000 cars / day) into patches smaller than 30 km2 and most of the remaining
valuable habitats are isolated and enclosed within the smaller patches. Thereby it is
noticeable that the size of the remaining undissected areas and their share of valuable
habitats are correlated slightly negative.
The investigations in the ETV are combined with experiments on grasshopper dispersal
along highways and on overpasses and with the analysis of habitat topology in SH at all.

155
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Results
Looking at processes or on habitat heterogeneity created by species themselves in the
ETV we found, that patch dynamics due to disturbance by cattle lead to habitat change and
to significantly increased patchiness in all investigation units (between 1999 and 2003 e.g.
from 8 to 14 vegetation types per 2500 m2) while the number of microhabitats increased even
more. Consequently the density e.g. of the egg clumps of grasshoppers (mainly the flightless
C. apricarius) were affected. Disturbance sites near vegetation borders mainly caused by the
makrofauna contain on average more than 40 times higher numbers of egg clumps (max.
188 egg clumps / 900 cm2) than the average grassland and the inhabited area of the
respective species has increased about 300 %. Higher grasshopper densities then can affect
vegetation structure and plant composition additionally and altogether the resulting
patchiness increases grasshopper emigration or dispersal respectively as well as effective
immigration. In comparison active grasshopper dispersal across frequently used roads is
completely inhibited so that isolated newly developed habitats remain deserted as long as
they are not connected e. g. by overpasses. But overpasses without combined landscape
development have low effect on insect connectivity again.

Fig. 2: Cascade or effects due to traffic infrastructure within the examined system

Modelling
Modelling habitat quality and habitat distribution in space and time and regarding the
influence of megaherbivores as well as modelling insect dispersal and immigration (as
indicator for the possibility of flightless species to react to changing landscapes) we can
show
1. the importance of accessibility of areas for big mammals or the importance of high
mobility of big mammals between habitats respectively and
2. the long-term risk of species loss on landscape level caused by barrier effects inhibiting
adaptive processes and
3. requirements on habitat connectivity in ecological networks.
Thereby we are suffering a lack of information about the barrier effect of roads with low
traffic density and about the effects of very narrow linear compound biotopes. But even more
important is the lack of information about the quantitative role of free living megaherbivores
as vectors for non-vertebrate species and their role as habitat builders since transhumance is
obsolete in Central Europe.

156
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Effects of road network density on abundance and road mortality of wildlife

populations

I. A. Roedenbeck

Biometry and Population Genetics – IFZ Research Centre for BioSystems, Land use and
Nutrition, Heinrich-Buff-Ring 26-32, 35392 Giessen, Germany.
e-mail: [email protected]

Introduction

Numerous indices have been developed in the latest past quantifying the degree of
landscape fragmentation by roads. One example is effective mesh size meff (Jaeger 2002),
being used in environmental monitoring programmes in Germany and Switzerland (Jaeger
2001). It is argued, that these indices indicate the state of wildlife populations, because local-
scale road effects on animals are proven (review in Glitzner et al. 1999, Trombulak & Frissell
2000), and hence, we could extrapolate these results to larger scales and populations.
However, this argumentation poses formidable extrapolation problems. Little is known about
the landscape-scale effects of roads on wildlife, thus far, and relationships between the
values of fragmentation indices and the abundance of populations have not been proven yet.

The aim of our study was to analyse landscape-scale road effects by comparing road
network density with abundance and road mortality of wildlife populations. We conducted our
study in Hesse, a 21´115 km² federal state of Germany. We used meff as index for the degree
of fragmentation, and official hunting statistics for roe deer to obtain abundances and
mortality rates of one of the last remaining mammal species in Germany with a large home
range. When conducting the analysis and interpreting the results we were confronted with
several problems, which we think are representative for empirical road studies at the
landscape scale. This paper should be a starting point for a discussion about how to deal
with problems of study designs at large scales. It is alarming that conducting studies and
generating results of high evidentiary weight is highly problematic at large scales, although
the most pressing policy and management decisions (e.g. decisions about the de-
fragmentation of road networks) take place at the landscape-scale.

Results

Landscape fragmentation in Hesse increased seriously since the 1930. The higher road
network density in an administrative district the smaller local populations of roe deer (Fig. 1).
Road mortality rates increased with increasing road network density (for details see
Roedenbeck & Köhler 2006). In a multiple regression analysis we included potential roe deer
habitat (area of forest, settlements, grasslands and agriculture) to correct our correlations.
Following the final model, roe deer abundances can be explained by the amount of forest
and settlements per administrative district, but meff was not included in the final model
equation.

Although roe deer populations seem to be larger in non-fragmented forest areas, we

cannot say whether this is due to few roads, much forest or the absence of settlements. It is
quite obvious that our study design failed to exclude habitat suitability as potential impact
factor. However, with the data in hand this was simply not possible. Base unit for regression
analysis were highly variegated administrative districts (base unit for hunting statistics), and
the small sample size of 27 districts did not allow any further classification.

157
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Figure 1. Effects of road network density (effective mesh size) on abundance and road
mortality of roe deer in the federal state of Hesse, Germany (published in
Roedenbeck & Köhler 2006).

Discussion

We argue that empirical studies about road effects at the landscape-scale are exposed to
many problems. First, it is highly problematic to find investigation sites with a comparable
amount of habitat. The larger the scale of investigation (e.g. due to dispersal ranges of target
species), the more difficult it is to find comparable sites. Second, large investigation areas
decrease the sample size. The smaller the sample size, the less replication is feasible, and
the lower evidentiary weight of the study results. Third, at large scales the researcher cannot
collect data by himself. Consequently, there is an urgent requirement for large-scale,
systematic wildlife monitoring programmes. These are seldom available for the species of
interest, and often of very low quality. Fourth, studies analysing potential impacts of human
activities on the environment should analyse the reaction of populations before and after an
intervention. In the context of road ecology this means monitoring populations over several
decades following the densification of road networks. But it is impossible to find sites in the
landscape, where all other potential impact factors (e.g. development of settlements,
intensification of agriculture) were similar over the entire time span of the study.
Consequently, results of large-scale studies have to be interpreted with caution, and we
urgently have to develop ways to incorporate such results into decision making albeit the
inherent limitations.

References
Glitzner, I., Beyerlein, P., Brugger, C., Egermann, F., Paill, W., Schlögel, B. & Tataruch, F. (1999)
Literaturstudie zu anlage- und betriebsbedingten Auswirkungen von Strassen auf die Tierwelt.
Endbericht. Magistrat der Stadt Wien, Abteilung 22-Umweltschutz. „G5“-Game-Management,
Graz, 178 S.
Jaeger, J., Esswein, H., Schwarz-von Raumer, H.-G., Müller, M. (2001) Landschaftszerschneidung
in Baden-Württemberg - Ergebnisse einer landesweiten räumlich differenzierten quantitativen
Zustandsanalyse. Naturschutz und Landschaftsplanung 33(10): 305-317.
Jaeger, J. (2000) Landscape division, splitting index, and effective mesh size: new measures of
landscape fragmentation. Landscape Ecology 15 (2): 115-130.
Roedenbeck, I.A. & W. Köhler (2006) Effekte der Landschaftszerschneidung auf Unfallhäufigkeiten
und Bestandsdichte und von Wildtierpopulationen. Naturschutz und Landschaftsplanung 38
(10/11): 314-32.
Trombulak, S. C., and C. A. Frissell (2000) Review of Ecological Effects of Roads on Terrestrial and
Aquatic Communities. Conservation Biology 14(1): 18-30.

158
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

How does the configuration of road networks influence the degree to which
wildlife populations are affected by roads?

J.A.G. Jaeger1
1
Swiss Federal Institute of Technology ETH Zurich, Department of Environmental Sciences
D-UWIS, Group for Ecosystem Management, Universitaetstr. 22, CHN F 73.2, CH-8092
Zurich, Switzerland.
e-mail: [email protected]

Introduction

Roads act as barriers to animal movement, increase wildlife mortality due to collisions
with vehicles, and reduce the amount and quality of habitat (Forman et al., 2003; Jaeger et
al., 2005). The persistence probability of wildlife populations in response to increasing road
density exhibits a critical threshold above which populations are prone to extinction (Jaeger
and Holderegger, 2005). To explore the potential of designing less detrimental road networks
such as the bundling of roads and traffic, I used a spatially explicit individual-based
simulation model of population dynamics. The purpose of this study was (1) to assess the
relative importance of the mechanisms through which the configuration of road networks
influences the degree to which roads detrimentally affect wildlife populations, (2) to better
understand how this influence depends on the behaviour of the animals at roads, and (3) to
identify characteristics of road network configurations that make road networks less
detrimental to the persistence of animal populations. My initial expectation was that the effect
of a road network on population persistence is the more detrimental the more patches it
creates.

Figure 1. Comparison of the effect of road network configuration (gridded pattern vs. parallel
configuration) on population persistence. The roads (black lines) are located between the
cells of the landscape model (indicated by the grey lines). L = number of roads, n = number
of patches (modified after Jaeger et al., 2006).

159
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Methods

I compared two groups of networks: (1) roads that were evenly distributed across the
landscape versus roads that were bundled together in one part of the landscape (close to
each other or combining all traffic on one larger road); and (2) a parallel pattern of roads
versus a gridded pattern where the patches or “meshes” form a checkerboard (Fig. 1). I
recorded persistence probability, times to extinction, and critical road densities in the runs of
a stochastic individual-based simulation model of population dynamics (Jaeger et al., 2006).

Results

The model results clearly supported the bundling concept: The overall barrier effect
created by a group of several roads bundled together, or by an upgraded road with more
traffic on it, was never more harmful to population persistence than the negative effects of an
even distribution of roads across the landscape (with the same overall traffic). However, the
results for a gridded versus parallel road pattern were surprising: The expectation that
fragmenting the landscape into more patches (gridded pattern) would be more harmful to
population persistence (while total road length is kept constant) was contradicted by the
model results when the degree of road avoidance by the animals was low (and traffic
mortality was the most relevant mechanism). Therefore, for animals that do not very strongly
avoid roads (e.g., amphibians) it is more important to preserve core habitats at sufficient
distances from roads than keeping the number of patches low. For animals that strongly
avoid roads, the more relevant mechanism is isolation, and the configuration with the higher
number of patches is more detrimental (as expected).

This implies that the relative importance of the mechanisms affecting the persistence of
populations strongly depends on the behaviour of the animals at roads, which, in turn,
depends on the road characteristics such as road width, traffic volume, and the surrounding
landscape. Road network configuration

The results from this model are an important step towards developing a network theory for
road ecology and towards the design of less detrimental road networks. Empirical studies
comparing landscapes with differing road network configurations (while total road length is
constant) should be conducted to validate the model results and provide a basis for
developing more practical models for use in planning and designing of highway networks.

References
Forman, R.T.T.; Sperling, D.; Bissonette, J.A.; Clevenger, A.P.; Cutshall, C.D.; Dale, V.H.;
Fahrig, L.; France, R.; Goldman, C.R.; Heanue, K.; Jones, J.A.; Swanson, F.J.; Turrentine, T.
& Winter, T.C. (2003) Road Ecology. Science and Solutions, Island Press, Washington.
Jaeger, J. & Holderegger, R. (2005) Schwellenwerte der Landschaftszerschneidung. GAIA 14(2):
113-118.
Jaeger, J.A.G.; Bowman, J.; Brennan, J.; Fahrig, L.; Bert, D.; Bouchard, J.; Charbonneau, N.;
Frank, K.; Gruber, B. & Tluk von Toschanowitz, K. (2005) Predicting when animal populations
are at risk from roads: an interactive model of road avoidance behavior. Ecological Modelling 185:
329-348.
Jaeger, J.A.G.; Fahrig, L. & Ewald, K. (2006) Does the configuration of road networks influence the
degree to which roads affect wildlife populations? C.L. Irwin, P. Garrett & K.P. McDermott (Eds).
Proceedings of the 2005 International Conference on Ecology and Transportation (ICOET), Center
for Transportation and the Environment, North Carolina State University, Raleigh, NC, pp. 151-163.

160
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Road effects on wildcats on different spatial scales

N. Klar1,2, M. Herrmann3, S. Kramer-Schadt1

1
Department of Ecological Modelling, Helmholtz Centre for Environmental Research - UFZ,
Permoser Str. 15, 04318 Leipzig, Germany,
e-mail: [email protected]
2
Department of Human Biology and Anthropology, Freie Universität Berlin,
Albrecht-Thaer-Weg 6, 14195 Berlin, Germany
3
OEKO-LOG.COM, Hof 30, 16247 Parlow, Germany

Introduction

Once widely distributed throughout Europe, wildcats (Felis s. silvestris) have suffered
significant reduction in their original range due to extensive hunting and trapping resulting in
fragmented and small populations (Piechocki, 1990). Despite a slow recovery of some
populations they are still a species of conservation concern throughout Europe (Stahl and
Artois, 1995). Their large home ranges and their high mobility make this species highly
vulnerable to traffic mortality in areas with a high human population density like in Central
Europe. This is especially important since knowledge on population dynamics is limited.

Small scale analysis

For the small scale analysis 12 wildcats were radio-tracked from 2001-2004 along a newly
built motorway in Southwest Germany. All road casualties were collected during the study
period on a 18 km long section of the road with different fencing (Table 1).
Between 23 and 35% of the estimated wildcat population living along the motorway was
killed per year in the sections without the complete wildcat proof fences. The wildcat proof
fence was effective. Most individual wildcats did not adjust their home ranges to the major
roads and crossed them regularly where possible .

Table 1: Road casualties of wildcats along the motorways A60 and A1/A48 in 2003 and
2004. Three different fence-types are compared.

Type of fence Length of roadNo of wildcatWildcat casualties

section (km) casualties / km / year
Ordinary wildlife fence 7.5 7 0.41
Wildcat fence without anti-overclimbing 4.0 4 0.44
Complete wildcat proof fence 6.4 1 0.07

Large scale analysis

On the larger scale a dataset of about 2400 wildcat sightings and 350 casualties from 30
years and the whole federal state Rhineland-Palatinate (RLP), Germany was used (Knapp et
al., 2000). Habitat models on different spatial scales were built using logistic regression.
Sink- and source-like areas were defined from the casualties to sightings ratio and proofs of
reproduction.
Wildcats occupy areas characterised by a high amount of forest and a low density of
roads and human settlements. Even so, more than 30% of the potential 2000 wildcat home
ranges in RLP include major roads like highways and motorways. 45% of the sinks were
found in areas with a road density of more than 1.1km/km², whereas only 25% of the sources
were found in these areas. The amount of road casualties per kilometre increased with traffic
amount (Figure 1).

161
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

3.0
Wildcat road casualties per 100 km

2.5

2.0

1.5

1.0

0.5

0.0
0 5000 10000 15000 20000 25000 30000 35000 40000
Vehicles per day (average of road category)

Figure 1. Number of wildcat casualties on different road categories (county roads, country
roads, national highways and motorways) with different average traffic amount.

Discussion

Since wildcats, unlike other carnivores (Kaczensky et al., 2003), seem not to adjust their
home ranges to major roads, single roads within good habitat can pose a sink-like situation
to the adjacent wildcat population. Especially areas with a high road density put such risks to
the wildcat population. On average 30% of the wildcats in RLP have to cross a major road on
a daily basis. The resulting high mortality rate could be one reason for the extremely slow
expansion of this species after the complete protection 70 years ago. So far the lack of
information on population dynamics in the wild is a major problem in wildcat conservation
and hinders us quantify the risk of road mortality on the population level. Our models can be
used to define conflict areas and target roads where wildcat-proof fences together with
crossing possibilities should be established.

This work was supported by the Dr. Joachim and Hanna Schmidt Stiftung für Umwelt und
Verkehr.

References

Kaczensky, P., Knauer, F., Krze, B., Jonozovic, M., Adamic, M., Gossow, H. (2003) The impact of
high speed, high volume traffic axes on brown bears in Slovenia. Biological conservation 111,
191-204.
Knapp, J., Herrmann, M., Trinzen, M. (2000) Artenschutzprojekt Wildkatze (Felis silvestris silvestris
SCHREBER, 1777) in Rheinland-Pfalz. Oppenheim: Landesamt für Umweltschutz und
Gewerbeaufsicht.
Piechocki, R. (1990) Die Wildkatze. 1st edn., A. Ziemsen, Wittenberg Lutherstadt.
Stahl, P., Artois, M. (1995) Status and conservation of the wildcat (Felis silvestris) in Europeand
around the Mediterranean rim, pp. 76. Strasbourg: Council of Europe.

162
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

A management dilemma – when habitat for threatened species occurs primarily

along and across roads

R. van der Ree1, C. Stewart2, S. Cesarini3, S. McCall2, A. Hahs1.

1
Australian Research Centre for Urban Ecology, Royal Botanic Gardens Melbourne, Care of
School of Botany, University of Melbourne, 3010, Victoria, Australia.
e-mail: [email protected]
2
School of Botany, University of Melbourne, 3010, Victoria, Australia
3
School of Biological Sciences, Monash University, Clayton, Victoria, 3800, Australia

Introduction
The agricultural landscape of much of south-eastern Australia is characterised by large
expanses of cleared farmland, interspersed by a few “blocks” of wooded habitat and
numerous linear strips of habitat along roads and watercourses. Dominated by open
Eucalyptus woodland, these linear strips of vegetation (in many respects analogous to
hedgerows and fencerows) represent more than 85% of the available wooded habitat within
some regions. For some threatened species, these linear strips are important for their
conservation because they comprise a large proportion of the best quality habitat available.
Indeed, many recovery plans have highlighted the role and importance of these linear strips
and promote their inclusion in the conservation network. However, a potential problem
arises when these strips are adjacent to roads and traffic (and cause mortality) or when
roads act as barriers that may limit their function as ecological corridors (e.g. van der Ree
2006).

In this paper we have evaluated the importance of the linear habitats for the conservation of
the Squirrel Glider, a small arboreal marsupial that is threatened with extinction. We then
incorporate the additional effect of roads on the viability of local populations. The overall
objective is to synthesise data from a number of different studies and evaluate the trade-off
between the habitat function of the linear strips of vegetation along roadsides and the
potential negative effects of increased mortality and fragmentation.

Methods and Results

This presentation is the combination of a number of ongoing studies which in broad terms
aim to identify and quantify the role and function of roads in the landscape, including their
habitat function, barrier effect and cause of human-induced mortality. We have combined
radiotracking, repeat trapping, GIS analyses and population viability modelling to investigate
this issue.

Habitat
The role of woodland adjacent to roads as habitat for the Squirrel Glider and other wildlife
has been demonstrated by extensive trapping and radiotracking previously (e.g. van der
Ree, 2002, van der Ree and Bennett, 2003). GIS analyses confirmed that a significant
proportion of the best quality habitat for the Squirrel Glider in south-east Australia occurs as
linear strips of woodland along roadsides, unused road reserves and watercourses.

Human induced mortality

In late 2005 trapping for Squirrel Gliders was conducted at 9 sites along minor country roads
and 12 sites along a major freeway (Hume Freeway with 10,000 vehicles per day, connecting
the two largest cities in Australia, namely Melbourne and Sydney), and a subset of 7 of the
sites with the highest capture rates were retrapped in late 2006. This allowed us to estimate
the difference in annual turnover of individuals at sites adjacent to a busy freeway and along
quiet country roads (typically unsealed, with approximately 100 vehicles per day). Annual

163
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

survival (calculated using recapture histories) along the Hume Freeway was approximately
30%, compared to approximately 50 – 60 % along the minor country roads. Previous
capture data and allometric regressions suggest that the likely survival rate for Squirrel
Gliders in continuous habitat would be about 50%.

Roads as barriers
Fifty six Squirrel Gliders were fitted with radiocollars and radiotracked for up to 6 months at
sites along the Hume Freeway and minor country roads. Squirrel Gliders regularly crossed
the minor roads, and only crossed the major freeway when mature trees were present within
the centre median. Analyses of gene flow across the freeway are currently underway.

Population viability
Modelling of the viability of Squirrel Glider populations is currently underway. Preliminary
results indicate that the increased annual mortality as a result of the Freeway is likely to pose
a significant threat to local populations of Squirrel Gliders. However, the extensive network
of linear strips adjacent to the highway appears to act as a source population, providing a
supply of individuals to fill what may be acting as a sink. Further modelling and field
research is required to clarify this.

Discussion
This study confirms that roadside vegetation has an important ecological and biological
function. These narrow strips of vegetation support populations of many species, including
those of conservation concern. We also show that different parts of the landscape contribute
differently to the functioning of populations. Annual survival of Squirrel Gliders was
significantly lower when populations occurred adjacent to major roads, when compared with
populations along minor unsealed roads. The identification of this potential “source-sink”
arrangement is significant because it may cause a rethink in the management of roadside
vegetation along major roads. The habitat adjacent to the major roads is high quality, and
thus supports (attracts) high numbers of individuals. However, the increased turnover
(presumably from increased mortality due to collision with vehicles) may negate its habitat
function. Further research is required to clarify the source-sink processes, and its
significance for roadside management. The regular crossing (by gliding) of the major and
minor roads by Squirrel Gliders demonstrates that they are willing to attempt the crossing.
Wide roads that cause treeless gaps of approximately 60 – 80 m or more appear to present a
barrier that most gliders are unwilling to cross. In contrast, if trees are provided in the centre
median, then regular crossing resumes. However, this provision of trees (and habitat) in the
median also increases the risk of collision of gliders with motor vehicles, because more
animals are residing close to the road.

References
van der Ree, R. (2002) The population ecology of the Squirrel Glider Petaurus norfolcensis,
within a network of remnant linear habitats. Wildlife Research. 29, 329-340.
van der Ree, R. (2006) Road upgrade in Victoria a filter to the movement of the endangered
Squirrel Glider Petaurus norfolcensis: Results of a pilot study. Ecological Management
and Restoration. 7, 226-228.
van der Ree, R. and Bennett, A. F. (2003) Home range of the Squirrel Glider Petaurus
norfolcensis in a network of linear habitats. Journal of Zoology (London). 259, 327-336.
van der Ree, R., Bennett, A. F. and Gilmore, D. C. (2003) Gap-crossing by gliding
marsupials: thresholds for use of isolated woodland patches in an agricultural landscape.
Biological Conservation. 115, 241-249.

164
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Impacts of roads and development on functional landscape connectivity for

wildlife in eastern Collier County, Florida, USA

D. J. Smith1, R. F. Noss2, M. B. Main3

1
Western Transportation Institute, Montana State University, Bozeman, MT USA 59717
e-mail: [email protected]
2
Department of Biology, University of Central Florida, Orlando, FL USA
3
Department of Wildlife Ecology and Conservation, University of Florida, Gainesville, FL USA

Introduction

Roads are one of the greatest threats to wildlife worldwide. Especially in areas with high
traffic volume, strategically placed wildlife crossing structures can assist wildlife to
successfully cross highways and maintain connectivity and gene flow within and among
populations.
South Florida has experienced explosive growth over the last 25 years with conversion of
agriculture and wildlands to residential and urban developments. Population increase over
the next three decades is projected to average 29%. Consequences that accompany such
growth include construction of new roads and widening of existing alignments, increases in
traffic volume, increases in invasive species, rapid levels of habitat loss and fragmentation,
and loss of native biodiversity. Evidence of these impending effects is demonstrated by the
increase in road mortality of Florida panthers Puma concolor coryi and black bears Ursus
americanus floridanus on rural roads in Collier County. Since 2000, the road mortality rate of
Florida panthers on rural roads in Collier County has quadrupled relative to previous
decades.
In 2002 a Rural Land Stewardship Area (RLSA) was designated in the northeastern portion
of Collier County, which contains approximately 80,743 ha of wilderness and rural
agricultural lands. The plan is to preserve 36,344 ha of environmentally sensitive land and
maintain approximately 30,279 ha of agricultural land over the next 25 years. Two primary
corridors were designated as Habitat Stewardship Areas and include restrictions on
development and land use. Concessions given to landowners in exchange for these
designations include allowances for increased development densities outside the
stewardship areas. The designated habitat corridors were identified by the U.S. Fish and
Wildlife Service as key landscape linkages for conservation of the endangered Florida
panther. Based on results of our preliminary study, we discuss the efficacy of the plan and
potential threats to functional habitat connectivity for the Florida panther and other species by
recently proposed developments. We also make recommendations for corridor
improvements and road crossings to ameliorate these impacts.

Methods

We used several methods to determine candidate sites for wildlife crossings by monitoring
and analyzing wildlife movement patterns along highway corridors (SR 29, CR 846 and CR
858) adjacent to designated stewardship areas. These included roadkill and track surveys,
and infra-red camera stations at selected sites. We also synthesized existing data from radio
telemetry, roadkill reports, and other studies, and conducted a landscape analysis using GIS.

Results and Discussion

For all road sections, we recorded 67 different species (from Dec 2005 through Aug 2006)
from roadkills, tracks, and cameras, categorized by faunal groups that included American
alligator, birds, carnivores, ungulates, domestic animals, meso-mammals, small mammals,
frogs, snakes, turtles, and river otter. A total of 136 tracks and 73 photos (focal species only:

165
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

bobcat, coyote, deer, panther, turkey, and wild pig) and 333 roadkills (all species) were
recorded.
We recorded no new roadkills of panthers in our study; yet, many panthers have been killed
on these roads in prior years. Nine panther roadkills occurred along CR 846 East between
1993 and 2006. In our study, Florida panther tracks were recorded on this same stretch on
five separate dates in April and May 2006; these tracks occurred in the same road segments
as previously recorded panther roadkills. Based on roadkill, track, telemetry, and landscape
information, six significant crossing areas (landscape linkages) exist along CR 846 East. The
most critical for Florida panther is the Okaloacoochee Slough and adjacent upland buffers.
Panther roadkills have also been recorded along CR 846 West on either side of the Camp
Keais Strand (a large cypress swamp). Several black bear roadkills were also recorded from
this general area. Significantly, we obtained photographs of an uncollared panther in this
area.
Significant wildlife crossing areas on CR 858 include the area east of SR 29 in the vicinity of
the Okaloacoochee Slough. The central portion of CR 858 produced two panther roadkills,
and we found a panther track in January 2006 along this stretch. Telemetry data also
indicate regular crossings in much of this area, which requires major restoration (conversion
of citrus groves) to improve functional connectivity. Also important is the western section of
CR 858 along the Camp Keais Strand and adjacent upland buffers. Significant restoration is
required to create upland buffers adjacent to the strand to improve the functionality of this
corridor.
Other hotspots for Florida panther roadkills are along certain segments of SR 29. Four
Florida panther roadkills have been recorded along the north section of SR 29 between 2003
and 2005, and eight panther roadkills were recorded along the south section of SR 29
between 1980 and 2003. Several black bear roadkills were recorded from these stretches of
SR 29. An important wildlife crossing area along the north section of SR 29 would connect
the Florida Panther National Wildlife Refuge to the Okalaocoochee Slough corridor.
Significant restoration is needed in both areas to restore functional connectivity for wildlife.
Based on roadkill locations and other data analyzed in this study, we identified significant
sections of each roadway that should be considered for retrofits. To improve habitat
connectivity within the RLSA, we propose a system of culverts, bridges, and barrier fences to
reduce roadkills and increase the permeability of each road for wildlife. Specific
recommendations are located in the final report at
http://swfrec.ifas.ufl.edu/pubs/pdf/imok_rpts/wildlife_movement.pdf.

Three critical components are necessary to establish a functionally connected reserve

system—core habitat areas of sufficient size, connecting habitat corridors of sufficient width,
and buffers that protect the interior quality of the primary network features (core areas and
linkages). The study area currently contains several large, protected, habitat-areas patchily
connected by extensive wetland corridors (Camp Keais Strand and Okalaocoochee
Slough). Road impacts such as wildlife-vehicle collisions, and habitat fragmentation and
isolation need to be addressed in plans to protect habitat connectivity if populations of wide-
ranging species such as the Florida panther and black bear are to remain viable. We
recommend conceptual design improvements to the RLSA habitat corridors that include a
tiered-buffer design and the addition of a central travel corridor, with particular reference to
the needs of the Florida panther. Implementation of this conceptual design would require
restoration of some agricultural areas to native habitat.

166
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Methodological and epistemological constraints on the estimation of the effects

of roads on ecosystem function

S. Findlay

Institute of the Environment, University of Ottawa, 555 King Edward, Ottawa, Ontario K1N
6N5, Canada

With the increased interest in the economic valuation of ecological goods and services,
there is a corresponding heightened interest in the ecological functions that sustain them
generally, and, in particular, the impacts of human activities thereon. De Groot et al. (2002)
note that ecological functions are of four general types: regulation functions, production
functions, habitat functions and information functions. At issue is the extent to which the
construction and operation of roads and road networks affects these functions.
Answering this question is problematic on several counts. First, road construction and
operation inevitably results in subsequent modifications to the landscape: forest removal,
residential or commercial development, agricultural development, etc., all of which are
expected to have independent effects on ecological functions. Thus disentangling the effect
of roads and road networks from the subsequent effects of other human activities presents a
formidable experimental design challenge.
Second, past road studies have been almost invariably been concerned with the
structural attributes of communities or ecosystems (e.g. population abundance, diversity,
etc.). Yet our interest is in ecosystem function, and there is ample evidence that, at least for
some important ecological functions, the structure-function link is highly reticulate.
Consequently, predicting the impacts of roads and road networks on ecological function
from their observed effects on structure will be problematic except in those (probably
comparatively rare) cases where the structure-function map is well-characterized,
Third, there is the question of scale and, in particular, the potential divergence of scale of
application versus scale of effect. The effect of a road on community or ecosystem structure
need not be restricted to locations in the immediate vicinity of a road: if, for example, a road
imposes a significant barrier to wildlife movement, the spatial effect may be felt at local to
regional scales, especially for populations with metapopulation structure. But for structural
attributes, if the (statistical) relationships between local and regional scales is known, then –
at least in principle – studies that evaluate local impacts can be used to estimate impacts at
larger scales. For ecosystem function, however, this is much more problematic because
spatial scaling relationships – if indeed they exists - are, in general, much more poorly
described. This despite the fact that certainly there exist ecological functions (e.g. water
regulation) for which local disturbance arising from road construction can have both direct
and indirect impacts over a range of spatial scales.
From these and other considerations, I contend, first, that for a given experimental
design, a priori inferential strength (see, for example, Roedenbeck et al. 2007) will inevitably
be greater for the estimation of road effects on ecosystem structure compared to ecosystem
function, and in some cases, considerably greater. Given that virtually all study designs used
to date in road ecology have rather low a priori inferential strength (Roedenback et al.,2007),
this implies that even in a Panglossian (experimental) world, the inference that roads have
caused (or will lead to) a specified effect on ecosystem function is/will be rather tenuous.
This in turn, begs the question of whether such studies are even worth doing in the first
place.
Should we then abandon the attempt to estimate the effect of roads on ecosystem
function? Not yet. But I contend that the universe of candidate studies worth prosecuting is
severely circumscribed along at least two dimensions. The first relates to the nature of the
functions themselves: (1) selected functions are those for which there exist strong, well
quantified empirical relationships with coarse resolution ecosystem structural attributes (e.g.
above ground biomass versus, say, the species-abundance distribution of trees); (2) these
structure-function mappings also have known spatiotemporal scaling relationships or, even

167
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

better, are scale-invariant. These two conditions will allow for the strongest inference from
the estimated impacts of roads on ecosystem structure at a given scale, to predicted impacts
on ecosystem function at the same – or possibly different – scales. Note that neither of
these conditions has anything to do with roads per se: indeed, they apply to any attempt to
estimate the effect of any stressor on ecosystem function.
The second dimension pertains to experimental methodology. Specifically, because of
the (likely) highly reticulate nature of ecosystem functional response to stress, road studies
should explicitly adopt a Bayesian experimental approach (Berry 2006; Stephens et al.
2006;); in which a set of ecosystem functions which differ dramatically in the a priori
likelihood of being affected by roads, are simultaneously considered. The pattern of
observed differences between “control” and “experimental” sites over the set of different
functional endpoints in relation to the specified prior then provides for a much stronger
inference about the impact of roads on ecosystem function because other factors which
might give rise to artifactual differences between control and experimental sites for a given
function are more effectively controlled through the between-function comparison.

References
Berry, D.A. 2006. Bayesian clinical trials. Nature Reviews Drug Discovery 5: 27-36.
De Groot, R.S., Wilson, M.A and R.M.J. Baumans 2002. A typology for the classification,
description and valuation of ecosystem functions, goods and services. Ecological Economics 41:
383-408.
Roedenbeck, I.A., Fahrig, L., Findlay, C.S., Houlahan, J., Jaeger, J.A.G., Klar, N., Kramer-
Schadt, S., van der Grift, E.A. (2007): The Rauischholzhausen-Agenda for Road Ecology. -
Ecology and Society 12(1): 11. [online] URL: http://www.ecologyandsociety.org/vol12/iss1/art11/
Stevens, P.A., Buskirk, S.W and C. Martinez del Rio 2006. Inference in ecology and evolution.
Trends in Ecology and Evolution. (2006), doi:10.1016/j.tree.2006.12.003

168
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Multi-scale analysis of wildlife crossing structures effectiveness in Spain

C. Mata, I. Hervás, J. Herranz, J.E. Malo, F. Suárez.

Departamento de Ecología, Facultad de Ciencias, Universidad Autónoma de Madrid.

Spain
e-mail: [email protected]

Several studies have dealt with the use by vertebrates and/or failure of wildlife crossing
structures (see review in Forman et al 2003). However, few attempts have been made to
analyze the question with a wider scope comprising the whole tiered spatio-temporal scales
embracing from planning (and its Strategic Environmental Assessment, SEA) to the
monitoring of corrective measures, and including the project design phase (and its
Environmental Impact Assessment, EIA). Here we present a multi-scale evaluation of the
question based on experiences developed by our research group, beginning with the fine
scale monitoring of corrective measures for fauna and attempting to foresee up to the effects
of strategic plans for infrastructure development along the next decade. It s important to note
that only a correct development and coupling of the different levels can lead to an effective
long-term conservation of natural systems correctly interconnected at a regional scale.
Our first approach is based on intensive monitoring the vertebrate use of 113 crossing
structures of the A-52 motorway (Zamora, North-western Spain) during 2001-2003. From this
work we can conclude that i) a large set of terrestrial vertebrate species use crossing
structures, with some selection patterns based mainly on structural features of passages.
Such patterns seem constant throughout the year while the intensity of use varies
seasonally. As a consequence, ii) different crossing structure types play complementary
roles in reducing the barrier effect of linear infrastructures. Thus, functional structures as
culverts, bridges and underpasses do increase road permeability, at least in locations with
low human use. In parallel, structures specifically designed for wildlife play a key role,
especially so for the most reluctant species such as ungulates (Mata et al., 2005; in prep).
Therefore, the implementation of an integrated set of crossing structures in an infrastructure
can be an effective way to maintain connectivity between populations splitted by it.
Widening the scope of our research, we have tested the possibility to predict vertebrate
use of crossing structures on an independent motorway 70-300 km away from first one (Mata
et al., 2006). Both areas share 80% of the vertebrate fauna and the relative frequency of use
by each species of the 5 structure types (including wildlife- and non-wildlife engineered
passages) was highly predictable (ANCOVA test, p<0.001) and unbiased among species
(p=0.752). Therefore, we can conclude that selection patterns shown by vertebrates can be
generalized and there is a real possibility to predict from the local fauna the type of crossing
structures needed by a new infrastructure during its design stage.
From this scale, we go up to the regional level by an analysis of present permeability of
linear infrastructures through an analysis of high capacity highways and railways whose
projects where passed after 1986. By these means, we focus only on infrastructures
subjected to the EU Directive on EIA (85/337/CEE) that include in their design mitigation
measures for fauna. So, 200 crossing structures (wildlife specific ones and adapted culverts)
from projects scattered through the Spanish territory were visited and checked for their basic
features of location, building and maintenance. Several mismatches were detected like the
fact that 32% structures defined in projects as wildlife specific were in fact passes of mixed
human-fauna use and scarce attention to adaptation for vertebrates. Regarding dimensions,
only 40.6% wildlife underpasses and 23.1% adapted culverts meet the minimum
requirements established for ungulates (Reed et al., 1975; Olbrich, 1984; Iuell et al. 2005).
This is rather noteworthy as wild boar (Sus scrofa) and roe deer (Capreolus capreolus) have
wide distributions and represent a serious threat to road/railway safety. Another key point is a
deficient revegetation of wildlife passes (both across and around them): only 25% wildlife
overpasses and 3.4% wildlife underpasses had been planted with vegetation. Revegetation

169
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

was almost non-existant in other pass types (over 90% without it) and only 11% adapted
culverts had prefabricated concrete walkways above the water level.
As a conclusion, projects under scrutiny showed a trend towards an increased attention to
environmental integration of infrastructures, though some recent projects were also found to
pay little care to fauna. Moreover, the implementation to the field of corrective measures
planned in the design phase shows several deficiencies that should be attended in order to
properly minimize the barrier effect on fauna.
Finally, under this framework of mitigation measures in Spain, it is necessary to assess
the main points of confrontation among the Spanish Strategic Plan for Infrastructures and
Transport ('PEIT 2005-2020') within conservation strategies. On the one hand, Spain faces
an expected exponential development of the transport network, with 6,000 km of new
motorways and 7,000 km of new high speed railways to be constructed. On the other hand,
the country has a social and political commitment with the European Nature conservation
policies as the EU action plan to Halting Biodiversity Loss by 2010 and Beyond. As a
pinpoint, it has been estimated that at least 180 Natura-2000 sites will be crossed by more
than 550 km of high capacity routes. Just this feature helps to clarify the need for a proper
evaluation and a coherent policy to balance the confronting interests of transportation and
nature conservation. However, the SEA carried out to the PEIT does not analyze in depth but
postpones the solution to this situation for later stages of planning and infrastructure design.
As a conclusion, we can state that basic knowledge for the mitigation of the barrier effect
of linear infrastructures on fauna exists and it is firmly established at the EIA stage. However,
the whole tiered process from planning to the implementation of new infrastructures shows
severe deficiencies that mainly affect the two extremes of the process: i) the strategic
planning for an environment-friendly development of infrastructures, and ii) the correct
development of projects during construction and a proper monitoring and maintenance of
mitigation measures.

References
Forman, R.; Sperling, D.; Bissonette, J.A.; Clevenger, A.P.; Cutshall, C.D.; Dale, V.H.; Fahrig, L.;
France, R.; Goldman, C.R.; Heanue, K.; Jones, J.A.; Swanson, F.J.; Turrentine, T. y Winter
T.C. (2003). Road Ecology. Science and Solutions. Island Press, Washington, DC. 479 pp.
Iuell, B.; Bekker, G.J.; Cuperus, R.; Dufek, J.; Fry, G.; Hicks, C.; Hlavác, V.B.; Rosell, C.;
Sangwine, T.; Torslov, N.. (2003). Wildlife and Traffic: A European Handbook for Identying
Conflicts and Designing Solutions. KNNV Publishers.
Mata, C.; Hervás, I.; Herranz, J.; Suárez, F. & Malo, J.E. (2005). Complementary use by vertebrates
of crossing structures along a fenced Spanish motorway. Biological Conservation 124: 397-405.
Mata, C.; Hervás, I.; Herranz, J.; Malo, J.E. & Suárez, F. (2006). A step forward in mitigation of
fragmentation by highways: predictability of terrestrial vertebrate use of crossing structures.
Conservation Without Borders. 20th. Annual Meeting. Society for Conservation Biology. San Jose,
CA, USA.
Mata, C.; Hervás, I.; Herranz, J.; Suárez, F. & Malo, J.E. (in prep.) Are motorway wildlife passages
worth building? Terrestrial vertebrate use of crossing structures and the implications for barrier
effect mitigation.
Olbrich, P. (1984). Study of the Effectiveness of Game Warning Reflectors and the suitability of Game
Passages. Zeitschrift Fur Jagdwissenschaft 30: 101-116.
Reed, D.F.; Woodward, T.N. & Pojar, T.M. (1975). Behavioral response of mule deer to a highway
underpass. Journal of Wildlife Management 39: 361-367.

170
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Road effect on fragmentation of forests and agricultural landscapes: examples

from Italy

E. Padoa-Schioppa, M.C. Poggesi, L. Bottoni

University of Milano-Bicocca, Department of Landscape and Environmental Science, –

piazza della Scienza 1, 20126 Milano, ITALY.
e-mail: [email protected]

Introduction

Fragmentation is a process that may be defined as “the breaking up of a habitat,

ecosystem or a land use type into smaller parcels” (Forman, 1995). It is widely recognized
that fragmentation is one of the major causes of the loss of biodiversity (Wilson, 2004).
Roads may be barriers and for some species roads verges may act as ecological corridors
but the major effect of roads is habitat fragmentation. Far from being just paved ribbons, with
a variable width, built to connect two or more areas, roads may affect not only their
surroundings but also the “horizontal” processes (water flow, fire spreading, plant dispersal
or animal movements) that dynamically contribute to landscape evolution. (Forman &
Alexander,1998). Road fragmentation causes biodiversity loss at the landscape level. The
effects of roads on animals are registered not only at the individual (traffic casualties) level
but, through the alteration of metapopulation dynamics and land transformation, they also
affect population persistence (Forman et al., 2003).
In Italy were road density and car density reach the highest levels of European Union
(Eurostat, 2005) may be useful evaluate the contribute of roads on fragmentation processes.

Materials and methods

Many fragmentation metrics have been proposed by landscape ecologists. We chose to

work on three metrics (division, splitting index, effective mesh size and the corresponding
auxiliary quantities, coherence, splitting density, net product) based on the probability that
two animals randomly placed in different areas somewhere in the region of investigation can
find each other (Jaeger 2000).
The use of mesh size and other related metrics is widespread in Germany (Jaeger et al.,
2001; Von Roedenbeck et al. 2005) and in other country (i.e. Switzerland and Canada): the
application of this metric in Italy too may appear useful.
We analyzed three regions (Lombardy, Sardinia and Piedmont). We divided the regions in
ecological units based on river basins (Padoa-Schioppa et al. 2006). For each ecological unit
we evaluated these fragmentation metrics in forest (in Lombardy also we carried out the
same analysis for agricultural landscapes) and then we elaborated a ratio between non-road
and road data sets in order to give a quantitative evaluation of the role of road network in
landscape fragmentation process.

Results and discussion

Results (see an example in figure 1) show that roads may be the most important cause of
fragmentation in many areas: in some ecological units the mesh size changes by 95% when
we add the road network. Were data on wildlife are available we were able to evaluate
effects on wildlife. As an example, in Lombardy, we observed “winners” and “losers” both for
forest and agricultural landscapes. Among winners there are Hooded crow (Corvus corone
cornix) in agricultural landscapes and starling (Sturnus vulgaris) in forest landscapes. Both
the species are more abundant in more fragmented areas. Loser birds are species that are
more abundant in less fragmented forest landscapes (i.e. the Buzzard, Buteo buteo but also
wren, Troglodytes troglodytes, and robin, Erithachus rubecola) or in less fragmented

171
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

agricultural landscapes (i.e. the Lesser Gray Shrike, Lanius collurio and a declining species
as sparrow, Passer domesticus).

Figure 1. Percentage of forest fragmentation caused by roads in Lombardy

From this analysis river basins appear a good ecological unit and fragmentation of
forested and cultural landscapes may be considered an element that contributes to the
overall quality of river basins, and may be measured easily by means of mesh size and other
related metrics. This approach allows us to identify the main areas where road construction
should be avoided. Furthermore we\are allowed to identify ecological units that are most
severely affected by road fragmentation and where therefore road construction is more
problematic or mitigation measures are needed.

References
EUROSTAT(2005) Europe in figures. Eurostat Yearbook 2005.
Forman, R.T.T., Alexander, L.E. (1998) Roads and their ecological effect. Annual Review of Ecology
and Systematics 29: 207-231
Forman, R.T.T., (1995) Land Mosaics Cambridge University Press, Cambridge (UK).
Forman, R.T.T., et al. (2003) Road ecology Island press Press, Washington.
Jaeger, J. (2000) Landscape division, splitting index, and effective mesh size: new measures of
landscape fragmentation. Landscape Ecology 15 (2): 115-130.
Jaeger J., Esswein H., Schwarz-von Raumer H.-G., Müller M. (2001) Landschaftszerschneidung in
Baden-Württemberg - Ergebnisse einer landesweiten räumlich differenzierten quantitativen
Zustandsanalyse). - Naturschutz und Landschaftsplanung 33(10): 305-317
Padoa-Schioppa, E. Poggesi, C., Bottoni L. (2006) River basins as ecological unit to evaluate
landscape fragmentation. In Landscape Ecology of Freshwaters
von Roedenbeck I.A, Esswein H, Köhler. W. (2005) Landschaftszerschneidung in Hessen
Naturschutz und Landschaftsplanung 37(10): 293-300
Wilson, E.O. (2004) The future of life. Knops Publishing Group, New York

172
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Landscape variables determining movements and road crossing by ungulates

J.E. Malo, C. Mata, F. Suárez.

TEG, Grupo de Ecología y Conservación de Ecosistemas Terrestres, Department of

Ecology, Universidad Autónoma de Madrid, E-28049 Madrid, Spain.
e-mail: [email protected]

Ungulates are key elements for landscape function in many ecosystems and they conflict
with road/railway function. Therefore, coupling knowledge of patterns of ungulate movement
through landscapes, their use of wildlife- and non wildlife-engineered crossing structures is
necessary. The main determinants of crossing structure design and location are also needed
for correct evaluation of both the extent of the problem and the possibilities for mitigation.
Addressing the coherence of information gathered from such different sources would enable
a) the statement of the possibility for generalization of patterns, b) the detection of their main
inconsistencies and c) the identification of the key points for research at the landscape scale
aimed at minimizing the future impact of roads and railways on large herbivores.
The knowledge about patterns of habitat use and movement through landscapes of
ungulates has broadened along the last decades thanks to advances of telemetry
instrumentation. Habitat selection by most ungulates of concern for road ecology (either due
to collisions or to population fragmentation) has been shown to favour forested areas
intermingled with scrubs, grasslands and/or crops. Within this mixed habitats, habitat
selection at finer scales led to a heavier use of the most productive patches but always in the
vicinity to sites with deeper shrub/tree cover that provides protection (potential or effective)
from predators (Spitz and Janeau, 1995; Tufto et al., 1996; Licoppe, 2006).
Research carried out independently on animal displacement in the wild has shown
biomechanical restrictions imposed by weight playing a key role (Richman and Aitchison,
1981), and that should lead to common trails reflecting minimum cost lines as it has been
shown for human mountain paths (Minetti, 1995). Thus, taking into account their weight,
ungulate trails are expected to form angles smaller than 10-15º with respect to contour lines
except when facing a steep slope.
In short, the studies referred above point to a theoretical location of ungulate trails and
thus the ideal location of wildlife crossing structures (or the fatal location of animal-vehicle
collision sites) in places that meet the following criteria: i) within forested areas and at a finer
scale ii) wherever some habitat heterogeneity exists, specially so iii) in the proximity of forest
edges. Moreover, iv) trails most commonly used should connect points with these
characteristics through lines of minimum displacement costs, in general flat.
Besides, studies on the location of ungulate crossing of roads are useful for contrasting
these patterns. In this sense, it is important to note that animal-vehicle collision sites in
Mediterranean patchy habitats seem to be linked to events of ungulate displacement meeting
a road (Malo et al., 2004) in a road-avoiding behaviour also described for reindeer under an
extremely different situation (Luell and Strand, 2005). The cases when animals move
towards roads in search of salt or other benefit are radically different and will not be a focus
of the present study.
Models developed to analyze landscape features of ungulate-vehicle collision sites (Malo
et al., 2004; Seiler, 2005) support the idea that trails reflect patterns of vegetation and
geographical features as pointed above based on habitat selection and displacement costs.
Thus, collisions usually happen within forested areas in points with some habitat
heterogeneity and where scattered trees or hedges get closer to the road. Moreover, the
presence of over- or under-passes, even if not specifically for fauna, reduces the probability
of collision. In the same vein, the study carried out by Dussault et al. (2006) in Quebec
shows moose-vehicle collisions happening in areas where the road crosses potential
corridors defined as those with slopes less tan 2%. Therefore, this kind of study point to
animal trails meeting the criteria previously exposed based on habitat selection and
biomechanics.

173
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Two elements should be taken into account regarding ungulate use of crossing structures
on motorways. First, some minimum requirements should be met by any structure in order to
be used by any large species as ungulates. In this sense, ungulates are noteworthy by their
use of large structures (over- or under-passes) and those specifically designed for wildlife
(Ng et al., 2004; Mata et al. 2005). Secondly, crossing structure location should be close
enough to natural animal paths in order to be used. Therefore, over-passes are expected to
be the most appropriate solution when forested areas and their edges appear on hillock tops,
as in this case passes (especially false tunnel-type ecoducts) can become a natural section
of trails. By comparison, under-passes can be expected to be the best solution in a wider set
of situations as their position in valley bottoms more easily coincides with sites where
animals are unconsciously funneled in search of minimum effort trails. Accordingly, a more
intense use of under-passes by ungulates than over-passes has been stated (Olbrich 1984).
Patterns of coherence among independent information sources have been shown above,
but several discordant observations exist that open new perspectives for applied research.
Thus, ungulates have been shown to use narrow culverts in some occasions (Krawchuck et
al. 2005) and in some cases they make a heavier use of wildlife over-passes than under-
passes (Mata et al., unpublished). What’s the underlying reason for these anomalies?
Though time since construction plays a role in animals getting used to crossing structures
(Olbrich, 1984; Clevenger and Waltho, 2005), even well after the 2-3 year adaptation time
described the structures show an extremely uneven or even infrequent use by ungulates.

References
Clevenger, A.P. & Waltho, N. (2005). Performance indices to identify attributes of highway crossing
structures facilitating movement of large mammals. Biological Conservation 121: 453-464
Dussault, C.; Poulin, M.; Courtois, & Ouellet, J.P. (2006) Temporal and spatial distribution of
moose-vehicle accidents in the Laurentides Wildlife Reserve, Quebec, Canada. Wildlife Biology
12: 415-425.
Iuell, B. & Strand, O. (2005) Monitoring effects of highway traffic on wild reindeer. C. Leroy, P.;
Garret, & K.P. McDermott. (Eds.) On the road to stewardship Proceedings of the International
Conference on Ecology & Transportation (ICOET-05). Center for Transportation and the
Environment, NC State University. Raleigh, NC, USA, pp. 292-300.
Krawchuk, A.; Larsen, K.W.; Weir, R.D. & Davis, H. (2005). Passage through a small drainage
culvert by Mule deer, Odocoilus hemionus, and Other Mammals. Canadian Field-Naturalist 119:
296-298.
Licoppe, A.M. (2006) The diurnal habitat used by red deer (Cervus elaphus L.) in Haute Ardenne.
European Journal of Wildlife Research 52: 164-170.
Malo, J.E.; Suárez, F. & Díez, A. (2004) Can we mitigate animal-vehicle accidents using predictive
models?. Journal of Applied Ecology 41: 701-710.
Mata, C.; Hervás, I.; Herranz, J.; Suárez, F. & Malo, J.E. (2005). Complementary use by vertebrates
of crossing structures along a fenced Spanish motorway. Biological Conservation 124: 397-405.
Minetti, A. (1995) Optimum gradient of mountain paths. Journal of Applied Physiology 79: 1698-1703.
Ng, S.J.; Dole, J.W.; Sauvajot, R.M.; Riley, S.P.D. & Valone, T.J. (2004). Use of Highway
Undercrossings by Wildlife in Southern California. Biological Conservation 115: 499-507.
Olbrich, P. (1984). Study of the Effectiveness of Game Warning Reflectors and the suitability of Game
Passages. Zeitschrift Fur Jagdwissenschaft 30: 101-116.
Reichman, O.J. & Aitchison, S. (1981) Mammal trails on mountain slopes: optimal paths in relation
to slope angle and body weight. The American Naturalist 117: 416-420.
Seiler, A. (2005) Predicting locations of moose-vehicle collisions in Sweden. Journal of Applied
Ecology 42: 371-382.
Spitz, F. & Janeau, G. (1995) Daily selection of habitat in wild boar (Sus scrofa). Journal of Zoology
237: 423-434.
Tufto, J.; Andersen, R. & Linnell, J. (1996) Habitat use and ecological correlates of home range size
in a small cervid: the roe deer. Journal ofAnimal Ecology 65: 715-724.

174
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Reflecting on the Big Picture for Road Ecology, Transportation, and Society

R.T.T. Forman

Harvard University, Graduate School of Design, Cambridge, USA.

e-mail: [email protected]

With roots in many nations, road ecology has just burst onto the scene for the benefit of
transportation and society. Its diverse principles are being put right to work for solutions to
numerous detailed problems at countless locations. The big picture however has yet to reach
our radar screens. Four overriding objectives for transportation and society, for which road
ecology is central, are highlighted. Then two of the many steps required are introduced.

The Four Road-Ecology Objectives for Transportation and Society

1. Improve the natural environment close to the entire road infrastructure.

2. Reestablish a sustained emerald network of large vegetation areas and effective major
wildlife corridors across the landscape through which roads pass.
3. Reestablish near-natural water conditions (e.g., groundwater and surface-water flows,
water-bodies, aquatic ecosystems, and fish populations) across the landscape through
which roads pass.
4. In new road construction, maintain the emerald network and near-natural water
conditions.

The first big-picture objective emphasizes a trajectory of improvement rather than a specific
end product. Location-by-location solutions fit, road-segment-by-road-segment solutions may
often be better, and road-network-by-road-network solutions are probably most efficient and cost
effective. The types of improvement, such as habitat enhancement, vegetated stormwater-
pollutant depressions, and reduction of the highway barrier effect, will vary from location to
location along the road system. This objective can be accomplished with ample flexibility for
transportation agencies, policymakers, and the public.

The second objective highlights the most important solution for sustaining biodiversity. Indeed
the major component of this is identifying, even creating, emeralds on the land undegraded by
roads, traffic, and other human effects. Connecting these large patches decreases the barrier
and habitat fragmentation effects of roads with traffic, which provides important insurance for
biodiversity in the face of an urbanizing and climate-changing world. Road network and
landscape ecological planning is a key to accomplishing this objective.

The third objective for transportation and society highlights water as the other major flow
across the land. Road construction has almost always significantly altered water flows and
water-bodies. Changing the balance to provide both near-natural water flows and traffic flows is
the key step. Much can be accomplished during ongoing maintenance work, in revising roadside
and ditch designs, and in all rehabilitation and upgrading projects involving roadbeds, bridges,
and culverts. Accomplishing the water flows and water-bodies goals is a major step, though
additional steps are required, toward accomplishing the goals of near-natural aquatic
ecosystems and fish populations.

The fourth objective for transportation and society should be the easiest to accomplish. It
accomplishes the second and third objectives without having to do them, a salutary cost-effective
and environmentally sensitive solution.

175
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Simple Spatial Models of Road Networks and Landscape Pattern

Existing model results suggest three principles in the search for an ecologically optimum
road-network form: (1) maintain a few large roadless natural areas; (2) concentrate the bulk of
the traffic on a small number of large roads; and (3) perforate or mitigate roads that separate the
few large natural areas (Forman 2006). The ecological effects of simple road networks with up
to sixteen enclosures were then compared, while varying the number and arrangement of
highways and small roads. A fixed degradation zone (Forman et al. 2002), plus interaction, was
used alongside a highway and not along a small road. Transportation geometry and ecological
patterns used roughly mimic a 10x25 km urbanizing landscape west of the Boston metro-area.
Dependent variables are related to the amount of the degradation zone in a network. Model
results suggest curvilinear patterns for (a) the arrangement of different road types, (b) the
number of undegraded tiny-enclosures in finer-mesh networks, and (c) the blockage effect on
wildlife movement using different permeability assumptions. The goal is to outline better and
worse road networks, based on succinct stated principles, in a form that policymakers and the
public can understand, explain, and use.

Roadside Woody Vegetation

Adding more woody roadside vegetation typically increases wildlife habitat, and also reduces
the barrier and fragmentation effect of busy roads (Forman 2005). The roadkill rate of some
species may increase, yet the increased habitat and increased quality of existing habitat behind
it should result in significant wildlife population increases. Increasing the amount of shrubs and
trees in many (not all) roadsides should help to reduce traffic speeds (driver’s perceived width of
vision ahead is narrower), and make safer highways (fewer, less-severe crashes per kilometer).
Many other environmental and societal benefits, with few disadvantages, can be provided.
Widspread pilot projects and research on greatly increasing roadside woody vegetation of
various types is needed.

References
Forman, R.T.T., Reineking, B. & Hersperger, A.M. (2002) Road traffic and nearby grassland bird
patterns in a suburbanizing landscape. Environmental Management 29: 782-800.
Forman, R.T.T. (2005) Roadside redesigns---woody and variegated---to help sustain nature and people.
Harvard Design Magazine Fall 2005/Winter 2006: 35-41.
Forman, R.T.T. (2006) Good and bad places for roads: effects of varying road and natural pattern on
habitat loss, degradation, and fragmentation. Proceedings of the 2005 International Conference on
Ecology and Transportation. C.L. Irwin, P. Garrett & K.P. McDermott, eds. CTE, North Carolina State
University, Raleigh, USA., pp. 164-174.

176
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Missing knowledge to reconcile growing road networks to viable

(meta)populations

C.F. Jaarsma1, F. van Langevelde 2

1
Land Use Planning Group, Wageningen University, Wageningen, The Netherlands.
e-mail: [email protected]
2
Resource Ecology Group, Wageningen University, Wageningen, The Netherlands.

Introduction

Infrastructure has large impacts on wildlife, as it is one of the principal causes for
mortality, and disturbance and fragmentation of their habitat (Van Langevelde and Jaarsma,
2004). The negative impacts on wildlife will increase due to the expected steady growth of
road networks and their traffic flows worldwide. We expect, however, that in industrialised
countries with a dense population not the network itself will expand (beside relatively small
adaptations), but the traffic volumes will explode. Where motorways now already suffer from
congestion, it is inevitable that especially the networks of lower order roads will take care of
these expanded flows. As a consequence of this shift to lower order roads, fragmentation in
the rural area will increase dramatically, because traffic volumes on these minor roads
largely determine the disturbance, resistance and mortality risk for animals. In human-
dominated landscapes lower order roads have a high density (in the Netherlands 1.55
km/km2, where primary roads including motorways cover 0.32 km/km2), so animals will
frequently encounter them. To mitigate this, another approach is needed than the traditional
one focusing mainly on major roads. This paper explores such an approach, aiming at a
balance between transportation and ecology and defines missing knowledge.

On the way to another approach

Traditionally, mitigating focused mainly on technical devices that change roads, such as
fences, ecoducts and lighting. However, such expensive technical interventions are only
realistic for major roads that have a limited length and no direct access for adjacent land use.
The dense network of lower order roads on the contrary offers frequent access to parcels,
farms and other rural buildings. These accesses generally prohibit an effective fencing with
related bridges, pipes and other facilities to allow wildlife to cross.
Another type of interventions, such as the reduction of traffic volumes (either temporary or
permanent) or vehicle speed, might drastically decrease the barrier and mortality effect of
lower order roads (Jaarsma et al., 2006). These interventions can result in traffic-calmed
areas (TCRAs) and may work for minor roads. The through traffic will use surrounding major
roads. In transportation planning, the mesh size of such TRCAs is considered to be roughly
between 20 and 50 km2. This is based on the time needed to leave the traffic-calmed area
and allows for a major network in grids with mesh width between 4 and 7 km.
The impact of traffic calming is a concentration of former diffuse flows on minor roads at
the surrounding major roads. This will result in somewhat higher volumes on major roads.
However, within the TCRA volumes and speeds will decrease substantially. Ultimately this
drastically increases traffic safety, the main reason for rural traffic calming so far.
However, another impact is a decrease of barrier effects and disturbance along minor roads.
Based on the assumption that roads with an average daily volume (AADT) below 1000
vehicles are not an ecological barrier, rural traffic calming virtually decreases the number of
roads. In Figure 1, the study area (50 km2) has currently 6 continuous landscape units (each
surrounded by roads with volumes > 1000). The smallest one is 0.2 km2, the largest one 19.4
km2. After traffic calming, 3 continuous landscape units remain (0.2, 9.3, and 39.7 km2
respectively).

177
Theme 2 Urban environment and transport
2.1 Symposium 2: Effects of roads and traffic on wildlife populations and landscape functions

Figure 1. The road network in the Ooststellingwerf region, The Netherlands, with a
distinction made between roads with an AADT above or below 1000 vehicles per day and the
resulting continuous landscape units in (A) the present situation and (B) the TCRA-
implemented situation (Jaarsma and Willems, 2002)

Missing knowledge

In our contribution, we explain how a concentration of traffic flows decreases road density.
We argue that planning of TCRAs increases habitat area for wildlife and decreases mortality
and disturbance due to traffic. For effective TCRAs, however, crucial knowledge is still
missing. For example, the choice for a volume of 1000 vehicles per day (as in Figure 1) is
arbitrary. Further research is needed to find threshold traffic volumes where animals can
safely cross a road, i.e. there is a sufficient low probability of being killed. So far, indications
for mesh size of TCRAs to provide sufficient habitat area for (meta)populations of wildlife is
also lacking. We argue that using the method to calculate metapopulation capacity (Hanski
and Ovaskainen, 2000) can be used as a tool to estimate parameters to plan TCRAs that
support viable (meta)populations of certain species.

References
Hanski, I. & Ovaskainen, O. (2000) The metapopulation capacity of a fragmented landscape. Nature
404: 755-758.
Jaarsma, C.F.; Van Langevelde, F. & Botma, H. (2006) Flattened fauna and mitigation: traffic victims
related to road, traffic, vehicle, and species characteristics. Transportation Research D 11: 264-
276.
Jaarsma, C.F. & Willems, G.P.A. (2002) Reducing habitat fragmentation on minor rural roads through
traffic calming. Landscape and Urban Planning 58: 125-135.
Langevelde, F. van & Jaarsma, C.F. (2004) Using traffic flow theory to model traffic mortality in
mammals. Landscape Ecology 19: 895-907.

178
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non
growing and shrinking urban landscapes

Variety matters! The planners’, lawyers’ and the landscape ecologists’ view on
land consumption combined in an interdisciplinary approach

H. Wittmer, H. Nuissl, D. Haase

Helmholtz Centre for Environmental Research UFZ, Permoserstr. 15, D-04318 Leipzig,
Germany, e-mail: [email protected]

Background

The conversion and consumption of natural or agricultural land into urban land poses a
major challenge to sustainable development worldwide. In Europe, although the population is
no longer growing the amount of land developed for urban uses continues to increase
(Figure 1 left). Contrariwise, many inner-city areas are shrinking and suffer from extensive
vacancies in housing and offices. Consequently, the task of reducing the conversion of open
land into urban land fares high on the political agenda at least in densely populated countries
such as Germany, Belgium, the Netherlands, Switzerland and the UK (Figure 1 right).

Figure 1. Left: Land consumption (settlement and transport) compared to population

development in Leipzig, Germany 1993-2005 (1993 = 100). Right: Residential density (=
inhabitants/residential km2) in selected European cities in the mid-1950s and late 1990s
(EEA 2006; MOLAND (JRC) and Kassanko et al. 2006)

Various attempts have been made in most countries at restraining the dynamics of land
consumption. However, whilst most of these strategies and instruments consist in defining
areas that must not be developed, they usually do not allow for the complex interaction of
environmental and socio-economic factors in a sufficient manner. This problem also holds for
“normative” land use strategies that simply aim at reducing the amount of land consumption
to a defined target, such as the German Federal Government’s goal to reduce the level of
land consumption from currently around 100 ha to only 30 ha in 2020. A review of existing
land use evaluation systems reveals that many questions with regard to the consequences of
land consumption are left aside. These questions are e.g.: How detrimental are specific land
use transitions for water balance and habitat integrity? What costs do they involve for
municipal budgets? This shortcoming is mainly a result of a lack of understanding of both
bio-physical and socio-economic mechanisms, concerning different landscape functions.

179
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

Approach and structure of the paper

Against the background outlined above the question arises how science and research
can support the development of more sophisticated strategies and instruments of urban land
use management than are currently available. What we think is needed is a methodology
which allows to evaluate the overall impact of land consumption and, consequently, to
determine which kind of land use is to be prevented at what places. Inspired by the DPSIR-
model we have developed a heuristic conceptual framework on the basis of which we are
now trying to elaborate such methodology (Figure 2). As an integral part of this methodology
we have designed an approach towards an integrated and interdisciplinary assessment of
land consumption and its various impacts. This approach differentiates between three
analytical levels of impacts of land use transition: plot, context and aggregated level. It
combines the (mostly quantitative) evaluation of impacts of land use transition on the
individual plot the use of which has actually been changed with an assessment, based on
bio-physical models and socio-economic estimates, of those impacts that occur beyond this
very plot. Both elements together provide a powerful data set that makes it possible to
disaggregate and downscale (supra-) national statistics so as to clarify the implications of
land consumption estimates. By pulling together the perspectives of different disciplines, the
framework facilitates the assessment and evaluation of impacts of land use transition
systematically. The paper concludes with some illustrative remarks on how the framework
could be used for decision support in land use management and spatial planning.

Figure 2. Conceptual framework for the investigation of land use transformation

References

Haase, D.; Nuissl, H. (2007) Does urban sprawl drive changes in the water balance and policy? The
case of Leipzig (Germany) 1870 – 2003. Landscape and Urban Planning, in print.
Haase, D.; Nuissl, H.; Lanzendorf, M.; Seppelt, R.; Wittmer, H.; Köck, W. (2007) Impact
assessment of land use transition in urban areas – a systematic, interdisciplinary approach from
an environmental perspective. Land Use Policy, in prep.

180
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

Landscape impact assessment for multifunctional land uses

K. Helming1, T. Kuhlman2, D. M. Wascher3, M. Perez-Soba3, S. Sieber1, P.

Tabbush4, O. Dilly5, H. Bach6, K. Tscherning1, B. König1, H. Wiggering1
1
Leibniz-Centre for Agricultural Landscape Research (ZALF), Müncheberg, DE.
e-mail: [email protected],
2
Agricultural Economics Research Institute (LEI), The Hague, NL, 3Alterra. Wageningen, NL,
4
Forestry Commission Research Agency. Alice Holt, UK, 5Chair of Soil Protection and
Recultivation, BTU Cottbus, DE, 6National Environmental Research Institute (NERI), DK

Land use and multifunctionality

Land use is one of the key pressures translating economic, political and other driving
forces into environmental impacts. The motivation of planners and policy makers to interact
with land use is devoted to the principle of sustainable development, which in turn is
considered to be intrinsically linked to the concept of multi-functionality. The rationale
addresses the interdependence of social, economic, and environmental effects of land use,
taking into account commodities and both negative and positive externalities (Wiggering et
al., 2003).

Impact Assessment at European level

In Europe, one of the important driving forces for land use decisions is policy making.
European policies aim at implementing the Sustainable Development Strategy (EC, 2001)
through the decoupling of economic growth from environmental degradation while fostering
social cohesion. To control progress, ex-ante Sustainability Impact Assessment was
introduced as instrument towards the fulfilment of this strategy. It is now a mandatory part of
the policy development process at European level. The “Guidelines for Impact Assessment”
(EC, 2005) of the European Commission require quantitative tools that are built upon
available data and provide indicator based information on the impact of any policy related
land use change on social, economic and/or environmental sustainability issues.

Approach to Impact Assessment for multifunctional land use

This paper reports on methodological approaches and some results of the EU Integrated
Project SENSOR (www.sensor-ip.org), which develops ex-ante Sustainability Impact
Assessment Tools (SIAT) to support decision making on policy options related to
multifunctional land use in European regions. SENSOR is based on 3 assessment streams:
(1) European indicator-based driving force and impact analysis of land use policy scenarios;
(2) Region-specific problem, risk and threshold assessment making use of spatial reference
systems and participatory processes; (3) Case-study-based sensitive area studies using
detailed information on sustainability issues. SENSOR addresses rural land uses for Europe
at NUTS-X level, which is spatially homogenised integration of the administrative
EUROSTAT units NUTS2 and NUTS3. Two additional spatial levels are utilised: (i) 1 km² grid
level for land use projections and environmental indicator analyses, (ii) a clustering of
European regions based on geo-physical and socio-economic variables.
Spatial impacts of six land use related sectors are addressed: agriculture, forestry, nature
conservation, tourism, transport and energy infrastructure. Scenario driven changes of
spatial shares of these sectors are simulated with a series of macro-economic and sectoral
models iterating with the land use model CLUE-S (Verburg et al., 2002). Scenario studies
representing narratives of optional global economic and policy developments for target years
2015 and 2025 are utilised to allow for ex-ante simulation of possible land use changes.

181
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

Land Use

Nature Transport
Agriculture Forestry Conservation Infrastructure Energy Tourism

Land use functions

Production Environmental Economic Social Cultural Ecological

Thresholds Sustainability
Choice Space
Targets

Region x Region y Region ... Region ... Region ... Region n

Stakeholder

Figure 1. SENSOR analyses impacts of spatial changes in six sectors on social,

environmental and economic indicators at regional level for Europe.

Based on an indicator system, land use change scenarios are analysed in their impact on
environmental, social and economic issues. For multifunctionality assessments, indicators
are grouped into a system of nine land use functions, which represent important goods and
services that are provided to society through land use. Simulated changes in land use
functions are used for a stakeholder inclusive approach of valuating impacts at regional and
local level. Accompanied with a framework for expert driven definition of regional limits and
targets, land use impact valuation can then lead to identify so called sustainability choice
spaces of land use changes in specific regions.
All the above described approaches are integrated into a Sustainability Impact
Assessment Tool (SIAT), which is a scenario driven meta-model based on response
functions describing relations between (i) policy options and land use changes and (ii) land
use changes and sustainability indicators. SIAT is a quick scan tool designed for policy
making purposes at European level. Validation of the tool is to be conducted in case study
areas, for which a more profound data and information basis exist, and where stakeholders
are consulted to valuate SIAT simulations. The approach is going to be extended to specific
issues of urban-rural linkages in the frame of a second project called PLUREL.

References
European Commission (2001) A Sustainable Europe for a better world: A European Union Strategy
for Sustainable Development. COM(2001)264final.
European Commission (2005) Impact Assessment Guidelines SEC(2005)791.
Helming K, Tscherning K, Wascher D, Kuhlman T, Sieber S, Bach H, Dilly O, Tabbush P. (2006)
SENSOR annual report, public part. In: Helming K, Wiggering H (Eds). SENSOR report series 1.
www.sensor-ip.eu. ZALF Germany.
Wiggering H, Mueller K, Werner A, Helming K (2003) The concept of multifunctionality in
sustainable land development. In: Helming, K. and H. Wiggering (eds) (2003): Sustainable
Development of Multifunctional Landscapes. Springer Berlin, Heidelberg, New York. p 3 – 18.
Verburg PH, Soepboer W, Veldkamp A, Limpiada R, Espaldon V, Sharifah Mastura SA (2002)
Modeling the Spatial Dynamics of Regional Land Use: the CLUE-S Model. Environmental
Management 30: 391-405.

182
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

Monitoring Urban Sprawl in Germany - Towards a GIS-based

Measurement and Assessment Approach

S. Siedentop

Institute of Regional Development Planning, Universität Stuttgart – Pfaffenwaldring 7, 70569

Stuttgart, Germany. e-mail:
[email protected]

Background

During the 1990’s, the phenomenon of urban sprawl received growing attention in the
international urban policy debate. In a recent report, the European Environmental Agency
addressed sprawl as one of the major challenges facing urban Europe in the next decades
(EEA 2006). However, a survey of the literature yields no agreement in terms of defining and
measuring urban sprawl. Some scholars denote sprawl simply as ongoing urbanisation or
low density development, others stress the level of suburbanization and deconcentration or
characteristics of land use patterns. The absence of a common understanding constrains the
analysis of sprawl’s causes, costs and consequences as well as the formulation of planning
strategies towards economically, ecologically and socially acceptable land use patterns.
Urban and environmental planning aiming to combat sprawl must have an agreed-upon way
to measure it in order to evaluate the progress of planning strategies and programmes
(Siedentop 2004).
Alongside the lacking definition and measuring convention, the specific process character
of urban land use change hinders more effective anti-sprawl-policies. Environmental, social
and economical problems caused by urban sprawl tend to be cumulative in nature. They
build up over a period of time and usually have more than one cause. In contrast, the legal
framework regulating the use of environmental resources generally addresses a single
pollutant or a single project but ignores multiple actions that can add up or interact to cause
cumulative effects. It has to be acknowledged that the mismatch between the scales at which
sprawl-related environmental degradation occurs and the scales at which regulatory
decisions are made is a significant obstacle to a more effective environmental management.
Therefore, the development of a multi-dimensional and multi-temporal measurement
approach is of crucial importance for urban and regional development policies.

Measurement Approach

On this background, a methodological framework for the measurement and assessment

of land use patterns and dynamics is presented. First, the approach should be able to
indicate the multiple driving forces and causative agents of land use change associated with
urban sprawl. Secondly, this framework should be qualified for considering and assessing
different impacts of land use and land use dynamics concerning ecological, social and
economical objectives of urban and regional development policies. It also should recognise
direct, indirect and cumulative cause and effect relationships. A third important requirement
is that the selected indicators should be applicable with available data sources. And last but
not least the approach has to deal with the varying scales on which land use policies are
carried out.
The GIS-based measurement tool is based on the assumption that urban sprawl is a
multidimensional phenomenon which can only be measured with a multiple-indicator
approach. Three general process characteristics of urban sprawl are being discerned: the
conversion from natural to artificial surfaces (surface-related impacts); the change of land
use patterns from a compact urban form to irregular, dispersed land use patterns (pattern-
related impacts) and the reduction of urban density (density-related impacts). Each sprawl-
dimension is related to specific environmental impacts as well as unintended social and

183
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

economic outcomes (Figure 1). These very different kinds of impacts can be addressed in
terms of their specific impact pathway (direct, indirect, cumulative impacts), their spatial and
temporal scale (local, regional and global impacts; short-term and long-term impacts) and
particularly affected environmental components, social groups or economic subsystems.

surface-related impacts

Figure 1. Three dimensions of

Loss/Degradation of
Natural Soils
urban sprawl and impacts
Surface
explained by one or more
Characteristics dimensions (examples)
Consumption of
Mineral Resources

Reduced Quality of
Reduced Quality of Recreation Areas
density-related impacts
Habitat Space
Change of Micro-/
Meso Climate

Land Use Pattern Urban Density

Low Efficiency of
Urban Infrastructures

High Automobile
Dependence
pattern-related impacts

The proposed indicator system contains about 20 core indicators. Some indicators can be
applied to the measurement of the (aggregate) land use pattern. Due to their low level of
elasticity, “pattern-indicators” are suitable for “global” sustainability appraisals of land use in
larger time intervals. Other indicators aim to address land use changes within a specific
period of time on patch level (see Table 1 with examples). The latter are notably qualified for
the continuous controlling of land use development in order to assess the success of
regulatory policies.

Table 2. Categorisation of selected sprawl indicators

Indicator Surface Pattern Density Pattern/Patch

Share of urbanised land (%) x Pattern
Consumption of prime agricultural land for urban uses x Patch
Per-capita availability of open spaces in a given radius X Pattern
Shape-index (of urbanised land patches) X Pattern
Integration of new development into already urbanised areas X Patch
Integration of new development in the public transport network X Patch
Urban density (inhabitants per hectare urbanised land) x Pattern
Mean Density of new development (inhabitants per hectare) x Patch
Per-capita length of network-related infrastructures (meter) x Pattern

References
European Environment Agency (2006) Urban Sprawl in Europe. The ignored challenge. EEA Report
No 10/2006. Copenhagen.
Siedentop, S. (2004) Urban Sprawl – verstehen, messen, steuern. DISP 160: 23-35.

184
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

A green infrastructure strategy to enhance the adaptive capacity of restructuring

city regions to climate change: the case of Greater Manchester

J. Handley1, S. Gill1, R. Ennos 2, S. Pauleit3

1
Centre for Urban and Regional Ecology, School of Environment and Development, The
University of Manchester
2
School of Biological Sciences, The University of Manchester
3
Centre for Forest, Landscape and Planning, The Royal Veterinary and Agricultural
University, Copenhagen, Denmark
e-mail: [email protected]

Climate change is already with us and its impacts are especially evident in towns and cities.
With this in mind the Engineering and Physical Sciences Research Council and the UK
Climate Impacts Programme established a research programme into Building Knowledge for
a Changing Climate (BKCC). One project within this programme is developing ways of
preparing for climate change through strategic planning and urban design. An important facet
of this work was to explore the potential of urban greenspace for climate adaptation. Greater
Manchester, a large polycentric urban region in North-Western Europe, was chosen as the
case study area. It is heavily influenced by structural changes with continuing out-migration,
and restructuring of peri-urban land use and economic functions.

Urban areas have distinctive biophysical features in comparison to surrounding rural areas.
For example, urbanisation replaces evapotranspiring vegetated surfaces with built surfaces
which store heat. This modifies energy exchange and contributes to the urban heat island
effect. The hydrological regime is also altered, with impervious surfaces increasing the
volume and rate of surface water runoff.

Climate change scenarios suggest that the UK will experience warmer wetter winters with
increased precipitation intensity, and hotter drier summers. This will impact on both people
and buildings, particularly within urban areas. Urban greenspace offers significant potential in
adapting cities for climate change through its role in ameliorating the urban climate, by
providing shade and evaporative cooling, and reducing surface runoff, through the
interception, storage and infiltration of rainwater. However, this potential has not been
explored. In addition, little is known about the impact of climate change on urban
greenspace, and how this may impact back on its functionality. This knowledge is critical for
the creation of adaptation strategies through planning, design and management.

Modelling work was undertaken to quantify surface temperatures during heat waves and
surface runoff during rainstorm events, in relation to greenspace cover. This was based upon
an urban characterisation which stratified the conurbation into 29 distinctive urban
morphology types (UMTs) and then estimated the surface cover of each by aerial photograph
interpretation. This data formed a main input into the surface temperature and runoff models,
which were run for the baseline 1961-1990 climate, as well as for UKCIP02 Low and High
emissions scenarios for the 2020s, 2050s, and 2080s time-slices.

Findings show that surface temperature and runoff are very dependent on the proportion of
green cover. This is increasingly important in the future as differences in surface
temperatures and runoff become more pronounced between densely built and well greened
areas. The surface temperature modelling found that in 1961-1990 the maximum surface
temperature of woodlands, the least built up UMT, is 18.4°C (on a day expected twice per
summer); 12.8°C cooler than that of town centres, the most built up UMT, at 31.2°C. By the
2080s the maximum surface temperature increases by 1.5°C-3.2°C in woodlands and 2°C-
4.3°C in town centres, depending on the emissions scenario. The surface runoff modelling
found that, in general, there was more surface runoff from the more built up UMT categories;

185
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

but soil type is also very important. The UMTs display the largest range of runoff coefficients
on high infiltration soils (e.g. sandy soils) and the smallest range on low infiltration soils (e.g.
clay soils). For example, for a one day per winter precipitation event with normal antecedent
moisture conditions in 1961-1990, there is 32% runoff from low density residential areas
compared to 74% from the more built up town centres on a sandy soil; on a clay soil this
changes to 76% and 90% respectively, much higher values and with a smaller difference
between them. Thus, surface sealing has a more significant impact on runoff on soils with
higher infiltration rates.

Following this, a series of ‘development scenario’ model runs were undertaken exploring
both the effects of current development trends as well as the potential of greening in
adapting for climate change. They included: residential and town centres plus or minus 10%
green or tree cover, greening roofs in selected UMTs, high density residential development
on previously developed land, increasing tree cover by 10-60% on previously developed
land, residential development on improved farmland, and permeable paving in selected
UMTs.

The findings suggest that urban greenspace offers significant potential in moderating the
increase in summer temperatures expected with climate change. For example, adding 10%
green cover in high density residential areas and town centres kept maximum surface
temperatures at or below 1961-1990 baseline levels up to, but not including, the 2080s High;
whilst greening roofs in areas with a high proportion of buildings was also an effective
strategy. On the other hand, the modelling work highlights the dangers of removing green
from the conurbation. For example, if green cover in high density residential areas and town
centres is reduced by 10%, surface temperatures will be 7°C or 8.2°C warmer by the 2080s
High in each, when compared to the 1961-1990 baseline case; or 3.3°C and 3.9°C higher
when compared to the 2080s High case where green cover stays the same.

The modelling work suggests that greenspace on its own is less effective at moderating the
volume of surface runoff under climate change. By the 2080s High, the one day per winter
precipitation event has 56% more rain than in 1961-1990, resulting in an 82% increase in
runoff from Greater Manchester. Whilst adding green cover can reduce runoff locally, this
effect is not enough to counter the extra precipitation from climate change. The use of
storage in combination with green surfaces will be required, and could be used to irrigate
greenspace in times of drought. There is significant potential to utilise sustainable urban
drainage (SUDS) techniques.

Another way of exploring possible climatic adaptations considered the green infrastructure of
the conurbation from the perspective of landscape ecology. The green infrastructure is thus
viewed as consisting of corridors, patches, and the overall matrix. These components play
different roles in terms of climatic adaptation. For example, flood storage is very important in
corridors, but has some importance as SUDS in the patches; rainwater infiltration is very
important in the matrix, as well as in patches; whilst evaporative cooling is crucial in the
patches as well as in the matrix. In the UK there is renewed interest in strategic planning at
the city-region level and a City-Region Spatial Strategy is currently being prepared for
Greater Manchester. The potential for strategic planning of the urban green infrastructure to
contribute to this emergent strategy through climate change adaptation will be discussed in
the presentation.

186
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

Urban restructuring in Central Eastern Europe – does the late phase of transition
still affect the environment?

A. Vaishar

Institute of Geonics, Academy of Sciences Ostrava,

Department of Environmental Geography, 602 00 Brno, Drobného 28, Czech Reoublic
e-mail: [email protected]

Introduction
Cities of Central Eastern Europe have been impacted by a very fast advance of
processes that took many decades in Western Europe. The social and economic transition
has been responsible for many changes in urban structures during the last 17 years. The
present state-of-art was investigated by international teams within the two projects:
Mobilizing re-urbanisation on condition of demographic change (5th EU framework
programme No. EVK4-CT-2002-00086) and Spatial consequences of demographic changes
in East Central European cities (Volkswagen foundation No. II/81150).

Present processes of urbanisation in the cities of Central Eastern Europe

Specific development within cities in the mentioned region can be characterised by the rapid
change of economic base from the secondary to the tertiary sector, by the mass use of
passenger cars as well as by the rapid social polarisation of the society. Such a development
has proceeded in relation with objective processes of the present phase of urbanisation:
globalisation, sub-urbanisation, counter urbanisation etc.
At the same time, a second demographic transition has taken place in the countries of
Central Eastern Europe. Fertility rate is high under natural reproduction. Life expectancy is
apparently increasing. Population ageing follows as a logical consequence. Opening of
borders and labour markets within EU encourages international migration. Immigrants often
settle in the former worker districts of big cities, where quarters (sometimes even ghettos) of
socially deprived population groups develop. This development affects the situation of inner
cities.

Empirical research

The cities of Central Eastern Europe included in the study are Brno and Ostrava (Czechia),
Gdańsk and Łódż (Poland), Leipzig (Germany), Ljubljana (Slovenia). Problems found were
identified in Vaishar et al. 2006.
The main source of environment pollution has mostly changed from industrial to car
traffic. It is not only the number of cars but also the way of their use that adds to the traffic.
The role of the car has changed from a source of recreation to the medium for enterprise and
prestige. The traffic has moved from the countryside back to the cities. Pollution, noise,
accidents, traffic jams, and parking problems illustrate the problem. The difficulties are
concentrated in inner cities which were built for substantially lower traffic needs. The extreme
situation can be found in Ljubljana that has become the capital of an independent country.
Conversely, most Central European cities have maintained a relatively efficient system of
public transport.
Withdrawal of heavy industries from inner cities has caused the problem of
brownfields. Socially and environmentally degraded areas impair the value of some parts of
the inner cities, which are otherwise quite attractive with respect to their location. Some
available examples of rehabilitation are e.g. shopping centres Łódż (Manufaktura), Brno
(Vaňkovka) or apartments (Łódż-Scheibler). Many of former industrial areas, landfills and
worker colonies are still waiting for their rehabilitation.

187
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

Urban greenery plays an incrementally important role in improving the urban environment.
Unfortunately, street greenery often gives away to new parking places or to the widening of
roadways.
The process of sub-urbanisation induces many environmental problems such as the low
efficiency of land and energy use and the mass use of cars. On the other hand, inner cities
lose the young, educated and socially stronger population. With the exception of individual
cases of gentrification, difficulties in the rehabilitation and maintenance of older city parts can
be observed.

Evaluation

It is more or less clear, that it is not possible to speak of the environment in general.
Target population groups seem to be one of crucial aspects of the problem. The question is
for whom is the milieu of inner cities attractive and what demands have such groups of
inhabitants on their environment. It shows that the community of inner cities starts to be
formed of young urban specialists (not fixed in their place till now) and together living mostly
childless couples on the one hand, and of old people or socially deprived and often
immigrant groups on the other.
The problem is that the environmental requirements of individual population groups are
different. Young educated people prefer almost contradictory aspects than old, less educated
people. Taking into account the environment as a part of life quality, we come to the
conclusion that the environment has to be measured as a set of objective and subjective
indicators.

Discussion

It follows from the research that the environment does not play a decisive role in the
settlement preferences of inhabitants. Rather, economic and social aspects are predominant.
It seems that not the environmental aspects but the awaked individuality after 40 years of
collectivism plays a decisive role even in the sub-urbanisation process. Environment can be
important as an impulse of secondary importance namely in the cases of serious problems
between special population groups.
In the future, the sustainable development of cities will greatly depend on the efficiency of
energy. From such a viewpoint, public transport should be preferred to individual transport
and the re-urbanisation of inner cities to the continuous process of sub-urbanisation. Under
such a pressure, the perception of big cities as places with extremely polluted environment
could change. On the contrary, inner cities are suitable for re-urbanisation as places of
effective land use and effective energy consumption.

Conclusion

The perception of environment has changed. People are increasingly concerned with the
issue. Unfortunately, the general interest in environmental matters is often not followed by
the discussion of individuals. The “not in my backyard” attitude sometimes complicates the
solution. Various green parties and movements do not help to make the problems found
easier. The unification of European legislation could bring some results in the long-term. A
crucial point appears however to be the need for a change in environmental behaviour.

References
Vaishar et al. (2006): Urban Environment in European Big Cities. Moravian Geographical Reports 14:
(1): 46-62.

188
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

Too much urban green? the challenge of shrinkage in cities for green space,
recreational ecosystem services and public acceptance of related policy
strategies

S. Schetke1, D. Haase², J. Breuste³

1
University of Bonn, Institute for Geoinformation and Geodesy, Devision Urban Planning,
Nussallee 1, D- 53115 Bonn, Germany
e-mail: [email protected]
2
UFZ-Centre for Environmental Research Leipzig-Halle, Department for Computational
Landscape Ecology, Permoserstr. 15, D-04318 Leipzig, Germany
3
University of Salzburg, Department of Geography and Geology, Division Physical and
Environmental Geography, Hellbrunner Strasse 34/III, A-5020 Salzburg, Austria

Background

Compared to the suburban growth at beginning of the 1990’s after the societal transition in
eastern Germany, today urban shrinkage processes due to demographic change, out-
migration and economic decline produce major vacancies in both housing and the
commercial sectors. Core city areas are affected as well as socialist housing estates.
Abandonment is followed by demolition impacting buildings and population density along the
rural-urban gradient. This trend in urban land use development is not only restricted to
eastern Germany; it also increasingly occurs in other city regions of Central European
countries as well as in western and southern European cities.

The challenge

As a consequence of shrinkage and demolition, there is a surplus of urban waste land and
open space which was never expected to become a reality, and furthermore needs to be
“designed” for the future. Here, urban landscape planning enters “undiscovered terrain”
because urban growth has dominated the agendas of urban planners for decades:
So, what are the effects of increasing potentials of urban open space in core and peripheral
districts? Both could be assumed: positive effects in terms of green and recreational quality
and negative effects when looking at emptiness and decreasing accessibility of social and
transport infrastructure of formerly compact urban bodies. Recently, there has been
observed a growth in green spaces accompanied by a decline of social functions within
several districts. There are no concepts on how to manage this development in urban parks,
which involve the following:
• What are the specific spatial pattern of urban shrinkage, decline of population and
related demolition of the urban fabric? How far does the perforation process
influence the social neighbourhoods?
• Do urban planning strategies exist to deal with the growth of green space and, what
is more, how can they be implemented?
• What are the expected ecosystem services of the newly gained urban green
spaces? How could their impact be assessed?

Research design, methods and spatial units of investigation

To find initial answers on the above questions, infrastructural and land use changes related
to vacancy and demolition have been detected. Additionally scenarios for the coming 10
years have been incorporated into the assessment. A multi-criteria indicator matrix
subsequently quantifies socio-spatio-environmental impacts of such land development
(scenarios). Seen from both sides, ecosystems and the urban inhabitant, different scenarios
of urban shrinkage thus could be assessed concerning “quality of life” and ecosystem
resources related to target values.

189
Theme 2. Urban environment and transport
2.2 Symposium 3: Beyond growth? Scientific and policy strategies for non growing and shrinking urban
landscapes

The evaluation of the urban green dynamics extension and ecological functionality is based
on urban structure types (Breuste, 2002). Urban structure types mark areas of physiognomic
homogeneous development, which are predominantly distinguished from each other by
characteristics in built up structures and open spaces (vegetation and soil sealing). These
are to a large extent homogeneous concerning density and portions of the built-up areas of
various forms and of different developments of the open spaces (soil sealing areas,
vegetation types and urban forest). Substantial ecological characteristics of a space can be
described by the land use form and the structural characteristics. Urban structure types
summarize spaces of similar environmental conditions and thus represent units that can be
up-scaled on empirical results on urban land use change.

Case Studies

For the case study city regions in eastern Germany, Leipzig and Halle, based on the land
use history since 1990 greenery and open land scenarios are presented. Further, for the
case of Halle, the current urban planning strategy and its implications had been evaluated.

Preliminary results

The results of the study gave strong evidence that firstly, eastern German cities possess a
“for-runner role” in urban restructuring and renewal processes arising from shrinkage.
Secondly, a surplus of urban greenery has ambivalent effects for both man and nature and
thirdly, that urban planning is not prepared to handle “too much” green (Schetke & Haase,
2007). On the one hand, demolition produces new open space and the potential of natural
plant succession and recreational value. In contrast it ignores the idea of “leaving nature to
itself” the social and “quality-of-life” dimensions of shrinkage in so far, that simply overgrown
areas in an environment of emptiness and decay are not necessarily perceived as attractive
for the urban residents (Rink, 2005).
The multi-criteria indicator matrix explores all these different dimensions of shrinkage
impacts in a quantitative form as well as permiting the transfer of the numbers into verbal
argumentation at both local and structure type level (Schetke & Haase, 2007).
Due to the properties regarding utilization type and building structure there exist direct
relations between the scientific framework of the urban structure types and the instruments of
urban planning such as a master plan, zoning plan or site/property planning. Scientific
findings might be thus easily enter administrative, political and legislation procedures and
documents for actions.
The results of both case studies Leipzig and Halle can be transferred to other urban
shrinkage examples in order to enhance urban planning strategies and policies in terms of
hindering and supporting aspects. This also encloses a better understanding of the
management of urban ecosystems.

References
Breuste, J., 2002. Urban Ecology. In: Bastian, O., U. Steinhardt (eds.). Development and
Perspectives of Landscape Ecology. Kluwer Academic Publishers, Dordrecht. pp. 405 – 414.
Schetke, S., Haase, D., 2007. Multi-criteria assessment of the socio-ecological development in
shrinking cities. Experiences from East Germany. Environmental Impact Assessment Reviews (in
prep.).
Rink, D., 2005. Surrogate Nature or Wilderness? Social Perceptions and Notions of Nature in an
Urban Context, In: Kowarik, I., Körner, S. (eds). Wild Urban Woodlands. New Perspectives for
Urban Forestry, Springer, Berlin et al., pp. 67-80.

190
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

2.3 Symposium 4: Applying landscape ecological principles in urban

environments

The efficacy of the urban-rural approach in studying the ecology of human

settlements

M. J. McDonnell, A. K. Hahs

Australian Research Centre for Urban Ecology, Royal Botanic Gardens Melbourne
c/o School of Botany, University of Melbourne, Victoria 3010, Australia,
e-mail: [email protected]

Introduction

The ever increasing human population is driving the expansion and creation of cities
and towns worldwide. Consequently, there is a tremendous call for more ecological
information in urban and exurban environments by natural resource managers, planners,
conservationists, scientists, and professionals associated with human health. In the early
1990’s, the urban-rural gradient approach was proposed as a useful method for studying the
ecology of cities and towns (McDonnell and Pickett, 1990). Since then, researchers have
characterised urban-rural gradients and have used the concept to better understand the
distribution of plants and animals as well as ecosystem processes in cities and towns around
the globe. Thus, we are at an appropriate stage in the development and use of the approach
to assess what we have learned, and what improvements can be made in the future to
achieve better research and management outcomes.

Urban-rural gradient studies

A review of the literature in May 2006 revealed 220 papers that had explicitly utilised
the urban-rural gradient approach. These studies examined the response of plants, animals,
insects, invertebrates, fungi, fish, reptiles and amphibians to gradients of urbanisation. The
most common organisms studied were birds. In addition, researchers interested in nutrient
cycling, conservation biology, landscape ecology, pollution, marine and freshwater
ecosystems and humans made use of the urban-rural gradient approach. The richness of
these studies, coupled with the quality of their results clearly indicates the value of the urban-
rural gradient approach to the ecological and social study of cities and towns.

Characterisation and quantification of urban-rural gradients

Urban-rural gradients were first characterised along transects using simple, easily
measured variables such as distance to central business district. Although it was a relatively
unsophisticated linear axis, the technique did reveal important trends in the distribution of
organisms and the rate of ecosystem processes suggesting that the concept was robust.
These studies were followed by more sophisticated variables which could be categorised into
three groups: 1) physical and chemical; 2) landscape structure and 3) human population
density. These variables were collected and presented along transects (Medley et al., 1995;
Luck and Wu, 2002), or across an entire city (Hahs and McDonnell, 2006). The
quantification of these variables was important in the development of the urban-rural gradient
approach for it allows for the comparisons between gradients within cities and between cities.
This was a significant step forward for researchers, planners and managers because it
allowed for the amalgamation of ecological and social science information (Theobald, 2004).

Representing urbanisation as a continuous variable also enables more subtle

gradients to be identified than when studies rely upon transects. For example, an index of
urbanisation (Index combined) that includes both demographic and physical data can be

191
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

presented for the entire city on two dimensional grids or as a three dimensional topographic
representation (Fig. 1). Representing urbanisation using three dimensions presents a more
accurate depiction of the non-linear, complex urbanisation gradients that occur in cities and
towns. The most intensely urban areas are represented by the peaks while the less urban
areas form the valleys (Fig. 1b). The depiction of urbanisation gradients in this manner
clearly illustrates the “mountain range” pattern of urbanisation which occurs in most cities.
This level of spatial resolution makes it possible to assess ecological and social responses to
similar levels of urbanisation along the gradients (i.e., peaks with peaks) or between different
levels of urbanisation (eg, peaks and valleys) along the gradients.

Figure 1. Values of Index combined across northern Melbourne, Australia represented as a) a two
dimensional grid, and b) a three dimensional representation of the same data. The lighter areas
represent high values of Index combined (45 – 55; intense urban), and the darker areas represent low
values of Index combined (0 – 5; rural). The direction of magnetic North ( ), and the location of the
Central Business District ( ) are marked.

The large body of work that has been conducted using the urban-rural gradient
approach has provided several insights into how urbanisation affects the ecology in and of
cities. Our recent work in Australia has also highlighted some areas where the theory and
techniques can be refined. First, it is important to explicitly define the urban-rural gradient
using quantitative measures, as this allows the study to be compared with equivalent
gradient studies, as well as ensuring the status of the study site is captured for the period of
the study (Hahs and McDonnell, 2006). Second, we need to move beyond viewing the
urban-rural gradient as a linear transect, and begin to represent gradients at the level of the
city, as this will allow us to take a more sophisticated approach to defining the location of our
study site within the landscape. Third, we need to improve our understanding of the
measures used to define the gradient, as well as the measures used for the response
variable, as the selection of specific measures can influence the findings of the study.

References

Hahs, A.K. & McDonnell, M.J. (2006) Selecting independent measures to quantify Melbourne’s
urban-rural gradient. Landscape and Urban Planning 78:435-448.
Luck, M. & Wu, J. (2002) A gradient analysis of urban landscape pattern: a case study from the
Phoenix metropolitan region, Arizona, USA. Landscape Ecology 17: 327-33.
McDonnell, M.J. & Pickett, S.T.A. (1990) Ecosystem structure and function along urban-rural
gradients: An unexploited opportunity for ecology. Ecology 71: 1232-1237.
Medley, K.E., McDonnell, M.J. & Pickett, S.T.A. (1995) Forest-landscape structure along an urban-
to-rural gradient. Professional Geographer 47: 159-168.
Theobald, D.M. (2004) Placing exurban land-use change in a human modification framework.
Frontiers in Ecology 2: 139-144.

192
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Beetle assemblages along urban to rural gradients: an international comparison

J. Niemelä

Department of Ecology and Systematics – PO Box 65, University of Helsinki, Finland.

e-mail: [email protected]

Introduction

Effects of urbanisation can be examined by using urban-rural gradients from densely built
city cores to rural surroundings (McDonnell & Pickett 1990). Here, I present results of an
international comparison – using the gradient approach – examining how generalisable the
ecological effects of urbanisation are around the world. I illustrate how comparative, research
can contribute to our understanding of urban ecology and to the development of theory.
The background of the GLOBENET project (http://www.helsinki.fi/science/globenet) is that
urbanisation creates modified landscapes that exhibit similar patterns around the world.
Nonetheless, little is known on whether or not these changes affect biodiversity in similar
ways across the globe, or depend more on local conditions (Samways 1992). A multi-
regional programme could potentially distinguish globally recurring patterns and convergence
from local phenomena. Such knowledge could foster collaboration among researchers and
managers in finding ways to mitigate the adverse ecological effects of urbanisation.
All over the world, urban landscapes consist of densely built cores surrounded by
decreasing intensity of development and increasing ‘naturalness’. These gradients can
provide a framework for ecologists to examine human-induced landscape changes and
compare the findings across the world to unravel generalities in biotic community structure in
relation to these changes. It is important to note that in the GLOBENET project we do not
compare cities per se, but patterns along the gradients between cities.
To assess changes in urban landscapes we have developed a programme that uses a
common field methodology (pitfall trapping), one taxonomic group (carabid beetles) in
visually-similar land-mosaics, in different parts of the world (Niemelä et al. 2000). Carabids
are sufficiently varied taxonomically and ecologically, abundant, and are sensitive to human-
caused disturbances to be a reliable monitoring group. They have been widely studied in
relation to land use throughout the world (e.g. Rainio and Niemelä 2003).
Hypotheses can be derived and tested using the gradient approach. For instance, one
can ask whether the intermediate-disturbance hypothesis (Connell 1978) apply to urban-rural
gradients or whether the predictions about how community structure responds to stressors
(Gray 1989) hold for carabids in urban environments: (a) diversity should decrease from a
high in rural areas to a low in urban areas, (b) opportunistic species should gain dominance
in urban areas, and (c) mean body size of the dominating species should decrease from less
disturbed to more disturbed habitat (Blake et al. 1994), here from rural to urban areas.

Results and discussion

To date the GLOBENET approach has been employed at least in Sofia (Bulgaria),
Edmonton (Canada), Helsinki (Finland), Debrecen (Hungary), Hiroshima (Japan) and
Birmingham (England) (Alaruikka et al. 2002, Niemelä et al. 2002, Ishitani et al. 2003, Venn
et al. 2003, Magura et al. 2004, Sadler et al. 2006). Generally, the carabid beetles collected
in forests in these cities showed evidence of an increase in overall abundance and species
richness from city centres to the rural surroundings supporting the prediction by Gray (1989),
but no evidence of elevated diversity at suburban sites, as predicted by the intermediate
disturbance hypothesis. However, there is variation in these patterns. Magura et al. (2004)
demonstrated that the number of carabid species was significantly higher in the rural and
urban areas compared to the suburban one. Thus, this result did not support the hypothesis
that overall diversity should decrease in response to habitat disturbance. Many of the above-

193
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

mentioned studies demonstrated, as predicted, that the dominance was usually higher in the
urban and suburban zones, i.e. urban areas were dominated by few species.
The proportion of large sized carabid beetles usually decreased towards the city centres
in the studies mentioned above, again as predicted by Gray (1989). Magura et al. (2006)
studied the variation in carabid body size along an urban-rural gradient, and in addition to
showing that small individuals were more prominent in urban areas, they demonstrated that
the inequality in body size of the carabid assemblages decreased (albeit non-significantly)
along the gradient from urban towards rural areas. Furthermore, many of the above studies
also showed that the proportion of short winged species decreased from urban to rural areas.
Moreover, the proportion of forest habitat species decreased significant from the surrounding
rural environments to the city centres, while the proportion of open habitat species increased
significant towards the city centres.
The GLOBENET results so far suggest that carabid beetles respond in a predictable
manner to urbanisation in different cities. The challenge now is to infer process from the
patterns. For example, Sadler et al. (2006) suggested that changes in assemblage structure
were related to woodland fragmentation, which led to variations in woodland size, location
and site disturbance due to trampling. However, species differ in their response to
urbanization. Large, flightless and specialist woodland species are susceptible to changes
associated with urbanization (e.g. fragmentation and disturbance), presumably due to their
longer life spans, lower reproductive rates, specialized niches and limited dispersal potential.

References
Alaruikka, D; Kotze, D.J; Matveinen, K. & Niemelä, J. (2002) Carabid beetle and spider
assemblages along a forested urban-rural gradient in southern Finland. Journal of Insect
Conservation, 6: 195-206.
Blake, S; Foster, G.N; Eyre, M.D. & Luff, M.L. (1994) Effects of habitat type and grassland
management practices on the body size distribution of carabid beetles. Pedobiologia, 38: 502-512.
Connell, J.H. (1978) Diversity in tropical rain forests and coral reefs. Science, 199: 1302-1310.
Gray, J.S. (1989) Effects of environmental stress on species rich assemblages. Biological Journal of
the Linnean Society 37: 19-32.
Ishitani, M; Kotze, D.J. & Niemelä, J. (2003) Changes in carabid beetle assemblages across an
urban-rural gradient in Japan. Ecography, 26: 481-489.
Magura, T; Tóthmérész, B. & Molnár, T. (2004) Changes in carabid beetle assemblages along an
urbanisation gradient in the city of Debrecen, Hungary. Landscape Ecology 19: 747-759.
Magura, T; Tóthmérész, B; Lövei, G.L. (2006) Body size inequality of carabids along an urbanisation
gradient. Basic and Applied Ecology 7: 472-482.
McDonnell, M.J. & Pickett, S.T.A. (1990) Ecosystem structure and function along urban-rural
gradients: an unexploited opportunity for ecology. Ecology, 71: 1232-1237.
Niemelä, J; Kotze, J; Ashworth, A; Brandmayr, P; Desender, K; New, T; Penev, L; Samways, M.
& Spence, J. (2000) The search for common anthropogenic impacts on biodiversity: a global
network. Journal of Insect Conservation 4: 3-9.
Niemelä, J; Kotze, D.J; Venn, S; Penev, L; Stoyanov, I; Spence, J; Hartley, D. & Montes de Oca,
E. (2002) Carabid beetle assemblages (Coleoptera, Carabidae) across urban-rural gradients: an
international comparison. Landscape Ecology, 17: 387-401.
Rainio, J. & Niemelä, J. (2003) Ground beetles (Coleoptera: Carabidae) as bioindicators. Biodiversity
and Conservation, 12: 489-506.
Sadler, J.P; Small, E.C; Fiszpan, H; Telfer M.G. &. Niemelä, J. (2006) Investigating environmental
variation and landscape characteristics of an urban–rural gradient using woodland carabid
assemblages. Journal of Biogeography 33: 1126-1138.
Samways, M.J. (1992) Some comparative insect conservation issues of north temperate, tropical, and
south temperate landscapes. Agriculture, Ecosystems & Environment 40: 137-154.
Venn, S.J; Kotze, D.J. & Niemelä, J. (2003) Urbanization effects on carabid diversity in boreal
forests. European Journal of Entomology, 100: 73-80.

194
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

The response of species to urban landscape patterns

U.M. Mörtberg

Dept of Land and Water Resources Engineering, Royal Institute of Technology, Stockholm,
Sweden.
e-mail: [email protected]

Introduction

Urbanisation is affecting biodiversity across the world and to achieve sustainable

development, the impacts of urbanisation on biodiversity must be considered on a landscape
level (Miller & Hobbs 2002). The urbanisation process causes loss, degradation and
fragmentation of natural habitats and at the same time creates new habitats. To explore and
compare the consequences of these changes on biodiversity, landscape patterns and the
urbanisation process need to be quantified. Further, in order to answer questions that are
posed by spatial planning, more research is needed on the response of species to these
changes.

Urbanisation can be seen as land use changes causing alterations in landscape patterns
in terms of habitat quality, quantity and connectivity, which in turn affects ecological
processes. In the same way as landscape pattern can be quantified, urbanisation can be
measured in parameters such as the distance to city centres, the density of buildings, roads
and of the human population, or as the gradient of fragmentation (e.g. Luck & Wu 2002). The
remaining fragments of natural habitat are affected by various measurable disturbances,
including management but also recreation, noise, pollution, etc. Examples of the spatial
pattern of such variables in and around the city of Stockholm are illustrated in Figure 1.
Further, the history of how such variables have affected the landscape is important.

Figure 1. The city of Stockholm, the capital of Sweden (a), with urbanisation parameters;
density of human population (b), noise level from road traffic, airport and other
sources (c), and recreation pressure measured as trail density in nature areas
(d).

Examples of urban landscape patterns affecting species’ distribution

Studies in urbanisation gradients in the hemi-boreal zone have for instance shown that
urban forest fragments have higher abundances of deciduous trees than the surrounding
rural landscapes, where commercial forestry promotes conifers (e.g. Jokimaki 1996,

195
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Mortberg 2004). In the Stockholm area, Sweden, forest-interior bird species of coniferous
and mixed forest avoid urban forest fragments, in particular those adjacent to busy roads
(Mortberg & Karlstrom 2005). Urban forest fragments were more beneficial to deciduous
forest birds, including several red-list species (Mortberg & Wallentinus 2000).

Both spatial and time dimensions have been addressed in studies of the large stands of
native, broad-leaved deciduous trees, particularly oaks, in and around the city of Stockholm.
This habitat supports a very high diversity of invertebrates, despite high disturbance levels,
even though the area is relatively isolated today. However, the long historic continuity of old
deciduous trees in the area, and the earlier wider distribution of the habitat may have led to a
situation where the high diversity of today reflects the historical landscape pattern
(Löfvenhaft & Ihse 1998). Large stands of old oaks are still left in highly urban areas, in
certain parts with high or recently broken connectivity, and the maintenance and
development of habitat quantity, quality and connectivity of these stands will be crucial for the
diversity of invertebrates in the area (Mortberg & Ihse 2006).

These studies have shown that species respond differently to urbanisation, and the
knowledge can therefore be used to plan for biodiversity in urbanised areas (Mortberg et al.
2006). For comparative studies of the consequences of urbanisation on biodiversity, more
knowledge is needed on the response of different species to the quantified landscape pattern
and processes in the urbanised landscape. Such knowledge is essential for planning and for
sustainable development.

References
Jokimäki, J. (1996) Patterns of bird communities in urban environments. Dissertation. Arctic Centre
Report N:o 16. University of Lapland. Rovaniemi, Finland.
Löfvenhaft, K. & Ihse, M. (1998) Biologisk mångfald och fysisk planering. Landskapsekologisk
planering med hjälp av flygbildsbaserad fjärranalys - metodstudie i Stockholm. Research Report
No. 108, Department of Physical Geography, Stockholm University, Stockholm. [In Swedish].
Luck, M. & Wu, J. (2002) A gradient analysis of urban landscape pattern: a case study from the
Phoenix metropolitan region, Arizona, USA. Landscape Ecology 17: 327-39.
Miller, J.R. & Hobbs, R.J. (2002) Conservation where people live and work. Conservation Biology 16:
330-337.
Mortberg, U. & Ihse, M. (2006) Landskapsekologisk analys av Nationalstadsparken. Underlag till
Länsstyrelsens program för Nationalstadsparken. Stockholm County Council, Report 2006:13. [In
Swedish].
Mortberg, U. & Karlstrom, A. (2005) Predicting forest grouse distribution taking account of spatial
autocorrelation. Journal for Nature Conservation 13: 147-159.
Mortberg, U. (2004) Landscape Ecological Analysis and Assessment in an Urbanising Environment –
Forest Birds as Biodiversity Indicators. Dissertation, Royal Institute of Technology, Dept of Land
and Water Resources Engineering, Stockholm. 50 pp.
Mortberg, U.M. & Wallentinus, H.-G. (2000) Red-listed forest bird species in an urban environment -
assessment of green space corridors. Landscape and Urban Planning 50: 215-226.
Mortberg, U.M., Balfors, B. & Knol, W.C. (2007) Landscape ecological assessment: a tool for
integrating biodiversity issues in strategic environmental assessment and planning. Journal of
Environmental Management 82: 457-470.

196
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Urban soil sealing – key indicator for urban ecological functionality and
ecological planning

J. H. Breuste

University of Salzburg, Department of Geography and Geology, Hellbrunner Strasse 34/III,

A-5020 Salzburg, Austria.
e-mail: [email protected]

Soil sealing as urban ecological problem

Cities and towns are characterized by sealed surfaces of different types and structures
(buildings, pavements etc.). Urban soil sealing is one of the main drivers of urban ecological
functions and ecosystem services (climate, water balance, soil functionality, biodiversity etc.).
There is not much knowledge available to quantify these functions depending and steered by
soil sealing in cities. Soil sealing is everywhere growing in cities and sealed surfaces grow
faster than the city itself (population, economy etc.). Urban soil sealing has been identified as
key indicator for de-naturalization and destruction of urban ecosystem services – without
many consequences. Planning instruments are proved to reduce soil sealing and its growth
rates (Boecker 1985, Pietsch & Kamith 1991, Duhme & Pauleit 1992, Münchow & Schramm
1997, Breuste 2002).

Objectives

The following topics will be investigated on examples of Central European cities:

1. Which functions (qualities) are influenced and in what way (quantities) they are
influenced?
2. For what is soil sealing a useful ecological indicator in urban ecosystems and how can
it be measured and monitored?
3. How can soil sealing be typified by ecological relevant characteristics?
4. Is there an urban pattern of soil sealing depending form other factors of land use (urban
structure)? Can soil sealing be indicated?
5. Why soil sealing is growing (steering processes, interest groups, stakeholders etc.) and
How can soil sealing be reduced or new soil sealing be avoided by using planning and
decision making instruments?

Methodology

Beside theoretically approaches and evaluation methodologies existing studies of

practical investigations in the Central European cities Greifswald, Leipzig and Salzburg will
be evaluated. This includes a 10 years long-term study of urban ecological changes on the
example of a German city in the economic transformation process of the 90th regarding soil
sealing growth. Soil sealing in urban ecosystems will be investigated on different levels
(whole city, districts and sites) to identify the indicator quality for ecological functions. In
Germany existing planning instruments will be proved for there ability to avoid soil sealing.
New economic steering instruments will be proposed.
The expected results of ecological functionality can be transferred to other cities and towns.
The specific planning instruments must be proved under different regional and local
conditions. This links to a better understanding of urban ecosystems and its management.

197
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Results

Hydrological functions (example)

The most strongly noticed risks of soil sealing are the effects on the urban water regime.
High storm water run-off on the one hand and low ground water table in cities on the other
hand are increasingly steered by soil sealing. This justifies extensive efforts for the analysis
of the expansion and characteristics of the soil sealing (e.g. by areal photographs and
satellite images) (Netzband 1998), measurements of influencing factors and urban programs
to reduce soil sealing (e.g. by application of adjusting and economic instruments). A general
goal of the ecological urban development must be to infiltrate a substantially larger part of the
precipitation in cities than so far and by this to reduce the storm water run-off. This should
reflect that the different kinds of soil sealing (pavement types) have a different infiltration
capacity.

Soil sealing management

For different soil sealing types descriptions and typical characteristics can be summarized
in “soils sealing catalogues”. These contain statements about infiltration capacity and run-off
and evaporation under different conditions of precipitation events and include the ability to
assess soil sealing to improve urban environmental management (Breuste et al. 1996, 2002.
Münchow & Schramm 1997).
Reduction of the sealed surfaces, improvement of their water permeability and additional
decentralized infiltration belong to the necessary measure of a complex management. A
necessity of this task is apart from the improvement of the monitoring methodology (remote
sensing, geographical information systems) also in the technical process analysis and its
spatial evaluation (experimental research and determination of sealing characteristics for
different functions).
The results of ecological functionality can be transferred to other cities and towns. The
specific planning instruments must be proved under different regional and local conditions.
This links to a better understanding of urban ecosystems and its management.

References
Böcker, R. (1985) Bodenversiegelung - Verlust vegetationsbedeckter Flächen in Ballungsräumen -
am Beispiel von Berlin (West). Landschaft und Stadt 17, 57-61.
Breuste, J. (2002) Urban Ecology. In: Bastian, O., U. Steinhardt (Eds.). Development and
Perspectives of Landscape Ecology. Kluwer Academic Publishers, Dordrecht. 405 – 414.
Breuste, J., T. Keidel, G. Meinel, B. Münchow, M. Netzband, M. Schramm (1996) Erfassung und
Bewertung des Versiegelungsgrades befestigter Flächen. Leipzig (=UFZ-Bericht 12)
Duhme, F. &. S. Pauleit (1992) Naturschutzprogramm für München: - Landschaftsökologisches
Rahmenkonzept. Geographische Rundschau 44, 554-561.
Münchow, B. & M. Schramm (1997) Permeable pavements – an appropriate method to reduce
stromwater flow in urban sewer systems? In: Breuste, J., H. Feldmann, O. Uhlmann (Eds.). Urban
Ecology. Leipzig. 183 – 186.
Netzband, M (1998) Möglichkeiten und Grenzen der Fernerkundung zur Versiegelungskartierung in
Siedlungsräumen. Ph.D. thesis, TU Dresden.
Pietsch, J. & H. Kamith (1991) Stadtböden. Entwicklungen, Belastungen, Bewertung und Planung.
Taunusstein

198
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Suburban habitats and their role in the urban-rural habitat network: Point of local
invasion and extinction?

R.B. Blair

Fisheries, Wildlife, and Conservation Biology, University of Minnesota, Saint Paul,

Minnesota, 55108 USA.
e-mail: [email protected]

Introduction

Urban and suburban habitats are becoming increasingly important to biodiversity

conservation efforts in the United States. Current land-use trends show that the acquisition
of parks and preserves has levelled off in the US while the amount of developed urban lands
is increasing exponentially. The tipping point between these competing land uses occurred
in 1990 when the amount of developed lands exceeded the amount of preserved lands
(McKinney, 2002). Urban- and suburban- ization is not only an issue of the amount of land
used but also of the types of habitats that it alters. Humans tend to settle in areas that are
biodiversity “hotspots” which may lead to species extirpation and extinction. An assessment
of economic associations causing species endangerment in the US found that urbanization
was the second most common factor associated with declines of species on the Endangered
Species List; only “interactions with invasive species” outranked urbanization in the number
of species that each endangered (Czech et al., 2000).

Because of this, suburban habitats may play an important role in biodiversity

conservation; however, they also present a conundrum. Typically, suburban habitats in an
urban-rural habitat network of naturally forested areas contain less than half of the native
woodland bird species that would exist at these sites if they were not developed. They also
contain more total bird species than if these sites were left in a natural state (Blair, 2002).
This apparent contradiction raises the questions “How do suburban habitats function in the
urban-rural habitat network?” Do these habitats serve as points of local extirpation for urban-
avoiding, woodland species? Do they function as points of local invasion for urban-exploiting
species? Are they potentially useful for the conservation of the species that are suburban-
adapters?

Methods

In this study, I analyze bird distribution data for three rural-to-urban gradients in different
cities and ecoregions of the United States: Palo Alto, California; Oxford, Ohio; and Saint
Paul, Minnesota. Each gradient consists of six sites of increasingly urban land use: a
preserve, an open-space recreational area, a golf course, single-family detached housing,
industrial park, and a business district. Each of these sites contains sixteen sampling points,
which were surveyed multiple times over two breeding seasons in order to obtain estimates
of the densities of all species. I used these density estimates to identify patterns in species
distributions and community measures including species richness, Shannon diversity, and
evenness. I also used these density estimates to characterize the bird communities in each
site with respect to life history characteristics and nesting attributes based in a literature
review. Finally, I used various community ordination techniques to compare bird
communities within individual gradients as well as across the three ecoregions to examine
landscape factors that play a role in bird distributions and to assess whether urbanization is
leading to a more homogeneous fauna nationwide.

199
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Results

The patterns of bird distribution in all three ecoregions were remarkably similar. Each
region had species that were distinct urban avoiders, species that existed only in the most
undisturbed woodlands. Each had a majority of species that were suburban adaptable with
highest densities at intermediate levels of development. Each had a small subset of species
that were urban exploiters having their highest-densities in the most urbanized sites. The
community compositions transitioned gradually across all three gradients with the highest
species richness and diversity occurring at intermediate sites of development. In contrast,
the most natural and the most developed sites had lower species richness and most urban
sites having the fewest species, the greatest number of exotics, and the lowest evenness.

This patterning suggests that suburban land uses, those represented by the intermediate
levels of development on the gradients, are a point of extirpation for woodland birds along
the gradient as well as an entry point for invasive species into urban systems.
Characteristics of the woodland species that go locally extinct include a narrow diet breadth,
a single brood strategy, strong territoriality, and a dependence on high nesting sites.
Characteristics of the invading species include broad diet breadth, a multiple brooding
strategy, a lack of territoriality in the breeding season, and the ability to occupy a variety of
nesting substrates.

Kentucky Warbler -- Ohio American Robin -- Ohio European Starling -- Ohio

0.05 6 5

5
0.04 4
Birds per Hectare

Birds per Hectare

4 Birds per Hectare

0.03 3
3
0.02 2
2

0.01 1
1
Business District

Business District

Business District
Open Space

Open Space

Open Space
Apartments

Apartments

Apartments
Golf Course

Golf Course

Golf Course
Residential

Residential

Residential
0 0 0
Preserve

Preserve

Preserve

Figure 1. Typical pattern of local extinction and invasion across a rural-urban gradient.

Discussion

Suburban habitat is a growing component of the landscape in the United States and is
consistently ranked as a major cause of endangerment. Consequently, understanding the
dynamics of extinction and invasion in suburbia is vital to comprehensive conservation
planning. Research from three different ecoregions of the United States (California, Ohio,
and Minnesota) suggests that urban-avoiding species, including many neotropical migrants
such as Kentucky warblers, are replaced by the urban-exploiting species, including many
invasives such as European starlings. This substitution of species appears to be driven by
differences in the life histories of these groups and suggests that reproductive strategies may
be the key to urban exploitation.

References
Blair, R.B. (2002) The effectsof urban sprawl on birds at multiple levels of biological organization.
Ecology and Society Volume:9 2, url www.ecologyandsociety.org/vol9/iss5/art2.
Czech, B.; Krausman, P.R. & Devers, P.K. (2000) Economic associations among causes of species
endangerment in the United States. BioScience Volume:50 593-601.
McKinney, M.L. (2002) Urbanization, biodiversity, and conservation. BioScience Volume:52 883-890.

200
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Suburban business and industrial sites: an unexplored opportunity for

biodiversity conservation?

R. Snep, W. Timmermans, V.Kuypers

Alterra – Wageningen UR, P.O. Box 47, 6700 AA Wageningen, the Netherlands.
email: [email protected]

Introduction

Business and industrial sites are exclusively designed and managed to provide firms an
optimal location for establishment. Often they are situated at the city’s edge, nearby
important road infrastructure and with extra land available for future expansions. As the
fabrication, storage and transport of commodities and goods are main activities, large
assembling halls and parking lots dominate the land use of business and industrial sites.
Their large portion of paved area and buildings makes them not the first place to think of
biodiversity, and earlier research (Blair,2001) confirmed the general idea that business and
industrial sites have little to offer for nature.
We, however, think that business and office sites have a certain value for biodiversity
conservation, at least in potential. In our research we found several clues that supported this
idea, and below we formulate and explain our findings:

I - The land use at business and industrial sites offers space for interesting habitats.
The high turnover rate of companies at business sites, the overcapacity of land as
reserved for future expansions and the aim to keep urban green maintenance as low as
possible create the perfect conditions for pioneer vegetations. After the business site has
been prepared, an open area with sandy soil is waiting for further development. If this
development takes several years to start, these vacant lots become habitat for pioneer
species that quickly colonize the bare soil. Among them are endangered plant and animal
species which habitat has become rare.
With respect to the buildings at business and industrial sites, most of them have flat roofs
(e.g. assembling halls). These roofs offer space for coastal birds (if covered with gravel) or
other bird species and many flying invertebrates (if designed as a green roof). Current data
show that plant and animal species easily colonize flat roofs if designed and managed in the
right way.

II – The suburban location of business and industrial sites adds to the habitat value these
sites have for biodiversity conservation
Most business sites are situated at the transition from urban to rural land, thereby being a
perfect stepping stone location for plants and animals that migrate between cities and their
surroundings. Snep et al. (2006) illustrated how habitat patches at the city’s edge could act
as source area for butterflies that colonize the city’s interior habitats. As expected the
dispersal capacity of the species and the size of the suburban source population determine
the range and impact of the suburban butterflies on the inner-city butterfly population. Green
business sites, designed and managed optimally to serve as butterfly habitat (without losing
their economic function) could as such strengthen butterfly populations in the adjacent
residential areas.

III –Business and industrial sites may already contain high biodiversity values
Preliminary results from a breeding bird survey of 23 business and industrial sites in the
Netherlands showed a list of 90 different bird species, of which 18 appear on the national
Red List of Endangered Bird Species. Among them the Grey partridge that occurred on at
least 6 business sites (Snep et al. in prep).

201
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

The largest population of the Natterjack toad in Flanders (northern Belgium) can be found
in the Port of Antwerp, where in the period 2003-2005 a chorus of approximately 1250 calling
males was recorded. Such large numbers are rarely observed in other habitat areas, and this
makes this port area a national hotspot for this endangered and highly protected amphibian
species (Snep and Ottburg in prep).
The data on breeding birds and amphibians illustrate the habitat value that current
business and industrial sites possess in some occasions, although it is clear that not all sites
contain high biodiversity values.

IV – Biodiversity conservation at business and industrial sites may provide benefits for
People, Planet and Profit
Cardskadden and Lober (1998) stated that nature conservation at corporate land result in
more benefits than for nature only. Firms that participate in nature conservation programs of
the Wildlife Habitat Council (WHC) improve their social status by showing environmental
awareness and concern for decreasing biodiversity levels and employees of those firms
enjoy their working environment much better. In addition, nature conservation measures (like
green roofs and green design) could reduce costs by better roof isolation and extension of
business site’s life time (urban green helps to hold the estate value of the business site).
Altogether green and ecological managed business and industrial sites may contribute to a
better quality of business and office sites for People, Planet and Profit.

References
Blair, R.B. (2001). Creating a homogeneous avifauna (Chapter 22) In: Marzluff, J.M., R. Bowman & R.
Donelly 2001. Avian ecology and conservation in an urbanizing world. Kluwer Academic
Publishers, Dordrecht, the Netherlands. 459-486.
Cardskadden, H. and D.J. Lober (1998). Environmental stakeholder management as business
strategy: the case of the corporate wildlife habitat enhancement programme. Journal of
Environmental Management 52 (2): 183-202.
Snep, R.P.H., P.F.M. Opdam, J.M. Baveco, M.F. WallisDeVries, W. Timmermans, R.G.M. Kwak &
V. Kuypers (2006). How peri-urban areas can strengthen animal populations in cities: a modeling
approach. Biological Conservation 127: 345-355.
Snep, R.P.H. et al. (in prep). Breeding birds of business and office sites and industrial areas in the
Netherlands. Determining land use factors that optimize business site habitats for birds.
Snep, R.P.H. and F.G.W.A. Ottburg (in prep). A habitat backbone as nature conservation concept
for industrial areas – the case of the Natterjack toad Bufo calamita in the Port of Antwerp
(Belgium).

202
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Biodiversity in urban habitat patches: habitat quality matters

J.P. Sadler, P.G. Angold

School of Geography, Earth and Environmental Sciences, Birmingham University,

Edgbaston, Birmingham B15 2TT, UK.
e-mail: [email protected]

Introduction

In cities the biological processes of dispersal interact with the landscape structure in
determining the distribution of populations of species present (Niemelä, 1999). Many cities
have a network of habitat fragments or ‘urban greenways’ comprising areas of semi-natural
habitats, secondary succession, ruderal and pioneer environments and open areas. These
habitats may be important features for biodiversity both as stable and as transient habitats
(McIntyre et al., 2000; 2001), and may also be valuable for their possible function as
‘corridors’ and ‘stepping stones’ to facilitate species dispersal (Kirkby, 1995).
This research aimed to: (i) examine the effects of habitat fragment size and connectivity
upon the ecological diversity and species distributions, and (ii) analyse the extent to which
the flora and fauna utilise the ‘urban greenways’ as corridors and as habitats.

Methods

Study area
The West Midlands conurbation comprises the City of Birmingham and several boroughs
collectively known as the Black Country. Birmingham City has a population of one million,
with an estimated 6 million others living within a 50 mile radius of the city. The city covers
27,000 hectares, 11% of the land cover is green space in the form of parks, and the city
includes approximately 4000 hectares of semi-improved neutral grassland, pockets of
ancient woodland, and 250,000 domestic gardens.

Study groups
The project comprised four components involving empirical survey work of both plants
and ground beetles (Carabidae) on a range of habitats (derelict, wetland and woodland), a
genetic study of four species of grassland butterflies, an examination of the plant
communities and hemeroby (Hill et al., 2002) and spatial modelling of mammal populations.
Landscape metrics (e.g. patch size, distance to corridor, area of similar habitat within 100,
1000 and 5000m zones around the sites) were captured in a GIS and related to species
metrics in GLM models and using multivariate analyses.

Results

Landscape context
Plants showed little relationships to the landscape metrics although site size (p<0.05) and
the proximity of the nearest similar sites were significantly for a few (25 out of 378 species)
(p=0.05). Similarly, for the invertebrates, habitat (patch) quality appears to be the significant
factor rather than landscape structure. For example, overall species richness and the number
of specialist wetland and derelict land beetles was related to site variables such as age
(p=0.036) and vegetation type (p<0.001), with only limited links to landscape variables (e.g.
the amount of similar habitat within 1000m of the sites, p=0.003) (Small et al. 2003).
The level of disturbance from recreational users heavily influenced woodland carabid
communities. This was most clearly illustrated by increasing dominance of a few ubiquitous
urban generalists (e.g. Pterostichus madidus) (df=2,21; F=12.345; p=0.001) and the decline
of the woodland (e.g. Cychrus carabioides) (df=2,21; F=8.842; p=0.002), and other large-

203
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

bodied species (e.g. Cararbus spp.) (df=2,21; F=19.858; p=<0.001). Although typically of
secondary importance landscape effects such as site size, the amount of built up and
wooded land within 5km of the sample sites were significantly (p<0.05) related to the
woodland beetle assemblages indicating some structuring of assemblages as a result of
woodland fragmentation and isolation (Sadler et al. 2006).

Corridor proximity
Plant communities of derelict sites showed no greater similarity to each other if adjacent
or close to designated urban corridors, suggesting that the corridors do not increase
connectivity. Out of a pool of 433 plant species only 7 were positively associated with
corridors (p<0.005) and these were either invasive wetland species (Impatiens glandulifera)
or associated with wetlands or disturbed areas (e.g. Lipinus polyphyllus). We found no
evidence that wetland specialist (r2 = 0.0077) nor derelict specialist (r2 0.005) beetle diversity
is greater on or near the corridors, and no evidence that corridors are necessary for dispersal
of butterflies (Wood and Pullin 2003).
Spatially explicit models suggest that dormice and water voles may depend on linear
habitats for dispersal and small and medium sized mammals (Muntjac deer) could use
corridors. However, this is not the case for plants and invertebrates (Angold et al. 2006).

Implications: planning for wildlife in the urban environment

Although a complex picture, this research provides little evidence that urban greenways
affect plants or invertebrate assemblage structure. Similarly, in only a few cases was the
landscape structure of the conurbation an influential factor in structuring the communities.
The results are consistent with the hypothesis that most urban species are able to move
freely around the city. However, there are groups of species (with particular traits) and
habitats with particular characteristics (e.g. more stable such as woodlands) where the
impact of urbanization is extremely negative. These findings indicate: (i) the importance of
identifying a target species or group of species for urban greenways intended as dispersal
routeways, and (ii) the maintenance of good quality habitat for other organisms.

References
Angold, P.G., Sadler, J.P., Hill, M.O., Pullin, A., Rushton, S., Austin, K., Small, E., Wood, B.,
Wadsworth, R., Sanderson, R., & Thompson, K. (2006) Biodiversity in urban habitat patches.
Science of The Total Environment, 360, 196-204.
Hill, M.O., Roy, D.B. & Thompson, K. (2002) Hemeroby, urbanity and ruderality: bioindicators of
disturbance and human impact. Journal of Applied Ecology, 39, 708-720.
Kirby, K. (1995) Rebuilding the English countryside: habitat fragmentation and wildlife corridors as
issues in practical conservation English Nature Science, No 10, Peterborough.
McIntyre, N.E. (2000) Ecology of urban arthropods: A review and a call to action. Annals of the
Entomological Society of America, 93, 825-835.
McIntyre, N.E., Rango, J., Fagan, W.F., & Faeth, S.H. (2001) Ground arthropod community structure
in a heterogeneous urban environment. Landscape and Urban Planning, 52, 257-274.
Niemelä, J. (1999) Ecology and urban planning. Biodiversity and Conservation, 8, 119-131.
Sadler, J.P., Small, E.C., Fiszpan, H., Telfer, M.G., & Niemela, J. (2006) Investigating environmental
variation and landscape characteristics of an urban-rural gradient using woodland carabid
assemblages. Journal of Biogeography, 33, 1126-1138.
Small, E., Sadler, J.P., & Telfer, M. (2006) Do landscape factors affect brownfield carabid
assemblages? Science of The Total Environment, 360, 205-222.
Wood, B.C. & Pullin, A.S. (2002) Persistence of species in a fragmented urban landscape: the
importance of dispersal ability and habitat availability for grassland butterflies. Biodiversity and
Conservation, 11, 1451-1468.

204
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Patterns of exotic species invasion in fragmented native grasslands along urban-

rural gradients in South Africa and Australia

S.S. Cilliers*1, N.S.G. Williams2 & F.J. Barnard1

1
School of Environmental Sciences and Development, North-West University
(Potchefstroom Campus), South Africa,
e-mail: [email protected]
2
Australian Research Centre for Urban Ecology, Royal Botanic Gardens, Melbourne,
Victoria, Australia

Introduction

Temperate grasslands are one of the world’s major biomes and include some of the most
diverse and productive terrestrial ecosystems (Suttie et al., 2005). In both South Africa
(Cilliers et al., 2004, Van Jaarsveld et al., 1998) and southeastern Australia (Williams et al.,
2005a, Kirkpatrick et al., 1995) native grasslands have been severely fragmented by
agricultural and urban development and few are protected in conservation areas. However,
high quality patches of native grasslands of conservation importance persist, in urban and
rural landscapes in both countries (Cilliers et al., 1999, Kirkpatrick et al., 1995). Fragmented
habitats have higher perimeter to surface area ratios than continuous habitats and are
thought to be vulnerable to invasion by exotic species via edge effects. Recent research in
Victoria, Australia indicates that grasslands surrounded by different landscapes exhibit
substantially different patterns of exotic species invasion. Rural grasslands had sharp well
defined edges dominated by exotic species which declined with distance from edge while
urban grasslands exhibited a variety of patterns but most lacked defined edges and had a
moderate to high exotic cover across the sites (Williams 2005). In this study we examine the
generality of these results by comparing structurally similar remnant grasslands dominated
by the same species (Themeda triandra) on two continents.

Methodology

Study sites were located in Victoria, Australia and North-West Province, South Africa.
The vegetation of rectangular, linear grassland fragments with straight edges adjacent to
roads and railways within two landscape matrix types (rural and urban) were surveyed.
Urban and rural areas were distinguished from each other based on the total length of roads
surrounding the fragment. Sites were surveyed with transects running perpendicular to the
two longest sides of the grassland remnant. Percentage cover of all vascular plant species
(distinguishing between natives and exotics) were recorded in quadrats spaced evenly along
the transect. A variety of edge detection methods were used to determine the depth of edge
influence.

Results

Edge effect boundaries at most rural grasslands were sharp and continuous with a
relatively short depth of edge influence (Fig.1a,b). Edges of urban grasslands were more
difficult to characterize but most had a greater depth of edge influence (Fig. 1 c,d). These
studies indicated that urban grassland edges have responded differently to fragmentation
than rural grasslands. Similar patterns were observed on different continents suggesting that
urbanization is a ubiquitous process that modifies edge effects in a similar manner despite
very different environmental, historical and social conditions in the cities examined.

205
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Figure 1. The mean percentage cover of native (●) and exotic (■) species across two
representative rural native grassland remnants in a) Western Victoria, Australia and b) North-
West Province, South Africa and across two representative urban grassland remnants in c)
Melbourne, Australia and d) Potchefstroom, South Africa.

References
Cilliers, S.S., Van Wyk, E. & Bredenkamp, G.J. (1999) Urban nature conservation: vegetation of
natural areas in the Potchefstroom municipal area, North West Province, South Africa. Koedoe 42:
1-30.
Cilliers, S.S., Müller, N. & Drewes, J.E. (2004) Overview on urban nature conservation: situation in
the western-grassland biome of South Africa. Urban Forestry and Urban Greening 3: 411-419.
Kirkpatrick, J., McDougall, K. & Hyde, M. (1995) Australia’s most threatened ecosystem: the
southeastern lowland native grasslands. Surrey Beatty & Sons, Chipping Norton.
Suttie, J.M., Reynolds, S.G. & Batello, C. (2005) Grasslands of the World. Plant Production and
Protection Series No. 34, Food and Agriculture Organization of the United Nations, Rome.
Van Jaarsveld, A.S., Ferguson, J.W.H. & Bredenkamp, G.J. (1998) The Groenvaly grassland
fragmentation experiment: design and initiation. Agriculture, Ecosystems and Environment 68:
139-150.
Williams, N.S.G. (2005) The ecology of fragmented native grasslands in urban and rural landscapes.
PhD thesis. University of Melbourne, Australia.
Williams, N.S.G., McDonnell, M.J. & Seager, E.J. (2005) Factors influencing the loss of an
endangered ecosystem in an urbanizing landscape: a case study of native grasslands from
Melbourne, Australia. Landscape and Urban Planning 71: 35-49.

206
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

Predicting land cover change and avian community responses in rapidly

urbanizing environments

J.A. Hepinstall1, M. Alberti2, J. Marzluff3

1
Warnell School of Forestry and Natural Resources, University of Georgia, Athens GA, USA
e-mail: [email protected],
2
Urban Ecology Research Laboratory and 3College of Forest Resources, University of
Washington, Seattle WA, USA.

Introduction

Predicting future landscape change is essential to inform the planning process and
conservation efforts. Land transformation increasingly is a major threat to biological diversity
in the USA and elsewhere in temperate regions of the globe (Sala et al. 2000). The influence
of rapid urbanization is particularly evident in the western USA (Hansen et al. 2005). While
we are beginning to understand how local urbanization processes influence biodiversity, we
know much less about how these altered processes of extinction and colonization will play
out through time.
Linking predictions from landscape change models to ecological models is generally
accomplished by developing spatially explicit habitat models (Schumaker 2005). We build on
such approaches by linking outputs from a sophisticated model of urban development
(UrbanSim) to a Land Cover Change Model (LCCM) we developed that incorporates
concepts central to landscape ecology in the prediction of land cover change 25 years into
the future for the Seattle, USA, metropolitan region - a region undergoing rapid urbanization
with 31% growth predicted (1 million people) in the next 25 years. Predicted land cover and
land use are then used as input into avian diversity models to predict the influence of
urbanization on biodiversity.

Methods

Land cover: The LCCM framework derives from the traditions of modelling landscape change
as a dynamic interaction between socio-economic and biophysical processes (e.g., Wear et
al. 1998). In the LCCM, transition probability equations are estimated empirically from
observed land cover in 1991, 1995, and 1999 for ten land cover classes as a function of 65
potential explanatory variables. Transition probabilities for each 30-m pixel to change land
cover class is hypothesized to be influenced by land use change and intensity of
development predicted by UrbanSim, a set of local and attributes of the pixel, the spatial
context of a site, and variables representing observed change.
Birds: Linear regression models were developed to predict both avian species richness (a
total of 57 common species and three habitat guilds) and individual species relative
abundance as a function of a series of landscape habitat metrics calculated for 139 1-km2
field sites. Ten landscape variables describing land cover and land use patterns were
developed for consideration. Land cover variables included: percent forest, percent urban,
aggregation of forest, number of patches of forest, and number and mean patch size of
urban patches. Land use derived from parcel data was used to calculate the percent, patch
density, and aggregation of residential parcels, and mean year built of all parcels within each
study site. Equations were applied to UrbanSim land use and LCCM land cover outputs to
predict changes in avian biodiversity at two scales.

Results

Land cover: Predictions of land cover from 2003 to 2027 show a decrease in mature forest
types (deciduous, mixed, and coniferous) from 60% of the study area to 38% and an
increase in developed land (heavy-, medium-, and low-intensity urban classes) from 17% to

207
Theme 2. Urban environment and transport
2.3 Symposium 4: Applying landscape ecological principles in urban environments

34%. Land in grass and agriculture use decreased from 14% to 10% of the area. Harvested
and regenerating forest increased from 9% in 2003 to 18% in 2027. Most landscape change
is concentrated in areas surrounding currently developed lands (i.e., expansion growth), with
little infill or outlying new patches of development – the latter primarily occurring along lower
elevations and up river valleys. Comparisons of land cover observed in 2002 and predicted
for 2003 indicate higher agreement in the amount and location of change when considered at
aggregate scales of 1km2, the scale of our avian models.
Birds: Mean total avian species richness for the study area is predicted to decline from 36 to
31. Within the region, urban and urbanizing zones are expected to see the greatest decline in
total species. Changes in species richness are concentrated in those regions of the study
area where land cover change is most dramatic – primarily in the urbanizing zone where
forest loss and aging of developments cause the greatest changes in land cover. We expect
future bird communities to be slightly less diverse and more vulnerable to future losses than
they are at present. We expect native forest birds to become increasingly reliant on higher
elevation forests as most low elevation forests will be converted to development too dense to
support viable populations. High elevation bird populations may be less sustainable due to
harsher winters and shorter growing seasons that may limit survival and reproduction.

Significance

Our results clearly indicate that landscape change in the Seattle metropolitan region is
likely to be extensive over the next 25 years and that both landscape composition and
configuration resulting from urbanization are important in determining avian species richness.
By showing how birds generally respond to the amount, configuration and age of
development we can provide planners with relevant tools to better understand how their
decisions concerning zoning, housing density, and designation of conservation areas affect
bird communities. With the increasing ubiquity of spatial data and GIS skills, even local
planning offices can develop and apply such tools. Fostering ecosystem functionality is an
important supplement to land use and land cover planning efforts requiring less manicured
yards with minimal grass, letting trees die and rot, enabling native predators like coyotes to
live in our neighbourhoods, and seeing the good that comes from natural disturbances.
Tending to our lifestyle as well as our land cover will be increasingly important to future bird
diversity if the changes we expect in the next 25 years occur.
We have shown how models of urban development can be translated into changes in land
cover and how these changes can be projected to affect the abundance and diversity of birds
in a rapidly urbanizing region. Models used and developed have been written in open source
code (primarily Python) and are transferable and reuseable in other urbanizing regions.

References
Hansen, A.J., Knight, R.L., Marzluff, J.M., Powell, S., Brown, K., Gude, P.H. & Jones, K. (2005)
Effects of exurban development on biodiversity: patterns, mechanisms, and research needs.
Ecological Applications 15:1893-1905.
Sala, O.E. et al. (2000) Global biodiversity scenarios for the year 2100. Science 287:1770-1774.
Schumaker, N.H., T. Ernst, T., White, D., Baker, J. & Haggerty, P. (2004) Projecting wildlife
responses to alternative future landscapes in Oregon’s Willamette Basin. Ecological Applications
14:381–400.
Wear, D.N., Turner, M.G. & Naiman, R.J. (1998) Land cover along an urban-rural gradient:
implications for water quality. Ecological Applications 8: 69-630.

208
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

2.4 Workshop on current and future research in urban ecology

Preservation of pastures as parts of the urban green infrastructure at Järvafältet,

Stockholm: risk of damage caused by man?

C. Florgård

Swedish University of Agricultural Sciences, Department of Urban and Rural Development,

Unit of Landscape Architecture, Box 7012, UPPSALA, Sweden
e-mail: [email protected]

Introduction

As cities grow, they usually spread out into their rural surroundings. Lately, interest in
preservation of the original vegetation in these surroundings (woodlands, meadows, pastures
etcetera) as parts of the future green infrastructure has increased. This planning and design
approach includes many functional, economic, biological, social and aesthetic advantages. In
Sweden, the preservation of pastures in cities has often been combined with the preservation
of ancient grave-fields. These grave-fields have often been used for grazing during centuries,
and have become a grass and herb vegetation with high biodiversity, great beauty and great
attractiveness to people (Hägerhäll 1999). In contrast to grazing in e. g. Africa and South
America, where grazing often is a threat to the vegetation (Primack 1993), grazing is a
prerequisite of the survival of these important vegetation types.
However, trampling might damage the vegetation. This might result in damage to the
ground layer vegetation, which in turn can result in damage to the ground. The heritage
authorities have claimed that the risk for erosion because of trampling is too high, and that
the areas have to be excavated instead of preserved, with loss of biodiversity and amenity.
The aim of this paper is to elucidate the risk of damage of pastures preserved as parts of
the urban green infrastructure.

Methods

Two pastures (called A and B) on ancient grave-fields in Stockholm have been studied.
The surroundings were developed for residential houses in 1954-55 and in 1974-75,
respectively. Area B was studied from the time before development, during development of
the surroundings of the pastures, and during later use by the local inhabitants. Both areas
are directly bordering on roads. The distances to residential houses and schools are 100 m
(area A) and 7 m (area B).
The evaluation method used was triangulation according to Yin (1994). The following sub-
methods were used: photographing from fixed points, analysis of aerial photos from the time
before and after development, and for area B observations documented on sketches and
vegetation analysis of fixed vegetation investigation plots. Area A was studied 1978-2006,
and area B was studied 1972-1982, 1987 and 2003-2006. The results of the analyses were
compared to the results of studies of areas where grazing has been abandoned.

Results

Grazing in both areas was heavy before development. After development, grazing in area
A was abandoned. Area B was established as a grazing area for a “city farm” with horses
and sheep. Grazing was assessed as heavy 1976, after that light, and in 2006 abandoned.
The intensity of trampling at area B and its surroundings has been presented in Florgård
& Forsberg (2007). This so called border area at a distance of less than 100 metres from
residential houses was not much used by the inhabitants, but nevertheless much more used
than large recreation areas at a distance of 1000-2000 m from the built-up area. Area A was

209
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

more used than area B. Damage caused by resident’s trampling and other wear and tear
was found to be rare in area A and could not be detected in area B.
Both areas are now more or less overgrown with bushes, shrubbery and trees,
concerning area A see Figure 1. Overgrowth has started from former forest edges, and has
been spreading throughout the area. The vegetation development will be analysed further,
but preliminary results imply that the species richness was lost and herbs were exchanged
by grass.

Discussion and conclusions

The pattern of the overgrowth is the same as have been found in many studies where
grazing was abandoned (see e. g. Bakker 1989, Glimskär & Svensson 1990). The conclusion
is that the wear by the residents is much less than that formerly caused by the cattle. The
trampling by the residents is so much less so these formerly open areas are overgrown by
bushes and trees. This overgrowth is a severe threat to the areas in many ways. Firstly, the
biodiversity will decrease and they will lose their biological value. Secondly, they will lose
their amenity. And thirdly, if big trees grow on the graves, they can be damaged if the trees
are felled and uprooted by wind. Instead of man’s wear being a threat, the lack of trampling
becomes a threat. Management plans for the areas are needed.

Figure 1. Distribution
of bushes, shrubbery and
trees 1974 (start of
development), 1987 (13
years after development,
and 2006 (32 years after
development). =
Bushes, shrubbery <2 m,
= Bushes, shrubbery,
trees 2-6 m, = Bushes,
shrubbery, trees >6 m,
= Fence,
= Vegetation analysis
plot.

References
Bakker, J. P. (1989) Nature Management by Grazing and Cutting. Kluwer, Geobotany 14,
Dordrecht/Boston/London.
Florgård, C. & Forsberg, O. (2006) Residents’ use of remnant natural vegetation at the residential
area of Järvafältet, Stockholm. Urban Forestry and Urban Greening 5(2):83-92.
Glimskär, A. & Svensson, R. (1990) Vegetationens förändring vid gödsling och ändrad hävd.
Swedish University of Agricultural Sciences, Department of Ecology and Environment, Report 38,
Uppsala.
Hägerhäll, C. (1999) The experience of Pastoral Landscapes. Swedish University of Agricultural
Sciences, Agraria 182, Uppsala.
Primack, R. B. (1993) Essentials of Conservation Biology. Sinauer, Massachusetts, USA.
Yin, R. K. (1994) Case Study Research. Design and Methods. Second Edition. Sage, Thousand
Oaks/London/New Delhi.

210
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Urban semi-natural vegetation and the ecological functioning of urban green

space at the landscape scale

A. Millard

School of Architecture, Landscape and Design, Leeds Metropolitan University, Hepworth

House, Leeds LS2 8AE, UK
e-mail: [email protected]

Introduction

It has long been recognised that urban habitats can support a higher diversity of plant
species than the surrounding countryside (Gilbert, 1989). Two primary reasons suggested for
this are the heterogeneity of the urban landscape and the relatively high number of alien
(non-native) species available (Kowarik, 1995). The potential pool of native and non-native
species is subject to a complex of natural and anthropogenic processes.
There is conflicting evidence as to how far urbanisation can increase native species diversity
or the extent to which alien species are displacing native ones. However, there is mounting
evidence for the importance of urban landscape heterogeneity in supporting greater species
diversity overall (Waniaa et al., 2006). Roy et al. (1999), however, found no evidence that
urban landcover increases overall plant species richness. Their data did suggest that
complete urbanisation doubles the proportion of alien species and that the availability of
urban habitats, together with high levels of disturbance, are important factors in maintaining
the urban flora.
Roy et al. (ibid) looked at Britain as a whole, sampling tetrads (2km squares) from across the
country and using the two categories of urban landcover from the Landcover Map of Great
Britain, suburban/rural and urban development. This abstract reports on work in progess
which also examines the relationship between species composition and urban land cover,
but at a smaller scale and utilising a greater number of urban land cover categories.

Method

Plant species' records have been obtained from a survey of tetrads (2x2km squares) across
a contiguous area of urban and peri-urban landscape in West Yorkshire, northern England
(Abbott, 2005). Species have been categorised into native, archeophyte, neophyte and
casual. The nature and extent of urban green space within a tetrad is being determined from
GIS vector datasets, supported by examination of aerial photographs at a resolution of
12.5cm. Analysis of data is focusing on spatial variations in the relative abundance of the
different categories of semi-natural flora in urban areas and how these might be related to
variations in the extent and nature of urban green space.

Preliminary results

Species richness and urban green space

Analysis of the data is currently at an early stage but preliminary observations suggest no
clear relationship between tetrad species richness and either the extent or the nature of
urban green space. Neither is there a clear relationship between species richness and the
distance of a tetrad from the city centre. However, three of the four richest tetrads do include
the largest areas of open space that have been designated for their nature conservation
value, within the built-up part of the study area

Categories of species

211
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

As with total species richness, the different categories of species (natives, archaeophytes,
neophytes and casuals) show no obvious relationship between extent or nature of urban
green space. However, the percentage of native species in those three species-rich tetrads
encompassing significant designated nature conservation areas is very close to the mean for
the study area as a whole. A combination of the additional archaeophytes, neophytes and
casuals has contributed markedly to the relative species richness of these three tetrads.

Management of urban green space and plant biodiversity conservation

There is some suggestion from the data that management of urban green space for
biodiversity conservation, implicit in the designation of particular areas, may have effects at
the landscape scale. However, this is far from conclusive and it is by no means certain that
further analysis of the available data will resolve this question convincingly. The complex of
natural and anthropogenic factors operating on fully urbanised- and suburban/peri-urban
areas is likely to mean that conservation management policies directed at particular sites will
have very limited effects on urban plant species composition at the landscape scale.

References

Abbott, P. (2005) Plant Atlas of Mid-west Yorkshire. Yorkshire Naturalists' Union.

Gilbert, O. (1989) The ecology of urban habitats. Chapman & Hall, London.
Kowarik. I. (1995) On the role of alien species in urban flora and vegetation. In: Pysek. P. ct al. {eds).
Plant invasions:general aspects and special problems, SPB Academic Ptibl.. pp, 85-103.
Roy, D.B.; Hill, M.O. & Rothery, P. (1999) Effects of urban land cover on the local species pool in
Britain. Ecography 22: 507-515.
Waniaa, A..; Kühnb, I. & Klotz, S. (2006) Plant richness patterns in agricultural and urban
landscapes in Central Germany—spatial gradients of species richness. Landscape & Urban
Planning 75 (1-2): 97-110.

212
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Can biological traits help to explain changes in plant communities along a rural-
urban gradient?

J. Vallet1,2, H. Daniel1,2, V. Beaujouan2, F. Rozé3

1
Institut National d’Horticulture, UMR A462 SAGAH, 2 rue Le Nôtre, 49045 Angers France. E
mail: [email protected]
2
Institut National d’Horticulture, UP Paysage, 2 rue Le Nôtre, 49045 Angers France.
3
Université de Rennes 1, UMR 6553 Ecobio, Campus de Beaulieu, Avenue du Général
Leclerc, Rennes France.

Introduction

The current growth of cities is regarded as one of the major threats to the conservation of
biological diversity (Czech, et al., 2000, McKinney, 2006). One important challenge in urban
ecology is to better understand the relationships between biological traits of species which
are more or less adapted to city environments and a defined set of environmental conditions
found in cities. In our study we analysed the relationships between biological traits of
spontaneous plants found in woodlands and a rural-urban gradient.

Material and methods

The study was undertaken in Angers, a city in western France. The Angers conurbation
has about 270 000 inhabitants. Fifteen woodland plots of about 1 ha were surveyed along a
rural-urban gradient stretching from the city centre to a distance of 10 kilometres. Plots were
comparable in term of their geological substrate, canopy layer (deciduous trees, mainly oak)
and their vegetation was dominantly indigenous (Florgard, 2000, Millard, 2004).
Data were in three tables (Fig. 1): R (environmental variables table), L (species composition
table), Q (biological traits table).
Environmental variables Spontaneous species
(land cover, edge or interior position, occurring in at least 5
disturbances…) quadrats (64 species)

Environmental variables Species

Sampling quadrats

Sampling quadrats

150 quadrats of 30 m²
=10 quadrats in each R L
woody stands, 5 at the
edge, 5 in the interior Hill and Smith
analysis CA
Species

Q Two categories of traits

Biological traits

RLQ 1. traits related to

dispersal ability
MCA 2. traits related to local
environmental variables
Figure 1. RLQ analysis of three tables (R, L, and Q)

Statistical analyses
Firstly, three separate ordinations were performed: a correspondance analysis (CA) on L,
a Hill and Smith analysis on R (Chessel, et al., 2006) and a Multiple Correspondence
Analysis (MCA) on Q (Fig. 1). A three table ordination method (RLQ analysis) was
subsequently carried out in order to link species traits directly with environmental variables
(Dolédec, et al., 1996). RLQ analysis combines the three separate analyses so as to
maximize the co-variation between environmental variables and species traits.

213
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Results and discussion

Three separate ordinations

The first ordination axis of environmental variable (table R) (explaining 59% of variance)
is negatively associated with 'cover of roads and buildings' (r=-0.96 and -0.95 respectively)
and positively with 'land under cultivation' (r=0.97). So it appears that the environment of
different sites is predominantly determined by their position on the rural-urban gradient.
The analysis of vegetation (table L) shows the existence of a gradient with clear
separation of ruderal species as Geum urbanum, Geranium robertianum, Alliaria petiolata
and species more typical of forests such as Deschampsia flexuosa, Pteridium aquilinum and
Teucrium scorodonia (Honnay, et al., 1998).
The structure of biological traits (table Q) is more complex but some groups of species
are distinguishable from other species principally on the basis of morphology and life-span.

RLQ analysis
The first axis of the three-table RLQ analysis explains 80% of the total variance. So we focus
on this first axis for a first exploration of the results. This first axis of the RLQ analysis
explains 90% of the inertia of R. Hence this axis is clearly related to the structure of the rural-
urban gradient. The relationship between species traits and the environmental characteristics
of their habitats is highly significant (p-value of Monte-Carlo permutation test<0.001).
Differences in the distribution of traits along the rural-urban gradient involve morphological,
phenological and physiological traits but not dispersal traits. Urban species are characterised
by a short life-span, rosette growth forms, phenologies largely independent of seasonality
and preferences for high pH and nitrogen levels. These characteristics of urban species are
relatively well known (Hill, et al., 2002) but have rarely been identified within a single habitat
type, in particular as regards forest communities. The lack of a clear relationship between
dispersal traits and landscape structure has often been noted with plants (Murphy and
Lovett-Doust, 2004). One reason for this might be that plants respond to resource quality
gradients which are difficult to analyse at landscape scales; secondly, plants rely on a range
of other agents for dispersal of pollen and seeds.

References
EN.REFLIST

214
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

The role of the landscape on the distribution pattern of animals along a gradient
of urbanisation

S. Croci 1,2, A. Butet 2, P. Clergeau 1

1
INRA SCRIBE, Campus de Beaulieu, Av. Général Leclerc, 35042 Rennes, France.
e-mail: [email protected]
2
CAREN UMR CNRS 6553 ECOBIO Univ. Rennes 1, Campus de Beaulieu, Av. Général
Leclerc, 35042 Rennes, France.

Introduction

The aim of our study was to evaluate the animal community responses to urbanisation.
We focused on animal communities of woodlots to avoid any source of covariation potentially
induced by the presence of different habitats within towns (Sadler et al. 2006). Urbanisation
impacted woodlots environment along a rural-urban gradient especially at two scales
(Clergeau et al. 2006): at the local scale (modifications of the vegetation of woodlots) and at
the landscape scale (modifications of the composition and configuration of the surrounded
landscape). These environmental modifications can affect animal communities differently
according to their dispersion abilities.
Consequently, we investigated the response of three animal communities having different
dispersion abilities to the urbanisation impacts on woodlots, the more common and natural
habitat patch in French towns. Urban impacts on woodlots were considered at three different
scales: local scale (within woodlots), close landscape scale (within 100m radius around
woodlots) and wide landscape scale (within 600m radius around woodlots).

Method

Studied sites
Our study took place in 13 small woodlots distributed along a gradient of urbanisation
extended over 10 km and ranging from the town centre of Rennes (Brittany, France) to its
more rural adjacent landscape.

Environment’s description
Using SIG and Fragstat v3.0, we investigated the landscape structure within 100m (close
landscape) and 600m (wide landscape) radius around each woodlot. In parallel, we
described the woodlot vegetation.

Animals’ surveys
In each woodlot, we carried out a survey of bird, carabid beetle and small mammal
communities during several months in 2004 and 2005. For each taxa, we took into account
variations of species richness, Shannon’s diversity index and Simpson’s dominance index
along the rural-urban gradient.

Statistical analyses
First, using PCA, we studied how the close and wide landscape and the vegetation of
woodlots varied along the gradient. Second, we performed variance partitioning analysis to
identify the environmental scale – local, close or wide- at which each animal community was
the most sensitive. Second, using Spearman rank correlations, we describe the response of
each taxa for its ‘sensitive’ scale.

215
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Main Conclusions

Birds, carabid beetles and small mammals showed different responses to urbanisation.
Birds were sensitive to urbanisation impacts at the local scale. Bird species richness
increases with the richness and the diversity of woodlots vegetation which were higher in
town than in rural landscape. Carabid beetle communities were sensitive to urbanisation
impacts at wide landscape scale. Surrounded landscape of woodlots were more fragmented
and characterized by mineral surfaces in town. This greater isolation of urban woodlots
induced an impoverishment and a diversity decrease of the carabid communities in town.
Nevertheless, about 50% of carabid beetle species were present both in urban and rural
woodlots. Small mammals seemed to be adversely sensitive to urbanisation impacts at the
local scale.
This study illustrated how important is to take into account 1) a single habitat type, 2) the
dispersion abilities of animals and 3) the different environmental scales potentially influencing
them to propose some assumptions on the mechanisms underlying the observed
distributions of animal communities along rural-urban gradient. In addition, we suggest that
urban woodlot is one of the urban habitats more prone to accommodate an important
biodiversity within town.

References
Clergeau, P.; Jokimakï, J. & Snep, R. (2006) Using hierarchical levels for urban ecology. Trends in
Ecology and Evolution 21: 660-661.
Sadler, J.P.; Small, E.C.; Fiszpan, H.; Telfer, M.G. & Niemelä, J. (2006) Investigating environmental
variation and landscape characteristics of an urban-rural gradient using woodland carabid
assemblages. Journal of Biogeography 33: 1126-1138.

216
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Modelling bird species distribution and abundance in urban areas

C.H. Nilon1,and P.S. Warren 2 .

1
Department of Fisheries and Wildlife Sciences, University of Missouri-Columbia, USA.
e-mail: [email protected]
2
Department of Natural Resources Conservation, University of Massachusetts-Amherst,
Amherst, MA 01003, USA

Introduction

As co-investigators on a long-term urban ecology research project in Baltimore, Maryland,

USA we want to understand how changes in the built areas of the city influence bird species
distribution and abundance. We are particularly interested in changes that occur at a lot,
street, and neighborhood scales that are initiated by residents and property owners. Our
goal is to develop models that predict bird species distribution and abundance using
variables from data that are used to measure changes in cities.

Methods

Study area
Our research takes places within the city limits of Baltimore City, Maryland. Baltimore is
has a population of 637,000 residents and its city limits encompass an area of 216 km2. Our
census points are a sample of 80 of the 202 Urban Forest Effect Model (UFORE) points in
Baltimore. The UFORE points represent the range of land uses found in residential census
tracts in the city (Nowak and Crane 2000).

Bird data
Our protocol for conducting bird counts is similar to a procedure developed for the Central
Arizona – Phoenix Long Term Ecological Research Project Grimm and Redman 2004). One
trained observer visited each point between 05:30 and 09:30. The observer recorded the
species and number of all birds seen or heard during a five minute counting period. Three
counts were made at each point between May and July in 2005 and 2006. We averaged the
number of detections on each count for each species and reported an index value of mean
detections / count.

Lot – street- and neighborhood-scale data

Lot-scale variables are measures of bird habitat features that occur within a single
residential lot and are under the control of a resident. We characterized lot-scale variables
using tree cover, shrub cover, and ground cover data that were collected at each bird census
point in 2001 as part of the UFORE analysis of Baltimore (USDA Forest Service
Northeastern Research Station 2005). Street-scale variables reflect bird habitat features that
are controlled by several residents and by local tree management and zoning decisions.
Neighbourhood-scale variables are measures of bird habitat features that are controlled by
local planning, zoning, and development decisions. We used data on building cover, building
density, vegetation cover, tree cover, and tree density compiled by the Baltimore City
Planning Department from 2000 LANDSAT imagery as measures of neighbourhood-scale
habitat features. These variables were measured within a 200 m radius of each bird census
point.

217
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Models
The abundance of seven bird species (Turdus migratorius, Thryothorus ludovicianus,
Sturnus vulgaris, Passer domesticus, Cardinalis cardinalis, Columba livia, Hylocichla
mustelina) are correlated (R2 > 0.25) with axes for a Bray-Curtis ordination of the bird census
points. We are using these seven species as dependent variables in regression models
based on lot, street, and neighbourhood –scale variables. We are using Krieging procedures
and these species and lot and neighbourhood-scale variables to develop spatial models that
predict species abundance values for all 202 UFORE points.

References
Grimm, N.B. & Redman, C.L. (2004) Approaches to the study of urban ecosystems: the case of
Central Arizona-Phoenix. Urban Ecosystems 7:199-213.
Nowak, D.J.& Crane, D.E. (2000) The Urban Forest Effects (UFORE) Model: quantifying urban forest
structure and functions. M Hansen and T. Burke (Eds.) Integrated Tools for Natural Resources
Inventories in the 21st Century. Proceedings of the IUFRO Conference. USDA Forest Service
General Technical Report NC-212. North Central Research Station, St. Paul, MN. pp. 714-720.
USDA Forest Service Northeastern Research Station. (2005) UFORE in action: Baltimore,
Maryland. Retrieved on 31 January 2007, from: http://www.ufore.org/action/02-00.html

218
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Spatial variation of soils and vegetation structure in urban landscapes of the USA

R.V. Pouyat1, I.D. Yesilonis1, W.C Zipperer2, K. Schwarz3, D. Nowak1

1
USDA Forest Service, Northern Research Station, c/o 5 Moon Library, SUNY-ESF, NY
USA,
e-mail: [email protected],
2
USDA Forest Service, Southern Research Station, P.O. Box 110806, Gainesville, FL 32611,
USA, 3Institute of Ecosystem Studies, P.O. Box AB, Millbrook, NY 12545

Introduction

When land is converted from forest, grassland, and farmland to urban land use, novel
management and disturbance regimes are introduced by humans. Most large-scale
disturbances such as site grading and vegetation removal occur in the construction phase of
urban development, while finer scale disturbances generally occur later. Horticultural
management introduces even finer scale disturbances and includes the establishment and
maintenance of lawns, shade trees, and planting beds on parcels of land that typically are
smaller than the parcels that were managed in the previous forest or agricultural landscape.
Horticultural management generally does not result in continuous physical disturbance of soil
or plant communities, so it has less impact on ecological processes than management of
agricultural systems, which continuously disturbs plant and soil systems throughout the year
(or annually). Urban environmental factors that could affect ecological processes overlay
these novel patterns of disturbance and management. These factors include the urban heat
island effect; increased atmospheric concentrations of carbon dioxide (CO2), and oxides of N
and sulfur; atmospheric N deposition, heavy metal, and organic chemical contaminants; and
the introduction of invasive plant and animal species. The net result of human disturbance,
landscape management, and environmental change associated with urbanization is a mosaic
of land patches that will vary in soil and vegetation conditions. In this presentation we will
make spatial comparisons of soil properties and vegetation structure at multiple scales in an
effort to characterize the urban patch mosaic. We use data mostly from the Baltimore
Ecosystem Study, but include comparisons with other cities in the USA.

Results and Discussion

Depending on the soil property measured, measurements in Baltimore at the city scale
showed that trace metals varied more by native parent material then proximity to potential
polluting sources or land use and cover, while Cu, Pb, and Zn were related to automobile
sources at both the city and regional scale regardless of vegetation structure. Vegetation
structure and some soil properties (Ca, K, P, and bulk density) were related to land use and
management regime. For example, land uses dominated by turf cover differed from tree
dominated cover types in having higher soil P and K concentrations and higher bulk
densities. Moreover, residential turf cover had more acidic soils and lower Ca concentrations
than turf cover related to commercial, travel rights-of-way, and institutional land use.
There were over 2.8 million trees > 15 cm in diameter in Baltimore City with more than
half of these trees established as volunteers (i.e., natural regeneration). Roughly a third of
the trees were growing in natural stand conditions, mostly in public park lands. The
percentage of area covered by tree canopies was 23.2%, which is roughly the same as other
cities in the northeastern USA (e.g., Philadelphia and Boston), but also surprisingly similar to
cities located in very different climates (e.g., Oakland, CA, USA, which was originally an oak
savannah cover type). The most abundant non-native species at the city scale was
Ailanthus altissima, which made up 5.2% of the tree population. The tree diversity index of
Baltimore City was 3.52, which is slightly higher than other eastern cities in the USA and
much higher than the range typically found in eastern deciduous forests (1.93-3.09).

219
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

At the neighborhood scale, invasive plant species and soil Pb and Cu contamination were
associated with disturbed areas, such as vacant lots in inner city neighborhoods, while
suburban areas had more formal horticultural landscapes with higher P levels in older aged
turf managed soils. The most abundant tree species in high-density residential
neighborhoods of Baltimore was A. altissima. Also in the suburban areas, organic matter
showed an increasing trend with age but was not statistically significant at P < 0.05. In
Baltimore, the number of tree stems (> 15 cm) per ha was 49 and 74 for high- and medium-
density residential land uses, respectively. Additional relationships between soil properties,
site factors (e.g., proportion of impervious surfaces), and vegetation structure are currently
being investigated and will be reported at the IALE meetings.
These results suggest that both natural and urban factors affect the spatial variation of
vegetation structure and soil properties at the city scale, while at finer scales urban factors
are the most important. For example, at the city scale some soil properties of the native soil
continued to persist. The continued strong relationship of these properties and the native
parent material suggests that the transport of soil within the Baltimore landscape occurred at
short distances. However, the fact that other soil properties did not vary by the native parent
material, such as the variables shown to differ among land-use and cover types (pH, P, K,
and bulk density), suggests that urban factors also had an affect on the spatial pattern of soil
characteristics in Baltimore. Urban factors also played a key role at the neighborhood scale.
For example, differences in heavy metals occurred between recently disturbed lots and grass
covered lots in inner city neighborhoods. This relationship suggests that heavy metal
contaminant sources may be more detectible at the neighborhood scale, e.g., deposition
gradients from roads and the remains of Pb based paints from preexisting structures.

220
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Rethinking urban waters from an ecosystem services perspective

D.G. Gledhill1, P. James 2, D.H. Davies 3

1
Research Institute for the Built and Human Environment, School of Environment and Life
Sciences, Peel Building, University of Salford, Salford, Greater Manchester, , UK.
e-mail: [email protected]
2
Research Institute for the Built and Human Environment, School of Environment and Life
Sciences, Peel Building, University of Salford, Salford, Greater Manchester, UK.
3
Biomedical Research Institute, School of Environment and Life Sciences, Peel Building,
University of Salford, Salford, Greater Manchester, UK.

Introduction

The use of the word green in phrases such as green space and green infrastructure may
inadvertently deflect attention from the blue aspects of the urban environment i.e. the ponds,
lakes, canals and rivers. Water plays an important part in modern cities e.g. sustainable
urban drainage, flood control and, not least, providing an essential aesthetic component to
everyday living and working environments.
Whilst freshwater habitats in rural settings have attracted research attention there is a
dearth of data relating to ponds in urban settings. One aim of this research is to address this
matter by a multidisciplinary evaluation of ponds within the Borough of Halton. Halton is
situated between Manchester and Liverpool in Northwest England. This survey makes a
contribution to addressing issues related to the ecological, social, cultural and economic
value of ponds in an urban setting. The initial focus of this paper lies with the data collected
on the ecological values and ecological services associated with these ponds. The authors
go on to raises issues for consideration as part of a wider, multidisciplinary research agenda
set within the context of sustainable urban development.

Urban pond ecosystems

There are an estimated 450 water bodies below two hectares within the 91 km2 area
occupied by the urban and industrial districts of Halton. Detailed ecological data from 30 of
these ponds surveyed between 2004 and 2006, revealed a total of 119 species of aquatic
invertebrates and 57 species of aquatic macrophytes. The median values per pond were 28
species of invertebrates and 10 plant species. The northern crested newt (Triturus cristatus),
a UK priority species (English Nature, 1997) was found to be breeding in 20% of sites. A
survey of 146 ponds in the City of Cardiff, Wales (Rich, 2000) found a median of 19.1
species of macrophyte per pond, a number significantly higher than the 9.6 species per pond
reported by the National Lowland Pond Survey (Pond Action, 1998). Between 1995 and 1998
the Pond Life Project surveyed 1000 mostly rural sites in the northwest of England (Guest
and Bentley, 1998). Median diversities of 32 species of invertebrates and 22 species of
aquatic macrophyte per pond were recorded. The botanical diversity of Halton’s ponds is
comparable with the national average, but lower than that for either Cardiff or rural northwest
England and compares favourably in terms of invertebrates with the Pond Life Project data.
Extrapolating from the mean area of ponds surveyed in Halton (581m2 ± 256 m2), it can
be estimated that the 450 or so sites cover an area of approximately 26.15 ha. With
approximately 10% of the total area of the Borough covered by the Mersey Estuary (and
major canal arteries), ponds potentially account for 4% of the remaining area. Given that
15% of Halton is open water (ODPM, 2005) then pond ecosystems account for almost a third
of that water area.

221
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Discussion

Ponds and reed beds are designated as habitats of UK conservation importance (English
Nature, 1997). Both the Halton and Cardiff studies illustrate the positive role that ponds have
within the context of urban ecosystems, exhibiting biodiversity similar to rural sites. Urban
habitats are frequently fragmented. Debate exists as to the relative merits of creating
corridors or mosaics of habitats (English Nature, 1997a). Pond habitats are by nature
fragmented, whether urban or rural and many pond species are adapted to living in a mosaic
type landscape. An urban park, therefore, presents little more of a challenge to dispersal
than open farmland. Nature conservation gains are not restricted to aquatic species.
Nationally important species such as the English bluebell (Hyacinthoides non-scripta), song
thrush (Turdus philomelus) and skylark (Alauda arvensis) have all benefited from land
protected from built development and made available for nature conservation as part of
urban drainage schemes.
While significant, biodiversity is not the only function of urban freshwater habitats. Ponds
also provide the capacity to drain a sustainable area of land providing flood regulation and
alleviation of heat island effects at the local scale. Surrounding edge habitats also increase
the amount of permeable service for rain water infiltration and may compensate for the loss
of soft surfaces in domestic front gardens (RHS, 2005). Blue/green infrastructure brings
people often divorced from the countryside into daily contact with nature on their own
doorstep (Miller and Hobbs 2002). This provides many of the social, cultural and, health and
well-being services associated with a greener or bluer environment (Lee and Evans, 2003).
Meeting the needs of future development will require a paradigm shift in current
management practices. Practitioners will have to move beyond single site strategies and
single objectives in favour multifunctional landscapes incorporating nature conservation,
sustainable development and human well being. In order to derive the maximum benefit from
these landscapes questions surrounding the people-nature relationship still need to be
answered. Questions such as; which groups of people utilise urban ponds? How they value
them? And how best can good ecological practice be reconciled with human expectations?
Addressing these questions will require closer engagement with stakeholders whose options
are as yet, not well understood.

References
English Nature (1997) The Urban Mersey Basin Natural Area: A nature conservation profile. English
Nature (now Natural England). Peterborough, England.
English Nature (1997a) A framework for the future: green networks with multiple uses in and around
towns and cities. English Nature, Research Report No 256, Peterborough, England.
Guest, J.P. & Bentley, D. (1998) Critical Pond Biodiversity Survey. Pond Life Project. Liverpool John
Moores University, Liverpool, England. Unpublished survey report.
Lees, S. & Evans, P. (2003) Biodiversity’s contribution to the quality of life. A research report for
English Nature, No 510. Environmental impacts team, English Nature. February 2003.
Miller, J.R. & Hobbs, R.J. (2002) Conservation where people live and work. Conservation Biology 16,
No 2 pp 330-337. April 2002.
ODPM (2005) Generalised land use data (GLUD) statistics for England. Office of the Deputy Prime
Minister, February 2005. London England.
Pond Action (1998) National Lowland Pond Survey 1996. Pond Action Oxford Brookes University,
England.
RHS (2006) Are we parking on our gardens? And do driveways cause floods? Garden Matters, Urban
Series, Royal Horticultural Society, May 2006. London, England.
Rich, T. (2000) A comparison of the ponds in the County of Cardiff with the national statistics from the
Lowland Pond Survey. In Pond Action (2000). Proceedings of the Ponds Conference 1998. Pond
Action, Oxford.

222
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Understanding everyday urban biodiversity – interpreting the ‘intrinsic value’ of

biodiversity through human perception and mutualism

M. Gyllin

Swedish University of Agricultural Sciences,

e-mail: [email protected]

Background

Positioning human beings in relation to biological diversity is a difficult but important task,
particularly considering the ‘everyday biodiversity’ that we more or less subconsciously
perceive and respond to when performing our daily routines. However, it is rather unclear to
what extent this ‘background biodiversity’ is actually perceived by different groups of the
public, to what extent it is believed to be beneficial and to what degree different organisms
can be identified and appreciated by the public, i.e. the perceptive resolution of biodiversity
experience. It could be argued that basic perception is the foundation for human interactions
with other life forms, that this connection is so important that it could probably be considered
a mutualistic relationship, and thus makes the idea that biodiversity has an intrinsic value
less controversial.

The perception of biological diversity

Although biological diversity has been defined ‘officially’ (UNEP, 1992), this definition is too
vague for scientific purposes, and several other definitions have been proposed more or less
ad hoc (Delong, 1996). The actual use in everyday life is a more intuitive practice, involving
values, feelings and professional background (Gyllin, 2004). Thus, what we spontaneously
regard as aspects of biodiversity probably differ very much among individuals. How much
detail we can perceive – the ‘biodiversity resolution’ – also varies with a number of factors,
most importantly perhaps our knowledge about plants, animals and ecological relationships.
In terms of ‘affordances’ (Gibson, 1986, 127-143), the number and quality of affordances
connected to biological objects would be different among individuals. To catch the inner core
of biodiversity as it is empirically experienced and valued by people, it is necessary to
establish in what way and to what extent it is perceived.

The intrinsic value of biodiversity and mutualism involving people

Decisions, planning policies and opinions regarding nature conservation often imply that
biodiversity has an intrinsic value, which is explicitly stated in the convention on biological
diversity (UNEP, 1992). Philosophically, this is difficult to defend (Randall, 1994; Oksanen,
1997), at least in its biocentric version, but to some extent it could be answered by
acknowledging that man is part of an integrated network of mutualistic relationships,
involving a large number of organisms that are somehow interconnected with humans, i.e. a
more holistic, ecocentric approach (Oksanen, 1997; Stenmark, 2000). In other words, what is
beneficial to (some aspects of) biodiversity is beneficial to man and vice versa. Evidence for
the existence of such relationships may be drawn from environmental psychology (Ulrich,
1986; Kaplan et al., 1998), and possibly explained by the biophilia hypothesis (Kellert &
Wilson, 1993). Thus, conservation and development of everyday urban biodiversity could be
motivated through general human well-being issues and the difference between the intrinsic
value and instrumental values of biodiversity would be less distinct.

The peri-urban environment

223
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

The peri-urban environment is an important, but largely underrated, interface between man
and nature. Particularly in densely populated and highly productive agricultural areas, such
as the southernmost parts of Sweden, the peri-urban areas provide some of the most visited
recreation areas, sometimes ephemeral and doomed to be built (Qviström & Saltzman,
2006), but sometimes made permanent owing to their popularity. Since they are often
unplanned and initially not intended for any particular purpose, and because accessible
green areas in such regions are limited, they may be heavily used for a number of different
recreational purposes. Experiencing biodiversity may be an important, however rarely explicit
function, and the peri-urban landscape provides an important refuge to a number of species
that do not fit in either the urban landscape or the sometimes even more hostile agricultural
landscape. It is reasonable to conclude that a substantial proportion of the encounters
between humans and other organisms in the agricultural landscape takes place in the peri-
urban environments.

Methods and expected results

Some of the aspects mentioned above are tested in an ongoing project, involving expert
groups as well as different groups of visitors, among them immigrants, local people, and
visitors. The main aim is to clarify the extents of biodiversity perception, and to relate this to
the subjects’ cultural, professional and social background, gender and age, and to the actual
vegetation structure and species composition regarding plants and animals. The study is
based on semantic questionnaires, simple questions and focus-group interviews. The
subjects are asked to identify as many species as they can, and to identify structural
variation in the surrounding vegetation, in motion as well as standing still at certain points in
the area. The study site is a newly established recreation area, Lake Arrie, an abandoned
quarry that has been used for a multitude of purposes, mostly spontaneous recreation, such
as horseback riding, fishing and bathing, but also more planned activities like paintball and
swimming lectures for small children. The area also has some biological qualities and
potential, and there is an ambition to integrate these qualities with recreation. Results of the
investigation are expected to provide new insights concerning the nature of human
perception and preferences regarding biological diversity, and also to deepen the knowledge
about the empirical meaning of biodiversity as such.

References
Delong, D. C., Jr. 1996. Defining biodiversity. Wildlife Society Bulletin 24(4): 738-749.
Gibson, J. J. 1986. The ecological approach to visual perception. Hillsdale, New Jersey, Lawrence
Erlbaum Associates.
Gyllin, M. 2004. Biological Diversity in Urban Environments - Positions, values and estimation
methods. Department of landscape planning. Alnarp, SLU.
Kaplan, R., Kaplan, S. & Ryan, R. L. 1998. With People in mind: design and management of
everyday nature. Wahington, D.C., Island Press.
Kellert, S. R. & Wilson, E. O., Eds. 1993. The Biophilia Hypothesis. Washington D.C., Island
Press/Shearwater Books.
Oksanen, M. 1997. The moral value of biodiversity. AMBIO 26(8): 541-545.
Qviström, M. & Saltzman, K. 2006. Exploring Landscape Dynamics at the Edge of the City: Spatial
Plans and Everyday Places at the Inner Urban Fringe of Malmö, Sweden. Landscape Research
31(1): 21-41.
Randall, A. 1994. Thinking about the value of biodiversity. Biodiversity and landscapes - A paradox of
humanity. Kim, K. C. & Weaver, R. D. Cambridge, Cambridge University Press: 271-286.
Stenmark, M. 2000. Miljöetik och miljövård. Lund, Studentlitteratur.
Ulrich, R. S. 1986. Human responses to vegetation and landscapes. Landscape and Urban Planning
13: 29-44.
UNEP 1992. Convention on biological diversity, UNEP.

224
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Applicability of metapopulation theory to biodiversity conservation efforts in

Greater Manchester

A.E. Kazmierczak1, P. James2

1
Research Institute for the Built and Human Environment, School of Environment and Life
Sciences, Peel Building, University of Salford, Salford, M5 4WT, UK
e-mail: [email protected]
2
Research Institute for the Built and Human Environment, School of Environment and Life
Sciences, Peel Building, University of Salford, Salford, M5 4WT, UK

Introduction

Natural and semi-nature habitats within urban areas are becoming increasingly
fragmented, isolated, disturbed and homogeneous (McKinney, 2006; Ahern, 1995).
Fragmentation and habitat loss are considered to be major causes of the decline in wildlife
numbers and diversity. Hence, to prevent further decline it is important to preserve and/or
develop areas that are large enough for populations of wildlife species to persist (Jongman,
1995). It is also necessary to recognise that conservation actions have to be taken outside
designated reserves (Walker, 1995) so that possibilities for exchange of individuals, and
hence genes, between sites are maintained and enhanced (Jongman, 1995).
An ecological network is a planning tool that can help maintain both structural and
functional connectivity in the landscape (e.g. Bouwma et al., 2002; Alterra, 2003) as it forms
an interconnected spatial framework of areas of high nature value. An ecological network
provides the physical conditions necessary for ecosystems and species populations to
survive in a human-dominated landscape (Jongman and Pungetti, 2004). One of the
rationales for grouping habitats into an ecological network of high connectivity is to provide
for the specific needs of species forming metapopulations which depend on exchange of
individuals for the survival of individual populations (Jongman, 2004). However, this rational
can be questioned with regards to its applicability in urban areas and with respect to the
observation that increased connectivity in the landscape may be unnecessary (Hobbs, 1988),
ineffective (Henein and Merriam, 1984) or in some cases even detrimental to species not
forming metapopulations (Simberloff et al., 1992). Addressing these issues leads, in turn to
questions about the application of metapopulation theory in urban areas, and about the
feasibility of directing conservation efforts into creating ecological networks in the wider
environment around designated areas.

Metapopulations and Ecological Networks in Greater Manchester – a case study

The authors report on a case study of the metropolitan area of Greater Manchester, UK.
Greater Manchester is a densely populated conurbation (19.4 people per hectare) in the
Northwest of England (National Statistics, 2007; ODPM, 2005). In this conurbation the
identification and enhancement of an ecological network is being championed by the local
planning authorities and promoted by ecological advisory bodies.
Following a critical review of the availability and relevance of data on the distribution and
dispersal behaviour of species of special conservation importance in Greater Manchester the
authors of this paper identify that there is a lack of empirical data relating to the occurrence
of metapopulations for these species. The implications of these findings are considered
within the context of the relevant wildlife conservation legislation and the planning process
currently operational in England. In particular local authorities may not be justified in requiring
a developer to conduct a protected species survey in the absence of a record for a particular
species in a particular place. Further, local authorities also experience difficulties when trying
to design and implement plans for an ecological network in the absence of empirical data.
There is, however, considerable evidence that structure and diversity within the green
space of cities is of greatest importance in determining the species diversity of these areas

225
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

(e.g. Smith et al., 2005; Sandström et al., 2006). The authors provide an analysis of
amount, distribution and ecological quality of publicly and privately owned green space within
Greater Manchester. Combining this analysis with information on the dispersal potential of
the species of conservation importance in Greater Manchester allows the authors to propose
a framework by which the ecological processes can be sustained and the biodiversity can be
maintained in densely populated conurbations.

References
Ahern, J. (1995) Greenways as a planning strategy Landscape and Urban Planning 33 131-155.
Alterra (2003) Ecological areas: linking protected areas with sustainable development Alterra,
Wageningen.
Bouwma, I.M; Jongman, R.H.G. & Butovsky, R.O. (2002) The indicative map of the Pan-European
Ecological Network for central and eastern Europe. Draft version European Centre for Nature
Conservation, Tilburg.
Henein, K. & Merriam, G. (1984) The elements of connectivity where corridor quality is variable.
Landscape Ecology 4 157-170.
Hobbs, E.R. (1988) Species richness of urban forest patches and implications for urban landscape
diversity. Landscape Ecology 1 141-152.
Jongman, R.H.G. (1995) Nature conservation planning in Europe: developing ecological networks.
Landscape and Urban Planning 32 169-183.
Jongman, R.H.G. (2004) Context and concept of ecological networks. IN Jongman, R.H.G.and
Pungett, G. Ecological networks and greenways. Concept, design, implementation Cambridge
University Press, Cambridge 7-33.
Jongman, R.H.G. & Pungetti, G. (2004) Introduction: ecological networks IN Jongman, R.H.G. &
Pungetti, G. Ecological networks and greenways. Concept, design, implementation. Cambridge
University Press, Cambridge 1-6.
McKinney, M.L. (2006) Urbanization as a major cause of biotic homogenization Biological
Conservation 127 247-260.
National Statistics (2007) Census 2007 – Greater Manchester Retrieved on 23 Jan 2007, from:
http://www.statistics.gov.uk/census2001/pyramids/pages/2a.asp.
ODPM (2005) Generalised land use data (GLUD) statistics for England Office of the Deputy Prime
Minister, London.
Sandström, U.G; Angelstam, P. & Mikusiński, G (2006) Ecological diversity of birds in relation to
the structure of urban green space Landscape and Urban Planning 77 39-53
Simberloff, D; Farr, J.A; Cox, J. & Mehlman, D.W. (1992) Movement corridors: conservation
bargains or poor investments? Conservation Biology 6 493-504.
Smith R.M; Gaston K.J; Warren P.H. & Thompson K. (2005) Urban domestic gardens (V):
Relationships between landcover composition, housing and landscape. Landscape Ecology 20
335-253.
Walker, B.H. (1995) Conserving biological diversity through ecosystem resilience. Conservation
Biology 9(4) 747-752.

226
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Modelling, assessing and monitoring urban socio-ecological systems: The new

challenge of shrinkage and perforating cities for urban green and nature
conservation.

D. Haase

UFZ-Centre for Environmental Research Leipzig-Halle, Department for Computational

Landscape Ecology, Permoserstr. 15, D-04318 Leipzig, Germany,
Email: [email protected]

Problem
Demographic transitions and individualisation as well as economic decline produce novel
pattern, densities and modified dynamics of urban land use along the rural-urban gradient.
Compared to the beginning of the 90ies right after the societal transition in East Germany,
today massive shrinkage and vacancies in both the housing and commercial sector, followed
by demolition and perforation, occur. Such processes of urban decline are known from UK
and the US in the 80ies and 90ies but did not face whole regions (most parts of East
Germany) in a tremendous short time interval (less than 15 years). Moreover, shrinkage is at
the same time interfered with splintered growth processes of the late transition phase from
the socialist to the market economy system.

Objectives
In consequence of the above mentioned processes, scientists and planners have to state a
massive surplus of urban brownfields and open space in the core city and at the periphery
which was never expected to become real in such quantity and, what is more, that have to be
“prepared” for the future. Normally, concepts and indicators of urban green aim in increasing
area and quality of greenery up to a target far away. Here, in case of shrinkage, common
urban landscape planning and monitoring systems enter “undiscovered terrain”. A series of
questions were asked, eg., should these areas be developed or to be left for natural
succession? Do such amounts of shrinkage have influence on normative targets of urban
ecological and human recreational values or targets?

Methodology
To find answers on the above asked questions, structural and land use changes related to
demolition and following perforation had been detected and quantified. Here, field mapping,
remote sensing and GIS-based data compilation had been applied. Further, urban waste
land remaining from vacancies and demolition had been analysed concerning their species
distribution and abundance related to (i) age and (ii) substratum of the land.
In consequence of the field work, a multi-criteria indicator matrix had been developed to
quantify environmental and ecological effects of urban open land and greenery for their
environments. Seen from both sides, the ecosystem species and the urban inhabitant,
different scenarios of urban shrinkage and perforation had been evaluated concerning
ecological and QoL related target values determined in the indicator matrix. Finally, the
FLAG multi-criteria evaluation model had been applied to produce a quick-“ample”-view to
select the best scenario depending on a pre-defined objective function.

Results
For the case study of Leipzig, situated in East Germany and representing a “fore-runner” in
terms of massive shrinkage processes and housing vacancies, results for the coming 10
years are presented and discussed. For two test areas, the prefabricated GDR-time housing
areas of “Leipzig-Grünau” and an old built-up district near the city centre “Leipzig-East”
recent demolition processes and planning framework are the base of the 3 scenarios.
In particular, existing target values for urban green per inhabitant and data of urban
biodiversity under growth conditions had been set against the data explored from the
shrinkage and demolition context (Figure 1)

227
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

450
water supply
400
seage disposal
350
district heating
300

costs (€)
250

200

150

100

50

0
stock -10% -20% -30% -40% -50%

demographic decline

Figure 1. Left: Share of public accessible urban green in Leipzig-Grünau. 2002 – Residential
vacancy >25%. 2005 and 2007 describe a demolition rate of 6 and 20% (broken line = target
value of 7.5m2/inh.). Right: Cost development of urban infrastructure under conditions of
shrinkage (modified according to Koziol, 2004).

Conclusions
What did we learn? The example of Leipzig gave evidence that firstly, East German cities
possess a “for-runner role” in urban restructuring processes such as shrinkage and
perforation arising from demographic change, secondly, a surplus of urban greenery offer
considerable positive effects for both man and nature, thirdly, that simultaneously costs of
marinating urban infrastructure has to be set against this positive environmental development
and endanger the development of green perforated districts, an, finally, urban planning is not
prepared to handle “too much” green or open space and therefore it needs methodological
input from science to cover these recent challenges. Last but not least, the results of this
study underline that urban landscape ecologists are able to valuably contribute to the debate
on urban shrinkage, a discussion originally and oftentimes exclusively driven by architects
and planners.

References
Haase D, Nuissl H, 2006: Spatial consequences and environmental impact of long-term land use
change. A case study for Leipzig (Germany), Landscape and Urban Planning (in press).
Haase D, Haase A, Bischoff P, Kabisch S, 2006. Guidelines for the ‘Perfect Inner City’ Discussing
the Appropriateness of Monitoring Approaches for Reurbanisation. Landscape and Urban
Planning (accepted).
Haase D, Seppelt R, Volk M, Lautenbach S, 2006. Landscape consequences of demographic
change – Insights from urban regions in Eastern Germany. Müller F, Zurlini G, Petrosillo I, eds.
Use of landscape sciences for the assessment of environmental security. Springer (in press).
Koziol M, 2004. The consequences of demographic change for municipal infrastructure. DfK 44 (1),
online publication.
Nijkamp P, Ouwersloot H 2003. A decision support system for regional sustainable development.
The FLAG Model. Amsterdam.
Schetke S, 2006. Multikriterielle Bewertung von Freiraumkonzepten und –szenarien in einer
schrumpfenden Stadt. Das Beispiel Leipzig. Diploma thesis, University of Leipzig, Department of
Geography (typescript).

228
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

The New Zealand urban ecology problem and some resolutions

C.D. Meurk

Landcare Research, PO Box 40, Lincoln 7640, New Zealand.

e-mail: [email protected]

Introduction

Basic urban ecology in New Zealand lags behind Europe. New Zealand ecologists have
preferred to study ‘pristine’ mountain, subantarctic and rain forest ecosystems, usually in
national parks. These systems are biased towards inhospitable environments (Kelly and
Park, 1986) remote from the experience of most people. Urban studies are increasingly
reductionist, generic or engaged with stormwater, waste and energy management. But in
New Zealand a healthy physical environment may not equate with biodiversity and so we
become further divorced from our highly endemic natural history (Conservation International,
2007). It must be accessible, identifiable and desirable for it to be relevant to sense of place;
enculturated nature is more likely to be protected (Fig. 1). This is taken for granted in older
densely populated nations where regulation generally prohibits removal of habitat. But an
easy link between indigenous nature and culture is thwarted in post-colonial New Zealand by
the overwhelming dominance of introduced species in cultural landscapes. The
biogeographic context and historic legacy, and the consequences for both our biota and
culture, are documented elsewhere (Meurk and Swaffield, 2000; Meurk and Hall, 2006).

Figure 1. The reinforcing cycle of visibility, familiarity and protectiveness - at least towards
that which is deemed useful (adapted from Meurk and Swaffield, 2000).

Reversing the Trends at a Cost

Conservation in New Zealand is achieved by seemingly endless killings of pests. This is

exacerbated by a culture out of tune with its natural history; that continues to spread
biosecurity risks. Although the calamaties facing our special biota (Wilson, 2004) are
publicised, this is often received with indifference and decision-makers continue to permit
removal of primary habitat. The replacement cost of this folly, even without factoring in the
complexities of mature foodwebs, etc., is ca. € 50 000/ha. Meanwhile, a pond ecosystem

229
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

may be restored in Europe by digging a hole, letting succession take place and translocating
non-vagile species. The same project in New Zealand requires painstaking planting then
fighting back the exotic smothering sward grasses that dominate seed banks. There are thus
severe ecological and cultural barriers to rebuilding sustainable indigenous lowland habitats.

The Role of the Urban Environment

New Zealand cultural landscapes have comparable species richness to National Parks,
but considerably higher numbers of naturalised species (Table 1). The statistics however
hide the important representation of lowland, often poorly protected, species in the cities.
Louv (2005) makes the empirically supported observation that the human relationship with
nature is paramount to a healthy, intelligent, and globally street-wise community. Exposure to
(indigenous) nature is therefore vital to not only understanding ecological processes (for
survival) but also to maintaining points of difference. This has to happen in towns as much as
anywhere. New Zealand is being introduced to this message but resistence to change comes
from influential people with attachment to a colonial past and often an extreme intepretation
of property rights, exacerbated by inadequate compensatory measures. As such, there is a
race to see if a maturing culture, with a sense of its unique place in the world, will evolve
before its underpinning indigenous biota dwindles below the point of no return. Gaining a
deeper appreciation of urban natural history in New Zealand, through information, online
recording of observations, and promoting innovative ways of integrating nature into living
landscapes (Meurk and Hall, 2006), is contributing to biodiverse futures that moreover
support the indigenous Maori people and their customary use aspirations.

Table 1. Summary floristics (±SE) of (all) New Zealand, National Parks or Reserves, and
Cities or Cultural landscapes (adapted from Given and Meurk, 2000)

NZ National Parks Cities

Indigenous Vascular spp 2500 583±76 491±65
Adventive Vascular spp 2500 148±21 475±120
Indigenous spp/000 ha 0.07 7±5.8 6±3.7
Indigenous spp as % of NZ 100 23±3 20±2.6

References
Conservation International (2007) Biodiversity Hotspots, Retrieved on 5th February, 2007, from:
http://www.biodiversityhotspots.org
Given, D.R. & Meurk, C.D. (2000) Biodiversity of the urban environment G.H. Stewart & M.E.
Ignatieva (Eds). Urban biodiversity and ecology as a basis for holistic planning and design.
Proceedings of a workshop held at Lincoln University, 28-29 October, 2000. Lincoln University
International Centre for Nature Conservation Publication 1. Wickliffe Press Ltd, Christchurch, pp.
22-33.
Kelly G. C. & Park G. N. (1986) The New Zealand Protected Natural Areas Programme: A Scientific
Focus, Department of Scientific and Industrial Research, Wellington.
Louv, R. (2005) Last child in the woods: saving our children from nature deficit disorder. Algonquin
Books of Chapel Hill, North Carolina.
Meurk, C.D. & Hall, G.M.J. (2006) Options for enhancing forest biodiversity across New Zealand’s
managed landscapes based on ecosystem modelling and spatial design. New Zealand Journal of
Ecology 30: 131-146.
Meurk, C.D. & Swaffield, S.R. (2000) A landscape ecological framework for indigenous
regeneration in rural New Zealand-Aotearoa. Landscape & Urban Planning 50: 129-144.
Wilson, K-J. (2004) The flight of the huia. Canterbury University Press, Christchurch, New
Zealand.

230
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Social functionality of urban green on the example of Karachi/Pakistan

M.M. Anwar

COMSATS Institute of Information Technology, Department of Environmental Sciences,

Division of Urban Landscape Ecology, University Road, Tobe Camp,
Abbottabad, Pakistan.
e-mail: [email protected]

Problem:
The impact of urbanization on the social function of green open spaces in the mega-cities of
South Asia, with special reference to Karachi/Pakistan. It is well recognized that urban green
space plays an important role in the social and natural sustainability of a city. However,
urban green spaces are one aspect often disregarded in the urban ecological studies. The
increased urbanization, associated with social-environmental degradation, has generated a
debate on what important role urban green spaces can play in social life of urbanities.

Objective:
To seek an answer to this question, this research hypothesized that urbanization and
development within the metropolitan regions in Karachi/Pakistan, are aimed only towards
accelerating economic growth. To test the main hypothesis, the study was carried out with
the following main objectives:
1. A comparison of two regions in Karachi, which have different socio-economic status.
2. Urbanization trend in metropolitan regions lead to loss of urban green space, and an
increase in environmental problems. Therefore, it is necessary to investigate the current
condition of urban green space.
3. The relationship between urban green space and environmental condition was
investigating by evaluated the ecological function of greenery and social functionality.

Methodology:
Beside theoretical linkages and approaches, the case study of a major city, Karachi, in
Pakistan was carried out to find out answers on the above mentioned hypothesis. On the
example of selected regions of the city, the significance and applicability of the above
mentioned questions will be discussed. The main idea of this paper is to show how much
urban green space is important for social needs in the megacity of Karachi/Pakistan.

Expected results:
The expected results can be used for development of urban green spaces for social
functionality for future generation. It is also important for the improvement planning strategies
identified.

231
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Wild in the city: urban lakefill becomes nature’s refuge. A case study of Tommy
Thompson Park, Toronto, Canada.

K. Landman, Y. Cardoso, W. Kehm

University of Guelph, Guelph Ontario Canada N1G 2W1.

e-mail: [email protected]

Redman et al. (2004) argue that scientists can no longer study social systems and
ecological systems as separate activities. Redman et al. (2004) and others (Grimm et al.
2000) argue for an integrative ecology that expressly integrates cultural institutions, policy
and economics. The study of urban ecosystems especially requires knowledge of the social
influences that act on ecosystems, as they generate diverse configurations of development
and land use (Alberti et al. 2003) and manifest obvious and subtle influences on ecosystem
processes both directly and indirectly, or remotely (Pickett et al. 2001). While humans play a
significant role in structuring landscapes, the structure tends to be the focus of ecological
research rather than the human behaviour that produced it (Nassauer 1995). Understanding
the spatial organization of urban ecosystems requires understanding hierarchically scaled
linkages of both spatial ecological heterogeneity and social organizational hierarchies
(Pickett et al. 1997).

Urban wild is the antithesis of the cultivated landscape, the opposite of a sterile green
park (Rink 2005), and is a concept that is gaining attention for studying urban ecology.
Typically, urban wild is a patch of nature set within a greater urban context. This patch may
be a remnant of a degraded ecosystem or a spontaneous, particular expression of conditions
in a brownfield (Urban Wild Group 2004). For this expression to occur, change is the norm,
history is important, and local conditions depend on what happens elsewhere and at other
times (Pulliam and Johnson 2002). To better understand the temporal and spatial linkages,
this research is set within the conceptual framework set out by the Long-Term Ecological
Research Network (Redman et al. 2004), and more specifically employs a single-unit,
longitudinal case study (Francis 2001).

This paper presents a case study of change, both in landscape structure and public
policy, on the Tommy Thompson Park (TTP) lakefill on Toronto’s waterfront. The current
configuration of the TTP extends 5 kilometres from the north shore of Lake Ontario. It is
comprised of a total land base of approximately 160 hectares and a water surface area of
100 hectares composed of embayments and disposal cells. TTP was created of construction
rubble and dredged material over the last half of the 20th century, with an ultimate end-use as
an outer-harbour headland and an urban waterfront airport. Engineers designed the human-
made spit to protect the Toronto harbour in response to increased shipping traffic. It is,
however, this spit formation that has contributed to this industrial site becoming important for
urban wildlife habitat.

Over the past five decades, significant ecological colonization has taken place at TTP,
resulting in an internationally-significant stepping-stone for migratory birds and a concomitant
public demand for a nature reserve on this site. Plans for the airport have long been
scrapped, and the care and management of TTP has been handed to the Toronto Region
Conservation Authority (TRCA). The TRCA is now grappling with wildlife habitat design and
species management, as well as the urban planning issues of a multi-cultural metropolitan
city which contributes to more than 300,000 TTP visitors each year, all of whom embody
conflicting views of nature. While conservation design and wildlife management is the
paramount mandate for the TRCA at TTP, success can only be achieved by integrating an
on-going participatory planning process.

232
Theme 2. Urban environment and transport
2.4 Workshop Current and future research in urban ecology

Wildlife management and urban planning can contribute to the sustainability of urban species
by using the principles of conservation design, maintaining flexibility in response to
unforeseen habitat successes and failures, and offering opportunities for advocacy groups
and lay people to participate along with scientists in shaping policy and practice. The TTP
case study offers place-based, integrative findings that address both societal concerns and
scientific questions about processes that change over the long term, and thereby an
opportunity for comparison with other urban wild sites.

References
Alberti, M.; Marzluff J.M.; Shulenberger, E.; Bradley, G.; Ryan, C. & Zumbrunnen, C. (2003)
Integrating humans into ecology: opportunities and challenges for studying urban ecosystems.
Bioscience 53(12):1169-1179.
Francis, M. (2001) A case study method for landscape architecture. Landscape Journal 20(1):15-29
Grimm, N.B.; Grove, J.M.; Redman, C.L. & Pickett, S.T.A. (2000) Integrated approaches to long-
term studies of urban ecological systems. Bioscience 70:571-84.
Nassauer , J.I. (1995) Placing Nature: Culture and Landscape Ecology Island Press, Washington
D.C.
Pickett, S.T.A.; Cadenasso, M.L.; Grove, J.M.; Nilon, C.H.; Pouyat, R.V.; Zipperer, W.C. &
Costanza, R. (2001) Urban ecological systems: linking terrestrial ecology, physical and
socioeconomic components of metropolitan areas. Annu Rev Ecol Sys 32: 127-157.
Pickett, S.T.A.; Burch, W.R. Jr.; Dalton, S.E; Foresman, T. W.;Grove, J. M.; & Rowntree, R.
(1997) A conceptual framework for the study of human ecosystems in urban areas. Urban
Ecosystems 1: 185-199.
Pulliam, H.R. & Johnson, B. (2002) Ecology’s new paradigm: what does it offer designers and
planners? B. Johnson and K. Hill (Eds). Ecology and design: frameworks for learning. Island
Press, Washington, D.C., London.
Redman, C.L.; Grove, J.M. & Kuby, L.H. (2004) Integrating social science into the long-term
ecological research (LTER) network: social dimensions of ecological change and ecological
dimensions of social change. Ecosystems 7(2): 161-171.
Rink, D. (2005) Surrogate nature or wilderness? Social perceptions and notions of nature in an urban
context. I. Kowarik & S. Korner (Eds). Wild urban woodlands: new perspectives for urban forestry.
Springer, Berlin, New York. pp 67-80.
Urban Wild Group (2004) What is urban wild? Retrieved September 13, 2006.
http://wwwurbanwild.blogspot.com/.

233
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

2.5 Open session 23: urban ecology and greenspace

Shifting the urban landscape paradigm – the ecosystem engineering and design
approach

A. Stokman

Institute for Open Space Planning and Design, Faculty for Architecture and Landscape
Sciences, Hanover University, Herrenhäuser Str. 2a, 30419 Hannover, Germany.
e-mail: [email protected]

Introduction

Due to economic growth combined with more and more refined technologies in the 20th
century humans were able to overcome the restrictions of natural environments. The
destruction of the “wild” nature in favour of a man-made, controlled image of nature has led
to standardized, costly and high-maintenance urban landscapes – environmental damages
related to it are being pushed downstream. Can we afford these urban landscapes in the
future? Challenges of modern urbanization trends and increasing environmental risks require
different forms of planning, design and management of urban landscapes. A discourse
between different disciplines is emerging, exploring on the design of urban landscapes by
linking design, engineering and ecological strategies (Mossop, 2006).

Design versus engineering versus ecology: segregated understanding of urban landscapes

Contemporary landscape and urban design still look at urban landscapes with mainly
aesthetic considerations, constrained by an attachment to the picturesque. The predominant
methods and techniques are based around an architectural paradigm of “final form” or a
static outcome which should be kept by maintenance and prevent wilderness to “invade” and
destroy the design (Yu, 2006). Engineers are designing the vital urban infrastructure – built
forms that manage movement, water, power and waste - judged against the only criterion of
technical efficiency of isolated systems (Picon, 2005). They generate huge structures related
to human settlements and yet do not meet broader human, aesthetic and ecological
objectives. Landscape ecology with its concern for the protection of landscape ecological
functions is closely related to biological sciences (Ahern, 2005) and often lacks an integrated
perspective considering the benefit also for people and economy rather than only nature.
This separation leads to isolated aesthetic, technical and ecological solutions creating
static systems that require a high level of control and maintenance and are unable to adapt
to changing conditions. However society, the environment and nature are complex dynamic,
changing and evolving systems. The underlying processes and overall spatial dimensions of
global simultaneous development trends are very hard to manage, control and foresee.

Challenges of today’s changing and dynamic urban landscapes

The first challenge is related to expanding, low-density urban territories blurring the
difference between the city and the countryside. More and more people follow the trend
towards living in a low-density, attractive green and watery environment outside the city
centres, longing for the presence of a different kind of landscape than what the dense urban
environment can offer. The expensive urban infrastructure has to be extended to serve the
same amount of people distributed in a larger area. The proportion of open space in urban
agglomerations is becoming higher as the cities are maximising their edges. How can these
extensive landscape and infrastructure systems still be afforded to be created and
maintained? How can these landscapes be made accessible for a large number of people
without destroying their ecological landscape functions?
The second challenge is related to shrinking cities. Due to the declining density there is
an over dimensioned technical infrastructure within the cities for too few people needing

234
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

extra high maintenance and also not functioning well because it not used within the limits of
its ideal capacity anymore. Also there is less money to maintain more and more open space
to prevent it from becoming neglected derelict land which makes the urban environment even
less attractive. Can the existing kind of urban landscape and infrastructure really be
maintained to keep the urban environment functioning and attractive in the future?
The third challenge is related to the fast growing and dense megacities where the
infrastructure development can’t keep pace with the rapid process of building construction.
There are technical obstacles of existing technologies being too expensive, not being flexible
enough to adapt to different circumstances and being quite susceptible to disturbances and
failures if they are not constructed, operated and maintained in the right way. At the same
time the sealing of huge areas leads towards big problems of groundwater subsidence and
increasing floods. While the ornamental landscape beautification of cities is increasing very
fast the ecological conditions within the urban environment are deteriorating even further.
What kind of infrastructure is flexible, efficient and feasible in this situation and how can the
urban landscape contribute to the sustainable urban development?

The integrated ecosystem engineering and design approach

Considering these challenges the existing models of urban design, infrastructure and
ecology have to be reassessed and integrated, using a transdisciplinary type of knowledge
production, a “mode 2” approach (Gibbons et. al., 1994). By the examples of projects in
Germany and China a transdisciplinary approach was developed to overcome the distinction
between natural and artificial systems towards integrated hybrid typologies. In these projects
the landscape contains a higher degree of ecological stability, making use of its dynamics for
urban infrastructure functions which require less intervention and technical control while still
offering attractive landscape experiences. These projects apply methods combining both
analytical and logical thinking with intuitive and creative thinking to gain a comprehensive
understanding of the landscape on a regional scale: its structure, organizational and financial
frameworks, natural and socio-economic dynamic processes, elements and functions. The
process of designing becomes at the same time a process of understanding towards a
recognition of locally appropriate, sustainable landscape patterns. This understanding shifts
the urban landscape paradigm: from a landscape that the society can bear the cost for
beautification or ecology towards an understanding where landscape becomes an inevitable
and constituent part of the cities infrastructure.

References
Ahern, Jack (2005): Theories, methods and strategies for sustainable landscape planning. In: From
Landscape Research to Landscape Planning. B. Tress et. al. (Ed.), Springer. pp. 119-131.
Gibbons, M.; Limoges, C.; Nowotny, H.; Schwartzman, S.; Scott, P. & Trow, M. (1994): The New
Production of Knowledge. Sage Publishers, London.
Mossop, E. (2006): Landscapes of Infrastructure. In: Waldheim, Charles (Ed.): The Landscape
Urbanism Reader. Princeton Architectural Press, New York. pp. 164-177.
Picon, A. (2005): Constructing Landscape by Engineering Water. In: Institute for Landscape
Architecture, ETH Zurich (Ed.): Landscape Architecture in Mutation. Gta Verlag, Zurich. pp. 99-
114.
Yu, K. (2006): Position Landscape Architecture. The Art of Survival. China Architecture and Building
Press, Beijing.

235
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

Escaped plants from garden into fallow lands in urbanizing rural area: influence
of local versus landscape factors

A. Marco 1, S. Oliveau 2, T. Dutoit 3, M. Deschamps-Cottin 1, J-F. Mauffrey, V.

Bertaudière-Montes1
1
UMR 151 UP/IRD Laboratoire Population-Environnement-Développement, Université de
Provence Centre Saint-Charles, 3 place Victor Hugo, 13331 Marseille CEDEX 3, France.
e-mail [email protected]
2 UMR 6012 ESPACE, UFR des Sciences géographiques et de l’aménagement, 29 avenue
Robert Shuman, 13621 Aix en Provence CEDEX, France ;
3 UMR 406 INRA-UAPV "Ecologie des Invertébrés", IUT Site Agroparc, Domaine Saint-
Paul, 84914 Avignon CEDEX 9, France ;

Context of the study

Over the last thirty years, rural areas of French Mediterranean countryside have been
subjected to a strong pressure of urbanization (Julien, 1999). Houses have spread,
essentially isolated residential houses and highly crowded housing estates, have reorganized
landscape mosaics in creating new ecological interfaces. The gardens of residential houses,
represent areas of voluntary introduction of native and alien horticultural species which are in
contact with a diversity of habitats (woodlands, fallow lands and fields under cultivation…)
which might be colonized by introduced species. Horticultural species, which are well-
adapted to mediterranean climatic and edaphic constraints, are able to escape, reproduce,
and become established outside gardens (Sukopp, 2004). They participate therefore in the
development of alien flora and are able to modify the floristic diversity and the ecosystem
functioning (biological invasions, hybridization, interspecific relations…) (Vitousek et al.,
1997). Considering the absence of recent anthropic perturbation on fallow land and the
increase of their cover in rural landscapes, following the abandonment of agricultural land
since 1950, fallow land offer suitable sites for the establishment of escaped plants.

Aim of the study

The aim of this work is to understand the mechanisms leading to floristic patterns of alien
species in urbanized landscapes in order to give information to urban planners to preserve
biodiversity in these areas. More precisely, we identified the explanatory factors of the
presence of escaped garden plants in the Mediterranean post-cultural fallow lands and their
relative importance with a spatial analysis on two hierarchical levels: the local structure of
fallow land and its surrounding landscape.

Methods

The study was carried out in the village of Lauris, in south-east France, in the
Mediterranean Basin region. It is located in the Natural Regional Park of Luberon, bordered
in the north by the watershed of the Petit Luberon and in the south by the Durance River.
This rural area consists of 53% of woodland, 37% of agricultural land and 10% of urban land
and is within meso-mediterranean bioclimate. It is in the zone of influence of two big towns
Aix-en-Provence and to a lesser extent of Marseille where urbanization has been spreading
into the surrounding agricultural land and natural countryside since 1975. In thirty years, the
population of Lauris has doubled, rising from 1,620 inhabitants in 1975 to 3,143 inhabitants in
2005.

During June and July 2006, 180 fallow land parcels were visited. In each parcel, we
sampled all the perennial escaped plants from gardens and noted on a scale plan their
abundance, using a reasonably consistent search effort depending on areas of fallow land.

236
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

We assessed the relative importance of 34 predictor variables for alien species richness
in the fallow land sampled. They were divided in three groups: 12 variables related to
gardens (introduction sites of horticultural species), 10 variables related to fallow land
characteristics (establishment site of horticultural species) and 12 variables related to the
surrounding landscape of fallow land. They were collected from the land near homeowners,
or estimated with GIS.

Data analysis

General Linear Modelling (GLM) was used to determine the combinations of independent
variables that best predicted the richness of escaped garden plants.

Results and Discussion

Overall 27 perennial from garden escapes, essentially ornithochorous species, were

found on the fallow land with highly variable abundances. We found that higher alien species
richness was mainly associated (73% of the variation) with 7 predictor variables related to
the introduction site, particularly with proximity and density of gardens around the fallow
parcels, but also establishment depended on the structure of vegetation of the parcel. These
results show therefore that both surrounding landscape and local factors are important
drivers for alien species richness. Firstly, the surrounding landscape determines condition
dispersal process of escaped garden plants. The closest juxtaposition between gardens and
fallow land, but also the highest density of gardens around fallow land, increase the risks to
colonization by escaped plants from a garden to a neighbouring habitat. The dispersal
process occurs mainly over short distance. Secondly, the local factors with the structure of
woody vegetation of fallow land may favour the establishment of ornithochorous species by
offering perching places for birds.
This study thus shows that the hierarchy theory applied in landscape ecology operates in
urbanized landscape and can be used to understand ecological process in these new
environments (Burel and Baudry, 1999).

References
Burel, F. & Baudry, J. 1999. Écologie du paysage. Concepts, méthodes et applications, Paris,
TEC & DOC.
Julien, P. 1999. Au-delà de l’urbanisation, l’étalement urbain caractérise la région, SUD INSEE
l’essentiel 23, 1-4
Sukopp, H. 2004. Human-caused impact on preserved vegetation. Landscape and Urban Planning
68, 347-355.
Vitousek, P.M.; D’Antonio, C.M.; Loope, L.L.; Rejmanek, M. & Westbrooks, R. 1997. Introduced
species: a significant component of human caused global change. New Zealand Journal of
Ecology 21, 1-16.

237
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

Towards strategies for the development of urban green spaces

C. Smaniotto Costa, J. Mathey, B. Edlich

Leibniz Institute of Ecological and Regional Development (IOER), Weberplatz 1,

01217 Dresden / Germany.
e-mail: [email protected]

Introduction

Urban green spaces are crucial elements of all cities. They shape the character and the
image of a city. They provide ecological diversity and form essential structural and functional
elements that make cities more liveable places for their citizens. Thus they assume a key
role of improving the quality of urban life not only because of their ecological functions but
also of their relevance for healthy citizens, societal well-being, economic benefits and the
central role that they perform for sustainable ideals. But frequent deficits in quantity and
quality all over Europe require appropriate management strategies for developing and
improving urban green systems (URGE-Team 2004). Backed on these premises IOER
carries out some international research projects on urban green spaces. Our contribution
presents and discusses the findings from the European research projects “URGE –
Development of Urban Green Spaces to Improve the Quality of Life in Cities and Urban
Regions” and “GreenKeys – Urban Green as a Key for Sustainable Cities” . Within these
projects networks with different European cities were established in order to obtain an
overview on green situation, development strategies and planning frame conditions as well
as to identify successful green development strategies and good practice examples.

Interdisciplinary instruments for analysing the performance of urban green systems

The traditional approach for the evaluation and development of green spaces is often
characterised by a rigid splitting up into sectoral analysis. The different departments of a
city’s administration gather green-related data in their own field of interest, which are later
analysed and evaluated. Based on this type of analysis, strategies for the development of
urban green spaces are often built upon a limited number of criteria related to a single
discipline where, generally, the opinions of other disciplines are not taken into account. This
process results typically in only quantitative targets (e.g. “create more green spaces”) with
little consideration being given to the needs of the local population, economic factors, wider
planning targets or even the qualitative aspects of green spaces (e.g. biodiversity). As a
consequence, the emerging strategies for urban green spaces are not in harmony with the
local circumstances. This effect is further compounded when other aspects of urban
development (housing, infrastructure etc.) displace green space planning and development
(URGE-Team 2004).
To solve this problem URGE developed an interdisciplinary approach, where criteria from the
four disciplines of ecology, economics, sociology and planning were selected and integrated
into a set of instruments called “The Toolbox”. With the application of the tools it is possible
to analyse and evaluate urban green systems and individual green spaces in a more multi-
dimensional way, which can support practitioners in achieving most favourable development
and management strategies for green spaces. The tools included in the Toolbox are: 1) City
Profile, 2) Interdisciplinary Catalogue of Criteria (ICC) and 3) Evaluation Methods. The City
Profile is a questionnaire designed for the compilation of general information about a city’s
urban green conditions. As a starting point for any necessary improvement, it enables the
user (e. g. municipalities, planning authorities) to carry out a self-assessment, to identify
existing strengths and weaknesses in policies and in provision of green spaces. The
Interdisciplinary Catalogue of Criteria (ICC) is a complex tool set which integrates ecological,
economic, social and planning aspects to assist practitioners in the interdisciplinary
assessment, development and management of urban green spaces. It contains criteria and
indicators and operates at two levels: The City Level deals with cities as geographical units

238
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

and aims to support the analysis the structure of green spaces of a city, and the Site Level
focuses on the evaluation of an individual green space. These two levels are not entirely the
same, since the different levels of interpretation require different questions, but they are
complementary. To support the analysis and the interpretation of the data collected by the
application of the ICC the URGE project suggests two evaluation methods: the Polyfunctional
Assessment Method (PFAM) and the FLAG Method. The URGE Manual provides a detailed
description of both methods, as well as their theoretical background.

Challenges for an urban green space strategy

As a result of the cross analysis within the cities
network in URGE Project we can state that one
challenge to be tacked is the low recognition of
green spaces as an indispensable component of
the urban infrastructure. This lack of recognition
is maybe a result of the low awareness about
the benefits offered by a good green space
structure. Green space development in many
cases is not linked with other planning activities.
This affects at the same time the improvement
of the urban social and economic environment.
The challenge means also putting green space
development high in the political agenda. An
Figure 1 The components of the intensive exchange and reflexion of urban
GreenKeys Strategy issues underpin the need for a Strategy for
approach Developing Urban Green Spaces, which shall
set out a collective vision for improved green
spaces, meet community needs and provide a reference point for allocating resources and
plans of action. In many cities there is a lack of a methodology to work out a green space
strategy. A target set by GreenKeys project is to develop a green strategy for all 12
participating cities and based on the experiences made, build and disseminate a pool of
green space strategies.
GreenKeys aims at developing and applying together with the city partners a strategy
approach, this based on a literature review and on the exchange of practical experiences,
methods and measures for initiating and strengthening the strategy development process
with different features and under different conditions (e.g. demographic trends, economic
transformation, land-use pressure, stage of development, structural policy).
This contribution aims at providing an analysis and an appraisal on the current situation on
the development of these strategies. As the project is still running the results represent an
interim situation, which is worth to be close assessed, then the problems and challenges
faced by GreenKeys cities network should be also the same posed to other cities.

References
Greenkeys-Team (2006) 'Guideline for the General Procedure of Developing and Implementing an
Urban Green Space Strategy'. http://www.greenkeys-project.net/media/files.
Smaniotto Costa, C. et al. (2005) 'Good Practice Examples for Green Space Development in
European Cities'. Life in the Urban Landscape. International Conference for Integrating Urban
Knowledge and Practice, Gothenburg, Sweden.
URGE-Team (2004) 'Making Greener Cities – A Practical Guide'. UFZ-Bericht. Nr. 8 (Stadtökologische
Forschungen Nr. 37). Leipzig.

239
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

Landscape Ecological Indicators to Evaluate Urban Greenland

W.-C. Su, C.-Y. Chang

Landscape Architecture Group, Department of Horticulture, National Taiwan University, 138

Keelung Rd. 10673 Taipei, Taiwan
e-mail: [email protected]

Introduction
How do people judge the correctness of the development and improvement of urban
greenlands toward the ecological way? For example, Ong (2003) offers an ecological
measure of the green plot ratio. Although it is proved effective in the evaluation of the
greenery in urban areas, this measure adopts a humanistic ecological perspective.
Land use and land cover changes associated with urbanization significantly affect the
composition of plant communities. Species are not only removed but also introduced (Wu &
David, 2002). In other words, vegetation composition and structure affect the habitat
capability directly (McComb et al., 2002). Based on biodiversity, it is reasonable to assess
the ecological quality through spatial pattern analysis, i.e. the landscape structure (Luck &
Wu, 2002; Kong & Nakagoshi, 2006; Abdullah & Nakagoshi, 2006). This research aims to
develop an indicator to assess the spatial character and ecological quality of a specific urban
greenland.

Method and material

This study was carried out in the eastern part of the main campus of National Taiwan
University (NTU). The analysis of landscape structure was based on aerial photographs. The
landscape elements were classified into seven classes: pavement, buildings, trees, shrubs,
grass, farm, and water with the software eCongnition version 4.0. The scale parameter in
multiresolution segmentation was set as 100 while shape factor as 0.1, compactness 0.9,
and smoothness 0.1.
In ArchView 3.2, the 200 × 200 m2 girds were applied on the digital data to defined the
landscape metrics. To detect the landscape patterns, the metric was computed using
FRAGSTATS (McGarigal et al. 2002). We chose the area-weighted mean shape index
(AWMSI), which was the key point because it contained the area and shape.

Results and discussion

In Figure 1, we learn pavement is the dominant patch in NTU campus, especially in the
main axis Palm Boulevard (grid 17, 18, 19, 20). Besides, it also happens to some of new
buildings with lager paved open space (grid 7, 8, 11, 25, 26). The AWMSI values of
pavement have a negative impact on bird populations (Chiang & Chang, 2006). It makes
sense because the human activities are relative intensive in these locations.
On the other hand, the AWMSI values of trees could be an indicator of bird habitats and
food sources. In the west side, named Black Forest, grid 9 and grid 15 have a better
performance than others. However, the other parts of Black Forest, grid 8 and grid 16, have
low AWMSI values because they are close to parking lot and dormitory.
Not only pavement but also the artificial grasses results in the isolation of patches.
Similarly, the woods and ponds are not connected. Because the farms have educational
function, the usage is very frequent. Therefore, the ecological quality is limited.
Data quality and accuracy assessment are both important points in landscape analysis
(Luck & Wu, 2002). In this study, we applied aerial high-resolution photographs taken in 2004
and detailed classification. Our results should be robust.

240
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

12

10

8 01tree
02shrub
03grass
AWMSI

6 04pavement
05building
06farm
4 07water

0
0 5 10 15 20 25 30 35
grid

Figure 1. Landscape metrics-the AWMSI values of landscape classification in each grid

which is 200 × 200 m2

Conclusion
One of the most ecological functions is to afford wildlife species food and habitats in the
urban environment (Attwell, 2000). However, the isolation of patches limits these functions in
ecific areas. The new development in NTU campus should avoid to destruct the existing
patches and corridors. Furthermore, since it is difficult to add new habitats, the more positive
way is to enhance the connection between patches.
Besides the spatial dimension, to adjust the density and strength of usage is another way
to reduce the impact. These analysis is helpful to assess the ecological quality and make
new strategy to improve the urban greenland.

Reference
Abdullah, S. A. & Nakagoshi, N. (2006) Changes in Landscape Spatial Pattern in the Highly
Developing State of Selangor, Peninsular Malaysia. Landscape and Urban Planning 3: 236-275.
Attwell, K. (2000) Urban Land Resources and Urban Planting - Case Studies from Denmark.
Landscape and Urban Planning 2-3: 145-163.
Chiang, Y. –C. & Chang, C. –Y. (2004) Exploring the Sustainable Environment Based on Landscape
Ecology and Landscape psychophysiology. Landscape Cognition and Preference, The Outdoor
Recreation Association of ROC, Taipei, pp. 153-167. (in Chinese)
Kong, F. & Nakagoshi, N. (2006.) Spatial-temporal Gradient Analysis of Urban Green Spaces in
Jinan, China. Landscape and Urban Planning 3: 147-164.
Luck, M. & Wu, J. (2002) A Gradient Analysis of Urban Landscape Pattern: a Case Study from the
Phoenix Metropolitan Region, Arizona, USA. Landscape Ecology 17: 327–339.
McComb, W. C., McGrath, M. T., Spies, T. A. & Vesely, D. (2002) Models for Mapping Potential
Habitat at Landscape Scales: An Example Using Northern Spotted Owls. Forest Science 2: 203-
216.
McGarigal, K., Cushman, S.A., Neel, M.C. & Ene, E. (2002) FRAGSTATS: Spatial Pattern Analysis
Program for Categorical Maps. Computer Software Program Produced by the Authors at the
University of Massachusetts, Amherst. Available at the following website:
www.umass.edu/landeco/research/fragstats/fragstats.html.
Ong, B. L. (2003) Green Plot Ration: an Ecological Measure for Architecture and Urban Planning.
Landscape and Urban Planning 4: 197-211.
Wu, J. & David, J. L. (2002) A Spatially Explicit Hierarchical Approach to Modeling Complex
Ecological Systems: Theory and Applications. Ecological Modelling 1-2: 7–26.

241
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

Floristic patterns and mediterranean exurban landscapes: Relative importance of

economic geography and landscape structure in functional diversity

E. Dumas1, C. Napoleone1, G. Geniaux1, T. Tatoni2

1
INRA. Ecodéveloppement, Site Agroparc Domaine St Paul, 84914 Avignon France.
e-mail: [email protected]
2
Institut Méditerranéen d’Ecologie et de Paléoécologie (IMEP, CNRS UMR 6116), Université
Paul Cézanne, Europôle de l’Arbois, Bâtiment Villemin, BP 80, F-13545 Aix-en-Provence

Introduction

Due to demographic growth over the last few decades, urban development and its
impacts on wildland areas (diversity loss, fire risk) has become a major concern in developed
communities in Mediterranean France. Especially, during the last decades, urban spread
around the city of Marseille (a large employment pole in the area), has extended over
agricultural and post-agricultural lands. Thus, urbanization and forests (results of abandoned
agricultural grounds) constitute two intermingling systems in interaction, linear and surface
urban configurations around or within forest islands, having recently re-structured the
landscape: Wildland and forested areas in the Mediterranean region are popular and
influence residential location choice. In this context, single houses answer a demand for
space that urban areas cannot offer. Mutually, urbanization adjacent to or within forested
areas modifies the structure and state of forest vegetation by increasing spatial
fragmentation. At a local scale, urbanization and the associated human activities modify the
composition and the structure of forest understorey and could then influence the functioning
of the urbanized forest ecosystems in the long term.
Thus, urban and wildland area relationships drive landscape dynamics and generate
various urban-forest interfaces. We therefore constructed research questions focused on
urban-forest interfaces and hypothesize that forest states can be explained by urban impact
via vegetation analysis at the landscape level. We tested this interdisciplinary landscape
level by comparing plant responses in urban-forest interfaces from ecological landscape and
socio-economic perspectives. This work was able to be made because of the approach of
the landscape proposed by landscape ecology and which answers a demand of economic
geography. With this goal in mind, we carried out floristic surveys in urban-forest interfaces.
We performed multivariate analysis to test the influence of landscape pattern and socio-
economic variables. Moreover, we built a Floristic Competitive Index (FCI), based on plant
functional traits to test which variables may explain the presence of more or less closed
forest (diversity, simplification of strata).

Floristic Competitive Index (FCI)

A Floristic Competitive Index (FCI) was derived from plant traits. Plant species are
appropriate indicators of environmental parameters (Lavorel and Kramer, 1999). Growth-
forms (Raunkier, 1934), adaptative stategies (Grime,2001) and seed dissemination (Muller
and Molinier, 1938) were chosen to describe functional diversity in urban-forest interfaces.
We used these plant functional traits because they are a means by which we can
characterize system function. This conceptual framework assumes that functional groups are
composed of species with shared responses to ecological processes. Every trait was
weighted by a coefficient. The FCI index is defined as follows:

n
FCI = ( ∑ α Cij / R) *√ N
i =1

242
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

where α is the coefficient, Cij is the coefficient of species i at site j, R is species richness and
N the total species richness (298 species). The square root was used to dampen the effects
of species richness (Wilhem and Ladd, 1988). The disturbance index was limited between 0
and 1.
Regressions involving the disturbance were built from socio-economic, ecological and
landscape variable groups. Local and regional levels were used in the model because both
local and regional factors influence ecological processes. Socio-economic variables were
recorded at the municipal level, ecological variables were recorded at local levels and
landscape variables were calculated from the local to landscape levels.

References
Grime J.P. (2001). Plant Strategies, Vegetation Processes, and Ecosystem Properties John Wiley &
Sons Ltd (Eds) Chichester
Lavorel, S. & Kramer, W. (1999). Functional analysis of plant response to disturbance. Journal of
Vegetation Science 10: 603–730.
Molinier, R. & Müller, P. (1938). La dissémination des espèces végétales. Revue Générale de
Botanique 50: 1–178.
Raunkiaer C. (1934). The life-forms of plants and statitical plant geography. Claredon Press (Eds),
Oxford
Wilhem, G. & Ladd, D. (1988). Natural area Assessment in the Chicago region. Pages 361-375 in
Transactions of the 53rd North Marican Wildlife and Natural Resources Conference (Louisville,
Kentucky), Wildlife Managment Institute (Eds), Washington, D.C.

243
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

The use of herbaceous vegetation to promote biodiversity in urban parks

J.C. García-Albarado1, N. Dunnett 2

1
Colegio de Postgraduados - Campus Córdoba. Km. 348 Carr. Fed. Córdoba- Veracruz.
Apdo. Postal 143, C.P. 94500. Córdoba, Veracruz, México.
e-mail: [email protected]
2
University of Sheffield – Department of Landscape. 3rd Floor Arts Tower. Sheffield England.

Introduction

Public parks have great potential in promoting biodiversity in towns and cities.
Traditionally, nature conservationists have concentrated on semi-natural habitats within a
park, and the wilder areas of a site as a focus for biodiversity enhancement. However, the
more intensively used and more ornamental areas also have much potential. This paper
addresses this potential, concentrating particularly on two key issues: public perceptions of
nature-like vegetation in high-visibility locations, and the positive role of non-native species in
ornamental contexts. The paper focuses on a study carried out in a well-used park in the city
of Sheffield, UK that investigated public responses to conventional and naturalistic plantings
at the entrance to the park.

Methodology

Two different plantings were established in two adjacent beds of a similar size in a highly
visible location. Planting 1 was a formal colourful summer bedding scheme of exotic (non-
native) species whilst planting 2 was a sown meadow containing both native and exotic
species (similar to natural vegetation), (Figure 1). The use of exotics for the latter was to
extend the flowering season and increase the structural complexity of the planting, when
mixed with native. 300 park visitors were interviewed about their level of preference, as well
as their likes and dislikes of each planting during summer 2002

Figure 1. Formal colourful summer bedding scheme

(right) and a sown meadow (left) at Endcliffe Park,
Sheffield, UK during summer 2002 in the vicinity of the
study plantings.

The evaluation of the beds was made using a questionnaire in which respondents were
able to point out their level of preference according to a Likert scale as well as their likes and
dislikes. The data obtained were processed using the SPSS version 11.0 software package.

The qualitative responses relating to respondents likes and dislikes were sorted into a
number of response categories. For example, all responses that related to the colour of any
of the plantings were assigned to the variable ‘colour’. Spearman correlations were used to
identify tendencies of respondents. These tendencies are presented in the main findings.

Main findings

According to the results, various characteristics were identified in both plantings to be

key aspects to consider as barriers and opportunities for promoting biodiversity in the more
intensively used areas of public parks: 1) Regardless the type of planting (formal or
informal) the public tends to be enthusiastic about the more creative use of colour of

244
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

flowers. Especially warm colours should be included in the composition of designed

vegetation, a fact that has been also documented by other studies (e. g. Atha, 2003). 2) The
natural, wild and informal appearance of the meadow was also valued not only for its overall
visual effect but also for the opportunity it offers in experiencing a sense of freedom, being
in touch with the natural world, etc. 3) The greater diversity of plant species and their
dynamic processes were also identified as positive characteristics. Its effect on the
promotion of niches is created by the multi layers of vegetation. Meadow type plantings
provide shelter and food for a great variety of insects, birds, and small mammals (Dunnett,
2004). 4) Height of plants was also highlighted by respondents in the meadow. Tall
herbaceous vegetations were usually disliked, and especially when the plants are all tall and
uniform, the preference decreases dramatically. Therefore a good mixture of species of
different heights is likely to enhance people’s preference.

In the case of bedding, the public tends to perceive this style as being too formal, an
aspect that was liked as well as disliked. There were people who tend to like its formality but
also to see it as dull. However, it was found that whilst both plantings were positively
received, a significantly greater number of respondents indicated that they would wish to see
more examples of the meadow in the future, compared with the formal planting type in the
park. There was also a tendency to have the meadow planting on a larger scale and
separated from the formal plantings within the park. It has been suggested by other research
(e.g. Hitchmough, 2004, Nassauer 1995) that the scale and location of ecological plantings
are likely to be very much related to the context of the place.

Finally, this research contradicts much of the perceived wisdom amongst public
landscape managers and suggests that there is a possibility of public acceptance for change
in the appearance and management of public parks to a less formal and more natural
aesthetic appearance. The research informs that both plantings, natural and formal may
have potential in any urban park that fulfils demands of its users.

References
Atha, J. (2003). Public perception of naturalistic herbaceous vegetation in urban environments:
Colour combination and distribution pattern. Unpublished MA dissertation. Department of
Landscape, University of Sheffield, UK.
Dunnett, N. (2004). The Dynamic Nature of Plant Communities. Pattern and Process in designed
plant communities. In: N. Dunnett and J. Hitchmough (Eds.). The Dynamic Landscape. London:
Spon Press. Pp. 97-114.
Hitchmough, J. (2004). Naturalistic herbaceous vegetations for urban landscapes. In: N. Dunnett
and J. Hitchmough (Eds.). The Dynamic Landscape. London: Spon Press. Pp. 130-183.
Nassauer, J. (1995). Messy ecosystems, orderly frames. Landscape Journal, 14 (2): 161-170.

245
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

Effects of land-use intensity and forest patch complexity on avian diversity in an

urbanized tropical island landscape

M. Suárez-Rubio1, J.R. Thomlinson2

1
Conservation and Research Center, National Zoological Park, 1500 Remount Road, Front
Royal, VA 22630, USA.
e-mail: [email protected]
2
California State University, Dominguez Hills, 1000 E. Victoria Street, Carson, California
90747, USA.

Introduction

Human population growth and the expansion and intensification of human land use are
among the most pressing environmental challenges of the 21st century (McKinney 2002).
Urban development in Puerto Rico has an accelerating and widespread pattern and
deforestation is causing the loss and fragmentation of its forests (Ramos-González 2001).
Only 1.2 % of the moist evergreen forests on the island are protected (Helmer et al. 2002),
generally, these forests occur at the lowest elevations where rates of land-cover conversion
to urban areas are highest. Puerto Rico provides a unique opportunity to assess how the
spatial arrangement of urban forest patches and the interior patch structure influence the bird
communities within a range of urbanization intensity.
1.0
P36
P18 P31 P32
P23 P30
P28 P29
0.5 P38 P27
P35 P22 P26
P16
P25 P33
P39 P20
NMS2 VH P24
P37
P21 Patch size
0.0 P34
P7
P14 P17
P19 P40
P4 P10
P13 P9
-0.5 P3
P2
P5 P11
Matrix
P15
-1.0 P1
P8

-1.5
-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
NMS1

Figure 1. Joint plot of NMS scores of forest patches scores with patch size, matrix, and
vertical heterogeneity (VH). The first and second NMS axes represent 44.0% and 22.6% of
the variance, respectively.

Methods

We randomly selected 40 forest patches in the San Juan metropolitan area, Puerto Rico
using IKONOS satellite images from 2002. To characterize the forest patches, we calculated
patch size, patch perimeter, boundary configuration index (using the shoreline development
index), matrix index (percent of constructed land in the matrix surrounding each forest patch),

246
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

isolation coefficient (Thomlinson 1995), and texture analysis (measured as the coefficient of
variation in normalized difference vegetation index (NDVI; Millward and Kraft 2004). In each
of the 40 forest patches, we censused birds thrice: in 2004 and 2005, and we determined the
foliage height profile in the 20 m radius plots centered at each bird census point.

Results

Both landscape and the interior patch structure are influencing the bird communities. Bird
assemblages differed along the urban-suburban gradient in Puerto Rico: some species were
relatively unaffected by urbanization, while several increased in abundance with increased
urbanization and some were sensitive to even minor urban disturbances. It is important to
understand the sensitivity of bird species to habitat degradation in the urban-rural interface,
areas that in Puerto Rico are used by both endemic and neotropical migrant species.
Identifying the importance of forest patches for these groups of birds will aid conservation.
We are also identifying species sensitive to fragmentation.

Table 1. Species identified as indicators of matrix urbanization, and vertical heterogeneity.

All indicator values (IV) were significantly larger than random values based on Monte Carlo
tests (1000 permutations, p < 0.05).
Indicator value

Species by landscape and vegetation variables Observed Random SD

Low-level urbanization matrix

Todus mexicanus 62.2 36.5 6.32

Columba squamosa 58.1 36.3 6.14

High-level urbanization matrix

Zenaida asiatica 56.8 31.3 6.28

Mimus polyglottos 47.7 23.1 6.41

Myiopsitta monachus 40.2 15.8 5.84

Spindalis portoricensis 49.1 26.1 6.5

High vertical heterogeneity

Todus mexicanus 56.9 36.4 6.17

Loxigilla portoricensis 52.5 28.9 6.38

Low vertical heterogeneity

Tiaris bicolor 32.3 18.9 5.9

References
Helmer, E.H.; Ramos, O.; López, T.d.M.; Quiñones, M. & Diaz, W. (2002) Mapping forest type and
land cover of Puerto Rico, a component of the Caribbean biodiversity hotspot. Caribbean Journal
of Science 38: 165–183.
McKinney, M.L. (2002) Urbanization, biodiversity, and conservation. BioScience 52: 883-890.
Millward, A.A. & Kraft, C.E. (2004) Physical influences of landscape on a large-extent ecological
disturbance: the northern North American ice storm of 1998. Landscape Ecology 19: 99-111.
Ramos-González, O.M. (2001) Assessing vegetation and land cover changes in northeastern Puerto
Rico: 1978-1995. Caribbean Journal of Science 37: 95–106.
Thomlinson J.R. (1995) Landscape characteristics associated with active and abandoned Red-
cockaded Woodpecker clusters in east Texas. Wilson Bulletin 107: 603-614.

247
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

The evaluation of the vegetation component of the urban landscape of Milan

V. Ingegnoli, S. Bresciani

Natural Sciences, University of Milan, 26 via Celoria, 20133, Milan (Italy)

e-mail: [email protected]

Introduction and methodological criteria

Ingegnoli (2002, 2006) directs his efforts towards the comprehension of the
landscape and of its main component -the vegetation mosaic- as a proper biological system.
Landscape ecology has been revised according to new scientific paradigms, ranging from
the Principle of Emerging Properties to the ‘order through fluctuation’ processes, proposing
new concepts (e.g. ecocenotope, ecotissue), new functions (e.g. biological and territorial
aspects of vegetation, BTC) and new applications (e.g. evaluation of vegetation, etc.).
The method utilised here, named LaBISV (Landscape Biological-Integrated Survey of
Vegetation), is able to integrate three different criteria (a biotic one, an environmental one
and a configurational one) with different temporal and spatial scales. It uses a parametric
standard form, a proper one for each type of vegetation, for the analysis and evaluation of a
vegetated tessera (Ingegnoli, 2002; Ingegnoli & Giglio, 2005). It helps in the definition of the
so called “normal state” for each specific type of tessera. The most important function
involved in this method is the BTC.
The biological territorial capacity or BTC (Ingegnoli 1991, 2002) is a synthetic
function, referred to the vegetation of an ecocoenotope (i.e. the integration of ecosystem,
community and microchore). It expresses the flux of energy an ecological system must
dissipate during a year to maintain its degree of organization and metastability. It is based
on: (1) the concept of resistance stability (Odum 1971); (2) the principal types of ecosystems
of the ecosphere (Whittaker 1975); (3) their metabolic data (biomass, gross primary
production, respiration, R/PG, R/B) (Duvigneaud 1977, Piussi 1994, Pignatti 1995).
All this improves vegetation science (Ingegnoli, 2005), allowing its capacity to analyze
and evaluate even an urban landscape.

The vegetation of the urban landscape of Milan

At this purpose, we present the case study of Milan and its hinterland. This landscape is
extended 1.103 km2 where Milan (municipality) covers only the 16,5% of this territory, and
vegetation is mainly concentrated in few remnant patches of woods (4,1%), in few urban
parks and gardens (4,2%) and in the remnant rural hedgerows (2-3%).
The vegetation summing only 11% of a so large landscape represents a big
environmental problem for 3.400.000 inhabitants of the “Great Milan”, which environment is
quite degraded, as shown in Table 1. Remember that with the same characters of the
sample average, a good ecological state of an oak forest can easily reach BTC = 8,00.
Moreover, the ratio BTC/BTC* which gives the level of maturity of a forest patch is less than
50%! On the other hand, in the city of Milan (1.300.000 inhabitants) the ecological state is
worst: only 5% of urban green, 1,5% of woods, 1% of rural hedgerows! Let us see the
reported figure 1. The large horizontal line in the figure represents the regional BTC: note
that all the urban parks have lower BTC values (about 1.2-1.5 Mcal/m2/year), while a
remnant patch of natural forest in Cusago reach 7.5. The decrease of the water table under
the town and the increase of the temperature forming a “thermal island” enhance the
degradation of the urban environment.

248
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

Table 1. Some examples of forest tesserae in the hinterland of Milan, demonstrating their low ecological condition.

Forest tesserae Q.T Q.F Q.E Q.U BTC BTC/BTC* H PB

Mcal/m2/yr m m3/ha

A-Quercus rubra, 53 84.9 31.4 32.8 5.33 52.2 29.2 649

Monza Park

B-Oak mixed wood, 43.9 59.1 32.7 49 4.57 44.7 24.4 300

Monza Park

C-QuercusCarpinetus, 53.8 59.1 39.1 52 5.08 49.7 22 280

Monza Park

D- Oak mixed wood, 53.8 53 37.7 36.4 4.53 43.2 20.5 290

Monza Park

E-Quercus-Carpinetus, 61.4 54.6 58.6 57.6 6.41 62.7 23 439

Parco Ticino, Robecco

F- Oak mixed wood 46.2 43.9 47.3 47 4.95 48.4 19.6 315

Parco Ticino, Robecco

G-Robinietum 18.9 53 35 47 3.48 34.1 9.5 120

B.Riazzolo, Cisliano

Sample average 4.9 48 21.2 300

BTC/BTC* = ratio to the threshold of maturity of a forested tessera;

QT= tessera quqlity; OF=quality of the phytomass; QE= quality of an ecocoenotope (sensu Ingegnoli 2002); QU= quality of landscape unit characters; PB =
plant biomass volume.

References
Duvigneaud P. (1977), Ecologia. in Enciclopedia del Novecento. Vol. II, Enciclopedia Italiana
Treccani, Roma.
Ingegnoli, V. (1993) Fondamenti di Ecologia del Paesaggio. Cittàstudi. Milano.
Ingegnoli, V. (2002) Landscape Ecology: A Widening Foundation. Springer. Berlin, New York.
Ingegnoli, V. & Giglio E. (2005) Ecologia del Paesaggio: Manuale per conservare, gestire e
pianificare l’ambiente. Sistemi Editoriali Se. Napoli.
Pignatti S. (1995) Ecologia vegetale. UTET, Torino
Piussi P. (1994) Selvicoltura generale. Torino, Ute

249
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

Managing feral exotic pets: decision-making process revisited

V. Servais1, P. Teillac-Deschamps2, R. Lorrillière2, V. Delmas2, A.C. Prévot-Julliard2

1
Département des Sciences et de la Communication. Université de Liège. Place du 20 Août,
Bât. A1. 4000 Liège. Belgium;
e-mail: [email protected]
2
UMR 8079 Laboratoire Ecologie, Systématique et Evolution. Univ Paris-Sud bat 362. 91405
Orsay Cedex. France.
mailto:
Exotic pets released in the wild occupy a particular position among human-driven
introductions: They are mostly introduced in urban areas, where human densities are the
highest and where habitats are no more natural. Moreover, in addition to being potential
invaders, these species are well known and appreciated by the general public. These
contradictory arguments may lead to potential conflicts between stakeholders in decision-
making process concerning management actions.
Slider turtles Trachemys scripta elegans have been introduced in urban wetlands since
1980’s, after massive pet trade from United States. They are notably present in managed
urban parks, where people use to go for walking or recreation goals. Despite naturalists’
fears and conviction that slider turtles have a great impact on natural communities, no
scientific studies supported this assertion.
As a part of a general project on slider turtle in Paris area (France), we studied the impact
of this exotic species on natural pond communities and human perception of slider turtles in
the urban context. In this talk, we propose a way to mix these biological and anthropo-social
questions through companion modelling and multi-agent system. We suggest how this
research process can be used as a guideline for decision-making, in order to integrate all
points of view and sensibilities, in other conservation programs in urban areas.
We studied the impact of slider on aquatic communities; through an experimental
protocol we weekly surveyed water composition and several compartments of the aquatic
communities of 12 semi-natural ponds (algae, vascular plants, zooplankton, arthropods,
gastropods and amphibians). These ponds contained 0, 1 or 3 adult female turtle
respectively. Preliminary results indicate that turtles impact the functioning of local
communities, but without decreasing neither species richness nor species abundance
(Teillac-Deschamp et al. in prep).
When asking the general public about their feelings on turtles, many people who spent
time hiking commented that they liked seeing turtles during their walks. For some urban
people, this exotic species is one of the few representations of nature to which they are
exposed in urban parks. This non-intuitive result is contrary to the conservationist’s beliefs
that every exotic species is a problem.
Finally, we studied the decision-making process of any turtle private owner wanting to get
rid of the pet. We found that the decision (releasing ii in nature or in recovery centre) strongly
depends on the general environmental perception of the owner.
Managers of public areas have generally both conservation and education aims.
However, the education tools they propose are only efficient for volunteers that are already
environmentally concerned. Nevertheless, many citizens go to urban parks for other reasons
than just nature-based and are not receptive to education messages. These users maybe
attracted to urban parks for aesthetic reasons, including the presence of exotic species in
these areas. In this context, the presence of slider turtles in urban parks could be associated
with a higher number of visitors and thus of a higher potential impact of nature-based
education tools developed by managers.
These results show that slider turtles present in urban parks should be considered not
only as a (potential) biological problem, but also as an element of a social-ecological urban
system. Our research findings will thus be used to propose management strategies for Slider

250
Theme 2. Urban environment and transport
2.5 Open session 23: urban ecology and greenspace

Turtles in urban wetlands. In particular, we ask whether it is worth systematically removing

turtles, considering the balance between impact, spreading potentialities and public
The fact that exotic species could attract people towards nature should be taken into
account in decision-making process, in the same way that the potential impacts of this
species on environment should also be considered.

Acknowledgements
This program is funded by the Regional Council Ile de France, the Direction de
l'Environnement Ile de France, the General Council Seine et Marne, the General Council
Essonne, the General Council Hauts de Seine, the General Council Seine Saint Denis. The
collaboration between the laboratory ESE (University Paris-Sud, France) and the Dept of
Communication (University of Liege, Belgium), is founded by CNRS and FNRS/CGRI (n° 18
220).

251
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

2.6 Open Session 24: The urban Landscape

The Crisis And Rehabilitation Opportunities Of Urbanization In Urban Fringe Rural

Settlements－A Case Study Of Taoyuan Region, Taiwan

M. Kuo

Chinese Culture University, 2F., No.1-8, Jinsi St., Jhongshan District, Taipei City 104,
Taiwan (R.O.C.).
e-mail:[email protected]

Recently the rapid growth of global population and its concentration in limited urban
areas have resulted in the changes of land use patterns and the fuzzy boundaries of urban
and rural areas. Regional economy, ecology, and culture affected by rural changes are
complex and diversified. The factors affecting the aforementioned rural agricultural
settlements are also connected to the global environment, politics, economics, and societal
changes.

In terms of large-scale national land development, the aforementioned changes can be

identified. However, due to the diversity and variation of these factors, the feedback of
changes to the intrinsic quality and formation of landscape ecological structures on a regional
scale is related to the value of resource changes, land capacity, and management policies
towards the environmental and ecological rehabilitation.

Taiwan is an island with 58.8% of high slope mountain areas covered by forests. The
flatlands are limited. The rapid population growth, 77% of the population concentrated in
urban areas (Yang, 2005), industrialization, and urbanization in the past 50 years have
significant impact on the ecological structure of lowland agricultural landscape. The urban
sprawl destructed agricultural wetlands and threatens river corridors. In addition, the
globalization in the last 20 years has resulted in a vast transformation in the agricultural
landscape. The internal structure of ecological systems in the existing farmland also faces
dramatic changes.

The Taoyuan Tableland area is the most distinct example of urbanization in Taiwan
farming settlements. It is located in the plains of northwestern Taiwan with only 40km from
the Taipei metropolitan area. Prior to 1949, the economic focus has been on the agriculture.
However, after 30 years of industrialization and the construction of freeways and
international airport, since 1980 the area economy has been rapidly converted to high-tech
industries and recreational tourism. Additionally, with the completion of the Taiwan High
Speed Railway system in 2006, it begins a fourth wave of revolutionary change in city and
country spatial development.

The Taoyuan Tableland possess a globally unique irrigation pond and canal system
which is an important lifeline of agricultural development in the past. In the recent years, the
irrigation ponds have been reduced from over 10,000 to 3,345, while the irrigation canal
system has been expanded to 2,184 km. The ecological habitat and water resource
conservation along with the rural landscape structure play an important role in developing the
cultural landscape in the area. The value of cultural landscape has reached the level of the
conservation of world cultural heritage.
This paper utilizes satellite images and GIS to explore the effect of urbanization on
special massive irrigation pond systems, the growth and decline of irrigation canal networks,
and the overall landscape ecological structure in the period of past 50 years. The landscape
ecology theory will be applied as a basic platform to examine past studies and collected data,
and utilize the spatial change history to explore and analyze the current states of regional

252
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

scale landscape ecology changes, to analyze the factors and driving forces behind the
landscape ecology change of the Taoyuan Tableland, to understand the pressure and impact
of urbanization on the life-support system, ecological system, and rural settlement ecological
system. Finally, the result of the aforementioned analysis will be used to examine the
responses and strategies of national land management policy implementation towards the
repair of the Taoyuan Tableland landscape ecology system.

Figure 1. Landscape ecological structure of current Taoyuan rural settlements, Taiwan

253
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

Investigating the relationship between the tree cover connectivity of the

metropolitan regions across the eastern United States and the number of
breeding occurrence of the three forest bird genera

S. Kato

University of Massachusetts/Amherst, Department of Landscape Architecture and Regional

Planning, 109 Hills North, University of Massachusetts, Amherst, MA 01003, USA.
e-mail: [email protected]

Introduction

In urban areas, even though greenspaces are often small and fragmented, they have a
potential to support important ecological as well as social functions. Thus, strategic planning
of urban greenspaces is important for those—human and non-human species—that benefit
from their services. This study focuses on the spatial configuration, especially connectivity, of
urban/suburban greenspaces and its effect on the number of occurrence of certain forest bird
species that could benefit from the increase in tree cover connectivity. By using certain bird
species that can act as the indicators of ecological health, the study hopes to claim that their
presence suggests an environment that can support related ecological functions.

Objective and Research Questions

The objective of the study is to investigate the relationship between the degree of
connectivity of tree cover of major metropolitan regions across the eastern United States and
the number of occurrence of breeding oriole species (Icterus spp.), tanager species (Piranga
spp.), and thrush species (Hylocichla spp.). Specific questions asked include: Can structural
connectivity metrics predict the target breeding bird abundance/occurrence in metropolitan
regions? Can they be surrogates for functional connectivity metrics? What is the range of
connectivity that best supports the breeding birds? Is there a generalizable relationship
between the appropriate degree of tree cover connectivity and the breeding bird
abundance/occurrence across different geographic/climatic regions?

Study Area and Land Cover Data

The study area will be the Eastern United States (i.e., the states east of the Mississippi
River). The study area contains the New York-Washington, D.C. megalopolis corridor—the
most densely populated region in the United States. Metropolitan Statistical Areas (MSAs) as
defined by the U.S. Office of Management and Budget will be used as the metropolitan
regions of interest (i.e., my samples). A MSA consists of one or more core urban area with
the population of at least 50,000 and neighboring areas with strong economic and social ties
to the core(s). Within the study area, I will compare the metropolitan regions with varying tree
cover connectivity. Land cover data will be acquired from the National Land Cover Database
2001, which includes other potentially useful data including a 30 m DEM, slope, aspect and a
positional index, and per-pixel estimates of percent imperviousness and percent tree canopy.
Per-pixel (30 x 30 m) estimate of percent tree canopy is of primary interest.

Bird Data

Three genera of bird species (i.e., Icterus spp., Piranga spp., and Hylocichla spp.) will be
chosen as target genera since they have a wide distribution across the eastern United States
and they are neither too rare nor to common in urban areas. The most important criterion of
choosing these genera is that they are forest birds expected to respond to changes in forest
structure (both composition and configuration) in urban areas. My assumption is that they

254
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

would increase in number if tree covers and their connectivity increase. Tanager species are
reported to be a good indicator of forest fragmentation (Rosenberg et al., 1999).
The bird data will be acquired from the Breeding Bird Atlas. It provides a listing of species
known to be breeding or suspected of breeding in each survey block at the time of the
survey. Each state is divided into 5 x 5 km survey blocks within which trained volunteers
count the evidence of breeding (“possible,” “probable,” or “confirmed”). To account for yearly
fluctuations in breeding occurrence, the number of breeding occurrence for each species will
be averaged over the past five years (2001-2005) and the average will be used as the
response variable.

Variables and Data Analysis

The number of breeding occurrence of the three bird genera is a response variable. The
degree of tree cover connectivity in each metropolitan region serves as a predictor variable.
A regression analysis will be conducted. Other predictor variables that may be included in a
stepwise regression model are: elevation, average annual temperature, and average annual
precipitation. The Akaike Information Criterion will be used to select the best fitting model
(Akaike, 1974).

Merit of the Study and Expected Results

It is important to characterize landscape heterogeneity in a way that is relevant to how

forest birds respond to the increase/decrease in tree cover connectivity. By using a
continuous explanatory variable (i.e., percent tree canopy), the study intends to capture the
effect of small patches and/or rows of trees that may provide a valuable food source and/or
serve as rest stops for forest birds to increase in number in urban settings. The traditional
patch-mosaic model (Forman, 1995) is not suited to represent these small patches of trees at
a coarse resolution. Since I hypothesize that forest birds would respond to these small forest
patches, there is a merit of using continuous explanatory variables that can relate to
continuous response variables such as organism abundance and occurrence (McGarigal and
Cushman, 2005).
The results will have implications for planning an appropriate degree of tree cover
connectivity in metropolitan regions to support breeding forest birds. The findings would lead
to the planning of urban greenspaces to support ecological as well as social functions. To
plan future urban greenspaces, it is crucial to consider their interrelatedness: both horizontal
(lateral) and hierarchical connections. The study will shed light on the important
environmental variables to predict the occurrence of the forest birds and suggest an
appropriate degree of tree cover connectivity in the metropolitan regions in the eastern
United States to sustain the breeding of these forest birds.

References
Akaike, H. (1974) A new look at the statistical model identification. IEEE Transactions on Automatic
Control 19: 716-723.
Forman, R.T.T. (1995) Land Mosaics: the ecology of landscapes and regions. Cambridge University
Press, Cambridge, UK.
McGarigal, K. & Cushman, S.A. (2005) The gradient concept of landscape structure J.A. Wiens &
M.R. Moss (Eds). Issues and Perspectives in Landscape Ecology. Cambridge University Press,
Cambridge, UK, pp. 112-119.
Rosenberg, K.V.; Lowe, J.D. & Dhondt, A.A. (1999) Effects of forest fragmentation on breeding
tanagers: a continental perspective. Conservation Biology 13: 568-583.

255
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

The spatial relation between “urban green” and “surface sealing” in semi-urban
areas

S. Meeus, H. Gulinck

Division of Forest, Nature and Landscape Research

K.U.Leuven, Department of Land Management and Economics
Celestijnenlaan 200E, 3001 Leuven, Belgium
e-mail: [email protected]

Urbanization is generally considered as a process of soil-sealing, fragmentation and

reduction of green elements and biomass. This presentation focuses on the spatial
interaction between the process of “hard” urbanization and the emergence of new green
elements and structures such as gardens and parks, green verges etc.

Measuring simultaneously urban green and surface sealing and investigation their spatial co-
occurrence can lead to the definition of indicators that not only help to deepen the insight in
landscape-ecological conditions of urbanization, but also to help in giving more land cover
and structural insight in urbanizing areas, which may eventually stabilize in semi-urban
landscapes. Semi-urban landscapes, however difficult to define or demarcate, deserve much
more attention as specific landscape settings next to rural, urban or natural.

Using various statistical techniques such as multivariate analysis, clustering and non
parametrical statistics, the relationships between (geographical) location of the site where the
indicators are measured and the distance to urban cores, as well as the effect of the cultural
landscape and scale effects are analyzed for a study area (Mechelen) located in the centre
of Flanders and two smaller study areas in more rural parts of Flanders.

For measurement of the properties of semi urban areas, a conceptual model based on a
possible range of combinations of properties of three main landcover / landuse classes
(sealed surfaces, urban green structures and agricultural land) was designed. This model
was successfully used to define “urban cores” in the semi urban area, using class area per
sample, these cores being different from dense city cores. The results were used for radial
proximity analysis.

The properties uses in the conceptual model are based on landscape metrics (as in the
FragStats applications), such as Class Area, Patch Size, dynamics and diversity measures.

As source data a very detailed landuse/landcover raster was used with resolution one meter,
based on the topographic Land Use maps on scale 1:10000 and compared, where
necessary with orthographic photos taken between 1990 an 2005.

The upper and lower boundaries for these specific properties for sealed surfaces, urban
green structures and agricultural land were statistically determined with a random sample of
sample areas covering the targeted population of ‘landscape type’ inside an assumed semi
urban area – for example “urban cores” or “hybrid agriculture”. For every sample area, 50
different extents were used and compared statistically, resulting in an “optimal extent” for the
sample size – 4 hectares in case of the urban cores analysis for example.

Combining the constraints for a landscape metric, for a given landuse/landcover class and
for the specific landscape type analysed, the multidimensional space of all possible
combinations of this property for the three landuse/landcover classes can be determined.

256
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

This conceptual model (recently called the

ASU model) can be expanded with other
(main) landcover classes, proven to be
found in semi urban areas, such as forest
fragments or water bodies. In addition to
this, a wide range of properties, patch and
landscape metrics, can be included in this
model.

Repeating the analysis as described

above, but for different properties, show
that for a big number of metrics, these
combinations of their values for different
landuse/landcover classes can be
investigated and used to describe typical
landscape configuration as found in semi
urban areas.

Combining these results in a proximity analysis (with for example the distance to the main
city in the center of the study area as key measure, the proximity of other urban cores, the
density of rural activity, demographic features or other ‘global’ properties), the dynamics of
the semi urban area can be investigated.

Results show that there are specific relationships between urban green and surface sealing,
that the location of the sample sites in relation to urban cores exist but are sometimes less
significant than the influence of the underlying older rural substrate – even in denser semi-
urban areas – and that surface sealing, as well as urban green are very dynamic variables or
indicators in the semi urban area.

Results will be presented for several semi urban areas in Flanders, a region know for a high
degree and diversity of semi urban areas.

257
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

Land-cover classification of Chongqing City and its surroundings using Landsat

data

Y. Zhao1, M. Tomita1, K. Hara1, M. Fujihara2, Y. Yang3, L. Da4

1
Department of Environmental Information, Tokyo University of Information Sciences, 1200-
2 Yato-cho, Wakaba-ku, Chiba 265-8501, Japan,
e-mail:[email protected]
2
Awaji Landscape Planning & Horticulture Academy/Institute of Natural & Environmental
Sciences, University of Hyogo, Japan.
3
Key Lab of Three Gorges Reservoir Region’s Eco-Environment (Chongqing
University) Ministry of Education, Chongqing / Faculty of Urban Construction and
Environmental Engineering, Chongqing University, China.
4
Department of Environmental Sciences, East China Normal University, China

Introduction

Chongqing is a mega-city located in the southwest part of China along the Yangtze
River. Numerous endangered species are found in and around the city, and the region
includes rich water and biological resources. Since 1997, however, Chongqing is being
redeveloped as the nucleus of an economic program designed to rejuvenate the inland
region, which includes the Three Gorges Dam. This development is suspected of having a
strong impact on the local biodiversity and ecosystems.
Remote-sensing is an effective tool for understanding landscape structure over a wide
scale, as well as changes in spatial patterns over time. This study utilizes Landsat ETM+
2001 data to establish a land cover classification for evaluating ecological conditions and
trends in and around Chongqing.

Study Area

Chongqing Municipality is situated on the upper reaches of the Yangtze River. The total
area of the municipality is 82,300km2, and the population 31 million. Chongqing has a
subtropical humid monsoon climate with four distinct seasons. Annual average temperature
is 18.4 centigrade and annual precipitation 1182.1mm. The local topography features a river
basin surrounded by mountains and hills, with elevations ranging from 150~3,000m. Rural
villages and mountainous areas account for 80% or more of the municipality.

Methods

The research employed Landsat ETM+2001/5/22 data, Path 128 & Row 39, obtained
from the Global Land Cover Facility of the University of Maryland. Eighty photographs taken
in August 2006 using a digital camera with GPSMAP 60CS GPS, were utilized to establish
ground truth. Based on this ground truth data, 7 classes of land cover were established, and
classification was implemented using the Maximum Likelihood Method by ERDAS IMAGINE
(Leica Geosystems GIS & Mapping, LLC).

Results and Discussion

The results of the land cover classification are shown in Figure.1 and Table 1. As can
be seen, cultivated (non-irrigated fields) land (35.2%) and rice paddy (13.4%) together
accounted for almost half of the total study area, showing that the land resources of
Chongqing are heavily used for agriculture. Cultivated land, as well as forest (18.3%), scrub
(24.2%) were concentrated at higher altitudes. Rice paddies were found mostly in the
valleys, but in some cases terraced paddies were built into the slopes. This land-use pattern

258
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

of closely following the local topography was evident even from the field observations, and in
this sense resembles the traditional land-use patterns in many areas of Japan. In contrast to
the agricultural valleys and hillsides, the level areas in and around the city, was mostly urban,
with development reaching as far as the foot of the mountains. The combined area of forest
and scrub accounted for 42.5% of the total. It is possible, however, that this substantial area
may at least in part be the result of a national policy for converting agricultural land to forest.
Stretches of forest along the higher slopes formed a continuous habitat corridor, which can
be expected to play an important role in conserving biodiversity in the study area.

People’s Republic of China

Chongqing

Figure 1. Results of land cover classification

Table 1. Results of land cover classification

Land cover class (%)

Bare land 1.5
Cultivated land 35.2
Forest area 18.3
Paddy area 13.4
Scrub area 24.2
Urban area 5.0
References Water area 2.4
M. A. Alrababah & M. N. Alhamad (2005) Land use/cover
of arid and semi-arid Mediterranean landscapes using Landsat ETM. International Journal of
Remote Sensing 27: 2703-2718.

259
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

Linking physical landscape character with open space aesthetics

Z. Wang, J.I. Nassauer, D. Brown

School of Natural Resources and Environment, University of Michigan.1550 Dana Building,

440 Church Street, Ann Arbor, MI 48109-1041,USA.
e-mail: [email protected]

Introduction

Ecological effects of exurban development are a growing global concern(Liu et al. 2003),
and American landscapes epitomize residential sprawl(Burchell et al. 2002). Aesthetic
preferences are one driver of sprawl in America, and preference for living near open space
has been identified as a driving factor for people moving to exurban areas (Kaplan and
Austin 2004; Vogt and Marans 2004). Yet, both the ecological health of exurban areas and
the long term availability of open space experiences there depend on planning to maintain
open space patterns even as more exurban development occurs. To help managers
incorporate both open space aesthetics and ecological concerns into planning, we
investigated the link between exurban residents’ preferences for open spaces and the
physical landscape character of their surroundings, as measured from mapped data.

Methodology

Our data about open space preferences were drawn from our web survey of 494 exurban
home owners in southeast Michigan. The web survey was conducted for a large research
project about spatial land use and ecological effects at the rural-urban interface. In this
paper, only open space preference data were drawn from the survey. Open space
preference was measured on a seven-point Likkert scale by asking respondents to indicate
how much they would value open space directly adjacent to their property.
To investigate the links between open space preferences and physical landscape
character, three aspects of nearby physical landscape character were explored. For
landscapes directly adjacent to survey respondents homes, we measured landscape
composition (the amount of thirteen landscape cover types), the shape of each landscape
cover type (core area index, shape index), and overall landscape configuration (core area
index, landscape evenness, landscape shape index). These indices were selected as they
have been indicated as relevant to individual perception or individual experience of their
surrounding environment from literature(Kaplan and Kaplan 1989; Palmer 2004; Zube et al.
1974). Home addresses of respondents were physically located in a GIS data layer. 468
respondents were used in this study since some of them did not provide valid addresses.
Physical landscape character was generated from 2001 landcover data at a 30-meter
resolution (State of Michigan). The amount of each landscape type adjacent to each home
address was generated from a circle with a radius of 400 meter and a center at each
respondent’s home location. 400 meter was used because of its acceptance in planning
literature as a comfortable walking distance (Atash 1994; Calthorpe 1989).
Statistical relationships between open space preference and the three types of landscape
character were examined using SPSS 13.0. Descriptive statistics and correlations were the
primary techniques executed to investigate the relationships.

Results and discussion

The respondent sample (n=468) reflects US Census of population descriptions of our

study area in most ways, including age, income, education and number of kids. In the study,
our respondents have tended to highly value open space adjacent to their homes with a
mean rating of 6.1(Min=1; Max=7).

260
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

Our research found that people living in areas dominated by urban built-up landcover
have significantly lower preferences for open space adjacent their home than do those living
in less built up areas. Those who living nearby deciduous forest, agricultural land or shrub
land have significantly higher preferences for nearby open space. However, coniferous
woodland within walking distance is linked negatively with exurbanites’ open space
preference, and those living near water do not have significantly different preferences than
others for adjacent open space.
Configuration of each surrounding landcover type strongly affected open space
preference. Larger core area and more complex shapes of agriculture, deciduous forests and
shrub link to significantly higher open space preference. For wetland and parks, their
complex shapes have significantly positive correlation with open space preference but not
their core areas. Cluster of roads and high density residential development have significant
negative relationships with people’s evaluation of their adjacent open spaces. Landscape
characterized larger patches of low density residential development had significant positive
correlations with resident open space preference while the complex shape of low density
development of had the opposite effect.
Our investigation of the surrounding landscape configuration suggests that exurbanites
tend to value more homogenous surroundings. Nearby landscapes that are characterized by
larger patch sizes (more core area) were significant positively related to resident open space
preference. In comparison, complex shape of nearby landscape links negatively with
people’s value of their adjacent open spaces. Landscape diversity is negative but not
significant to open space preference.
Our research revealed that open space landscape character is strongly related to open
space aesthetics of exurban residents. We conclude that the landscape character of nearby
open spaces is linked with residents’ preferences In addition, residents prefer to live in areas
with more complex shapes of open spaces. Planning to maintain woodland character and
agricultural land as a part of exurban landscapes while providing good access may help to
satisfy residents’ preferences and to enhance ecological quality.

References
Atash, F. (1994) Redesigning suburbia for walking and transit: emerging concepts. Journal of Urban
Planning and Development 120, 48-57.
Burchell, R., Lowenstein, G., Dolphin, W., Galley, C., Downs, A., Seskin, S., Still, K., and Moore,
T. (2002) Cost of Sprawl-2000. National Academy Press, Washington, DC.
Calthorpe, P. (1989) Pedestrian pockets: New strategies for suburban growth. In The pedestrian
pocket book: A new suburban design strategy (D.Kelbaugh, ed., pp. 7-20. Princeton Architectural
Press, New York.
Kaplan, R., and Austin, M. E. (2004) Out in the country: sprawl and the quest for nature nearby.
Landscape and Urban Planning 69, 235-243.
Kaplan, S., and Kaplan, R. (1989) The Experience of Nature: A Psychological Perspective.
Cambridge University Press, New York.
Liu, J., Daily, G. C., Ehrlich, P. R., and Luck, G. W. (2003) Effects of household dynamics on
resource consumption and biodiversity. Nature 421, 530-533.
Palmer, J. F. (2004) Using spatial metrics to predict scenic perception in a changing landscape:
Dennis, Massachusetts. Landscape and Urban Planning 69, 201-218.
Vogt, C. A., and Marans, R. W. (2004) Natural resources and open space in the residential decision
process: a study of recent movers to fringe counties in southeast Michigan. Landscape and Urban
Planning 69, 255-269.
Zube, E. H., Pitt, D. G., and Anderson, T. W. (1974) Perception and Measurement of Scenic
Resources in the Southern Connecticut River Valley, pp. 1-191. University of Massachusetts,
Amherst.

261
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

Applying a landscape ecological approach in the analysis of prospective

development for low-rise housing in Tyumen suburbs (Tyumen region, Russia)

N.R Rakhimova

Tyumen State University, Department of Ecology and Geography, ul. Semakova 12, Tyumen
625003, Tyumen region, RUSSIA

Introduction

It is increasingly recognized that in the second part of twentieth century the acceleration of
urban population growth, city development, urban sprawl, and megalopolises leads to the
spreading of urban life-styles to rural areas or the movement of urban populations to
suburban areas, i.e. suburbanization. Suburbs first appeared in the big cities of the USA from
the end of nineteenth to the beginning of twentieth century (Varivonchnik, 2004). The reason
for this was the arrival of new types of transportation like automobiles and electric trains. This
phenomenon then spread to Europe (Antrop, 2000), at first to the west and then to the east.
It was associated with the advantages of having a privately-owned home, independence and
a high standard of living.

Tyumen suburbanization

Post-perestroika Russia too experienced the effects of suburbanization. The initial boom in
country house building involved the central regions of Russia. Later this process became
more popular not only in the European part of country but in peripheral regions as well.
Tyumen region is one of them. However, it was only in recent years that real legislation
regulating land use in suburbs came into force. In response to the growing demand for out-
of-town housing, the city administration made a decision to assign relevant areas for low-rise
housing.
Now planning aspects are regulated by the Architectural Code of the Russian Federation,
and issues of suburban planning fall within the purview of regional planning. So far there is
not enough attention to ecological aspects in these official documents. Unlike in developed
countries, debates about the use of landscape ecological approaches in planning procedures
are not seen as germane to this process in Russia. Nevertheless, in this research I would
like to apply some existing approaches of landscape ecology to Tyumen.

Figure 1. Areas of prospective development for low-rise housing (Tyumen general

layouthttp://multitran.ru/c/m.exe?a=sa&t=217432_2_1&sc=0)
Assessment of the attraction of the developing landscape

262
Theme 2. Urban environment and transport
2.6 Open Session 24: The urban Landscape

Based on the general layout of Tyumen I have picked out five prospective areas (Figure 1).
These areas have been chosen because they reveal more potential in terms of sustainable
landscape planning. I have chosen two parameters to assess landscapes in prospective
areas for low-rise housing: sustainability and value. These parameters are more important for
residential area planning: sustainability is the ability of the landscape (landscape complex) to
maintain its rate of functioning and spatial structure under changing external influences
(Kurbatova, 2004); and value is the aggregate of aesthetic attractiveness, uniqueness,
comfort and recreational resources of the landscape (Drozdov et al., 2006). I have composed
a matrix of these parameters and identified an integral coefficient for the attraction (κi) of the
developing landscape in order to offer a more opportune comparative assessment and for
further application of this data in suburbs planning.

Table 1. Matrix of integral coefficient for the attraction of the developing landscape (κi)

Value Highly Valuable Not

Sustainability valuable valuable
Highly sustainable 9 8 6
Sustainable 7 5 3
Unsustainable 4 2 1

Having defined the integral coefficient for the attraction of each developing landscape for
individual suburban housing, I suggest using this formula to assess the attraction of some
areas comparatively:
∑ S ×κ
i
i i
Km = 1
,
Sc
where Km – average coefficient of attraction for area, κi – integral coefficient of attraction of
particular landscape, Si – square area of landscapes with the same value of κi, Sc – total
square area of land under assessment.

Having taken into account this research, area number 4 is the most felicitous for low-rise
housing in Tyumen suburbs in terms of sustainable landscape planning. But if we want to
get a complex assessment, we should take into account other factors in these areas, such as
pollution, accessibility of communication, transportation and so forth.

Different level plans with landscape ecological approaches like this can form one of the axes
of a sustainable development strategy, both for individual districts as well as for the country
as a whole over the long-term period. My research is just one of the attempts to apply the
landscape ecological approach to a real Russian city where there are challenges about
developing low-rise housing in order to create a sustainable interconnected natural and man-
made system for comfortable living.

References

Antrop, M. (2000). Changing patterns in the urbanized countryside of Western Europe. Landscape
Ecology 15: p.257-270.
Drozdov, A.V. (2006). Landscape planning with engineering biology elements. Association for
scientific publications, Moscow.
Kurbatova, D.S. (2004). Landscape ecological basis of town planning structures forming. Madzhenta,
Moscow-Smolensk.
Varivonchnik, I.V. (2004). Country of suburbs: ”American dream” today. USA – Canada. Economics,
politics, culture 7: p.72-85.

263
Theme 2. Urban environment and transport
2.7 Posters

2.7 Posters

Change analysis for planning sustainable landscape in the Urban Biosphere

Reserve of Rome

C. Blasi, M. Marta , G. Capotorti, M. Marchese

Department of Plant Biology, University of Rome “La Sapienza”Piazzale Moro 5, Rome, Italy.
e-mail: [email protected]

The extent that urban growth affects the local ecosystems can be controlled by high
quality land management, an essential ingredient in all urban growth yet in most cities there
have been virtually no effective measures to control land development. This is the case of
Rome where master plans, including guidelines on land development and the future
directions for urban growth, have rarely been realised. The most rapid growth of built up
areas happened from the mid 1950s to the late 1960s, during the mass movement of
population from country to the city. In the following decades, the population growth slowed
down, while the urban spread still continued affecting the high biodiversity and landscape
diversity characterizing Rome municipality. These trends underscore the urgent need to
support the capacity of local governments to manage urban development. In this framework,
the project of Rome as Urban Biosphere Reserve promotes the increase of environmental
quality and the improvement of urban sustainability. The Urban Biosphere Reserve,
encompasses representative areas of the variety of Rome landscapes defining areas with
different functions and conservation of Rome ecosystems.
Knowledge of land cover transitions derived from three temporal data sets (1954; 1980;
2001) assessed the efficiency of the UB Reserve zoning and provided valuable information
for understanding the effects of changes in land uses. The emphasis was given to the
transition zones, more affected by significant changes in land uses. Between 1954 and 2001
about 30% of agricultural lands were transformed in built up areas. In addition to the
incredible growth of built up areas, about 80% of discontinuous urban fabric were converted
to continuous urban fabric. Core areas, representing the best conservation areas, and buffer
zones, contiguous to the core and outstanding for specific landscape components, showed
less changes in land use even if there was a significant increase in landscape fragmentation.
It is worth noting that a positive effect of land use change was the little but significant
increase of natural areas, mostly happened between 1954 and 1980.
The information derived from the three temporal data sets allowed to define the
guidelines for a sustainable urban development, mostly based on the protection of sites of
conservation interest, the increase of ecological connectivity between natural areas, the
implementations of urban requalification actions, the safeguard and increase of traditional
and low intensive agricultural practices. Strategic priorities and actions were identified for
each of the 15 areas of the Urban Biosphere Reserve, giving attention to the specific
landscape components characterising the different areas.

References
Blasi C., Capotorti G., Di Pietro R., Ercole S., Filesi L., Fortini P. & Celesti-Grapow L. (2002)
Natural landscape in the area of Rome. EuroMAB 2002 Meeting, Rome, 2002.
Blasi C., Capotorti G., Di Pietro R. Filesi L. & Fortini P. (2002) Natural landscape in the area of
Rome and a new proposal for the urban biosphere reserve. EuroMAB 2002 Meeting, Rome,
2002.
Capotorti G., Marta M, Marchese M. & Blasi C. (2005) Programma MAB per la definizione di Roma
come Riserva Urbana della Biosfera. Informatore Botanico Italiano, vol. 37: 122-123.
Capotorti G., Marta M, & Blasi C. (2005) Vegetation and landscape diversity for defining the Urban
Biosphere Reserve of Rome. European IALE Congress, 2005: 62-63.

264
Theme 2. Urban environment and transport
2.7 Posters

Woody biomass in semi-urban landscapes

S. Meeus, H. Gulinck

Division of Forest, Nature and Landscape ResearchK.U.Leuven, Department of Land

Management and Economics Celestijnenlaan 200E, 3001 Leuven, Belgium
e-mail: [email protected]

As semi-urban landscapes form the transition area between densely urbanized areas and
much less urbanized rural areas, they enclose land cover types from both the urban as the
rural landscapes. In particular the ‘green’ land cover classes, such as urban green
(roadsides, parks, gardens, …) linear landscape elements (wooded banks and verges,
hedgerows,…) and non-linear and mostly non-defined green structures such as various
forms of thicket and shrubs, fallow grounds with green elements and smaller natural patches
whether managed as a natural area or not. The semi-urban area is assumed to be an
important carrier of biomass and its associated cycles. In order to develop further the
ecological role of these green components, the actual amount of biomass in the semi-urban
areas needs to be known.

Measuring biomass for these highly hybrid green structures is, however, very complex. The
biomass of herbacous and grassy vegetation, as well as high and woody structures (trees)
can be estimated and calculated relatively easy and with good accuracy, for shrublike green
structures – such as hedges and verges, biomass analysis is very difficult and assessment
can only give a very rough picture of the amount of biomass.

As alternative measure – with a specific landscape morphological goal as background – an

estimate of the volume used by these green structures can provide interesting information.

The method consists of a planimetric and if possible volumetric measurement of the green
structures, a comparison with a classification algorithm and a statistical assessment of the
biomass amount.

The ‘closeness’ of an area is an important property for viewshed analysis and results show
an important influence of the urban green in this landscape analysis for semi urban areas
and is related to the geographical location of these green structures.

265
Theme 2. Urban environment and transport
2.7 Posters

Plant trait patterns: differences between urban and rural areas?

S. Knapp, I. Kühn, S. Klotz

UFZ, Helmholtz Centre for Environmental Research – UFZ, Dep. of Community Ecology –
Theodor-Lieser-Str.4, 06120 Halle, Germany.
e-mail: [email protected]

Introduction

Cities differ from rural landscapes: human densities peak in urban areas. Pollutants,
waste, energy and nutrients concentrate there. Temperatures and precipitation are higher, air
moisture is reduced. Foreign species are more frequent, land-use is often more
heterogeneous. With some of these aspects, cities anticipate probable future conditions
under global change. We hypothesize that the differences between urban and rural areas act
as environmental filters resulting in different plant trait patterns.

Methods

We tested our hypothesis at national scale (Germany, urban, agricultural and forested
grid cells of 10’ latitude x 6’ longitude) and at regional scale (city of Halle and rural
surroundings, Central Germany). The latter was differentiated into protected areas and
randomly selected 250m²-plots (Wania et al., 2006). We analysed the proportions of trait
states per study unit for several leaf traits, floral traits, reproductive traits and whole plant
traits (Klotz et al., 2002). As all states of a trait add to 100%, the decline in one proportion
causes the increase of at least one other proportion. Therefore, we used log-ratios of
proportions. To correct for the effects of environmental parameters, we used multiple
regression with the log-ratios vs. environmental parameters. We compared the resulting
residuals of urban and rural study units using t-test and z-values. Due to differences in
sample size (nurban = 59, nagricult = 1365, nforest = 312), we used a resampling technique for the
national scale (59 grid cells, 999 times).

Results and conclusions

Most traits show significant differences between urban and rural areas at national scale
but not at regional scale. Urban areas have for example more plants pollinated by wind than
by insects, less plants with hygromorphic leaves, more annual but less biennial plants, more
therophytes, more plants reproducing by seed instead of vegetatively and more stress-
tolerant ruderals (sensu Grime, 1979). These results reflect the patchiness and climate of
urban areas, disturbance regimes and the high proportion of ruderal habitats in cities. As the
protected and randomly selected areas at regional scale do not differ in plant trait patterns,
an effect of nature protection does not seem to exist. This might be different at other scales.
The different reactions of plant trait patterns on different scales shall be further investigated.

References
Grime, J.P. (1979) Plant strategies, vegetation processes, and ecosystem properties. Wiley,
Chichester.
Klotz, S.; Kühn, I. & Durka, W. (2002) BIOLFLOR – Eine Datenbank zu biologisch-ökologischen
Merkmalen der Gefäßflanzen in Deutschland. Schriftenreihe für Vegetatiosnkunde 38, Bundesamt
für Naturschutz, Bonn.
Wania, A; Kühn, I. & Klotz, S. (2006) Plant richness patterns in agricultural and urban landscapes in
Central Germany – spatial gradients of species richness. Landscape and Urban Planning 75: 97-
110.

266
Theme 2. Urban environment and transport
2.7 Posters

Landscape change in the municipality of Rome (1954-2001)

R. Frondoni, B. Mollo, C. Blasi

Department of Plant Biology, University of Rome “La Sapienza”, Piazzale A. Moro 5,

00185 Rome, Italy.
e-mail: [email protected]

Introduction
Knowledge of land cover and land use change is crucial for assessing the stability and
dynamics of landscapes and for gaining options and suggestions for management and
conservation activities. In this context, an analysis of land cover transitions in the Municipality
of Rome (about 128,700 hectares) has recently been carried out, with the objective of
understanding the significance and spatial extent of change processes in the last 50 years.
The analysis employed the overlay of three 1:25,000 land cover layers within a GIS
environment. These data were gained by the interpretation of multi-temporal aerial
photographs (years 1954, 1980 and 2001), and were iteratively checked for thematic and
spatial consistency (Kayhko & Skanes, 2006). Land cover change was analysed over the
whole Municipality area and also within ecologically homogeneous land units, defined by a
hierarchical land classification process (Blasi et al., 2005).

Results and discussion

Important transformations occurred in the study area, with significant magnitude and rate
particularly in the 1954 -1980 period. In the 1980 -2001 interval, change processes
maintained the same direction but generally occurred at a slower pace and with smaller
spatial extent (affecting only 16% of the territory).
Urbanisation is by far the major process of change: artificial areas cover today about 32%
of the study area versus 11.6% in 1954. Most of the gained surface derives from pre-existing
artificial areas. There was also an expansion of urban areas and industrial units over coastal
habitats (which have declined by 60% since 1954) and agricultural land (declined by 27%),
particularly heterogeneous agricultural areas.
A major change is represented by the natural dynamics of vegetation, which led to a
constant increase in woodlands and to a significant expansion of scrub over abandoned
agricultural areas and quarries.
Despite these changes, at the landscape level results show a significant relative stability
of landscape composition between 1954 and 2001, which is mainly due to: an urbanisation
model largely based on the compaction of pre-existing urban areas; the overall conservation
of natural and semi-natural habitats (except in coastal areas); and the maintenance of
agricultural activities (agricultural land still represents the major land cover type).
Analysis of landscape change within land units in the 1954-2001 period identified
ecologically relevant areas characterised (when compared with the average situation over
the Municipality as a whole) by severe urbanisation processes or by considerable natural
dynamics towards mature vegetation. This approach is helpful for providing planning
guidelines and detecting particularly vulnerable areas that require special protection and
restoration measures.

References
Blasi, C; Capotorti, G. & Frondoni, R. (2005) Defining and mapping typological models at the
landscape scale. Plant Biosystems 139(2): 155-163.
Kayhko, N. & Skanes, H. (2006) Change trajectories and key biotopes – Assessing landscape
dynamics and sustainability. Landscape Urban and Planning 75: 300-321.

267
Theme 2. Urban environment and transport
2.7 Posters

Apomictic species, apophytes and anecophytes as elements of phytodiversity of

urban-industrial regions and their applicability as indicators

G.H. Loos

Biological Station of Western Ruhrgebiet, Ripshorster Str. 306, D-46117 Oberhausen,

Germany.
e-mail: [email protected]

By studies of phytodiversity and ecology of the compartments of the diversity (genera,

species and lesser ranked taxa) in the old urban-industrial region Ruhrgebiet (Germany,
Northrhine-Westphalia), it was pointed out that a lot of the apparent native species are
neogenic (anecophytes, so called “Indigenophyta anthropogena”, see Sukopp & Scholz
1997) or occurred at new habitats (apophytes). A quantitative and qualitative overview of
these taxa in relation to the other taxa could be given (see also Keil & Loos 2005). On the
one hand, there are obvious anecophytes like primary hybrids and hybrid derivatives with
stable characters (hybrids especially in genera Epilobium, Populus and Salix), on the other
hand, there may be only differences in genotypes while the phenotype is identical (probably
many species which are recognized as native in general). Apophytes grow in natural
landscapes especially at river banks and its surroundings. An important migration way of
them were railways (with similar ecological conditions as stony river banks). From these
habitats further dispersal took place, without dependence on line structures.

A lot of taxa of both status types are apomicts (and within this group in most cases
agamospermous taxa). In cultural landscapes apomictic species as well as other
anecophytes and apophytes were generated by farming. In urban-industrial ecosystems, the
ecological factors are partly different from agricultural landscapes (e. g. higher grade of
fragmentation), but at part these factors are very similar, too (different niches in more or less
small areas). The apomictic species of Taraxacum sectio Ruderalia (Asteraceae,
Cichorioideae) are preferably inhabitants of grasslands, some species are most frequent in
urban-industrial habitats – but the grassland species are also present in settlements and the
urban-industrial species can be found in wet meadows in some cases.

Present studies show that there are some possibilities to use such taxa as indicators.
Indicator attributes concern land-use intensity, primarily succession stadiums, soil types,
hemerobic stages and therefore the conservation value. A detailed overview of the significant
indicator species was ascertained and their indicator attributes were pointed out. The
occurrence of not apomictic anecophytes and apophytes together with apomictic species
leads to the reflection that in a phase of an unsteady and ambiguous Global Change
evolution processes provide for all contingencies.

References
Keil, P. & Loos, G.H. (2005) Anökophyten im Siedlungsraum des Ruhrgebietes – eine erste
Übersicht. Conturec 1: 27-35.
Sukopp, H. & Scholz, H. (1997) Herkunft der Unkräuter. Osnabrücker Naturwissenschaftliche
Mitteilungen 23: 327-333.

268
Theme 2. Urban environment and transport
2.7 Posters

Development of trees and soils of the forested area in Expo’70 Park 30 years after
reclamation

T. Sasaki1, Y. Morimoto1, J. Imanishi1

1
Graduate School of Agriculture, Kyoto UniversityKitasirakawa Oiwake-Cho, Sakyo-ku,
Kyoto, Japan 606-8502
e-mail: [email protected]

Introduction
The Expo’70 Commemorative Park was constructed at the site where a world
exposition in 1970 was held. It is located in Suita city in Osaka Prefecture, Japan, and
belongs to warm-temperature zone and lucidophyllous forest zone. The exposition site was
made by large scale reclamation; therefore, the original surface stratum was perfectly lost.
After the exposition, the site was mounded with imported local soils from the neighbouring
hills. Then, revegetation with mainly evergreen broad-leaved trees was conducted. Intensive
studies about vegetation and soil were conducted in 1982 and 1995-97.
In this study, we selected twelve sample plots in the forested area where the previous
studies were conducted, and surveyed about trees and soils to reveal the effect of
revegetation for 30 years.

Materials and Methods

In each plot, tree height and DBH were investigated for all trees (DBH ≥ 1cm). In
addition, a pit was excavated in each site, and undisturbed and disturbed soil samples were
taken from 0-5, 20-25, 50-55 cm depth for investigating physical and chemical properties.

Results and Discussion

The canopies reached over 10m and most of canopy species were evergreen broad-
leaved trees. Regeneration of the canopy dominant species did not successfully occur.
Tree density has declined; indicating self-thinning has occurred but seemed insufficient.
The soil physical properties were improved in most of the plots, to the level that the root
system was not affected. The yield per hectare was maximum at the plot where trenches
were dug for improvement of drainage after the monitoring in 1982. It was suggested that
appropriate management after revegetation can lead to satisfactory forest development even
under difficult soil environmental conditions.

References
Njoroge, J. & Morimoto, Y. (2000) Studies on Soil Development as Influenced by the Method of
Large Scale Reclamation of a Sub-urban Forest. Journal of the Japanese Society of Revegetation
Technology 25(3):184-195
Morimoto, Y & Kobashi, S. (1985) On the pedogenic process in the forest areas of the
Commemorative Park of Expo’70. Journal of the Japanese Institute of Landscape Architecture
48(5):115-120

269
Theme 2. Urban environment and transport
2.7 Posters

Vegetation cover where there is none: the third dimension of trees over sealed
surfaces and their influence on the city environment

A. Walz

Department of Geography, Marshall University, Huntington, WV 25755, USA.

e-mail: [email protected]

Introduction

In most climates cities are warmer than surrounding areas (urban heat island, Arnfield,
2003). Sun-exposed pavement absorbs radiation and releases it overnight; keeping the
neighbourhoods warmer than if the surface was vegetated or shaded. Mature large-growing
road-side trees overlap streets to a degree where a large proportion of the pavement is
shaded, intercepting radiation before it reaches sealed surfaces. Existing literature
recognizes the cooling effect of trees in urban areas (Oke, 1989, Rosenzweig et al., 2006).
No reports exist on the specific influence of road-side trees.
Land cover in cities is generally highly fragmented with small patches of vegetation
interspersed with buildings, roads, parking lots, and other features. Coarse scale satellite
information does not pick up these details (Lo et al., 1997). In this study high resolution
satellite images are used to map vegetation in a city.

The Study

In a twelve-block neighbourhood trees’ canopy width and height were measured manually
and pavement temperatures collected at locations of various levels of sun exposure. Analysis
revealed that before sunrise on mornings after hot summer days the surface temperatures in
exposed locations were still two degrees warmer than in shaded sites (Walz, 2006).
The above mapping technique is time consuming and only suitable for small study areas.
A new procedure to map the canopy that overlaps with roads was tested using high
resolution satellite images. The difference in spatial distribution of high fractional vegetation
cover between spring images (green lawns but bare canopies) and summer images (canopy
fully developed) indicates the location of tree canopies over sealed surfaces. This method of
determining tree cover over roads simplifies canopy mapping, and much larger areas in cities
can be processed. In future steps thermal infrared data will be used to examine the
relationship between the proportion of exposed sealed surfaces and temperature. This
technique could be adapted by city planners in order to ameliorate the environmental quality
in cities, and to identify gaps in urban forests.

References
Arnfield, A. J. (2003) Two decades of urban climate research: a review of turbulence, exchanges of
energy and water, and the urban heat island. International Journal of Climatology 23: 1-26.
Lo, C. P.; Quattrochi, D. A. & Luvall, J. C. (1997) Application of high-resolution thermal infrared
remote sensing and GIS to assess the urban heat island effect. International Journal of Remote
Sensing 18: 287-304.
Oke, T. R. (1989) The micrometeorology of the urban forest. Philosophical Transactions of the Royal
Society of London Series B 324: 335-349.
Rosenzweig, C.; Solecki, W.; Parshall, L.; Gaffin, S.; Lynn, B.; Goldberg, R.; Cox, J. & Hodges,
S. (2006) Mitigating New York City’s heat island with urban forestry, living roofs, and light
surfaces. 86th American Meteorological Society Annual Meeting, Jan. 31, 2006, Atlanta, Georgia.
Walz, A. (2006) The Influence of Trees in Residential Neighborhoods on Pavement Temperature.
Annual Meeting of the Association of American Geographers, March 7-10, 2006, Chicago, IL.

270
Theme 2. Urban environment and transport
2.7 Posters

Landscape ecological study on urban green spaces in Hanoi, Vietnam

U.D Pham, N. Nakagoshi

1
Graduate School for International Development and Cooperation, Hiroshima University, 1-
5-1 Kagamiyama HigashiHiroshima City, Hiroshima 739-8259, Japan
e-mail: [email protected]

Urbanization is a vital process and is increasing rapidly throughout the world, and
consequently is one of the main causes of ecological fragmentation, which is a serious threat
to biodiversity. Urban green spaces are an important component of urban ecosystems,
providing urban dwellers with many benefits such as environmental, aesthetic, recreational,
and economic values. However, as a result of increased building density, green spaces in
urban areas are reduced in both area and the quality of ecological services. It is found that
methods and principles of landscape ecology are useful and especially suitable for studying
urban green spaces.

This study aims at assessing the status and the temporal change of green spaces
based on landscape metrics in Hanoi in the period from 1996 to 2003; by applying the graph
theory and gravity model to study connectivity and eco-network of these green spaces. This
will assess the fragmentation of urban green spaces; and then we apply suitability analysis
based on GIS to identify appropriate sites for green spaces and landscape ecological
principles in planning a comprehensive green structure in Hanoi up to 2020. We
implemented this study by using Erdas 8.6 software for processing satellite images and Arc
GIS 9.0 for doing the GIS work. Assessing the temporal changes of green spaces was then
based on landscape ecological metrics by using Fragstats 3.3. In this period, the most
significant potential network for biodiversity conservation in Hanoi will be identified. It is a
basis for building an inner greenbelt in Hanoi planning up to 2020. Application of landscape
ecological principles and land suitability analysis also shows that green structure, including
green wedges, greenways, greenbelts, green fingers, green hearts and other green spaces,
is considered as a network of green elements and is more than the sum of these green
spaces. It expresses the connectivity and network or the inter-connection of green spaces in
the urban area instead of separate green spaces. Thus, comprehensive green structure
planning is proposed at two scales. At the city scale, an inner greenbelt and three green
wedges surrounding peri-urban have been proposed to enhance a skeleton of green
structure. At the neighborhood scale, greenways and other green spaces are planned to
allow nature to permeate into in the built-up area. The green network of this study will not
only help slow the extension of the urban sprawl process, but also improve urban
environment and maintain biodiversity in Hanoi City, Vietnam. Finally it will also provide a
basis for building a garden city or an eco-city in the near future.

271
Theme 2. Urban environment and transport
2.7 Posters

Settlement and urban areas along motorways: Changes in area and composition

K. Müller1, C. Steinmeier 1, M. Küchler 1

1
Swiss Federal Research Institute WSL – Zuercherstrasse 111, Birmensdorf, Switzerland.
e-mail: [email protected]

Landscapes in Central Europe are constantly undergoing changes due to different land
use claims. It is expected that the landscape is and will be influenced more by the increasing
traffic, especially apparent in the proportion of the settlement and urban areas (Steinmeier
and Müller 2007).
The purpose of the present investigations is to quantify the changes of the settlement and
urban areas along the main motorways through the Swiss Alps. Since changes in developed
land usually run in only one direction – it expands and does not revert to other land use
classes (Turner et al. 2001) – we are interested in its composition, e.g. which sub-classes of
it are expanding.
For the analyses we used two surveys of the Land Use Statistics of Switzerland (LUSS)
from 1985 and 1997. These statistical data sets consist of sample points which are regularly
spaced over Switzerland on a hectare grid and which are classified into land use categories.
In this study we used the major class settlement and urban areas, as well as its sub-classes
building areas, industrial areas, transportation areas and special areas. We divided the
landscape according to the five ecoregions of Switzerland, as they vary in terms of natural
and cultural preconditions. We further divided them into distance zones to the motorway exits
of 1km width and used relief characteristics as the outermost boarder. In these sub-regions
we then calculated the indicators, e.g. change of proportion of building area, and made
significance tests for every indicator in every sub-region.
The main findings are that the zones close to the motorways show a higher growth of the
settlement and urban areas, as well as their sub-classes building areas, industrial areas and
transportation areas. We can find this distance trend in all ecoregions. However, the analysis
of the relative area-changes, which means changes compared to the previous proportion of
the land use classes, only confirms a significant distance trend for the overall class
settlement and urban areas, and only in the ecoregions Northern Alps, Central Alps and
Southern Alps. In these regions the distance correlates strongly with the relief. If this is
considered, the variation explained by the distance diminishes even more.
Overall, these analyses allowed us to reveal a distance trend in the changes of the
settlement and urban areas and its sub-classes within the first few kilometres of the
motorway exits. By analysing the relative changes, such a trend can also be confirmed for
the settlement and urban areas as a whole in the mountain regions. However, further
analysis shows that a great part of the changes is explained by the relief. Therefore the relief
is an important factor, which should not be neglected in Switzerland, especially in the
mountain regions.

References
Steinmeier, C. & Müller, K. (2007). Monitoring of Supporting Measures - Environment (MSM-E).
27.01.2007,http://www.wsl.ch/forschung/forschungsprojekte/monitoring_flankierende_massnahme
n2/index_EN
Turner, M.; Gardner, R. H. & O’Neill, R.V. (2001). Landscape Ecology in Theory and Practice:
pattern and process. New York, Springer.

272
Theme 2. Urban environment and transport
2.7 Posters

Ecological research on Riga city street greenery

G. Čekstere1, A. Osvalde1, O. Nikodemus2

1
Laboratory of Plant Mineral Nutrition, University of Latvia Institute of Biology, Miera 3,
Salaspils, LV-2169, Latvia.
e-mail: [email protected];
2
Faculty of Geography and Earth Sciences, University of Latvia, Alberta 10, Riga, Latvia.

The street greenery is a very significant landscape’s design element with important
ecological value in the high building density area of the Riga City central part.

The first planned street trees in the city’s green structure appeared during the 18th century
as alleys connecting the old city’s gates with the recreation area. The street greenery
currently compiles 39 % from the overall street length (90 km) of the city’s central part. The
spatial arrangement of the street greenery is irregular and not completely corresponds to the
historical planning. The smallest amount of the street trees has been found in the Old Town
due to the highest building density.

Out of 16 tree species occurring in Riga’s street greenery nowadays Tilia x vulgaris and
Tila cordata are the most frequent. Tilia x vulgaris is avowed as the most appropriate species
for the urban environment in Riga.

The elimination of the greenery as the result of the carriageways expanding or trees
decaying has been stated in comparison to the beginning of the 20th century. The decrease
in the total amount of the street trees has been observed since 1979. The strongest street
trees decrease (more than 10 trees per street section) has been stated in 5 main streets.
Reconstruction of some green areas recently has been done in separate street’s sections.
Currently trees in the containers have become a new street landscape’s element as well.

The results of the bioindication research of the street greenery in Riga reveal that the
ecological status of Tilia x vulgaris could be characterized as seriously injured. The damage
to the deciduous trees typically appears as a necrosis of leaves. During the latest years, the
defoliation of leaves has been already observed at the beginning of August.

The investigation demonstrates that the physiological status of the trees has been
negatively affected due to systematic use of de-icing agents on the streets during the winter
period, as well as soil compaction, air pollution and probably alteration of the city’s
microclimate. As concluded from the research results, for the visual observation of the
necrosis of Tilia x vulgaris leaves the critical level concentration for Na is 0.18 % to 0.24 %
and for Cl - 0.62 % to 0.66 %.

273
Theme 2. Urban environment and transport
2.7 Posters

Restoration of multifunctional forests in north-west suburban area of Milan:

scenario and proposals

E. Padoa-Schioppa, A.B. Digiovinazzo, L. Bottoni

1
Università degli Studi di Milano-Bicocca, Department of Environment and Landscape
Sciences Piazza della Scienza 1 – 20126 Milano Italy.
e-mail: [email protected]

Suburban area of Milan (1.982 Km2, with about 3.869.037 inhabitants) represents one of
the most important European metropolitan area. That is, agricultural land use represents
49% of the territory, urbanisation 38%, natural vegetation and uncultivated fields 11%, pits
and redevelopment areas 1%, streams 2%.
The aim of this study is the identification of a green belt (“MetroBosco project”) around the
city of Milan, to improve woods and forest patches that already exist.
The functions of these restored woods should be: i) mitigation of roads impact; ii) CO2
absorption; iii) fruition; iv) increase of forest biodiversity; v) ecological network.
We carried out some scenarios, each of them maximizing one of the mentioned functions.
The term “scenario” is used to describe different situations in the future as well as a series of
events lead from the current state to the future state (Van Den Berg & Veeneklaas F.R. 1995
in Tress B., Tress G. 2003). By this definition, our scenarios are all realistic the same,
depending on the specific function maximized.
We used D.U.S.A.F regional maps to calculate different land uses (ArcView GIS 3.2
software): these are layers that cover all suburban area of Milan, with a meaning similar to
Corine Land Cover. We also overlapped aerial images to correct possible mistakes.
Then we used fuzzy logic method to carry out each scenario. This approach is a logic, in
which membership functions are graded from 0 to 1 (Zadeh 1965). This method allows us to
better describe ambiguous concepts that are not well modelled by traditional logic.
The approach of scenarios can help the stakeholders during the planning processes,
because it shows different images of a possible future depending on their decisions.

References
Tress B., Tress G., 2003. Scenario visualisation for participatory landscape planning-a study from
Denmark. Landscape and Urban Planning 64: 161 – 178.
Zadeh L.A., 1965. Fuzzy sets. Information and Control 8: 338 – 353.

274
Theme 2. Urban environment and transport
2.7 Posters

Urban influence in bird communities. A case in three neighbourhoods of Buenos

Aires city, Argentina

P.V. Perepelizin 1, A.M. Faggi 1,2

1
Universidad de Flores, Facultad de Ingeniería – Av. Nazca 274, , Buenos Aires, Argentina.
e-mail: [email protected]
2
Museo de Ciencias Naturales Bernardino Rivadavia, CONICET – Av. Ángel Gallardo 490,
(C.P. C1405DJR), Buenos Aires, Argentina

Introduction

Studies about the effects on urbanization on bird communities in Latin America and,
particularly in Argentina, are almost inexistent, knowledge essential for sustainable urban
planning and management tending on the conservation of their communities.
This paper, based mainly in two dominant ecological theories: urban gradients and
spatial-structuring analysis, compared bird richness and abundance with local habitat
characteristics of green spaces and land use of each neighbourhood. Bird habitat
preferences are linked to habitat structures inside the urban matrix.

Methods

In 3 neighbourhoods: “San Telmo”, near down town; “Almagro”, in the geographical center
of the city; and “Versalles”, in the West limit, land use was characterized through satellite
images for assessing the habitat quality. The existing green areas were classified in five
categories: Forest, Park, Savannah, Grassland and Artificial. Bird abundance in the green
areas was assessed in 34 visit point counts of 0.33 ha each. Richness in the rest of the
neighbourhoods and presence of nests were considered as well. Species were categorized
in: Urban-exploiters, Undetermined and Urban-avoiders. In addition, indexes for richness,
diversity, similarity and Reciprocal Average ordination were used for data analysis.

Urban impacts on bird communities

Twenty four bird species have been registered; 85 % were native from the pampean
ecoregion and nesting in the city. The exotic birds counted were Sturnus vulgaris, Aratinga
leucophthalma, Passer domesticus and Columbia livia.
The neighbourhood “Versalles”, with moderate levels of disturbance, has high presence of
breeding species, lower influence of exotic, higher richness and abundance of Undetermined
birds and Urban-avoiders and the highest level of richness outside green areas. The
existence of big green areas in “San Telmo” compensates somehow urban pressure. It has
the highest level of breeding species and richness of exotics. Open places contribute to the
presence of raptors. “Almagro”, due to the lack of green areas and the dominance of high
edification, behaves as a centric neighbourhood. It is characterized by high density of exotic
species and poor richness.
Decreases of bird richness and diversity and increases of exotic dominance trough
urbanization could be mitigated creating green areas, especially woodlands and conserving
and constructing new residential areas of low edification. These are suggestions to take into
account for a future urban planning and management.

275
Theme 2. Urban environment and transport
2.7 Posters

Landscape-ecological conditions for the development of Bratislava

T. Hrnčiarová
Institute of Landscape Ecology of Slovak Academy of Sciences, Štefánikova 3, P. O. Box
245, 814 99 Bratislava, Slovak Republic
e-mail: [email protected]

The basis of the landscape ecological evaluation of the area was the LANDEP methodology
of landscape ecological planning. It was aimed at the evaluation of settlements as well as the
latest methodological procedures. The presented work includes the methodological approach
of evaluation of urban landscape on the basis of ecological and partially also on
environmental principles. One of the main causes of disharmony in land use is the
misunderstanding of the features of landscape elements and their interrelations. For this
reason we elaborated the methodological procedure of optimum spatial arrangement and
functional land use (landscape ecological plan) with respect to landscape ecological, cultural-
historical and socio-economical conditions. The procedure of elaboration of the landscape
ecological plan is based on the methodology LANDEP and has five basic steps. A high
proportion of Bratislava is now built-up, therefore decision making as well as proposals will
include areas now used for agricultural purposes and partially for forest management. The
presented proposals can be divided into two outputs:
(i) Proposal of optimum land use for seven required activities and land use within the
whole town area. Suitability of use follows from abiotic, ecological and hygienic limits. For
each activity an independent map was elaborated: suitability of land use for the required
activity according to abiotic limits; suitable of land use for the required activity according to
abiotic and ecological limits; landscape ecological suitability of land use for the required
activity involving landscape ecological potential.
(ii) Landscape ecological plan of urban areas is based on landscape ecological suitability
and the establishment of territorial potential and it gives an alternative proposal on how to
use the given area. It is a synthesis of three previous proposals. Proposals are completed
with landscape ecological measures to mitigate the main negative impacts. The presented
methodological procedure represents how landscape ecological limits ought to be taken into
account in decision making about the development of the town in the future. Intensive further
construction in the town does not allow decisions always to be made according to landscape
ecological conditions of land use. In many cases the land use is unchanging, so it is possible
only to define or complete the intensity of use and proposal of different measures.
This work was supported by the Ministry of Education of the Slovak Republic under the
contract No. 2/7027/27 Evaluation of landscape diversity changes and by the Slovak
Research and Development Agency under the contract No. APVV-51-035102 Creation of
environmental limits for sustainable development (on example of model territories).

276
Theme 2. Urban environment and transport
2.7 Posters

Socio-cultural influences on urban ecosystems: A case study of Tommy

Thompson Park

Y. Westerveld Cardoso, K. Landman

University of Guelph - Guelph, Ontario, Canada.

e-mail: [email protected]

This study highlights the socio-cultural influences on urban ecosystems with a particular
focus on the achievements, setbacks and challenges in the regeneration and management of
Tommy Thompson Park (TTP) in Toronto, Canada. TTP is a lakefill created of construction
rubble and dredged material, and today is being managed by the Toronto and Region
Conservation Authority (TRCA) as an urban wilderness. Its current configuration extends 5
kilometres into Lake Ontario, with total land base of 160 hectares and a water surface area of
100 hectares composed of embayments and disposal cells. The site context and history are
complex. The park has been shaped by a web of influences such as political will, scientific
rationale, natural forces, and the passion of citizens and park users. Ecological regeneration
of culturally-modified urban ecosystems is a challenge and requires strategies for re-
development. For TTP, these strategies are affected by the uncertainty of biophysical
processes, socio-cultural influences and constant change, and are amplified by value-laden
issues advocated by stakeholders. This study features an in-depth, longitudinal case study of
this urban wild, situated on this metropolitan, post-industrial site. It presents the history,
planning and management at varying spatial and temporal scales, which uncovers social and
ecological patterns and processes and reveals process complexities from pre-planning to
master plan implementation and management.

277
Theme 2. Urban environment and transport
2.7 Posters

Feasibility of trail development along secondary urban watercourses in Tucson,

AZ, USA

M. Livingston1, J. E. Jones2, J. L. Patton3

1
University of Arizona, School of Landscape Architecture, Tucson, AZ, USA.
e-mail: [email protected]
2
CFA, 1150 Corporate Blvd., Reno, NV, , USA.
3
Norris Environmental, 100 N Stone Ave, Tucson, Tucson, AZ, USA.

Introduction

Vegetated buffers along urban watercourses are typically well-suited for recreational uses
such as biking, walking trails and wildlife-viewing. However, development pressures in cities
can lead to fragmented and redirected watercourses to accommodate other land uses. In
some cases, unplanned access increases erosion, and loss of vegetation and potential
wildlife habitat in these areas. It is critical that appropriate access to these semi-natural areas
be addressed for future conservation efforts. Such efforts would also provide a basis for
creating passive recreational opportunities where currently few exist in urban areas, except
along larger waterways.

Methods and results

This research evaluated the feasibility of access and trail development along secondary
watercourses within the urban core of Tucson, AZ. The following question was posed, “How
feasible are existing smaller urban watercourses in Tucson for access and trail
development?” Feasibility of individual corridors was evaluated and ranked based on
watercourse composition and bank treatment, adjacent buffer and streambed vegetation,
streambed surface, path corridor, barriers (road crossings, underpass properties and
elevation changes), and linkages to larger watercourses. Results indicated that the majority
of the watercourses studied have natural banks, accessible dry streambeds, and
predominantly exotic vegetation. Two-lane road crossings were the predominant barrier, and
linkages with larger watercourses were not prevalent. Results suggested that planned
access is possible for particular watercourses within the study area, but linkages between
secondary and primary watercourses is limited and needs to be addressed.
Recommendations related to trail development including design applications addressing
circulation and urban wildlife habitats along these urban watercourses will be discussed.

278
Theme 2. Urban environment and transport
2.7 Posters

Development of a future scenario model of compact growth in the Charlotte,

North Carolina region of the U.S.

D.H. Yankee1, M.H. Mehaffy 2, E.R. Smith 2

1 Tennessee Valley Authority – 400 W. Summit Hill Drive, Knoxville, TN 37828, USA. email:
[email protected]
2 U.S. Environmental Protection Agency – 109 T.W. Alexander Drive
Research Triangle Park, NC 27711, USA

The negative impacts of urban sprawl on natural resources are a growing problem for many
regions across the United States. A solution was investigated for the 15-county Greater Charlotte
Bi-State Region in North Carolina and South Carolina. A partnership among the Environmental
Protection Agency, local Council of Governments, and University of North Carolina at Charlotte
was established to assess how alternative development patterns could protect environmental
quality while also bolstering economic growth. Input was obtained from multiple types of
stakeholders to formulate two future development scenarios. Spatial representations of current
conditions and the two future growth scenarios were created from which water, terrestrial, and
socioeconomic variables were derived. One scenario was a “growth as usual” scenario while the
“Center’s” scenario explored the feasibility of directing the majority of projected growth toward
existing municipalities and other regional centers and thereby preserving the outlying area from
unchecked sprawl. This poster will address the methods and techniques used to model the
“Center’s” scenario as well as compare the resulting environmental impacts from “growth as
usual” and “Center’s” compact development scenario. The model suggests that high-density
growth along these re-developed corridors would accommodate the high level of projected
population growth. Growth directed in such a manner would reduce development pressure in the
areas outside of the corridors and prevent the capacity of the infrastructure in these areas from
being exceeded.

279
Theme 2. Urban environment and transport
2.7 Posters

Effects of traffic noise on acoustic communication in birds

K.M. Parris

School of Botany, University of Melbourne VIC 3010, Australia.

e-mail: [email protected]

Introduction
Birds use acoustic signals for a number of purposes: to attract and bond with mates,
defend territories, beg for food, and warn of danger from predators. Background noise
reduces the distance over which an acoustic signal (a song or call) can be detected, and
birds can use a variety of strategies to maximise the audibility of their signals in naturally
noisy habitats (Brumm and Slabbekoorn 2005). Anthropogenic noise such as traffic noise
provides an additional challenge for acoustically-communicating animals. Field studies have
shown reduced densities of some bird species in habitats close to roads (e.g. Reijnen et al.
1995; Rheindt 2003), and changes in the characteristics of bird song in noisy urban areas
(e.g. Slabbekoorn and den Boer-Visser 2006; Wood and Yezerinac 2006).

Model & Data

I will present a mathematical model of the effect of traffic noise on the distance over which
acoustic signals of varying frequency (pitch) can be detected by a conspecific bird, based on
the frequency distribution of traffic noise, the structure of the avian auditory system
(Langemann et al. 1995; Lohr et al. 2003), and the attenuation of sound in open forests. The
model demonstrates that birds with lower-frequency signals will experience more acoustic
interference from traffic noise than birds with higher-frequency signals, despite an increase in
the bandwidth of auditory filters with frequency. I have used the model to predict the effect of
traffic noise on detection of calls by eight species of birds from south-eastern Australia. I
expect all species to demonstrate a frequency shift (Warren et al. 2006) by singing/calling at
a higher pitch at noisy sites than at quiet sites, thereby reducing acoustic interference from
traffic noise. However, I expect species with lower-frequency signals to demonstrate a larger
frequency shift. To test these predictions, I have collected high-quality recordings of bird
calls and songs at 60 survey sites south-east of Melbourne, Australia. All sites are in strips
of remnant vegetation next to roads, but the roads vary in size from small dirt roads with little
traffic to multi-lane freeways. I am currently analysing the recorded signals to assess the
presence and size of a frequency-shift in populations of birds exposed to traffic noise.

References
Brumm, H. & Slabbekoorn, H. (2005) Acoustic communication in noise. Advances in the Study of
Behavior 35: 151-209.
Langemann, U., Klump, G.M. & Dooling, R.J. (1995) Critical bands and critical-ratio bandwidths in
the European starling. Hearing Research 84: 167-176.
Lohr, B. Wright, T.F. & Dooling, R.J. (2003) Detection and discrimination of natural calls in masking
noise by birds: estimating the active space of a signal. Animal Behaviour 65: 763-777.
Reijnen, R., Foppen, R., ter Braak, C. & Thissen, J. (1995) The effects of car traffic on breeding bird
populations in Woodland. III. Reduction of density in relation to the proximity of main roads.
Journal of Applied Ecology 32: 187-202.
Rheindt, F.E. (2003) The impact of roads on birds: does song frequency play a role in determining
susceptibility to noise pollution? Journal für Ornithologie 144: 295-306.
Slabbekoorn, H. & den Boer-Visser, A. (2006) Cities change the songs of birds. Current Biology 16:
2326-2331.
Warren, P.S., Katti, M., Ermann, M. & Brazel, A. (2006) Urban bioacoustics: it’s not just noise.
Animal Behaviour 71: 491-502.
Wood, W.E. & Yezerinac, S.M. (2006) Song sparrow (Melospiza melodia) song varies with urban
noise. The Auk 123: 650-659.

280
Theme 2. Urban environment and transport
2.7 Posters

Human-Nature Symbiosis -- Landscape Ecological Planning for the Renewal of

Suburban Park

N. Yen, F. C. Yu

Graduate Institute of Building and Planning, National Taiwan University

No.1, Sec. 4, Roosevelt Rd., Da-an District, Taipei City 106, Taiwan (R.O.C.)
e-mail: [email protected]

Introduction

Landscape ecological planning has been paid much attention by planners over urban
and nature areas, but cases in suburbs have been rarely mentioned. The suburban
landscape is a mosaic of human buildings and nature habitats, in suburbs of Taipei, Grant
Lake Park serve as a stepping stone between urban parks and forested basin edge. The
ecological function of the park is getting more concern because of the traversing MRT (Mass
Rapid Transport System). Therefore, we propose the park renewal plan with principles of
landscape ecology. In the local scale, the ecological network plan can be made through
delicate observation: birds’ flying routes, extent of noise, and green coverage, etc; instead of
accurate estimation of landscape ecology indices. The results proved that for a local park in
the human-nature symbiosis suburbs, planning principles and process are the same for
human and fauna: to make a better place for living, feeding and transportation.

Landscape ecology concepts

In Grant Lake park renewal plan, have we used landscape ecology principles on
planning the edge and boundary of water and woodland patches, greenway corridors, and a
vegetation network.
In Grant Lake Park, we have used bird behaviour and movement as a model for the
ecological network. The avian species observed are: Little Egret (Egretta garzetta), Cattle
Egret (Bubulcus ibis), Black-crowned (Nycticorax nycticorax), Taiwan Blue Magpie (Urocissa
caerulea). The movement network for birds were also studied and mapped.

Planning features

In order to create isolated habitats, we planned several floating islands in the middle of
the Grand Lake; birds can then have more resting and nesting sites. The vegetation cover
should be replaced by native plants which will create diverse habitats of islands, woods,
hedges, wetlands and sandbanks, to increase vegetation diversity of the water bank, and to
make abundant microhabitats. Most important, we suggest a green tunnel for the MRT line,
which the vegetation covering will fill up the ecological gap. The renewed Grand Lake park
will be no more just a neighbourhood park but a thriving human-nature symbiosis resort.

References
Dramstad. W. E.; Olson, J. D.& Forman, R. T. T. (1996) Landscape Ecology Principles in landscape
Architecture and Land-use Planning. Island Press, Washington DC, US
Forman, R. T. T. (1995) Landscape Mosaic. Cambridge University Press, Cambridge, UK. P.273-274
Jongman, R. & Pungetti, G (2004) Ecological Networks and Greenways. Cambridge University
Press, Cambridge, UK.

281
Theme 2. Urban environment and transport
2.7 Posters

Searching for participatory planning in the urban-rural fringe in Spain

V. Hernández-Jiménez, B. Ocón, D. Pereira

Dpto. Proyectos y Planificación Rural. ETSI Agrónomos. Universidad Politécnica de Madrid.

Av/ Ciudad Universitaria s/n. 28040 Madrid, Spain.
e-mail: [email protected]

The institutional and landscape changes in Spain over the last three decades have shaped
current decision-making processes in the land planning system. Natural protected areas,
strategic farming areas (i.e. typical Mediterranean agro-ecosystem) and new or potential
urbanized areas shape and share the territory. The intensive use of land in the urban-rural
fringe and the wide range of stakeholders involved, have generated social and environmental
conflicts. Different themes such as territorial planning and governance need to be explored
taking into account the role of integrative research. The combination of quantitative methods
and discursive approaches could harmonize the complexity of these emergent conflicts in
regional planning and management.

This research group deals with these issues in the framework of the Integrative Systems and
the Boundary Problem project (ISBP) funded by the EU FP6 (NEST-2005-Path-CUL). ISBP
emerged from a former research (TiGrESS - MADRID) which was instrumental in helping
stakeholder communities to adapt to emerging circumstances and changes in the regional
planning process. This work suggested a set of future scenarios, under a common
framework of sustainability. These scenarios were explored enhancing stakeholder
participation. ISBP will continue this former study exploring how to negotiate new institutional
boundaries to reduce tension between antagonized communities of stakeholders.

The Madrid Autonomous Community and Barcelona (one of the four regions of Catalonia)
are the case studies chosen for this work due to their remarkable moment as the two
greatest metropolitan areas in Spain.

Carried out by the research group on Sustainable Planning at the Polytechnic University of
Madrid (Spain) in collaboration with Newcastle University (UK).

References
Healey, P. (1997a). Collaborative Planning: Shaping Places in Fragmented Societies. Planning,
Environment and Cities. Palgrave McMillan. London
Pike, A., Rodriguez-Pose A. and J. Tomaney (2006). Local and Regional Development. Routledge.
New York
Winder, N. (Ed) (2006). TiGrESS Final Report. Retrieved: June 2006 from http://www.tigress.ac
Winder, N. (Coord.) (2006). ISBP. Retrieved: February 2007 from http://www.tigress.ac/isbp/

282
Theme 2. Urban environment and transport
2.7 Posters

Green space strategy (Dublin)

D. Sgouros

Landscape Architect, Zygomalli 7, 45332 Ioannina, Greece.

e-mail: [email protected]

The study area is in the triangle between Richmond Road, Clonliffe Road and Drumcondra
Road. Most of the land is covered by the premises of Holly Cross
College, where there is the Archbishop’s House, the main buildings of the historical college,
and other buildings that are recorded as National Monuments. The open area of the site is
approximately 20 acres. There is no use or specific activity in the site. Tolka river runs in the
site towards southeast.
The river is the boundary of the college land. On the north bank there are houses but the
land is dominated by old buildings that are currently being used as garages, warehouses and
light industries. At the southeast part of the study area there are houses. The character of
this part of the city is very degraded.
Especially on the north bank the land is almost derelict and the people who live in the
adjoined houses are subjected to views of very low aesthetics.
The aim of the project is to change radically the character of the study area by creating new
housing and a new park. It will be part of a "chain" of green areas which will include the
existing surrounding parks. This principle could subsequently be followed in the broad area
of Glasnevin, Whitehall and Marino where there is abundance of open institutional land.
The river Tolka is a natural feature of significant proportion. It can be the focal point of this
part of the urban environment.
The geomorphology of the site presents a uniformity that offers a broad potential of
The existing mature vegetation offers the opportunity of a plan that can be
Implemented in short term, and be integrated with a long term planting plan.
The access to the river corridor is presently almost non existent. The current use of land
does not step with the creation of a new amenity and recreation space
of vast size. At the north bank there are residencies in low density, warehouses and small
industries. The south bank is covered mostly by the land of Holly
Cross College. Holly Cross College is an institution of great historical importance for Dublin
and Ireland that dates in the middle 19th century. The site includes the Archbishop’s House
and other buildings (i.e. The Red House) which are recorded as National Monuments. Any
intervention into the very premises of the institution should not be undertaken without serious
consideration.
The adjacent land of Holly Cross College can become a new area of amenity. It can be part
of an integrated green space strategy for the broad area
(Drumcondra, Glasnevin, Whitehall and Marino) that will:
1. Create more publicly accessible open spaces
2. Provide recreational routes for walkers, joggers and cyclists
3. Increase the accessibility of both existing and new public parks
4. Create links between pedestrian origins and destinations for journeys to work, to
schools, to public transport.
There is a challenge to increase residential density and redevelop derelict lands. Such a
task will improve the living conditions and quality of life of the inhabitants, by providing new
housing and creating an adjacent amenity area (neighbourhood park) of significance along
the river. The project includes the design of anti-flooding measures to tackle with the
problem of flooding that can occur in the area after heavy raining, as it happened during
2002.

283
Theme 2. Urban environment and transport
2.7 Posters

Landscape fragmentation of urban green spaces in Hong Kong

Y. H., Tian, C. Y., Jim

School of Geography, University of Hong Kong, HKSAR, China.

e-mail: [email protected]

A large amount of literatures has assessed the complicated process of landscape

fragmentation. However, few studies tackle the urban green spaces (UGS) system in the
most compact city in the world such as Hong Kong. Inspired by related studies and assisted
by the statistical, GIS and RS techniques and FRAGSTAT 3.3 software. The study area
include 13 districts of Hong Kong: 4 old districts of Hong Kong Island, 5 old districts of
Kowloon and 3 new towns of the New Territories which represent first generation, second
generation and the latest generation. The Fragmentation Index of green patches in the study
area is modeled with the hexagon of 160 ha as appropriate sample unit. The model of
fragmentation includes almost all the characteristics of landscape pattern: AREA_MN, PD,
SHAPE_MN, ENN_MN, CONNECT, DIVISION, MESH and SPLIT. Results indicate that the
fragmentation level of UGS are increasing from the countryside to the center of the districts,
and built-up areas especially those in the center of Kowloon districts have the highest
fragmentation of UGS. New towns always have lower fragmentation and higher ecological
functions of UGS than old towns because of the increasing concerns on the construction of
green spaces on spacious lands. This model is validated to be applicable in the study. The
uneven distribution of fragmentation of green spaces indicates the unbalanced distribution of
spatial pattern of green patches in the study area and in the built-up areas of different
districts. More measures could be introduced to increase the continuity of green patches and
the green cover. These measures may include constructing greenways or green corridors,
setting up green stepping stones, enlarging the existing green patches and strengthening the
management.

284
Theme 3 Ecological Networks, fragmentation and connectivity

Theme 3. Ecological Networks, fragmentation and

connectivity

285
Theme 3 Ecological Networks, fragmentation and connectivity

286
Theme 3. Ecological Networks, fragmentation and connectivity
3.1 Symposium 5: Ecological infrastructure: theory and application

3.1 Symposium 5: Ecological infrastructure: theory and application

Green Infrastrucure for Cities: The Spatial Dimension

J.Ahern, S. Kato

University of Massachusetts, Department of Landcape Architecture and Regional Planning,

Amherst, MA, USA.
e-mail: [email protected]

Planning for sustainable cities is a complex process addressing the fundamental areas of
economic, environmental and socially-equitable sustainability. This paper focuses on the
environmental area, with theories, models, and applications illustrating possible spatial
configurations of a green infrastructure to support ecological and physical processes in the
built environment including: hydrology, biodiversity, and cultural/human activities. Green
infrastructure is an emerging planning and design concept that is principally structured by a
hybrid hydrological/drainage network, complementing and linking relict green areas with built
infrastructure that provides ecological functions. Green infrastructure plans apply key
principles of landscape ecology to urban environments, specifically: a multi-scale approach
with explicit attention to pattern:process relationships, and an emphasis on connectivity.
The paper provides theoretical models and guidelines for understanding and comparing
green infrastructure approaches. International examples at multiple scales are discussed to
illustrate the concepts and principles introduced.

Key Ecological Processes and Functions

Ecological processes are the mechanisms by which landscapes function – over time, and
across space - and are therefore appropriate to use as the goals for - and the indicators of -
sustainability. Landscape ecology provides a theoretical perspective and the analytical tools
to understand how complex and diverse landscapes, including urban environments function
with respect to specific ecological processes (Pickett et al. 2004).

The Ecological Society of America defines ecological functions as those that provide
"services" that moderate climatic extremes, cycle nutrients, detoxify wastes, control pests,
maintain biodiversity and purify air and water (among other services) (ESA 2006).

Landscape Ecology Principles for Green urban Infrastrucutre

Key ideas from landscape ecology that are relevant to green urban infrastructure for
sustainable cities include: 1) A multi-scaled approach is based on hierarchy theory that
addresses the structure and behavior of systems that function simultaneously at multiple
scales. 2) The pattern:process dynamic is arguably the fundamental axiom of landscape
ecology because the spatial composition and configuration of landscape elements directly
determines how landscapes function, particularly in terms of species movement, nutrient
and water flows (Turner 1989). 3) Connectivity is a property of landscapes that illustrates the
relationship between landscape structure and function. In general, connectivity refers to the
degree to which a landscape facilitates or impedes the flow of energy, materials, nutrients,
species, and people across a landscape.

Guidelines for Planning and Designing a Green Urban Infrastrucutre

Articulate a Spatial Concept

Spatial concepts guide, inspire and communicate the essence of a plan or planning
strategy to provide for specific ABC functions. Spatial concepts are often articulated as

287
Theme 3. Ecological Networks, fragmentation and connectivity
3.1 Symposium 5: Ecological infrastructure: theory and application

metaphors that are highly imaginable and understandable by the public, but which also can
support and inspire the planning process (Zonneveld 1991). Spatial concepts are well
understood in planning, but less so in science. They represent an important interface of
empirical and intuitive knowledge through which rational knowledge is complemented with
creative insights..

Strategic Thinking
When the existing landscape supports sustainable processes and patterns, a protective
strategy may be employed. Essentially, this strategy defines an eventual, or optimal
landscape pattern that is proactively protected from change while the landscape around it
may be allowed to change. When the existing landscape is already fragmented, and core
areas already limited in area and isolated, a defensive strategy can be applied. This
strategy seeks to arrest /control the negative processes of fragmentation or urbanization. An
offensive strategy is based on a vision, or a possible landscape configuration that is
articulated, understood and accepted as a goal. The offensive strategy differs from
protective and defensive strategies in that it employs restoration, or reconstruction, to re-
build landscape elements in previously disturbed or fragmented landscapes. The
opportunistic strategy is conceptually aligned with the concept of green infrastructure by
seeking new or innovative “opportunities” to provide ABC functions in association with urban
infrastructure.

The Greening of Infrastructure

To achieve sustainability in urban landscapes, infrastructure must be conceived of, and
understood as a genuinely possible means to improve, and contribute to sustainability. If
one only thinks about avoiding or minimizing impact related to infrastructure development,
the possibility to innovate is greatly diminished.

Learn by Doing
The adaptive approach is promising for green infrastructure because the knowledge to
plan and implement these systems is evolving. If experimental applications can be
practiced routinely, the potential to build empirical knowledge, while exploring sustainability
is quite profound.

References

Ecological Society of America, 2006. Retrieved on June 30, 2006

http://www.actionbioscience.org/environment/ esa.html
Pickett S T A, Cadenassso M L and Grove J M , 2004. Resilent cities: meaning, models,
and metaphor for integrating the ecological, socio-economic, and planning realms.
Landscape and Urban Planning: 69:4, 369-384.
Turner M G , 1989. Landscape Ecology: the Effect of pattern on process. Annual Review of
Ecological Systematics. 20:171-197.
Zonneveld W , 1991. Conceptvorming in de Ruimtelijke Planning. Universetiet van
Amsterdam.

288
Theme 3. Ecological Networks, fragmentation and connectivity
3.1 Symposium 5: Ecological infrastructure: theory and application

Why patches of fragmented woods in the city? - Contemporary uses of

satoyamas in Japanese urban areas as a source of renewable resources

M. Yokohari1, T. Terada2
1
Graduate School of Frontier Sciences, The University of Tokyo, Kashiwa, Chiba, Japan;
e-mail: [email protected]
2
Graduate School of Systems and Information Engineering, Univeristy of Tsukuba, Tsukuba,
Ibaraki, Japan

Satoyama and the Kyoto Protocol

The majority of forests in Japan, which cover nearly 70% of the land area, are on mountains,
however a number of forests patches surrounded by farmlands and urban areas can still be
found on hills, plateaus and lowlands. These fragmented woods remaining in the areas close to
human settlements are known as satoyamas, a landscape type that represents up to 20% of
Japan’s land area. Most satoyamas have a common historical background having been
maintained for agricultural uses; harvesting firewood and charcoal, and collecting litter for
making organic fertilizer. However, the introduction of fossil fuels and chemical fertilizers
deprived satoyamas of their economic viability, leading to their neglect.

The importance of maintaining satoyamas for safeguarding ecological values has recently
been discussed. However, because the current management schemes are mostly conducted
by volunteers, those satoyamas that enjoy continuous maintenance are extremely limited. A
sharp increase of maintained satoyamas may never be expected unless new incentives for
maintenance are introduced.

Restoration of satoyamas can be expected by utilizing renewable resources, especially wood

and litter harvested from satoyamas. The Kyoto Protocol to the UN Framework Convention on
Climate Change has set the goal for Japan to reduce collective emissions of greenhouse gases
by 6% compared to the 1990 emission rate, which should be achieved by 2010, and the
reduction of the CO2 emission is regarded as the key issue to achieve the goal. Substituting
biomass as an alternative energy source for fossil fuels is expected to be an indispensable
issue for the reduction of CO2 emission. If we can utilize woods and litter from satoyamas as
biomass resources it may add a substantial incentive for satoyamas to be maintained.

Data collection and the estimation of potential biomass production

To estimate the potential amount of biomass that may be harvested, we have taken Tsukuba
City, Japan as a case study. Although Tsukuba is located on the fringe of Tokyo, land use of
the city is still fairly rural where approximately 12% of the land area is covered by satoyamas.
The actual stock of biomass in satoyamas in Tsukuba was estimated by measuring the number
and sizes of trees and ground cover in the selected satoyama patches, and applying the data
sets to the models derived from precedent studies.

The amount of biomass which can be harvested from satoyamas is determined not only by
the actual stock of biomass but also by the way the satoyamas are maintained. In this study
maintenance schemes to realize four goals named as “scenic beauty”, “quiet recreational use”,
“active recreational use”, and “bio-diversity” were set, and the amount of biomass which may be
harvested according to the four schemes were estimated by developing estimation models. As
the result, “Bio-diversity” scheme with clear cutting of trees is expected to provide the highest
amount of biomass at 4,000 tons annually, while “scenic beauty” scheme only with the forest
bed management is estimated to result in the lowest amount as 1,600 tons.

289
Theme 3. Ecological Networks, fragmentation and connectivity
3.1 Symposium 5: Ecological infrastructure: theory and application

Generating electric power by biomass to reduce CO2 emission

Among various schemes for CO2 reduction, generating electric power by biomass as an
alternative and renewable energy source is one practical means to substitute for fossil fuels.
We therefore estimated the amount of electric power that can be generated by biomass
harvested from satoyamas in Tsukuba to identify how the biomass will be contributing to the
reduction of CO2 emission.

The amount of electric power generated by biomass has been estimated as 6 to 28 million
kilowatts annually, sufficient for 1,300 to 5,200 households, 2 to 7%, in Tsukuba. The Ministry of
Environment Japan is currently entering discussions to reduce the collective greenhouse gas
emissions from each local municipality. Accordingly, in 2002 Tsukuba set a goal to reduce the
CO2 emission by 2,132 tons per year. From precedent studies it is estimated that 1 ton of
biomass will reduce 0.42 tons of CO2 emission when the biomass is used for generating electric
power. According to this estimation the biomass harvested according to the “scenic beauty”
management scheme will achieve 90% of the goal set by Tsukuba when the biomass is used for
generating electric power, while the “bio-diversity” scheme will clear nearly 400% of the goal.
Thus Tsukuba may easily exceed the goal only by utilizing biomass provided from satoyamas.
Satoyamas remaining in and around urban areas are no longer useless fragmented woods
waiting for future urban developments, but may become indispensable uses of land for the
sustainable future of the city.

290
Theme 3. Ecological Networks, fragmentation and connectivity
3.1 Symposium 5: Ecological infrastructure: theory and application

Green infrastructure: a strategic approach for landscape ecology in urban areas

P. Pellegrino

University of São Paulo, Faculty of Architecture and Urbanism, Design Department, Rua do
Lago 876. 05508-900. São Paulo SP Brazil
e-mail: [email protected]

Introduction

The application of landscape ecology principles in landscape architecture and planning as

part of an overall green infrastructure are yet to be fully tested. The use of the concept of
green infrastructure is presented as being strategically effective to support ecosystem
services in the urban fabric. The opportunities and restraints for the implementation of this
green tapestry of different kind of open spaces is reviewed through two case studies in São
Paulo City. They encompass different scales: an urban watershed plan in the border of a
forest reserve and a design proposal for an inner city neighborhood watershed; and, in this
way, providing the opportunity to look for the real viability of maximizing the overall
performance and sustainability of the urban open spaces.

Development of an urban landscape approach

Why Green Infrastructure?

The interconnected network of natural areas and other open spaces that conserves
natural ecosystems values and functions inside and around the urban areas can be seen as
the backbone of their green infrastructure (Benedict and McMahon, 2006). The abundance
and distribution of natural features in the urban landscape like forests, wetlands, and streams
should be seen as an infrastructural capacity just as built infrastructure like roads and utilities
are, since they provide the ecosystem services that are equally necessary for the
sustainability of the built areas as also of the remnant natural elements (Weber et al, 2005).

Green infrastructure strategy

Green infrastructure planning

Landscape Ecology principles can be more easily understood and applied in the scale of
a plan for an urban watershed, like the example chosen, that is located in the fringe of the
urban region with part still covered with forest, confronting the other half characterized by the
growing overall impervious coverage of the city, in the process of encroaching the watershed
with the continuous canalization of disconnected urban creeks and rivers, and the
implementation of urban detention basins. The implementation of a green infrastructure was,
then, viewed as an opportunity to counteract these trends, buffering the natural ecosystems
and creating a distinctive local landscape identity and a more sustainable urban environment.

291
Theme 3. Ecological Networks, fragmentation and connectivity
3.1 Symposium 5: Ecological infrastructure: theory and application

Figure 1. View of the proposed Green Infrastructure for the Urban Watershed Plan.

Green Infrastructure design

Can landscape design be viewed as much more than mere beautification of the urban
environment? An emergent palette of landscape design typologies – like rain gardens,
bioswales, wetponds and green grids - plus the concept of the urban forest, are part of this
effort, that goes in the direction to integrate pieces of a high-performance infrastructure that
protects and even improves urban hydrology, climate, and ecology. Adopting these
principles in the adaptation of the cycles and processes found in nature in the scale of an
inner city neighborhood watershed, that has been totally urbanized and deprived of its former
natural features, was the goal of the example shown here at the design scale.
.

Figure 2. Bioswales and rain gardens are connected as a green grid with the urban forest.

References

Benedict, M. & McMahon, E. (2006) Green Infrastructure: linking landscapes and communities.
Island Press, Washington, DC.
Frischenbruder, M. & Pellegrino, P. (2006) Using greenways to reclaim nature in Brazilian cities.
Landscape and Urban Planning 76: 67-78.
Weber, T; Sloan, A. & Wolf, J. (2006) Maryland’s Green Infrastructure Assesment: development of a
comprehensive approach to land conservation. Landscape and Urban Planning, 77:1-2:94-110.

292
Theme 3. Ecological Networks, fragmentation and connectivity
3.1 Symposium 5: Ecological infrastructure: theory and application

Patch Dynamics and Urban Design

S.T.A. Pickett1, B.P. McGrath2, M.L. Cadenasso3, J.M. Grove4

1
Institute of Ecosystem Studies, Millbrook, NY, USA,
2
Columbia University Graduate School of Architecture, Planning and Preservation, New
York, NY, USA and Urban-Interface, LLC, Newark NJ, USA,
3
University of California, Davis, USA, and
4
US Department of Agriculture Forest Service, Burlington, VT, USA.

Patch dynamics is both an evocative term for interacting with urban design professions, and
the shorthand for a well developed theoretical approach to landscape ecology. This theory
addresses the structure and function of spatial heterogeneity in ecological systems at any
scale. Patch dynamics theory highlights the mosaic or graded structure of spatial
heterogeneity, the flows among patches, the role of patch boundaries, and the temporal
changes in individual patches as well as the entire mosaic. This ecological theory applies
quite well to human ecosystems, such as the varied areas of urban regions, and can
accommodate biophysical, social, buildings, and infrastructural mosaics. These different
mosaics are in fact linked and interacting, and the dynamics of each on is important to the
others. We present examples of these mosaics, using metropolitan Baltimore, Maryland,
USA, as a case study. Furthermore, we present how patch dynamics can be translated to
the fields of urban design, where it encourages a connected, systems approach to design,
and stimulates designs that recognize the dynamic nature of the urban mosaic.
Effectiveness and relevance are demonstrated with two cases. One, an attempt to use
landscape as energy machine/infrastructure, in the Northern Netherlands, and two,
restructuring Rotterdam as ‘Water City’ to enhance its attractiveness, identity and
sustainability by revealing its ecological and land/historical heritage that has been severely
disturbed by a modernist approach to Post-war urban construction.

293
Theme 3. Ecological Networks, fragmentation and connectivity
3.1 Symposium 5: Ecological infrastructure: theory and application

Ecological Infrastructure and Landscape Urbanism

J. Koh

Wageningen University, Netherlands

There is growing awareness among planners and designers that large-scale design must be
more adaptive and open-ended than determined, that design at the urban scale involves
more repair and restructuring than new construction, and that city and landscape must form
an integrated and complementary whole.

Ecological infrastructure (1) provides a useful tool that gives robust structure and continuity
to the open-ended approach to large-scale and long-term design, and (2) demonstrates how
a landscape approach to urban and regional design, landscape urbanism, is relevant,
effective, and contemporary.

Effectiveness and relevance are demonstrated with two cases. One, an attempt to use
landscape as energy machine/infrastructure, in the Northern Netherlands, and two,
restructuring Rotterdam as ‘Water City’ to enhance its attractiveness, identity and
sustainability by revealing its ecological and land/historical heritage that has been severely
disturbed by a modernist approach to Post-war urban construction.

294
Theme 3. Ecological Networks, fragmentation and connectivity
3.1 Symposium 5: Ecological infrastructure: theory and application

Design Ahead of Time: Urban Growth Pattern Based on Ecological Infrastructure

K. Yu, D. Li
1
The Graduate School of Landscape Architecture, Peking University, Beijing China;
e-mail: [email protected]

The Negative Approach — Landscape as Infrastructures for Urban Development

How landscapes change is related to the issue of time. Recognizing that urbanization
and globalization processes are fast and overwhelming, a “negative approach” should be in
contrast with the conventional development planning approach. By “negative”, We mean
landscape architects and planners should plan and guide urban development by identifying
and designing an ecological infrastructure (EI) before the development plan evolves. The EI
is defined as a structural landscape network composed of the critical landscape elements
and spatial patterns that are of strategic significance in safeguarding the integrity and identity
of the natural and cultural landscapes and securing sustainable ecosystem services,
protecting cultural heritages and recreational experiences.
In a conventional model of urbanization, time can be visualized as concentric annual-
ring sprawl. Earlier, greenbelts and green wedges were seen as landscape structures to stop
and prevent sprawl, and they were included in the comprehensive master plan. Current
evidence shows that these greenbelt and wedge dreams have often failed (Ahern, 2002;
2003; Kühn, 1995; Yokohari, 2002).
Landscape ecology, provides a new ecological planning model which focuses on
landscape patterns, horizontal processes and change, provides fundamentals for developing
green infrastructure that can be used to integrate the horizontal processes of urban
development with ecological protection.
At the regional scale, ecological infrastructure is represented as permanent regional
landscape of flood prevention, ecological networks, heritage corridors and recreational
corridors, which are planned for protection and used to define the urban growth pattern and
city form. At the intermediate scale, the regional ecological infrastructure is to be integrated
into the interior urban structure, and become the urban green space system that integrates
various functions such as commuting, cycling, heritage protection and recreational activities.
At the fine scale, the ecological infrastructure is to be used as the defining structure for urban
land development, and to be used to guide the site-specific design.
EI becomes an integrated medium of various processes, bringing nature, man and
spirits together. It is the efficient landscape security pattern to safeguard ecological and
environment integrity, cultural identity and to provide for people’s spiritual needs.

The Case study: Urban Growth Based on Ecological Infrastructure

Taizhou is located at the south east coast of China, with a total area of 9411 square
kilometers and a population of 5.5 million. Of these 0.7 million live in urban areas; the urban
population will increase to 0.9 million in 2010, and 1.3 million in 2020, and 1.5 million in 2030.
Although quite rural and agricultural in character, it is now one of the fastest growing areas in
China due to the booming of small private industries. Under the influence of the monsoon
climate and being adjacent to the east sea, flooding has been a major hazard. In response to
stormwater and flood problems, the landscape has been shaped into a unique form with a
network of water courses that integrate natural water systems, wetlands and man made
ditches, as well as cultural heritage features. This water network landscape, which has been
effective in safeguarding the agricultural processes for thousands of years, is now facing the
challenge of being destroyed by the rapid urbanization process beginning in the 1990s. The
wetlands have been filled, rivers have been straightened and channelized, cultural heritages
that are not listed as protected historical relics have been destroyed, visual and recreational
experiences have been totally ignored. In addressing the above situations, ecological

295
Theme 3. Ecological Networks, fragmentation and connectivity
3.1 Symposium 5: Ecological infrastructure: theory and application

infrastructures at various security levels were then identified and planned to safeguard the
sustainability of the living landscape and guide the urban sprawl according to alternative
development scenarios.
Defining ecological infrastructure at the large scale: The regional EI was planned
through the identification of critical landscape patterns (security patterns) for the targeted
processes (Yu, 1996). The security patterns are composed of elements and spatial positions
that are strategically important in safeguarding the different processes of the landscape.
Models including suitability analysis, minimum cost distance and surface models were
applied to identify security patterns for specific processes. Three security levels - low,
medium and high - are used to define the quality of the security patterns in safeguarding
each of the targeted processes.
Using the three EI alternatives as framing structure, scenarios of regional urban growth
patterns were simulated using GIS: the Adjusted Sprawl Scenario, the Aggregated Scenario,
and the Scattered Scenario. Comparative impact evaluations were made for these scenarios
by a planning committee composed of decision makers of the city, planning experts from all
over the country, stakeholders include officials from various functional departments of the
Taizhou city government, representatives of individual villages who originally owned the land,
representatives of real estate developers and representative of investors who are eagerly
waiting for the development rights for the land. One of the three urban pattern scenarios was
finally selected as the most feasible by the decision makers, after a long time and multiple
brainstorms among the planning committee. As expected, the Aggregated Scenario, which
is based on the medium quality EI, was considered the more balanced and less difficult to be
realized.
Defining ecological infrastructure at the medium scale: Based on the aggregated
Scenario and the green lines of the regional EI, overall design and management guidelines
were developed for the medium quality EI, and especially for the green corridors that function
as critical EI elements in water management and biodiversity conservation, heritage
protection and recreation. During this process of making design guidelines, interactions were
made between the planners and local people.
(6) New models of urban land development: testing EI at the small scale: At a selected
site (10 square kilometers in size), following the EI guidelines developed above, alternative
urban development models were designed to test the possibility of building an EI-based city.
In these, ecosystem services safeguarded by EI are built into the urban fabric so that typical
urban sprawl can be avoided. These new urban land development models were presented
to the developers and investors, as well as the city decision makers, to let them know that
the business-as-usual models of land development can be avoided. The new way of
development by building the EI into their land use scheme will not only help the whole city,
but will also benefit future development ecologically and economically.

References

Ahern, J., 2002. Greenways as Strategic Landscape Planning: Theory and Application.
Wageningen University, The Netherlands.
Manfred Kühn, 2003, Greenbelt and Green Heart: separating and integrating landscapes in
European city regions, Landscape and Urban Planning, 64:19–27.
Yokohari, Makoto, Takeuchi, Takeuchi, Watanable, Takashi, and Yokota, Shigehiro
2000, Beyond greenbelts and zoning: A new planning concept for the environment of
Asian mega-cities. Landscape and Urban Planning, 47:159-171.
Yu, K.-J., 1996. Security patterns and surface model in landscape planning. Landscape and
Urban Planning, 36(5):1-17.

296
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

Does the matrix matter? Landscape structure and population genetic architecture
of amphibians

A.G. Nicieza1, M. Choda1, S. García1, and D. Álvarez1,2

1
Ecology Unit, Department of Biology of Organisms and Systems, University of Oviedo,
Oviedo, Asturias, Spain
e-mail: [email protected]
2
Genetics and Natural Resources Group, Department of Functional Biology, University of
Oviedo, Oviedo, Asturias, Spain

Introduction

The increasing anthropogenic impact on natural ecosystems urges for the design and
application of programmes for monitoring genetic diversity at the appropriate spatial and
temporal scales (Schwarzt et al., 2007). Ideally, these programmes should be based on the
knowledge of fine-scale genetic structures of target species in specific landscapes, which
allow for the identification of evolutionary significant units (ESUs) and delineation of
management units (MUs) (Palsbøll et al., 2007). Given the interdependence between fine-
scale genetic structure and gene flow, and the potential influence of gene flow on population
genetic diversity and long-term population viability (Whitlock and McCauley 1999), assessing
the spatial and geographic scales over which substantial gene flow occurs is a critical step
for the conservation and management of threatened species.
Over the last decades, a significant number of amphibian species have undergone severe
declines or extinction (Stuart et al., 2004). Surprisingly, despite many amphibian species are
particularly deme-structures organisms, very few studies have been conducted to produce
data on gene flow across a fine geographic scale. The paucity of information may be even
worse because one can expect different genetic structures for a species inhabiting different
landscapes. Similarly, differences in reproductive modes or dispersal abilities among species
or populations occupying equivalent landscapes can generate substantial differences in the
spatial organization of genetic diversity. Here we present a study aimed to explore how the
‘interaction’ of spatial landscape patterns and the species’ attributes can mould the structure
of subdivided populations.

Study system

We examine how landscape patterns can affect genetic diversity and its spatial structure
by using a model system consisting of three contrasting landscapes and amphibian species
differing in dispersal and reproductive modes, and dependence on the aquatic habitat. In a
first step we used two species, Rana temporaria and Salamandra salamandra, as model
organisms. In R. temporaria, the sites for breeding and larval development can be viewed as
small islands or archipelagos of aquatic habitat within a surrounding non-habitat matrix of
land. In contrast, in the study area S. salamandra can breed either in small creeks and rivers
(ovoviviparous reproductive mode) or in the absence of aquatic habitat (viviparous
reproductive mode). In the Cantabrian Range, the two species can be found along a 2200 m
altitude gradient. To identify the mechanisms involved in generating different population
structures (“continuous” population, metapopulation, small isolated populations) we sampled
over 30 breeding areas allocated among three different types of landscape: i) a coastal area
with smooth topography, ii) a mountain area with substantial barriers and a soft matrix
(expected low resistance), and iii) a karstic, mountain area with large barriers and a hard
matrix (expected high resistance). We used a hierarchical sampling design which included
four rings with decreasing sampling intensities. Landscapes i) and iii), and the transition zone

297
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

between them, form the first ring, a 2000-km2 area, characterised by a relatively complex
topography. Landscape ii) is in the third ring, but the scale of sampling was equivalent to that
of the coastal and karstic areas. Within these landscape structures, distances between
breeding sites ranged from 0,5-20 km.

For R. temporaria we are examining variability for 12 microsatellite loci to determine

whether a model of isolation by distance or a model incorporating the effects of physical
barriers can explain the observed variation in genetic diversity and the pattern of population
differentiation. In the coastal area, breeding sites have a fine-grained distribution, with a
relatively large number of relatively neighbouring, small units. In contrast, in mountain areas
the species tends to a coarse-grained distribution, with large but more isolated populations.
We hypothesize that, in different landscapes, differences in the degree of isolation between
population units and overall connectivity will generate contrasting types of population
structure and large variation on overall and local genetic diversity. A parallel approach is
being conducted for S. salamandra: we expect the same landscape can generate diverse
population structures in species differing in dispersal and reproductive modes. Our ultimate
objective is to produce a general framework to predict the spatial scales, or the topographic
structures, for evolutionary significant units from data on species life-history and geographic
information.

References

Palsbøll, P.J., Bérubé, M. & Allendorf, F. (2007) Identification of management units using population
genetic data. Trends in Ecology and Evolution 22: 11-16.
Schwartz, M.K., Luikart, G. & Waples, R.S. (2007) Genetic monitoring as a promising tool for
conservation and management. Trends in Ecology and Evolution 22: 25-33.
Stuart, S.N., Chanson, J.S., Cox, N.A., Young, B.E., Rodrigues, A.S.L., Fischman, D.L. & Waller,
R.W. (2004) Status and trends of amphibian declines and extinctions worldwide. Science 306:
1783-1786.
Whitlock, M.C. & McCauley, D.E. (1999) Indirect measures of gene flow and migration: FST ≠
1/(4Nm+1). Heredity 82: 117-125.

298
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

What can landscape genetics contribute to landscape ecology?

V. H. Dale1, G. M. Crutsinger2, A. Classen3

1
Environmental Sciences Division, Oak Ridge National Laboratory,
P.O. Box 2008, Oak Ridge, TN 37831-6036, USA.
e-mail: [email protected]
2
Department of Ecology and Evolutionary Biology, University of Tennessee,
Knoxville, Tn 37996, USA. 3Environmental Sciences Division, Oak Ridge National
Laboratory,P.O. Box 2008, Oak Ridge, TN 37831-6422, USA.

Landscape genetics deals with how genetic variation can affect the expression of
phenotypic variation over space, the extended consequences of phenotypic variation on
communities and ecosystem processes, and in turn, how spatial variations can influence
genetic conditions (Holderegger and Wagner, 2006). Therefore, the new field and techniques
of landscape genetics enrich the types of questions and relationships that can be explored by
landscape ecologists. In particular, they allow a focus on processes (causes and implications
of observed patterns) as well as the patterns themselves. The types of questions that can
now be addressed include constraints and implications of hierarchical changes in diversity
(ranging from genetic to landscape diversity), interactions between diversity within or among
trophic levels, and how landscape structure can affect genetic variability via dispersal and
gene flow. Some of these issues can be explored by using models, and the tools of
landscape genetics provide the ability to consider these issues in quantitative controlled
experiments.

Landscape ecology relates patterns of land cover and land use to environmental
conditions over a variety of geographic scales and provides important theoretical principles
that can be applied to the genetics conditions (Holderegger and Wagner, 2006). For
example, landscape changes (e.g. habitat fragmentation, edge effects between dissimilar
land uses, land cover corridors, and changes in hydrology) can concentrate or distribute a
population, species, or community, thereby affecting ecological processes such as dispersal,
migration, predation, etc. Changes in weather over time or longer-term changes in climate
(including ecological disturbances such as fires and hurricanes) can also impact habitat,
hydrology, and landscape dynamics in ways that affect the genetic variation within a
landscape.

On the other hand, genetics conditions are increasingly being recognized as affecting
landscape patterns and processes. Whitham et al. (2006) show that heritable genetic
variation within individual species has community and ecosystem influences (such as leaf
litter decomposition and N mineralization) in examples from microbes to vertebrates.
Furthermore, the interactions among genotypes within a local population (i.e. genotypic
diversity) can determine the diversity or associated species and drive ecosystem level
processes. For example, Crutsinger et al. (2006) experimentally showed that increasing
population genotypic diversity in a dominant old-field plant species, Solidago altissima,
determined arthropod diversity across trophic levels and increased net primary production
(Figure 1). Surprisingly, the magnitude of the effects of plant genotypes on communities and
ecosystem processes were directly comparable to the effects at the plant species level.
Recent evidence also shows how genetic interactions link terrestrial and aquatic
communities and may have significant evolutionary and conservation implications (Leroy et
al. 2006).

299
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

Fig. 1. The positive

relationship between
genotypic diversity of S.
altissima and ANPP. Each
circle represents plot-level
biomass at peak-growing
season.

But, to date, no one has developed a conceptual model for how genetic influence
may play out at the landscape level or used an organized approach to define and prioritize
the key research questions in the field and how they can be addressed. Holderegger and
Wagner (2006) compiled a guide on the disparate disciplines may best interact. Manel et al.
(2003) reviewed available tools for such studies (although the tools and technologies for
such analyses are rapidly expanding).

Therefore, we have designed a new conceptual model for understanding how genetic
factors contribute to landscape patterns and how landscape conditions affect genetics. A key
focus is on the interactions between these two perspectives and fields of study. We also
demonstrate the benefits of an interdisciplinary approach in addressing the mechanisms
underlying landscape genetics and develop the key research questions and conceptual
model of how landscape ecology interacts with genetics. Our perspective is that landscape
ecology and genetics interact in a complex manner to explain hierarchical patterns and
processes at the genetic-, species-, and ecosystem-level. Linking these disparate fields of
study allows innovations within each field as well as a commitment to a new holistic
perspective on analytic tools that may enable scientists to better answer complex global
questions.

References

Crutsinger, G.M.; Collins, M.D.; Fordyce, J.A.; Gompert, Z.; Nice, C.C. & Sanders, N.J. (2006)
Plant genotypic diversity predicts community structure and governs an ecosystem process.
Science 313: 966-968.
Holderegger, R. & Wagner, H.H. (2006) A brief guide to landscape genetics. Landscape Ecology 21:
793-796.
LeRoy, C.J.; Whitham, T.G.; Keim, P. & Marks, J.C. (2006) Plant genes link forests and streams.
Ecology 87: 255-261.
Manel, S.; Schwartz, M.K.; Luikart, G. & Taberlet, P. (2003) Landscape genetics: combining
landscape ecology and population genetics. Trends in Ecology & Evolution 18: 189-197.
Whitham, T.G.; Young, W.P.; Martinsen, G.D.; Gehring, C.A.; Schweitzer, J.A.; Shuster, S.M.;
Wimp, G.M.; Fischer, D.G.; Bailey, J.K.; Lindroth, R.L.; Woolbright, S. & Kuske, C.R. (2006)
Community and ecosystem genetics: A consequence of the extended phenotype. Ecology 84:
559-573.

300
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

Molecular genetic testing of landscape ecological hypotheses: the dynamics of

amphibian species distributions

J. W. Arntzen

National Museum of Natural History | Naturalis - P. O. Box 9517,2300 RA Leiden, The

Netherlands,
e-mail: [email protected]

Introduction

Most amphibian species in the Holarctic realm (frogs, toads and salamanders) reproduce
in standing water. Water bodies such as ponds are mostly man-made and spaced out (water
supply to cattle, reservoirs for fire-fighting etc.) and not randomly distributed over the
landscape. In spring adult amphibians migrate to the ponds for breeding after which they
return to the terrestrial habitat where they will spend most of the year for shelter and
foraging. Given that adult amphibians show breeding site fidelity with low dispersal, ponds
represent localized breeding populations (or demes) and the matrix of occupied ponds in the
landscape represents a metapopulation. Therewith, amphibians provide us with a well-
defined and tangible system for spatial ecological research (Jehle et al., 2005a).
Dispersal comes into effect with the successful reproduction of an individual in a pond
other than the one it was born. Under this definition dispersal amounts to gene-flow. In
amphibians the spread from the natal pond takes mostly place during the juvenile phase, in
between metamorphosis and maturation. At the population level, amphibian survival depends
on the metapopulation structure, in particular on the dispersal among local populations
(Halley et al., 1996) and dispersal is thought to be a function of the distance between ponds
and the characteristics of the intervening habitat. This provides an excellent setting to study
the effect of landscape features on amphibian population persistence, for the purpose of
nature conservation and management. Since juvenile amphibians are small and not
amenable to tagging the inferences on dispersal (gene-flow) will be indirect, using molecular
genetic markers (Jehle and Arntzen, 2002).

Materials and methods

Surveys were conducted on the pond presence/absence of toads (Bufo bufo) and
salamanders (Triturus cristatus, Triturus marmoratus) in agricultural landscapes of Britain
(259 ponds, Leicestershire) and France (300+ ponds, Mayenne). Landscape features such
as agricultural field use, forestation, waterways, presence of competing species etc. etc.
were gathered from topographical maps and satellite remote imagery. Logistic regression
was used to identify landscape features associated with amphibian occurrence. Genetic
variation at 3, 8, and 5 mini- and microsatellite loci (amounting to 9500 genotypes) was used
to estimate pairwise genetic distance between occupied demes. I worked under the
assumption that good habitat to live in is good habitat to disperse through (and the converse)
and I tested with partial-Mantel tests the hypothesis that dispersal between demes is not only
affected by distance (i.e., isolation by Euclidean distance) but also influenced by the quality
of the intervening habitat (i.e., 'ecological distance').

Results

Firstly, positive and statistically significant effects were found in salamanders between the
connectedness of populations and local population size. Secondly, genetic distance and
Euclidean distance between populations were significantly associated in each species
investigated. The relationship between connectedness and population size was frequently
asymmetric, with gene-flow directed from large to small populations. Additional associations
with habitat features were insignificant, except when ecological distance included the

301
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

presence of rivers as barriers to dispersal in Triturus cristatus in Britain and the presence of a
competing Triturus species (T. cristatus vs. T. marmoratus) as a barrier to dispersal in
France.

Discussion

Effects of isolation by distance (Euclidean, ecological) were demonstrated. The ecological

effects operated for relatively large scale phenomena (rivers, interspecific competition) and
not for small landscape elements (hedgerows, spinneys, ditches) or land use, that are often
hypothesized to affect metapopulation connectivity and thought to be crucial to amphibian
survival. Asymmetric gene-flow is probably related to reproductive output and the numbers of
juveniles available for dispersal (Jehle et al., 2005b). However, reproductive success
fluctuates dramatically from pond to pond and from year to year, rendering the system fairly
intractable at small spatial and temporal scale.
As a by-product of the amphibian survey in France, I discovered pockets of T.
marmoratus completely surrounded by T. cristatus (cf. Arntzen & Wallis, 1991). Because
amphibians have limited terrestrial dispersal capability and because the cristatus -
marmoratus species distribution is largely parapatric, I argue that the enclaves are the
signature of a T. cristatus range expansion at the expense of T. marmoratus. Elsewhere I
found enclaves for fire bellied toads in central Europe (Bombina bombina, B. variegata), for
marbled news in central Portugal (Triturus marmoratus, T. pygmaeus) and for crested newts
in Serbia (Triturus cristatus superspecies). The enclaves vary in size from several to
hundreds of km and vary in age from decades to thousands of years. I argue that in
amphibians and other species with low dispersal capability enclaves help to reconstruct past
distributions and that pattern (distribution) implies process (dispersal). GIS-based analysis of
landscape ecological parameters help to fine-tune the inferred biogeographical scenarios.
The generated hypotheses can be tested with the help of molecular genetic data. In my
congress presentation I will illustrate this approach, with examples from toads and
salamanders.

References
Arntzen, J.W. & Wallis, G. (1991) Restricted gene flow in a moving hybrid zone of newts (Triturus
cristatus and T. marmoratus) in western France. Evolution 45: 805-826.
Halley, J; Arntzen, J.W. & Oldham, R. (1996) Predicting the persistence of amphibian populations
with the help of a spatial model. Journal Applied Ecology 33: 455-470.
Jehle, R. & Arntzen, J.W. (2002) Microsatellite markers in amphibian conservation genetics.
Herpetological Journal 12: 1-9.
Jehle, R.; Burke T. & Arntzen, J.W. (2005a) Delineating fine-scale genetic units in amphibians:
probing the primacy of ponds. Conservation Genetics 6: 227-234.
Jehle, R.; Wilson, G.A.; Arntzen, J.W. & Burke, T. (2005b) Contemporary gene flow and the spatio-
temporal genetic structure of subdivided newt populations (Triturus cristatus, T. marmoratus).
Journal Evolutionary Biology 18: 619-628.

302
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

Inferring the influence of landscape on roe deer gene flow using connectivity
estimates based on a weighted combination of several landscape features

A. Coulon1, J.F. Cosson 2, N. Ray 3, A.J.M. Hewison1

1
Comportement et Ecologie de la Faune Sauvage, Institut National de la Recherche
Agronomique, BP 27, Castanet-Tolosan Cedex, F 31326, France ; present : Cornell
Laboratory Of Ornithology,159 Sapsucker Woods Road, Ithaca, New York 14850, USA.
e-mail: [email protected]
2
Centre de Biologie et de Gestion des Populations, Institut National de la Recherche
Agronomique, Campus International de Baillarguet, CS 30016, Montferrier-sur-Lez, F 34988,
France
3
Computational and Molecular Population Genetics, Zoological Institute, University of Bern,
6, Baltzerstrasse, CH-3012 Bern, Switzerland

Introduction

Landscape genetics has been touted as a powerful approach to infer the influence of
landscape on animal dispersal movements. One method consists of comparing how different
types of geographic distances between populations or individuals explain genetic distances
between them. This approach usually compares straight-lines and least-cost paths, the latter
being the shortest pathways which follow or avoid a given landscape element hypothesised
to influence dispersal movements (e.g. Michels et al. 2001, Berthier et al. 2005). A better
correlation between genetic distances and geographic distances along least-cost paths
indicates that they are closer to the real dispersal paths than the straight-lines, and that the
landscape feature they are based on influences gene flow and dispersal movements. Until
now, most least-cost distances have been calculated in relation to the distribution of a single
type of landscape feature. However, dispersal movements of a given species are likely to be
influenced to varying degrees by several landscape features. It would hence be valuable to
compute least-cost paths reflecting this combined influence of several landscape elements. A
few studies combined several landscape features but without taking into account the fact that
they probably influence dispersal movements to different degrees (Spear et al. 2005, Vignieri
2005). Indeed, the difficulty is to set the relative weight of each element. Here, we used direct
data of roe deer (Capreolus capreolus) movements sampled by GPS collars to model multi-
feature connectivity and test its influence on roe deer gene flow in a fragmented landscape.

Estimation of multi-feature connectivity

Roe deer movement analysis

We fitted 20 roe deer with GPS collars recording each animal’s location at regular time
intervals (one fix every 2 to 6 hours). We used these locations and a GIS of the study area to
model with a Step Selection Function (SSF, Fortin et al. 2005) roe deer preferences for
landscape features during movements. An SSF describes the probability that a given
segment of the landscape is used by an individual during a movement step, as a function of
several variables. Here, we included the extent of woodland around the step, the distance of
the step to the nearest road, and the topographic position of the step (i.e. valley bottom or
ridge). The selected model indicates that in the study area, moving roe deer tend to avoid
roads, but only in relatively open landscapes; it also suggests that they avoid most forested
sectors and valley bottoms.

Connectivity modelling
The model estimated from the GPS data enables to estimate the probability of use of a
landscape segment as a function of the simultaneous and weighted presence of several
landscape elements. We extrapolated multi-feature connectivity values from this model by

303
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

applying it to the full extent of the study area. In this setting, the probability of use of a
landscape unit (pixel) is directly related to its connectivity.

Influence of connectivity on roe deer gene flow

We sampled 800+ roe deer from two genetically differentiated populations covering an
area of 40 × 55 km (containing the smaller area where the GPS study was performed). We
genotyped them at 11 microsatellite loci. We used the multi-feature connectivity map to
elaborate least-cost paths between genotyped roe deer within each population. To test the
influence of this connectivity on dispersal movements, we then carried out Mantel tests
between inter-individual genetic distances and several variables reflecting length and
landscape composition of alternative paths.
Preliminary results indicate that genetic distances between roe deer are not always better
explained by least-cost distances compared to Euclidean distances (Table 1). However, tests
involving potentially more informative characteristics of alternative paths (i.e. incorporating
information on the type of landscape crossed by least-cost paths) are currently being
performed. We will also estimate the gain of combining several landscape features to
estimate connectivity, compared to a single-variable approach.

Table 1. Correlation coefficients and P-values (in brackets) of the Mantel tests between
pairwise genetic and geographic distances, for two types of geographic distances.

Population Sex Euclidean distance Least-cost distance

North-western (N=667) Males 0.0162 (0.081) 0.017 (0.078)
Females 0.0325 (0.003) 0.0344 (0.001)
South-western (N=173) Males 0.041 (0.03) 0.0398 (0.038)
Females 0.0342 (0.073) 0.0333 (0.083)

References
Berthier, K.; Galan M.; Foltête J. C.; Charbonnel N. & Cosson J.-F. (2005) Genetic structure of the
cyclic fossorial water vole (Arvicola terrestris): landscape and demographic influences. Molecular
Ecology 14: 2861-2871.
Fortin, D.; Beyer H. L.; Boyce M. S.; Smith D. W.; Duchesne T. & Mao J. S. (2005) Wolves
influence elk movements: behavior shapes a trophic cascade in Yellowstone national park.
Ecology 86: 1320-1330.
Michels, E.; Cottenie K.; Neys L.; De Gelas K.; Coppin P. & De Meester L. (2001) Geographical
and genetic distances among zooplancton populations in a set of interconnected ponds: a plea for
using GIS modelling of the effective geographical distance. Molecular Ecology 10: 1929-1938.
Spear, S. F.; Peterson C. R.; Matocq M. D. & Storfer A. (2005) Landscape genetics of the blotched
tiger salamander (Ambystoma tigrinum melanostictum). Molecular Ecology 14: 2553-2564.
Vignieri, S. N. (2005) Streams over mountains: influence of riparian connectivity on gene flow in the
Pacific jumping mouse (Zapus trinotatus). Molecular Ecology 14: 1925-1937.

304
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

Influences of landscape features on gene flow of Martes Americana in northern

Idaho, USA

T. N. Wasserman¹, S.A. Cushman², M.K. Schwartz²

¹Huxley College of the Environment, Western Washington University, Bellingham, WA., USA
e-mail: [email protected]. 360-650-7353
²USDA Forest Service, Rocky Mountain Research Station, 800 E Beckwith, Missoula, MT.

Introduction

Forest fragmentation can have a dramatic effect on landscape connectivity and

dispersal of animals, potentially reducing gene flow within and among populations. American
marten populations (Martes americana) are sensitive to forest fragmentation and the spatial
configuration of patches of remnant mature forest has an important impact on habitat quality.
Habitat fragmentation leads to a decrease in landscape connectivity, thus hindering
movement among resource patches. This alteration can result in a reduction in gene flow
between populations, leading to a loss of genetic diversity within fragments.

Methods

In this study, three landscapes were surveyed covering the Selkirk, Purcell, and
Cabinet Mountain ranges in Northern Idaho over four winters where genetic information from
Martes americana were collected via non-invasive methods. Hair snares were set along
transects across the Idaho Panhandle Nat’l Forest during the winter months of January,
February, and March of 2004, 2005, 2006 and 2007. Each trap was baited with deer meat,
beaver castor, and gusto, a commercial call lure. Each trap was lined with 5 gunbrushes to
non-invasively obtain hair samples from animals visiting the snare, from which genetic data
was obtained. Bait stations were set for 2 weeks after which each station was revisited to
collect hair samples and re-baited for another 2-week cycle. During the check, hair was
collected from the gunbrushes. Each gunbrush is considered a single sample. Hair samples
were sent to the RMRS Wildlife Genetics Lab in Missoula, MT. Genetic relationships were
determined using diagnostic restriction enzyme patterns followed by amplification of a region
of cytochrome b on mitochondrial DNA.
The study area lies within the Idaho Panhandle Nat’l Forest and is a 4,000 square
kilometer area encompassing the Selkirk, Purcell, and Cabinet mountains. The topography is
mountainous, with steep ridges, narrow valleys and many cliffs and cirques at the highest
elevations. Elevation ranged from approximately 700m to 2000m above sea level. The
Kootenai River trench runs down the middle of the study area, separating the Selkirk
Mountains on the west from the Purcell Mountains on the east, with a five to seven mile wide
unforested, agricultural valley and a broad, deep river between. The climate is characterized
by cold, wet winters and mild summers. The area is heavily forested, with Abies lasiocarpa
and Picea engelmannii codominant above 1300 meters, and a diverse mixed forest of
Pseudotsuga menziesii, Pinus contorta, Pinus ponderosa, Pinus monticola, Abies grandis,
Tsuga heterophylla, Thuja plicata, Larix occidentalis, Betula papyrifera, Populus tremuloides,
Populus trichocarpa dominating below 1300 meters.

Results

Over 100 individual marten were detected across the 4,000 square kilometer study
area. The genetic similarities were based on the pair-wise percentage dissimilarity among all

305
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

individuals based on 9 microsatellite loci. We compared their genetic similarities with several
dozen landscape resistance hypotheses. The landscape resistance hypotheses describe a
range of potential relationships between movement cost and landcover, slope, elevation,
roads, Euclidean distance and a putative movement barrier. These hypotheses were divided
into several organizational models.

The degree of support for each model was tested with causal modeling on
resemblance matrices using partial Mantel tests. Quantifying the relationship between
landscape structure and gene flow can give biologists insight into connectivity of populations
and metapopulations through space and time. Correlating genetic similarity of individuals
across large landscapes with hypothetical movement cost models can give reliable
inferences about population connectivity. By linking cost modeling to the actual patterns of
genetic similarity among individuals it is possible to obtain rigorous, empirical models
describing the relationship between landscape structure and gene flow, and to produce
species-specific maps of landscape connectivity.

306
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

Corridor network connectedness does not necesseraly enhance biological

connectivity: the case of the landscape genetics and pollen flows of Primula
vulgaris in a hedgerow network landscape

P. Campagne, A. Baumel, L. Affre, P. Roche, T. Tatoni

IMEP, CNRS UMR 6116 (Institut Méditerranéen d'Écologie et de Paléoécologie), Université

Paul Cézanne (Aix-Marseille III), Europole méditerranéen de l'Arbois, Bâtiment Villemin, BP
80. F - 13545 Aix-en-Provence cedex 04 FRANCE.
e-mail: [email protected]

Introduction

Landscape connectivity is generally considered to facilitate gene flow between

populations. Hence, highly connected hedgerow networks are often thought to support higher
gene flow than disconnected hedgerow networks or open landscapes.
In order to test this hypothesis, we studied gene flow between Primula vulgaris populations at
the landscape level. This species is restricted to wooded elements (patches and hedgerows)
and we considered two contrasted study sites (dense and connected hedgerow networks vs.
disconnected, open hedgerow networks).

Material and Methods

We used AFLP markers to study the spatial population genetics and a fluorescent powder
as a pollen analogue to study spatial pollen flow. We also examined ant dispersed seed flow.

Results and Discussion

Figure 1. Moran’s Istd correlograms for AFLP markers in dense and sparse hedgerow
networks. Error bars indicate 95% confidence intervals. To simplify, upper distance classes
(more than 1200 ml) were excluded from analysis.

307
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

Figure 2. Predicted values for a GLM model (link: logarithm; Quasi-poisson distribution) of
fluorescent powder frequency as a function of distance to powder source.

Our results indicated a strong spatial genetic autocorrelation that is significantly more
pronounced in the dense and connected hedgerow networks than in the open and
disconnected ones up to 400 m (Figure 1). This means that for a given distance, two
individuals are more similar in the open landscape than in the dense connected landscape.
Seed dispersal by ants is very short scale (about few metres), while pollen flow estimated by
pollen dye analogues appeared to occur up to 400 m. When considering the decay of
fluorescent dye frequency with distance from sources, it appears that pollen analogues are
dispersed to significantly longer distances in the open and disconnected landscape (Figure 2;
Table 1).

Table 1. Parameter estimation by the GLM model (link: logarithm; Quasi-poisson distribution)
of fluorescent powder frequency as a function of distance to powder source.

Estimate Std Error t value p value

Intercept 2.75 0.078 35.03 < 0.001
Distance -0.04 0.005 -8.03 < 0.001
Type of network (Sparse) -0.33 0.113 -2.87 < 0.01
Interaction 0.02 0.005 4.48 < 0.001
These results lead us to conclude that while landscapes with dense and connected
hedgerow networks offer a continuous distribution of suitable habitat for Primula vulgaris, in
the mean time it leads restricted gene flow, mainly through reduced pollen dispersal.

308
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

A comparison of current methods for linking genetic variation to landscape

heterogeneity using simulated data.

N. Balkenhol, L. P. Waits

University of Idaho, Department of Fish & Wildlife Resources, Moscow, ID, , USA.
e-mail: [email protected]

Introduction

The goal of landscape genetics is to describe and explain how landscape attributes affect
genetic variation of plant and animal populations. Various statistical methods for linking
genetic information to landscape data exist, and many of them are described in Manel et al.
(2003) and Storfer et al. (2006). There are, however, very few published studies that test the
utility of different approaches, or compare them under realistic scenarios. Due to the rapid
rate of analytical development, no consensus exists on the best approach for relating genetic
variation to landscape attributes. Although it is unclear whether current statistical methods
produce comparable and repeatable results, a large number of researchers have applied
these methods to a variety of research questions in molecular ecology and conservation
biology (Holderegger and Wagner, 2006; Storfer et al., 2006).
To evaluate the performance of different methods, we analyzed simulated data with
known landscape genetic relationships using eight of the most commonly used statistical
approaches in landscape genetics. This simulation study represents the first rigorous
comparison of the multitude of current approaches in landscape genetics.

Methods

Landscape genetic simulations

We identified and simulated six landscape genetic scenarios commonly encountered in
empirical datasets. These scenarios include:

1. Isolation-By-Distance (IBD). In this scenario, spatial (i.e., Euclidean) separation

distance influences genetic differentiation among populations.
2. Landscape resistance (LR). In this scenario, the landscape matrix shows varying
degrees of movement resistance, thereby influencing gene flow and genetic
differentiation among populations.
3. IBD and LR. A combination of spatial separation distance and matrix resistance
influences population genetic differentiation.
4. Habitat relationships. Genetic differentiation is influenced by habitat dependent
dispersal, leading to greater gene flow within than between habitat types.
5. Barrier. Genetic differentiation among populations is influenced by one or more
(linear) barriers to movement.
6. Panmixia. Genetic differentiation is not influenced by space or landscape attributes.

The different scenarios were simulated using a combination of Easypop (Balloux, 2002)
and ArcGIS (ESRI, 2005). We first simulated the spatial distribution of populations and the
landscape factors influencing genetic differentiation in the GIS. We then used this spatially
explicit data for the genetic simulations in Easypop. A set of control criteria was calculated
for each simulation, to verify that the desired landscape-genetic relationships were
successfully simulated. All landscape scenarios and the population genetic data were
simulated multiple times with varying configurations, so that valid conclusions could be drawn
from the analyses.

Compared statistical approaches

309
Theme 3. Ecological Networks, fragmentation and connectivity
3.2 Symposium 7: Landscape genetics: testing landscape effects on gene flow

The simulated data were analyzed with different statistical approaches that can be used
to link genetic information to landscape attributes. These approaches include Mantel tests
(Mantel, 1967), RELATE (Clarke and Warwick, 2001), partial Mantel tests (Smouse et al.,
1986), multiple matrix regression (Legendre et al., 1994), BIOENV (Clarke and Ainsworth,
1993), a hierarchical Bayesian approach (Foll and Gaggiotti, 2006), distance-based
redundancy analysis (dbRDA; Legendre and Anderson, 1999), and canonical correlation
analysis (CCA; ter Braak, 1986).
We used these statistical approaches to test different landscape genetic hypotheses that
correspond to the simulated scenarios. We then compared results obtained with the different
methods, and the conclusions that researchers would draw from them, using a causal
modeling framework.

Results & Discussion

Our results suggest that the different approaches do not always identify the same
landscape factors as having an effect on genetic differentiation, even under relatively simple
scenarios. This could potentially lead to erroneous conclusions about landscape genetic
relationships, if only a single method is used for the analysis. We therefore recommend that
several methods should be applied to a given empirical data set, and that the statistical
approaches should carefully be matched to the ecological hypotheses tested. The results
will help scientists choose appropriate methods for particular research questions and data.
Our analyses are also useful for teaching landscape genetics, and can help to identify further
research needs in this young field.

References
Balloux, F. (2002) EasyPop version 1.8: A software for population genetics simulations. I.C.A.P.B.
(Institute of Cell, Animal and Population Biology), University of Edinburgh, Edinburgh, Scotland
UK.
Clarke, K. & Ainsworth, M. (1993) A method of linking multivariate community structure to
environmental variables. Marine Ecology Progress Series 92: 205-219.
Clarke, K. & Warwick, R. (2001) Change in marine communities: an approach to statistical analysis
and interpretation, 2nd edition. PRIMER-E, Plymouth.
ESRI (2005) ArcGIS - ArcInfo. Environmental Systems Research Institute, Inc.
Foll, M. & Gaggiotti, O.E. (2006) Identifying the environmental factors that determine the genetic
structure of populations. Genetics 174: 875-891.
Holderegger, R. & Wagner, H.H. (2006) A brief guide to Landscape Genetics. Landscape Ecology
21: 793-796.
Legendre, P. & Anderson, M.J. (1999) Distance-based redundancy analysis: Testing multispecies
responses in multifactorial ecological experiments. Ecological Monographs 69: 1-24.
Legendre, P; Lapointe, F.-J. & Casgrain, P. (1994) Modeling brain evolution from behavior: A
permutational regression approach. Evolution 48: 1487-1499.
Mantel, N. (1967) The detection of disease clustering and a generalized regression approach. Cancer
Research 27: 209-220.
Smouse, P.E; Long, J.C. & Sokal, R.R. (1986) Multiple regression and correlation extensions of the
Mantel test of matrix correspondence. Systematic Zoology 35: 627-632.
Storfer, A; Murphy, M. A; Evans, J. S; Goldberg, C. S; Robinson, S. J; Spear, S. F; Dezzani, R;
Delmelle, E; Vierling, L. & Waits, L. (2006) Putting the “Landscape” in Landscape Genetics.
Heredity advance online publication 1 November.
ter Braak, C. (1986) Canonical Correspondence Analysis: A new eigenvector Technique for
multivariate direct gradient analysis. Ecology 67: 1167-1179.

310
Theme 3. Ecological Networks, fragmentation and connectivity
3.3 Symposium 16: Animals on thermal landscapes

3.3 Symposium 16: Animals on thermal landscapes

Global climate model data for the Predictive Modeling in Seascapes: Pelagic
Seabird-Habitat associations revisited

F. Huettmann1
1
EWHALE lab, Institute of Arctic Biology, Biology and Wildlife Department, University of
Fairbanks-Alaska, Fairbanks Alaska 99775 USA.
e-mail: [email protected]

Introduction

Seabirds occur on all oceans; they are among the most abundant bird species in the
world. Besides many well-known breeding populations on colonies, the majority of birds is
actually found at sea where they spend by far most of their lives (Schreiber & Burger 2002;
Gaston 2004). Most habitat associations of pelagic seabirds are poorly known, and often
vaguely described quantitatively and without confidence as being ‘associated’ with prey
patches, shelf-edges, and proximity to colonies. Much confusion on seabird-habitat
characteristics stems from soft or inappropriate concepts to describe seabird-habitat
associations (Jones 2001; Gottschalk et al. 2005 for review), and from ignoring scale
(Huettman and Diamond 2006 for review and application). Often, such studies are done
directly on variables measured in the field (‘in situ’) using an opportunistic small scale study
design, rather than using generalized and well-assessed data, or ones collected that address
specific questions with the goal that findings should be generally applicable (Aebischer &
Robertson 1993; Braun 2005). Many of such investigations deal with prey and space, but
ignored the importance of temperature layers.
Based on six intensive seabird-habitat models, here I investigate whether the online
publication of climate data like the World Ocean Atlas (WOA Levitus 1994: namely air, sea
surface SST and water body temperature) provided for a major change in our knowledge of
seabird habitat.

Methods

The models assessed here used besides WOA additional ocean climate datasets such as
COADS (e.g. air pressure) or the NSCAT instrument (e.g. wind speed and wave height). The
analysis of the climate layers was done by the author and colleagues in concert with
traditional habitat descriptors using progressive modelling software tools (GLM, Mixed
Models, CART, MARS, TreeNet and Ecological Niche Modeling) applied in the North Atlantic
and Gulf of Maine (>15 seabird species), Falkland Islands (Black-footed Albatross), coastal
British Columbia (Marbled Murrelet) and Bering Sea (Short-tailed Albatross) (Huettmann &
Diamond 2001, Yen et al. 2004, Pittmann & Huettmann 2006, Huettmann et al. unpublished
for Albatross)). These studies were then analysed for providing progress in our
understanding of seabird-habitat relationships by ranking climate layers vs. traditional habitat
predictors.

Results

Results suggest that ‘Seascape Ecology’ offers large opportunities for study (see also
Huettmann & Diamond 2006), especially when climate layers are incorporated. In many
instances, climate layers are driving the large-scale seabird distribution by forming entire
ecosystems with a specific species set-up. However, other than plain correlations, it is
currently not well addressed how SST exactly links with air and waterbody temperatures.
Virtually all concepts from Landscape Ecology and terrestrial wildlife can be applied to

311
Theme 3. Ecological Networks, fragmentation and connectivity
3.3 Symposium 16: Animals on thermal landscapes

seabirds and how they use seascapes. The advent of climate layers has also added a new
dimension to seabird-habitat studies and relationships. The fact that these climate layers are
provided online and for free has created a global role model for data delivery in general (e.g.
www.worldclim.org).

Discussion

Results indicate that seabird habitat preferences are robust in space when implementing
general at-sea weather patterns. Climate drives much of the species set-up in an ecosystem.
Specific climate-related topics such as the ‘tail wind’ hypothesis or the value of using Di-
Methyl Sulfid (DMS; a plankton-based climate-change aerosol affiliated with clouds) are not
used to their full potential, yet. It is obvious that SST affects predatory performance through
‘energy’ questions, e.g. burst speed of the prey, and thus affects the large-scale distribution
of seabirds (Cairns et al. in review, Newton 2003). Therefore, the current global warming
trend will change seabird communities drastically, specifically in high latitudes (Schreiber &
Burger 2002). This will add further to a stressed ocean environment (Myers & Worms 2003)
and requires careful consideration for the sustainable management of seabirds and their
seascape habitats (Nettleship 1991).

References
Aebischer, N. and Robertson P.A. (1993) Testing for Resource Use and Selection by Marine Birds:
A Comment. Journal of Field Ornithology 65: 210-213.
Braun, C. E. (2005) Techniques for Wildlife Investigations and Management. The Wildlife Society
(TWS), Bethesda, Maryland USA.
Cairns, A. J., Gaston, T. & Huettmann, F. (in review) Endothermy, Ectothermy, and the Global
Structure of Marine Vertebrate Communities. Marine Ecology Progress Series
Gaston, A.J. (2004) Seabirds: A Natural History. Yale University Press, London.
Gottschalk, T., Huettmann, F. & Ehlers, M. (2005) Thirty years of analysing and modelling avian
habitat relationships using satellite imagery data: a review. International Journal of Remote
Sensing 26: 2631-2656.
Huettmann, F. & Diamond, A.W. (2006) Large-scale effects on the spatial distribution of seabirds in
the Northwest Atlantic. Landscape Ecology 21: 1089-1108.
Huettmann, F. & Diamond, A.W. (2001) Seabird colony locations and Environmental determination of
seabird distribution: A spatially explicit seabird breeding model in the Northwest Atlantic.
Ecological Modelling 141: 261-298.
Jones, J. (2001) Habitat Selection Studies in Avian Ecology: A Critical Review. Auk 118:557-562
Levitus S.(1994) World Ocean Atlas 1994, 4 volumes, Washington, D.C., National Oceanic and
Atmospheric Administration, National Oceanographical Data Center.
Myers, R.A. & Worm. B. (2003) Rapid worldwide depletion of predatory fish communities. Nature
423: 280-283.
Newton, I. (2003) The Speciation & Biogeography of Birds. Academic Press, London
Nettleship, D.N. (1991) Seabird management and future research. Colonial Waterbirds 14:77-84.
Pittmann, S. & Huettmann, F. (2006). Chapter 4 - Seabird Distribution and Diversity. In: An
Ecological Characterization of the Stellwagen Bank National Marine Sanctuary Region:
Oceanographic, Biogeographic, and Contaminants Assessment. Battista, T., R. Clark, S. Pittmann
(eds). NOAA Technical Memorandum NCCOS 45. Silver Springs, MD.
Schreiber, E.A. & Burger, J. (eds). (2002) Biology of Marine Birds. CRC Marine Biology Series
(P.L.Lutz, ed.) CRC Press, Washington D.C.
Yen, P., Huettmann, F. & Cooke, F. (2004). Modelling abundance and distribution of Marbled
Murrelets (Brachyramphus marmoratus) using GIS, marine data and advanced multivariate
statistics. Ecological Modelling 171: 395-413.

312
Theme 3. Ecological Networks, fragmentation and connectivity
3.3 Symposium 16: Animals on thermal landscapes

Amphibian physiology and landscape permeability: Modeling thermal landscapes

to predict amphibian movements

P.E. Bartelt1, R.W. Klaver2, W.P. Porter3, D.S. Pilliod4, C.R. Peterson5, A.L. Gallant2
1
Department of Biology, Waldorf College, Forest City, IA 50436 USA.
e-mail: [email protected]
2
U.S. Geological Survey, Center for Earth Resources Observation and Science (EROS),
Sioux Falls, SD 57198 USA
3
Department of Zoology, University of Wisconsin, Madison, WI 53706 USA.
4
U.S. Geological Survey, Snake River Field Station, Boise, ID USA.
5
Herpetology Laboratory, Department of Biological Sciences, Idaho State University,
Pocatello,ID 83209 USA.

Introduction

Calculating estimates of landscape permeability can be an effective way of

understanding patterns of distribution and movements of animals across landscapes (Ray et
al., 2002; Chardon et al. 2003). Amphibians require resources that often are dispersed
across a landscape. For example, breeding ponds, foraging areas, and hibernacula may be
separated by distances of kilometers. Costs associated with overland movements can affect
the persistence of amphibian populations and help explain metapopulation dynamics (Funk
et al., 2005). We used thermal and moisture landscapes and cost-surface analysis to model
amphibian movement corridors through terrestrial habitats. Because amphibians are closely
coupled with their physical environment, the costs of amphibians' overland movements
reflect their ability to conserve body water while moving through suitable thermal gradients.
Land use patterns that alter habitat structure can increase the costs of overland movements
for amphibians, which can contribute to population declines.

Methods

Terrestrial habitats were modeled with first principles models of microclimate, heat and
mass transfer, and amphibian physiology (Porter and Mitchell 2003). Spatial input for these
models included topographic data, spatially interpolated meteorological data, and maps of
vegetation cover types and density. We parameterized the models with measures of animal
morphology and physiology, such as preferred core body temperatures and tolerance of
dehydration. Model outputs included animal core body temperature, rates of evaporation, and
discretionary energy. Using a cost-path function, we produced maps of least-cost paths and
corridors of accumulated costs between source and destination habitat patches. These paths
and corridors are measures of landscape permeability. We also produced sets of random
paths and straight line paths to test the hypothesis that actual paths used by amphibians
were less costly than random and straight line paths between the same source and
destination habitats.

Results & Discussion

We applied this analysis to populations of Western toads (Bufo boreas) and Columbia
spotted frogs (Rana luteiventris), both of which inhabit forested landscapes in Idaho, USA.
We used our results to map movement corridors that reflected actual paths of telemetered
individuals. Initial analyses showed that actual paths fell within some of the least costly
movement corridors and avoided high-cost areas. However, total average accumulated costs
for six female toads showed that actual paths were about 50% more costly than least-cost
paths and were not significantly different from the cost of straight-line paths. In a previous
study, these toads were highly terrestrial, active during the day (e.g., foraging), and they
selected habitats that increased their ability to maintain warm body temperatures. For

313
Theme 3. Ecological Networks, fragmentation and connectivity
3.3 Symposium 16: Animals on thermal landscapes

example, shrub habitats with little canopy cover had the warmest and most humid conditions
among habitats used by toads (Bartelt, 2000) and had the highest selection measures
(Bartelt et al., 2004). However, the toads traveled primarily at night (when humidity was high)
and this may help explain why costs of actual paths were not closer to that of least-cost
paths. Oriented random paths were about twice as costly as actual paths, although some of
these differences may be a function of the parameters set for random paths. Therefore, the
applicability of this modeling approach may not fully explain amphibian movement corridors,
but the approach still has value for modeling habitats suitable for retreat and daily activities.
Analysis of similar data for spotted frogs continues at the time of this writing.

Several studies provide evidence that certain land cover changes – e.g., forest
harvests (deMaynadier and Hunter, 1998) and habitat fragmentation (Funk et al., 2005) –
can affect the distribution and movements of amphibians. Because global climate change
also would alter microclimates among habitats, it, too, may affect amphibian populations.
Refining this approach to modeling amphibian terrestrial habitats may help us better
understand these relationships and minimize disruption to movement corridors, thereby
improving amphibian conservation.

References
Bartelt, P.E. 2000. A biophysical analysis of habitat selection in Western toads (Bufo boreas) in
southeastern Idaho. Ph.D. Dissertation. Idaho State University. Pocatello, Idaho.
Bartelt, P.E., C.R. Peterson, and R.W. Klaver. 2004. Sexual differences in the post-breeding
movements and habitats selected by Western toads (Bufo boreas) in southeastern Idaho.
Herpetologica 60: 455-467.
Chardon, J.P., F. Adriaensen, E. Matthysen. 2003. Incorporating landscape elements into a
connectivity measure: a case study for the Speckled wood butterfly (Pararge aegeria L.).
Landscape Ecology 18: 561-573.
deMaynadier, P.G. and M.L. Hunter, Jr. 1998. Effects of silvicultural edges on the distribution and
abundance of amphibians in Maine. Conservation Biology 12: 340-352.
Funk, W.C., A.E. Greene, P.S. Corn, F.W. Allendorf. 2005. High dispersal in a frog species suggests
that it is vulnerable to habitat fragmentation. Biological Letters 1: 13-16.
Funk, W.C., M.S. Blouin, P.S. Corn, B.A. Maxell, D.S. Pilliod, S. Amish, F.W. Allendorf. 2005.
Population structure of Columbia spotted frogs (Rana luteiventris) is strongly affected by the
landscape. Molecular Ecology 14: 483-496.
Porter, W.P. and J.W. Mitchell. 2003. Method and system for calculating the spatial-temporal effects
of climate and other environmental conditions on animals. US 2003/0040895 A1 U.S. Patent
Office. Washington, D.C.
Ray, N., A. Lehmann, and P. Joly. 2002. Modeling spatial distribution of amphibian populations: a
GIS approach based on habitat matrix permeability. Biodiversity and Conservation 11: 2143-2165.

314
Theme 3. Ecological Networks, fragmentation and connectivity
3.3 Symposium 16: Animals on thermal landscapes

Thermal heterogeneity and river fish distribution in a groundwater-influenced

high desert landscape

C.E. Torgersen 1, D.S. Bateman 2, D.P. Hockman-Wert 3, M.D. McSwain 4, R.E.

Gresswell 5
1
USGS FRESC, University of Washington, College of Forest Resources
Box 352100, Seattle, WA 98195 USA.
e-mail: [email protected]
2
Oregon State University, Department of Forest Science, Corvallis, OR 97331 USA.
3
USGS FRESC, 3200 SW Jefferson Way, Corvallis, OR 97331 USA.
4
BLM, Prineville District Office, 3050 NE Third, OR 97754 USA.
5
USGS NRMSC, Montana State University, Bozeman, Montana 59717 USA.

Introduction

The Lower Crooked River is a remarkable groundwater-fed stream flowing through

vertical basalt canyons in the Deschutes River Valley in central Oregon. The 15-km section
of the river between the Crooked River National Grasslands boundary near Ogden Wayside
and river km (RKM) 13 is protected under the National Wild and Scenic Rivers Act (16 U.S.C.
1271-1287) for its outstandingly remarkable scenic, recreational, geologic, hydrologic,
wildlife, and botanical values (ORVs), and significant fishery and cultural values.
Groundwater springs flow directly out of the canyon walls into the Lower Crooked River and
create a unique hydrologic setting for native coldwater fish, such as inland Columbia Basin
redband trout (Oncorhynchus mykiss gairdneri). To protect and enhance the ORVs that are
the basis for the wild and scenic designation, the Bureau of Land Management (BLM) has
identified the need to evaluate, among other conditions, fish presence and habitat use of the
Lower Crooked River. The results of this and other studies will provide a scientific basis for
communication and cooperation between the BLM, Oregon Water Resources Department,
Oregon Department of Fish and Wildlife, and all water users within the basin. These
biological studies initiated by the BLM in the region reflect a growing national awareness of
the impacts of agricultural and municipal water use on the integrity of freshwater ecosystems.
The goal of this project was to examine spatial patterns in fish assemblages, aquatic
habitat, and water temperature in the Lower Crooked River during summer conditions.
Specific objectives were to (1) characterize the spatial distribution of native and non-native
fishes, (2) describe variation in channel morphology, substrate composition, and water
temperature, and (3) evaluate the associations between fishes, aquatic habitat, and water
temperature.

Methods

The survey of fishes and aquatic habitat in 19 km of the Lower Crooked River was
completed on July 29 - August 3, 2004. Fish and aquatic habitat sampling was conducted
from RKM 13 to 32. Visual assessments of fishes were conducted with mask and snorkel by
a two-person crew. The snorkelers conducted the survey in a downstream direction but
surveyed fishes in individual channel units (pools, glides, riffles, and rapids) in an upstream
direction. Fish were identified and counted in either the entire channel unit, or in a portion of
the channel unit, depending on the length of the unit. Units longer than 100 m were sampled
for fish only in the upper portion of the unit. Variation in stream temperature was assessed at
six spatially dispersed locations in the Lower Crooked River. Spatially continuous
longitudinal patterns of stream temperature were evaluated using forward-looking infrared
(FLIR) remote sensing (Torgersen et al. 2001).

315
Theme 3. Ecological Networks, fragmentation and connectivity
3.3 Symposium 16: Animals on thermal landscapes

Results and discussion

Stream temperature and water quality parameters changed dramatically in a downstream

direction in the Lower Crooked River. Longitudinal patterns in stream temperature indicated
substantial cooling from groundwater inputs in a downstream direction. The distribution of
fishes was marked by an upstream–downstream gradient dominated by salmonids in
downstream reaches and cyprinids in upstream reaches. Spatial patterns in the distribution
of pools, glides, riffles, and rapids were highly variable among the electrofishing/angling
sites, as revealed by the spatially continuous survey of aquatic habitat conducted by
snorkelers. Changes in water temperature and turbidity corresponded with distinct spatial
structuring of fish assemblages in the Lower Crooked River. In only 18 km, the river
transformed from a warm, turbid cyprinid stream to a clear, cold trout river. Longitudinal
patterns in fish distribution, aquatic habitat, and water temperature suggested that
temperature and, potentially, turbidity were the primary physical drivers of fish assemblage
structure in the Lower Crooked River.
The unique groundwater-influenced thermal landscape of the Lower Crooked River
created an environment in which coldwater fish (salmonids and cottids) were distributed
among species that typically characterize a coolwater fish assemblage (cyprinids and
catostomids). Fishes in the Lower Crooked River exhibited thermal preferences that were
generally similar to those described in the literature for Pacific Northwest fish assemblages
(Zaroban et al. 1999). However, some species exhibited anomalous distribution patterns
with respect to their thermal preferences described in the literature. High water temperatures
have a significant influence on salmonid behavior and growth in high desert streams (Li et al.
1994). Several species of coldwater fishes have been shown to behaviorally thermoregulate
by locating thermal refugia several degrees cooler than ambient water temperatures
(Torgersen et al. 1999, Ebersole et al. 2003). Spatial heterogeneity in water temperature is
particularly pronounced in the Lower Crooked River due to groundwater inputs and may
provide multi-scale thermal refugia important for the existence of coldwater fishes such as
rainbow trout, mountain whitefish, and sculpins

References
Ebersole, J.L; Liss, W.J. & Frissell, C.A. (2003) Thermal Heterogeneity, stream channel
morphology, and salmonid abundance in northeastern Oregon streams. Canadian Journal of
Fisheries and Aquatic Sciences 60: 1266-1280.
Torgersen, C.E; Price, D.M; Li, H.W. & McIntosh, B.A. (1999) Multiscale thermal refugia and stream
habitat associations of chinook salmon in northeastern Oregon. Ecological Applications 9:
301-319.
Li, H.W; Lamberti, G.A; Pearsons, T.N; Tait, C.K. & Buckhouse, J.C. (1994) Cumulative effects of
riparian disturbances along high desert trout streams of the John Day Basin, Oregon.
Transactions of the American Fisheries Society 123: 627-640.
Zaroban, D; Mulvey, M; Maret, T; Hughes, R. & Merritt, G. (1999) Classification of species
attributes for Pacific Northwest freshwater fishes. Northwest Science 73: 81-93.
Torgersen, C.E; Faux, R.N; McIntosh, B.A; Poage, N.J. & Norton, D.J. (2001) Airborne thermal
remote sensing for water temperature assessment in rivers and streams. Remote Sensing of
Environment 76: 386-398.

316
Theme 3. Ecological Networks, fragmentation and connectivity
3.3 Symposium 16: Animals on thermal landscapes

The role of thermal landscapes in resource selection by black-tailed deer

J. G. Kie, R. T. Bowyer

Department of Biological Sciences, Idaho State University, Pocatello, ID 83209-8007 USA

e-mail: [email protected]

Introduction

An important question in ecological studies of large mammals is to what degree and in

what manner do thermoregulatory constraints influence habitat selection individuals (Parker
and Gillingham 1990)? Animal response undoubtedly varies among different populations
across broad geographic areas and likely depends on pelage resistances, degree of
acclimation, and other factors (Mysterud and Østbye 1999). To illustrate this question, we
studied habitat use by black-tailed deer (Odocoileus hemionus columbianus) on winter range
in northern California, USA. Vegetation types used by deer included open woodland-
grassland savannas, deciduous woodlands, and dense thickets of live oak (Quercus
wislizenii) cover. The climate was Mediterranean, with temperatures rarely dropping below
freezing.

Methods and results

During warm, sunny days, temperatures from black-body devices were higher than
ambient air temperatures in habitats other than live oak, but were closer to ambient
temperatures in dense live-oak cover (Fig. 1). Deer were more likely to occur in or near (<
150 m) live oak than were random locations, spending 80% of their time in or near live oak
during day and 70% of their time in that habitat at night. Deer were more likely to occur in or
near live oak when weather was warm, dry, sunny, and calm.

The most parsimonious logistic regression model for whether deer would occur in or near
live oak was based on the interaction between wind speed and relative humidity (AIC = 283),
which had an 18% probability of being the best model (AIC weight, wi = 0.18). Five other
models, however, could not be distinguished from the best model based on the criterion
ΔAIC < 2 (Table 1).

We formulated a resource selection function (RSF) using conditional logistic regression,

blocking by individuals, to distinguish between deer locations and random points based on
whether the point was in or near live-oak cover. The RSF was significant (P = 0.004), and
indicated that a point in or near live-oak cover was more likely to be a deer location than a
random point.

We then calculated similar RSFs after splitting the data into 2 subsets: calm, where wind
speed was < the median value, and windy, where wind speed was > the median value.
When it was windy, habitat was not a significant predictor of deer locations versus random
points (P > 0.05). When it was calm, not only was habitat was a significant predictor (P =
0.035), but the model preformed substantially better than did the model for all wind
conditions based on AIC scores. We concluded that black-tailed deer in winter pelage were
using live-oak thickets to ameliorate heat gain on calm, sunny days with low humidity,
although recognizing that other factors such as risk of predation may play a role as well.

317
Theme 3. Ecological Networks, fragmentation and connectivity
3.3 Symposium 16: Animals on thermal landscapes

TEMPERATURE (oC) 30

20

10

AIR TEMPERATURE
BLACK-BODY TEMPERATURE (LIVE-OAK COVER )
0
BLACK-BODY TEMPERATURE (OTHER HABITATS)

00:00 04:00 08:00 12:00 16:00 20:00 00:00

TIME (HR:MIN)

Figure 1. Comparison of air temperature and thermocouple

temperatures inside black-bodies placed in live-oak cover and non-cover
habitats on a representative sunny day (15 February 1991).

Table 1. Best 6 of 22 different logistic regression models distinguishing deer locations in or

near live oak cover versus other habitats, ranked by AIC values and AIC weights (wi).
Variables were: Air (air temperature, 0C), Rh (relative humidity, %), Solar (solar radiation,
watts * m-2), and Wspeed (wind speed, m * min-1). These 6 models were indistinguishable
based on the criterion ΔAIC < 2. Individual (Id) was forced into each model to account for
variability among individual animals.

Model AIC wi
Rh*Wspeed Id 283.139 0.180

Wspeed Id 284.274 0.131

Air Id 285.442 0.129

Air*Wspeed Rh*Wspeed Id 283.781 0.111

Wspeed Air*Wspeed Id 283.810 0.102

Solar*Wspeed Rh*Wspeed Id 284.118 0.057

References
Mysterud, A., & E. Østbye. (1999) Cover as a habitat element for temperate ungulates: effects on
habitat selection and demography. Wildlife Society Bulletin 27:385-394.
Parker, K. L., & M. P. Gillingham. (1990) Estimates of critical thermal environments for mule deer.
Journal of Range Management 43:73-81.

318
Theme 3. Ecological Networks, fragmentation and connectivity
3.4 Open Session 9: Ecological networks

3.4 Open Session 9: Ecological Networks

Assessing functional connectivity by simulating migration areas for amphibians:

a tool for conservation management of landscapes

A forest habitat network approach to directing native woodland improvement and

expansion on the Argyll islands, Scotland

D G Moseley1, R Worrell2, B Black3, D Ray1

1
Forest Research, Ecology Division, Northern Research Station, Roslin, Midlothian, Scotland,
e-mail: [email protected]
2
Upper park, Aberfeldy, Perthshire, Scotland, PH15 2EH
3
Argyll Woodlanders, Whin Bank, Lunga Drive, Ardfern, by Lochgiphead, Argyll, Scotland,

Introduction
Many of the native woodlands along the west coast of Scotland are biologically rich in
bryophytes, lichens and filmy ferns, owing to the mild oceanic nature of the climate. Yet the
resilience of these communities may be threatened by the effects of fragmentation and
climate change. The impact of habitat fragmentation and isolation on woodland biodiversity is
widely recognised (UK Biodiversity Group, 1995; Peterken, 2002), requiring a range of
methods of buffered consolidation, restoration and improvement, and new woodland
expansion to deal with habitat decline. This paper describes a study located on the coastal
islands of Argyll, Scotland, to provide information on native woodland improvement and
expansion opportunities that may increase the functional connectivity of native woodlands,
that will not jeopardise open ground habitats and functional networks, and which targets land
most suitable for woodland establishment.

Methodology
An assessment was made to identify woodlands with a high biodiversity taking particular
account of three areas, 1) the floristic quality of lower plants, 2) the structural diversity of the
stands, and 3) a high deadwood component. This methodology allowed the identification of
woodland habitats which will provide a source for dispersal into neighbouring habitat, and
which must be protected and enlarged by buffered expansion.
A habitat network approach, using a spatial model from Forest Research’s suite of tools
called “BEETLE” (Biological and Environmental Evaluation Tools for Landscape Ecology)
(Watts et al. 2005) was used to assess the distribution of native broadleaved woodland
habitat patches within the landscape, and predict the dispersal ability of woodland species to
other woodland patches through the intervening landscape matrix. Where the woodland
patches are sufficiently close to allow dispersal, they are assumed to form a functionally
connected network. The model calculates distribution of functional networks comprised of
woodland habitat patches, plus parts of the matrix through which woodland species are able
to disperse, determined by a permeability weighting profile of the matrix mosaic. This
represents a focal species (Lambeck, 1997) approach to model habitat requirements and
dispersal abilities of species encountered in different woodland types. In this study a 1000 m
maximum dispersal distance was set, representing moderately mobile species, and was
reduced according to the matrix permeability, with more modified habitat types having a
greater resistance to dispersal.

Results
The approach determined the functional connectivity of existing broadleaved woodlands
and indicated possibilities for their improvement, expansion, or linkage, and opportunities for
converting coniferous stands to native broadleaved woodland. The model outputs are a
series of maps indicating opportunities and constraints:

319
Theme 3. Ecological Networks, fragmentation and connectivity
3.4 Open Session 9: Ecological networks

Opportunities
1. Target management to improve the biodiversity quality of woodlands adjacent to high
quality woodland to form larger networks of high biodiversity quality woodland (Table 1).
2. Restore and convert conifer stands to native woodland, where conifers separate patches
of broadleaved woodland. This will reduce the isolation of broadleaved woodland and,
over time, link patches of woodland within larger networks (Table 1).
3. Expand existing broadleaved woodland by natural regeneration on open ground to link
neighbouring woodland patches into a network across the zone extending up to 500 m
from the edge of each patch. This represents the area and limit of new woodland
establishment by regeneration (the first 50 to 100 m) and planting (beyond 50 to 100 m).
4. Link neighbouring networks by planting new “stepping stone” woodland between patches
within existing networks to allow dispersal through the matrix and the new woodland
patch.

Table 1. Opportunities for improvement of broadleaved (BL) woodland and conversion of

conifer woodland to native broadleaved woodland to reduce broadleaved
woodland fragmentation on a selection of the Argyll Islands.
Island Improvement Conversion
Total % BL woodland Total % BL woodland
area (ha) area on island area (ha) area on island
Mull 803 14.6 1915 34.8
Islay 80 3.0 358 11.7
Shuna 60 20.0 0 0.0
Lismore 7 5.0 <0.1 <0.1

Constraints
Land considered unsuitable and excluded from the expansion and linkage opportunities
included land with high wind exposure, peat bog, and sites designated for conservation (e.g.
wetland, geological, and aquatic habitats). Site suitability of remaining land areas would need
to be identified through an Ecological Site Classification (Pyatt et al., 2001; Ray, 2001),
followed by a site survey.

Discussion
This work has highlighted the opportunities for improving native broadleaved woodland
networks on the larger islands, many of which have large established broadleaf and conifer
woodlands (Mull and Islay). It has also determined where resources can be targeted to
achieve large gains in native broadleaved woodland biodiversity on the smaller islands with
smaller areas of woodland (Shuna). The results of the study are now being used in a
potential bid for state (Forestry Commission) funding for native broadleaved woodland
expansion on the Argyll island and will aid strategies to improve native broadleaved habitat
connectivity, helping to address issues of habitat fragmentation and isolation.

References
Lambeck, R. J. (1997) Focal species: a multi-species umbrella for nature conservation. Conservation
Biology, 11, pp. 849-856.
Peterken, G.F. (2002) Reversing the Habitat Fragmentation of British Woodlands. WWF-UK,
Godalming, Surrey.
Pyatt, D. G; Ray, D; & Fletcher, J. (2001) An Ecological Site Classification for Forestry in Great
Britain :Bulletin 124, Forestry Commission, Edinburgh.
Ray, D. (2001) Ecological Site Classification Decision Support System V1.7, Forestry Commission -
Edinburgh.
UK Biodiversity Group (1995) Biodiversity: The UK Steering Group Report - Volume I: Meeting the
Rio Challenge (Annex A-E). English Nature, Peterborough.
Watts, K; Humphrey, J; Griffiths, M; Quine, C. & Ray, D. (2005) Evaluating Biodiversity in
Fragmented Landscapes: Principles. Information Note 73, Forestry Commission, Edinburgh.

320
Theme 3. Ecological Networks, fragmentation and connectivity
3.4 Open Session 9: Ecological networks

Landscape management approach – vital to conservation in India

P.K. Mathur

Dept. Landscape Level Planning and Management, Wildlife Institute of India,

P.Box # 18, Chandrabani, Dehra Dun – 248 001 (Uttarakhand), India.
e-mail: [email protected]

Introduction

India owing to its strategic geographical location at the confluence of three

biogeographical realms, varied landforms, and climatic conditions harbors diverse natural
ecosystems. Natural areas along with associated rich floral and faunal assemblages make
India as one of the 12 mega diversity countries. With just 2.4% land area, >one billion
people and >15% world’s livestock, India accounts for 7-8% of recorded species and covers
20.6% area under various forests (FSI, 2005). Country has established a network of
protected areas (PAs) comprising 94 national parks and 501 wildlife sanctuaries covering
4.7% area (Anon., 2006) and is committed to maintain its natural heritage while pursuing
economic development. Wide ranging measures at policy as well as programmatic levels
have been adopted those are instrumental in ensuring sustainable livelihoods and
conservation of biodiversity. However, PAs amidst human dominated landscapes pose
management challenges for wildlife and local communities due to rigid enforcement of
wildlife law, small average size, traditional resource dependence and developmental needs.
Much of the endangered wildlife (e.g. elephant, lion, tiger) exists outside PAs in the adjacent
natural habitats or in human dominated matrix. Local communities often suffer due to
prevailing strict PA rules and contentious wildlife damage problems. Landscape approach to
planning and management of wildlife resources is being validated. Paper summarizes select
pioneering case studies carried out in difficult landscapes across the country and highlights
the relevance of this approach to conservation in the context of country’s wide array of
natural resources, land use, growing demands and developmental needs.

Case studies

Study on the transhumance and silvopastoral dependence in the Great Himalayan

National Park Conservation Area (GHNPCA) adopting a landscape level assessment
approach found that (a) mountain people rely on the whole landscape for their livelihoods;
consequently, policies and institutions for mountain forests and agro-forestry must recognize
interactions between agricultural land uses, forests and livestock; (b) every strategy for
ensuring that mountain people derive sustainable livelihoods from their forests must be tailor-
made for the local physical, biological, cultural and political environment – and ways of
responding to change must always be included (Mathur and Mehra, 2005). Exclusion of
traditional silvopastoral use in GHNP and subsequent active protection may lead to an
overall recovery of forests and pastures. This way, the legal role of a national park would be
fulfilled. However, it is difficult to predict how this recovery would affect the overall diversity.
Further, the restriction of access to nearly 65% area of GHNPCA is likely to increase
livestock pressure on the remaining area and this is expected to accelerate degradation of
those remaining silvopastoral resources.

Recent researches and management experience in managing Gir forests, the last
surviving wild abode of Asiatic lion (Panthera leo persica) amply illustrated that the
conservation of last surviving wild population of lion along with the management of newly
established meta-populations in the peripheral forests seems to be possible by adopting a
regional planning approach on a larger landscape while addressing requirements of local
communities and rapid development in the surrounds of Gir.

321
Theme 3. Ecological Networks, fragmentation and connectivity
3.4 Open Session 9: Ecological networks

Findings of a recently concluded major project aimed to address the requirement of

managing Indian forests for biological diversity and productivity based on ecological
assessments in four pilot conservation areas (CA) that represent priority Indian ecosystems
revealed that conservation of biodiversity along with sustainable livelihoods is possible by
thinking broadly across major landscape areas i.e. beyond the boundaries of the specific PA
or forest areas of interests (Lehmkuhl et al., 2006). Think “outside the box” explains that the
broad context within which the management area resides that greatly affects conditions
within the area and success of management actions. Further, study recommended for
considering the spatial or temporal cumulative effects of land use, conditions, and trends
within, among, and beyond the boundaries of immediate interests. Study also advocates
integrating management plans across administrative boundaries and between forest and
wildlife management. Extensive study on vegetation assessment in Terai Conservation Area
as a part of the project further concluded that the adoption of landscape management
approach is critical for maintaining diversity of vegetation in TCA as floral diversity varied
considerably in individual constituent areas (PAs and managed forests) and each forest type
assessed.

Conclusion

Selected case studies amply illustrate that small PAs neither conserve the entire
array of representative biodiversity nor they provide adequate habitat to wide ranging faunal
species. It is clear that India is moving towards wildlife and forest conservation at a broader
spatial scale and by incorporating more land use conditions than ever before. Better use and
focus of concepts of landscape ecology and conservation biology within boundaries of PAs
and managed forests, and outside to community, revenue, and other lands are of paramount
significance to help ensure protection and restoration of forests and other natural areas in
India for future generations.

References
Anonymous (2005) State of Forest Report 2003, Forest Survey of India (Ministry of Environment &
Forests), Dehra Dun, pp. 1–134.
Anonymous (2006) India’s Third National Report to Convention on Biological Diversity, Executive
Summary, Ministry of Environment & Forests, Government of India, New Delhi, pp. 1–21.
Lehmkuhl, J.F.; Mathur, P.K.; Sawarkar, V.B.; Holthousen, R.S.; Marcot, B.G. & Raphael, M.G.
(2006) Managing Indian forests for biological diversity and forest productivity. J.A. McNeely, T.M.
McCarthy, A. Smith, L. Olsvig-Whittaker & E.D. Wikramanayake (Eds). Conservation Biology in
Asia, Society for Conservation Biology Asia Section and Resources Himalaya Foundation,
Kathmandu, Nepal, pp.92–114.
Mathur, P.K. & Mehra, B.S. (2005) Transhumance and silvopastoral dependence in the Great
Himalayan National Park Conservation Area – a landscape level assessment. M.R. Mosquera-
Losada, J. McAdam and A. Rigueiro-Rodriguez (Eds). Silvopastoralism and Sustainable Land
Management, CABI Publishing, Oxfordshire, U.K., pp. 357–358.

322
Theme 3. Ecological Networks, fragmentation and connectivity
3.4 Open Session 9: Ecological networks

Designing a general plan for reserved territories based on landscape cover

analysis

M.Y. Puzachenko1, I.A. Onufrenya2, D.N. Kozlov3, Y.G. Puzachenko1

1
A.N. Severtsov Institute of Ecology and Evolution RAS, 119071 RF Moscow,
Leninsky prospect 33,
e-mail: [email protected]
2
World Wide Fund for Nature, 109240, RF Moscow, Nokoloyamskaya, 19-3
3
Lomonosov Moscow State University Faculty of Geography, 119899 RF Moscow,
Leninskiye Gory 1

The modern concept of territorial forms of nature reservation is based on ecological

network, which has its theoretical basis on island biogeography – large “islands” connected
with each other by a system of smaller islands of various size which among other play a role
of “stones” which overcome areas useless for sustainable survival. IUCN defines an
Ecological Network as a coherent system of natural and/or semi-natural landscape elements
that is configured and managed with the objective of maintaining or restoring ecological
functions as a means to conserve biodiversity while also providing appropriate opportunities
for the sustainable use of natural resources (IUCN & Syzygy, 2005).
In accordance with existing perceptions Econet includes the following structural parts: 1)
core areas, which conserve general biodiversity in natural landscapes, 2) corridors or
stepping stones which allow species to disseminate and migrate between core areas,
decreasing isolation and improving connectedness of natural systems, 3) recovery territories,
usually close to core areas, which extend core area network to optimal size, 4) buffer zones
protecting network from potential negative influence and permitting all types of activities not
contradicting with biological diversity conservation aims. Evidently, the main idea on which
this construction bases needs hierarchical organization of network. Three levels of eco-
network hierarchical organization are: 1) objects of international and national significance, 2)
national and regional level and 3) local level. According to general aims and criteria Econet is
component of approach to sustainable development organization on the basis of landscape
planning and management. Operational criteria for selection of all four types of Econet are
very diverse and determined by: geographic conditions, history of region economy
development, how fully a region is studied, availability of scientific staff, which can realize
wide range of work for specific projects and funding possibilities. When solving this task for
Russia and CIS countries, we have to consider certain difficulties.
The approach developed uses the following criteria: 1) landscape cover as a multiple
landscapes reflection is organized hierarchically and this hierarchy should naturally be
reflected in Econet, 2) on any hierarchical level, species biodiversity is a function of spatial
diversity of ecosystem types and their territorial combinations, 3) due to fractal properties and
hierarchy the relatively homogenous complexes of ecosystems exist on any level and they
coincide with areas of ecological optimum for most of habiting species, 4) global
geodynamics determines structures at different scale which generate quasi-linear forms with
ecosystems that may differ from their surroundings and are usually expressed by relief and
by remote sensing data. They form a network of natural corridor ecosystems. The crosses of
corridors are “nodes” which are characterized by maximum diversity of landscapes (we
should state here that most cities are situated in such “nodes”).
These criteria, basing on the theory of landscape and landscape ecology allow the
development of a system of measurement that enables us to select quantitatively functional
elements of Econet and to assess their potential quality.
Hierarchical levels of organization were extracted using 2-dimensional Fourier spectral
analysis of digital elevation model and satellite imagery bands. Basing on classified
multispectral remote imagery (SPOT, VEG2000, GeoCover, MODIS) we develop maps of
ecotope types. Usually it is possible to determine their physical meaning but in is not
necessary in general case. In all cases, natural ecosystems within specific climatic conditions

323
Theme 3. Ecological Networks, fragmentation and connectivity
3.4 Open Session 9: Ecological networks

are characterized by the ultimate absorption of solar radiation or by specific absorption

spectrum. Such classifications can be used to assess landscape diversity with
implementation of various metrics (diversity, evenness, fragmentation, number of borders,
fractal dimension etc.). Each of these metrics reflects landscape spatial structure various
aspects which have different functional meaning. At the same time, many of them correlate
with each other. Leaving aside methods of their integration we should note that each
hierarchical level of organization contains ecosystems with various potential landscapes
diversity and with various functional values. Further, using special procedure we can outline
linear elements (potential corridors) and nodes. The status of nodes can be determined by
the number of linear elements which form them. Integration of measured properties of spatial
structures including territories of potential reservation and linear systems allows the design of
a general scheme. Further, individual analysis of each potential object is processed and
descripted. Proposed technology has been applied in designing general Econet scheme for
Southern Siberia Mountains, south of Eastern Siberia, south of Far East, South Urals,
Yakutia, North Caucasian region, vast Asia region, including all CIS countries. Detailed
description can be found in paper about Yakutia Econet (Puzachenko, 2004, 2005).
When estimating the efficiency of the practical use of developed technology we can say
that on the first stage technology and its results evokes a more or less negative reaction.
After, the work begins and the study becomes progressively more useful and leads to a
design stage. Finally, this experience demonstrates wide possibilities in landscape planning
because of designing on the basis of remote sensing information, digital elevation models
and field data. Algorithms of relief and remote sensing information were realized in programs
made by Gleb Aleschenko. The research is made with support of WWF grants.

References
Puzachenko Yu.G., Puzachenko M.Yu., Onufrenya I.A., Alishenko G.M. (2004) Elaboration of
scheme of protected area distribution on the basis of the remote information (by the example of
Yakutia). Geography and natural resources, №1, pp. 10-24.
Puzachenko M., Puzachenko Yu., Kozlov D., Krever V., Onufrenya I., Alishenko G. (2005) Use of
the SPOT VGT-S10 product to discriminate and evaluate ecosystems for ecological aptness and
for the design of an ecological network. Proceedings of the Second International VEGETATION
User Conference, European Communities. pp. 439-449.
IUCN & Syzygy (2005) Ecological network. from: http://iucn-ce.org/ecological-
networks/GLOBALECONETCD

324
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

3.5 Open Session 10: Landscape modelling and population ecology

Absent or non-detected? an application of occupancy models in fragmented

landscapes

A. Mortelliti, L. Boitani ,

University of Rome “La Sapienza”, Department of Animal and Human Biology – Viale
dell’Università 32, 00185, Rome, Italy.
e-mail: [email protected]

Introduction

Research focusing on species distribution (e.g. metapopulation dynamics) often deals

with presence/absence data. Whilst presence is easy to confirm, species absence rather
than non-detection is problematic, since detection probability – the probability of detecting at
least one individual of the species on a particular sampling occasion - is often less than one.
These issues have received growing attention, a major contribution was brought by
Mackenzie (2002, 2003) who incorporated detection history into a maximum likelihood
estimation model, permitting researchers to model detection probability and proportion of
sites occupied as a function of site/sampling covariates. We applied these models on data
collected during a survey on the distribution of Insectivores, Carnivores and Rodents in a
fragmented landscape in the Province of Siena, central Italy.
Our objective is to determine a sampling protocol to infer species absence in a site
(woodland patch). We begin by modeling detection probability and an estimate of the
proportion of sites occupied as a function of site characteristics, time or environmental
variables. Once we determine detection probability we calculate the sampling effort, required
to infer a species’ absence. We will here show results concerning three species Bank vole
(Myodes glareolus), Red squirrel (Sciurus vulgaris) and Beech marten (Martes foina).

Materials and Methods

Rodents were livetrapped using Sherman traps transects, with number of transects
proportional to patch size. Carnivores distribution was studied combining various techniques:
scent station surveys, track and camera trap surveys and interviews with local people. Baited
hair tubes were used to determine red squirrel distribution (number of tubes proportional to
patch size).
We ran analyses using program PRESENCE (available for download at http: // www.mbr-
pwrc.usgov/software.html), utilising multiple season and single season models, including
covariate effects. For each species we defined a set of a priori models with varying
covariates that could explain the patterns of patch occupancy. Models were first ranked
according to AICc values. A multimodel inference approach was followed to account for
model selection uncertainty, thus model averaging was used to estimate parameters
(Burnham & Anderson, 2002). Values of detection probability were converted into an
estimate of sample size (number of repeated visits) using the formula proposed by Reed
(1996): N = ln (α level / ln (1 – p). α was fixed at 0.05 and p being detection probability.
Reported values refer to our sampling effort, so are not meant to be generalised.

Results and Discussion

Terrestrial rodents
Myodes glareolus showed a seasonal variation in detection probability (p spring = 0.44; p
summer = 0.31, p autumn = 0.08; p winter = 0.09; Table 1) that could reflect many possible
causes, such as variation in density. Some sites were only seasonally occupied, due to both
colonisations and local extinctions (colonisation probability = 0.19; extinction probability =

325
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

0.13). From a sampling design perspective, the high values for required sampling size
(number of days of trapping) obtained for autumn and winter (37 and 33 days respectively;
versus 5 and 8 for spring and summer respectively) suggest that reliable data at reasonable
costs can be collected only during spring-summer. Overall, the main point is that combining
information on detection probability and sample size estimation (together with
colonisation/extinction probability) we were able to identify colonisation/extinction events in
each habitat patch, differentiating them from cases of non-detection.

Arboreal rodents
Red squirrel detection probability, using baited hair tubes, was relatively high (p = 0.56);
consequently the distribution estimate we obtained was relatively unbiased since the naïve
estimate (number of sites where the species was found) and estimate of proportion of sites
occupied are very close (Table 1). With our sampling protocol, 40 days of sampling with 4
visits at regular intervals are sufficient to infer squirrel absence.

Carnivores
Despite low values of detection probability and consequently high required sample size
values, our estimate of beech marten distribution was relativelly unbiased since the naïve
estimate and proportion of sites occupied are very close (Table 1).
Overall, we strongly suggest the utilisation of this modeling approach since it provides a
good framework to determine sampling protocols and to evaluate the reliability of
presence/absence data, fundamental for a non-biased species distribution in patchy
landscapes.

Table 1. Summary of the parameter estimates for the first 2 ranked (according to AICc)
occupancy models obtained through analyses with program PRESENCE. Psi= proportion
of sites occupied; naïve estimate = proportion of sites where the species was actually
found; w = Akaike weights.

Model AICc ∆AIC w naive Psi

Myodes glareolus
Psi, gamma(),eps(),p(seas.) 284.62 0.00 0.20 0.24 0.24
psi,gamma(dist500),eps(),p(seas) 285.38 0.76 0.14 0.24
Sciurus vulgaris
Psi(logha)p() 54.44 0.00 0.88 0.32 0.32
Psi()p(logha) 60.27 5.83 0.04 0.73
Martes foina
Psi(sum500),p() 149.7 0.00 0.66 0.4 0.5
Psi()p() 152.67 2.97 0.11 0.5

References
Burnham, K.P. & Anderson, D.R. (2002) Model selection and multimodel inference – a practical
information – theoretic approach. 2nd ed. Springer-Verlag, New York.
MacKenzie, D. I., Nichols, J.D., Lachman, G.B, Droege, S., Royle, J.A. & Langtimm, C.A. (2002)
Estimating site occupancy when detection probability is less than one. Ecology 83, 2248-2555.
MacKenzie, D. I., Nichols, J.D., Hines, J. E., Knutson, M.G. & Franklin, A.B. (2003) Estimating site
occupancy, colonisation, and local extinction when a species is detected imperfectly. Ecology 84,
2200-2207.
Reed, J. M. (1996) Using statistical probability to increase confidence of inferring species extinction.
Conservation Biology 10, 1283-1285.

326
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

The importance of movement behavior in setting landscape connectivity

B.J. Goodwin

University of North Dakota 1 – Department of Biology, 10 Cornell Street Stop 9019, Grand
Forks, ND, 58202-9019,USA.
e-mail: [email protected]

Introduction

Interest in landscape connectivity has increased over the last few decades (Goodwin,
2003). Despite the proliferation of landscape connectivity studies, few published papers have
explored how organism movement behavior impacts landscape connectivity (but see Bowne
et al., 2006; Stevens et al., 2004). Many published papers equate landscape connectivity
solely with measures of landscape structure and ignore the influence of movement behaviour
(Winfree et al., 2005). To apply connectivity metrics judiciously, it is important to understand
the influence of movement behavior on different connectivity metrics.

Approach

To explore how measures of connectivity respond to changes in movement behavior I

used an individual based, spatially explicit simulation model of the movements of individuals
through landscapes. Landscapes consisted of habitat and two matrix patch types. For all
simulations habitat was held at a constant proportion (5%) and in a constant pattern (5
equally sized patches evenly distributed across the landscape) and the amount of one of the
matrix patch types was varied. Connectivity was measured as dispersal success, search
time, landscape wide average habitat patch immigration rate, central habitat patch
immigration rate, and landscape wide average habitat cell immigration rate. I varied
movement behavior (propensity to move, step lengths, turning angles, edge crossing
behavior, perceptual range, conspecific attraction) in one of the matrix patch types. By
simulating the movements of individuals through landscapes I could determine the influences
of movement behavior on connectivity. Furthermore, since habitat amount and arrangement
were held constant impacts of movement behaviour on connectivity would suggest that
measures of connectivity based solely on landscape structure will be unreliable.

Findings and conclusions

Results from the simulations illustrate how variation in movement behavior can
dramatically impact measures of landscape connectivity, by as much as 500% (Fig. 1).
Furthermore, movement behavior in the matrix element had a stronger impact on
connectivity than movement behavior in the habitat element. Finally, movement behaviours
that led to either very slow movements or very fast movements through the matrix element
had significantly greater impacts on landscape connectivity. Based on these simulation
results it seems prudent to consider movement behavior when determining landscape
connectivity since connectivity measures based solely on landscape structure may be
misleading. These results argue for more effort investigating organism movement in
landscapes and, particularly, which aspects of the landscape might greatly aid or hinder
movement.

327
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

Figure 1. Influence of the movement behvior in a matrix element upon the impact of
changing the amount of that element in the landscape on connectivity. Impact refers to the
maximum change in connectivity (in this case immigration) as the proportion of the
landscape consisting of the matrix element varies. In all cases habitat proportion and pattern
does not vary. Movement behaviour consisted of: move probability (probability of an
individual moving during a time step), move distance (median step-length drawn from an
exponential distribution), K (concentration parameter in a von Mises distribution where K = 0
produces random motion and higher values of K concentrate turn angles in the forward
direction producing straighter movement paths), and cross (probability of crossing between
matrix elements).

References
Bowne, D.R.; Bowers, M.A. & Hines, J.E. (2006) Connectivity in an agricultural landscape as
reflected by interpond movements of a freshwater turtle Conservation Biology 20: 780-791.
Goodwin, B.J. (2003) Is landscape connectivity a dependent or independent variable? Landscape
Ecology 18: 687-699.
Stevens, V.M.; Polus, E.; Wesselingh, R.A.; Schtickzelle, N. & Baguette, M. (2004) Quantifying
functional connectivity: experimental evidence for patch-specific resistance in the Natterjack toad
(Bufo calamita) Landscape Ecology 19: 829-842.
Winfree, R.; Dushoff, J.; Crone, E.E.; Schultz, C.B.; Budny, R.V.; Williams, N.M. & Kremen, C.
(2005) Testing simple indices of habitat proximity American Naturalist 165: 707-717.

328
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

Effects of landscape complexity in agent-based population models

J. Nabe-Nielsen1,4, R. Sibly2,4, C. J. Topping3,4, M. C. Forchhammer1,4, K.

Sudharsan3,4

1Section for Climate Effects and System Modelling, National Environmental Research
Institute, University of Aarhus, Frederiksborgvej 399, PO Box 358, Roskilde, Denmark
e-mail: [email protected]
2School of Animal and Microbial Sciences, University of Reading, Whiteknights, Reading, UK
3Department of Wildlife Ecology & Biodiversity, National Environmental Research Institute,
University of Aarhus, Grenåvej 14, DK-8410 Rønde, Denmark
4Centre for Integrated Population Ecology (CIPE), www.cipe.dk

Introduction

How are animal populations affected by the size, shape and spatial arrangement of
landscape patches? This central ecological question has proved intractable to classic
experimental methods and hard to parameterise when using theoretical approaches. Agent-
based models (ABMs) provide a promising way forward, allowing detailed modelling of the
behaviour of individual animals in realistic landscapes (Grimm et al. 2005; Wiegand et al.
1999). Changes in landscape composition affect landscape fragmentation and patch quality,
which in turn causes animals to change behaviour. Here we analyze how animal population
dynamics are affected for four contrasting species using an agent-based model where the
behaviour of each individual is directly affected by realistic local temporal variations in patch
quality. Spatially explicit Agent-based models allow us to experimentally alter landscape
composition and directly measure effects on population dynamics while controlling for other
sources of variation.
The aim of this study is to investigate how populations of four evolutionarily distinct species:
field vole (Microtus agrestis), skylark (Alauda arvensis), a carabid beetle (Microtus agrestis)
and a spider (Erigone atra) are influenced by variations in landscape complexity in an
existing agricultural landscape in Denmark. Our analyses reveal effects of landscape
functional connectivity (Bélisle 2005) and edge effects (Urban 2005) on the population
dynamics of these species.

Methods

The population dynamics is modeled in a series of increasingly simple landscapes using the
agent-based model ALMaSS (Topping et al. 2003). In the 10x10-km reference landscape
around Bjerringbro in Denmark all fields, forests, creeks etc. are mapped at a 1-m2
resolution. The behaviour and physiological status of each individual is modelled to reflect
dynamic variations in the landscape, whereas the over-all response of each population is an
emergent property.
The landscape modifications we study are designed to address three important aspects of
landscape complexity: the effects of patch shape, patch location and patch size. First the
patch shape is randomized by creating artificial patches with random shapes around the
original patch centres and with the same size as the original patches. This allows us to study
whether the linear elements that may serve as corridors (e.g. hedgerows and field
boundaries) affect population dynamics. The second simplification is to randomize patch
location while retaining the original sizes of the patches. This permits us to analyse whether
populations are affected by the spatial arrangement of the patches, as may occur in species
that are affected by landscape fragmentation or that depend on synergetic effects between
different types of patches. Finally both patch types and sizes are allowed to vary. This allows
us to quantify the importance of the present allocation of land to fields, forests and other
patch types in the agricultural landscape.

329
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

Here we characterize stability of populations in terms of the rates at which they return to
equilibrium sizes after disturbance. This is a property of the relationship between population
size and population growth rate (Sibly et al. 2005) A landscape that allows a population to
rapidly return to equilibrium after being perturbed to low numbers is favourable for the
species.

Results and discussion

Variations in landscape complexity result in changed equilibrium population sizes, but the
amount and direction of the change varies among species. The relationship between
population growth rate and population size is relatively unaffected by variations in landscape
complexity, and it is always negative in the vicinity of the equilibrium population size. If a
species is perturbed to a particular density in two different landscapes, its population size
therefore increases faster in the landscape where its equilibrium population size is highest.
Both vole and skylark have higher equilibrium population sizes, and higher return rates, in
randomized landscapes. This is because areas that were only suitable as corridors in the
original landscape are transformed to primary habitat when patch shape is randomized. Our
analyses also suggest that high return rates may not be sufficient to ensure long-term
survival of a species. If corridors are transformed to primary habitat this may result in
increased population growth in many patches, while preventing the long-distance dispersal
that is necessary for re-colonizing remote patches after large local disturbances.
By predicting population consequences at different spatial and temporal scales the study
helps advance ecology science. Our results contribute to the on-going discussion about the
importance of local population survival for the maintenance of species on the landscape-
scale. The methods developed here can be used to guide management decisions in order to
increase land use sustainability.

References
Bélisle, M. (2005) Measuring landscape connectivity: The challenge of behavioral landscape ecology.
Ecology 86: 1988-1995.
Grimm, V.; Revilla, E.; Berger, U.; Jeltsch, F.; Mooij, W.M.; Railsback, S.F.; Thulke, H.H.; Weiner,
J.; Wiegand, T.; DeAngelis, D.L. (2005) Pattern-oriented modeling of agent-based complex
systems: Lessons from ecology. Science 310: 987-991.
Sibly, R.M.; Barker, D.; Denham, M.C.; Hone, J. & Pagel, M. (2005) On the regulation of populations
of mammals, birds, fish, and insects. Science 309: 607-610.
Topping, C.J.; Hansen, T.S.; Jensen, T.S.; Jepsen, J.U.; Nikolajsen, F. & Odderskær, P. (2003)
ALMaSS, an agent-based model for animals in temperate European landscapes. Ecological
Modelling 167: 65-82.
Urban, D.L. (2005) Modeling ecological processes across scales. Ecology 86: 1996-2006.
Wiegand, T.; Moloney, K.A.; Naves, J. & Knauer, F. (1999) Finding the missing link between
landscape structure and population dynamics: a spatially explicit perspective. American Naturalist
154: 605-627.

330
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

The use of Mahalanobis Typicalities to model the distribution of species with

different response curves to environmental gradients.

F. Sangermano, J. R. Eastman

Clark University, Graduate School of Geography and Clark Labs - 950 Main St., Worcester
MA, 01610, USA.
e-mail: [email protected]

Introduction

Several studies have compared the performance of different algorithms for modeling
species distributions, with varying results among species, areas and scales of analysis (eg.
Elith et al. 2006). Review papers on species distribution modeling (e.g. Austin, 2002; Guisan
and Zimmermann, 2000) indicate the necessity of taking into account ecological theory in the
modeling process. In this work we present a procedure for modeling species distributions
called Mahalanobis Typicalities and discuss the results of a test in which it is used to model
the habitat suitability and distributional range of virtual species with different response curves
(normal, skewed, bimodal and mixed) to environmental gradients.

Methods

Mahalanobis Typicalities (Eastman, 2006) express the degree to which the values of a set
of environmental variables at a location are typical of known instances of a specific species.
They are derived from the Mahalanobis Distance [ M2 = (x- µi)t Vi -1 (x- µi) ]; where x is the
vector of environmental measures at a location, µ is the vector of the mean environmental
measures for all known instances of the species in question and V is the variance/covariance
matrix. Typicality is the probability of any location having a Mahalanobis Distance greater
than or equal to that observed at the location of interest. Locations having attributes identical
to the multivariate centroid have a Typicality of 1.0, with less typical locations approaching a
Typicality of 0.0 at the limits of the distribution (Eastman, 2006). The other techniques used
in this paper are Maximum Entropy (MaxEnt), Genetic Algorithm for Rule-set Production
(GARP), Ecological Niche Factor Analysis (ENFA), BIOCLIM and DOMAIN. Details of these
techniques can be found for example in Elith et al. (2006), Phillips et al. (2006), and Hirzel et
al (2002). MaxEnt was run with both the Linear option (MaxEnt-Linear) and the Automatic
options (MaxEnt-Auto). GARP habitat suitability was calculated as the average of 100 runs.
Four virtual species were generated with normal, skewed, mixed and bimodal response to
environmental gradients. The root mean square error (RMS) and mean absolute error (MAE)
were calculated to evaluate the performance of the model in predicting habitat suitability.
Maximum Kappa (KIA max) and the area under the Receiver Operating Characteristic (ROC)
plots (AUC) were calculated to assess the performance of the models in predicting the
species distribution.

Results and discussion

The performance of Mahalanobis Typicalities under a Normal response curve was almost
perfect (RMS = 0.03, MAE = 0.02), and decreased with non-normal responses. The worst
prediction from this method was for the bimodal response (RMS = 0.28, MAE = 0.25),
followed by the skewed response (RMS = 0.11, MAE = 0.07). Typicalities proved to be fairly
tolerant to mixed skewed/normal responses (RMS = 0.07, MAE = 0.04). Comparing the
results of the Mahalanobis Typicalities procedure with the other methodologies, significant
differences were found. In all cases, MaxEnt-Linear was not capable of identifying the degree
of suitability. MaxEnt-Auto however, had a better performance. Typicalities perform best for
mixed and normal response curves with RMS = 0.07 and 0.03 respectively, while MaxEnt-

331
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

Auto was the best in the case of Skewed responses (RMS = 0.05, MAE = 0.03). In the case
of bimodal response curves, none of the methodologies had as good a performance.
BIOCLIM had the lowest RMS (0.19) and MAE (0.16) for this response curve, followed by the
Typicalities and MaxEnt-Auto (RMS = 0.36). GARP and DOMAIN were the worst performers
in predicting the habitat suitability in all cases.
For normal and mixed responses, both AUC and maximum Kappa were the highest for
Typicalities. For skewed response curves MaxEnt-Auto performed better than the other
models. Bimodal responses are again the least accurate, ranking BIOCLIM higher for AUC
and MaxEnt-Linear higher for maximum Kappa (Table 1)

Table 1: AUC and KIA max for each model under the different responses to environmental
gradients. In bold the higher agreement

AUC KIA max

Normal Skewed Mixed Bimodal Normal Skewed Mixed Bimodal
BIOCLIM 0.997 0.984 0.997 0.795 0.877 0.764 0.897 0.127
DOMAIN 0.951 0.964 0.928 0.726 0.472 0.653 0.515 0.065
ENFA 0.988 0.993 0.890 0.718 0.735 0.832 0.375 0.088
GARP 0.875 0.969 0.891 0.717 0.290 0.674 0.419 0.088
Typicalities 0.999 0.990 0.999 0.768 0.930 0.851 0.936 0.097
MaxEnt Auto 0.997 0.999 0.988 0.605 0.859 0.940 0.848 0.063
MaxEnt Linear 0.704 0.907 0.743 0.792 0.161 0.507 0.261 0.196

Conclusions

Mahalanobis Typicalities is a powerful method for modeling species distributions with

presence only data. It is shown here to have excellent performance in predicting both habitat
suitability and species range for normal distributions although the method is highly tolerant of
mixed non-normality conditions.

References
Austin, M. P. (2002) Spatial prediction of species distribution: an interface between ecological theory
and statistical modelling. Ecological Modelling 157: 101-118.
Eastman, J.R. (2006) Idrisi 15.0 The Andes Edition, Help System. Clark University-Clark Labs,
Worcester MA.
Elith, J.; Graham, C.H. & the NCEAS. (2006) Novel methods improve prediction of species’
distributions from occurrence data. Ecography 29: 129-151.
Guisan, A. & Zimmermann, N. (2000) Predictive habitat distribution models in ecology. Ecological
Modelling 135: 147-186.
Hirzel, A.H.; Hausser, J.; Chessel, D. & Perrin N. (2002) Ecological-niche factor analysis: How to
compute habitat- suitability maps without absence data? Ecology 83: 2027-2036.
Phillips, S.J.; Anderson, R.P. & Schapire R.E. (2006) Maximum entropy modeling of species
geographic distributions. Ecological Modelling 190: 231-259.

332
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

An automated procedure for the bulk re-processing of species range polygons

J.R. Eastman, F. Sangermano

Clark University, Graduate School of Geography and Clark Labs - 950 Main St., Worcester
MA, 01610, USA.
e-mail: [email protected]

Introduction

A fundamental resource for the assessment of priorities for biodiversity conservation has
been the archives of species range maps drawn by specialists based on combinations of
field observations and historical accounts. However, it is known that there are omissions in
the extents of some of these ranges and the polygonal boundaries are drawn with varying
degrees of precision. In this work we present an automated procedure to re-draw the
polygons based on species distribution modeling using environmental variables. Using
independent validation data, the results for a selection of species are presented

Methods

In order to develop and test the modeling procedure, we selected four species for which
we had reasonably large sets of observation points for validation: Microryzomys minutus
(n=88), Bradypus variegatus (n=99), Alouatta seniculus (n=67) and Tapirus terrestris (n=41).
Polygon data was obtained from NatureServe’s InfoNatura database (Peterson et al., 2005)
and observation points from the Global Biodiversity Information Facility (www.GBIF.org).
Fourteen environmental variables at 1km resolution were used to model the species (percent
herbaceous, percent trees, mean annual NDVI, NDVI seasonality, maximum precipitation of
the wettest month, mean annual precipitation, minimum precipitation of the driest month,
precipitation seasonality, temperature annual range, mean temperature diurnal range,
maximum temperature of the warmest month, minimum temperature of the coldest month
and mean annual temperature. An additional input was the map of ecoregions developed by
the World Wildlife Foundation (Olson et al. 2001).
In the first step of the process, the original polygon is cleaned through the calculation of
the empirical probability of the ecoregion being part of the range polygon. This is calculated
as the area of the ecoregion that intersects the range polygon divided by the total area of the
ecoregion. This map of probabilities was used as the weight for a Weighted Mahalanobis
Typicality modeling method. Areas with probabilities less than 0.5 are removed for the
polygon, and the resulting refined polygon is used in the process of variable selection.
Mahalanobis Typicalities are sensitive to irrelevant variables, so we developed an
additional procedure that looks at the ratio of the Standard Deviation on each variable within
the refined polygon to that over the entire study region (South America in this context). If the
species is constrained by an environmental variable, the variability of that variable inside the
polygon should be smaller than the variability of the variable in the entire region. An arbitrary
value of 0.6 was used as a threshold to select the important variables for the species in
question, keeping only the variables with values less than 0.6.
The empirical probability of the ecoregions was also added to the group of environmental
variables selected, as a way to help limit the predicted distribution of species to the
ecoregions they are known to belong to. Then, the Weighted Mahalanobis Typicalities
procedure implemented in the IDRISI software system (Eastman, 2006) was used as the
method to model the species distribution. It differs from the standard Mahalanobis
Typicalities procedure in that is based on a weighted mean and weighted
variance/covariance matrix. It was run using the original polygon as training, the weights
based on the ecoregions and the subset of variables selected. The result is a continuous

333
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

map from zero to one that represents the habitat suitability for the species modeled. We
threshold this suitability to generate the final range polygon considering habitat suitabilities
less than 0.01 as absences of the species. A second threshold at 0.1 was also applied to see
how threshold selection influenced the results. Finally, a procedure was developed to
automatically evaluate the result of each species modeling by comparing the area of the new
polygon to that of the original, facilitating the detection of large discrepancies between
original and modeled polygon.

Results

For A. seniculus there were significant extensions as well as contractions of the original
range. Of particular note is the fact that the original polygon followed the border of Bolivia,
while the new polygon extends the range into Bolivia as it should. The high Andes are
removed from the polygon. Two points in Bolivia were not captured in the original polygon
but are captured in the predicted one. With a threshold of 0.01, 85% of the points are
correctly classified (75% using a threshold of 0.1). For T. terrestris, 95% of the points were
classified correctly with the 0.01 threshold. The major difference between the predicted and
the original polygon is that the high Andes and parts of the southern range were removed. In
addition, the range was extended into northern Argentina, where it is known to have occurred
and is being re-introduced. This northern part of Argentina is missed with the 0.1 threshold,
which generates a smaller polygon with an accuracy of 85%. For B. variegatus, there is a
considerable reduction of the range, especially in the areas of the high Andes, where for
temperature constraints the species is known to not occur, and the llanos of Colombia. Using
the 0.01 threshold, 91% of the points were correctly classified, reducing to 78% in the case of
a threshold of 0.1. For M. minutus, 99% of the points were correctly classified under the 0.01
threshold and 95% under the 0.1 threshold. However the lower threshold in this case
produces over-prediction of the range of the species.

Conclusions

The procedure has significant merit for the automated processing of species polygons.
The most sensitive parameter was the threshold used in creating a hardened polygon from
the continuous typicalities. Although a 1% threshold worked well in general, careful attention
must be placed on false positive errors (over-prediction). In cases were large over-prediction
exists; increasing the threshold usually solves the problem.

References
Eastman, J.R. (2006) Idrisi 15.0 The Andes Edition, Help System. Clark University-Clark Labs,
Worcester MA.
Olson, D. M; Dinerstein, E.; Wikramanayake, E,D.; Burgess, N.D.; Powell, G.V.N.; Underwood,
E.C.; D'amico, J.A.; Itoua, I.; Strand, H.E.; Morrison, J.C.; Loucks, C.J.; Allnutt, T.F.;
Ricketts, T.H.; Kura, Y.; Lamoreux, J.F.; Wettengel, W.W.; Hedao, P. & Kassem, K.R. (2001)
Terrestrial Ecoregions of the World: A New Map of Life on Earth. BioScience 51: 933-938.
Patterson, B. D.; Ceballos, G.; Sechrest, W.; Tognelli, M.F.; Brooks, T.; Luna, L.; Ortega, P.;
Salazar, l. & Young, B.E. (2005) Digital Distribution Maps of the Mammals of the Western
Hemisphere, version 2.0. NatureServe, Arlington, Virginia, USA.

334
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

Testing the ecological significance of gradients in habitat quality for fauna in

Australian savannas

B. Price1, C. McAlpine1, S. Phinn1, D. Pullar1 and J. Ludwig2

1
Centre for Remote Sensing and Spatial Information Science, School of Geography,
Planning and Architecture, The University of Queensland, St Lucia 4072, QLD, Australia.
e-mail: [email protected]
2
CSIRO Sustainable Ecosystems, PO Box 780, Atherton 4883, QLD, Australia

The paradigm of habitat patches as islands surrounded by a matrix of unsuitable

habitat, originating from MacArthur and Wilson’s (1967) theory of island biogeography, has
had a pervasive influence on conservation biology, metapopulation theory and landscape
ecology. Human-modified landscapes form the basis of this model where landscape structure
ranges from intact, through fragmented, to relictual. The discrete patches paradigm has
provided useful insights into the ecological impacts of human-induced habitat destruction on
biological populations (Fahrig, 2003). However, many landscapes are not characterised by
discrete patch boundaries, but instead exhibit gradients in cover type, vegetation density and
structure with spatially and temporally variation. At present, ecology is lacking a quantitative
approach to measuring continuous gradients in landscape structure and to test the ecological
importance of these gradients for the breeding, foraging and movement behaviour of fauna
species.
Intact landscapes which have not been cleared for agricultural and urban
development, often exhibit continuous local scale (<10km) variation in structure related to
factors such as topography, geology, moisture availability, fire history, as well as some large
scale (100s-1000s of km) climatically induced gradients (Pearson, 2002; Woinarski et al.,
2005). Many of the tropical savanna landscapes of northern Australia which have not been
converted to intensive agriculture vary continuously with no clear boundaries between
vegetation types (Pearson, 2002; Woinarski et al., 2005).
The inadequacy of the discrete boundary model to quantify gradients in habitat
structure, or to differentiate between structure and function, suggests the need for continuous
metrics to quantify landscape structure and function.
The continuous landscape concept differs from the discrete patch paradigm in that it
conceptualises landscapes as having diffuse rather than discrete boundaries. The amount of
habitat and its spatial configuration within the landscape remain relevant measures of
landscape pattern, but must be conceived in a fundamentally different manner. Keeping in
mind that species may require more than one habitat type (Fahrig, 2003; Law& Dickman,
1998; Wiens, 1997), and following Fischer et al. (2006), our approach is to break species
habitat requirements down into needs for food, shelter, movement and reproduction. The
amount of habitat equals the sum total of suitable habitat elements within a given landscape
which are required throughout a species’ lifespan. The spatial configuration of habitat is a
measure of the distribution of those habitat elements across the landscape extent, defined
from an organism perspective. Similar to the discrete model, certain landscape elements act
as filters to a species’ movement such as those with increased risk of predation,
physiological stress and intense competition (Wiens, 1997). Habitat elements which offer
food, shelter or mating opportunities represent benefits to a species while those with risk of
predation, competition or physiological stress represent costs (Wiens, 1997).
In recent years remote sensing technology has been advancing at a great rate,
however, ecological studies have not taken full advantage of the capabilities of this
technology. Modern high spatial resolution (down to 0.6m and 0.05m) satellite and airborne
remote sensing is capable of capturing fine-scale heterogeneity in habitat structure in
continuously varying landscapes (Turner et al., 2003). Advanced spatial analysis techniques
are available to analyse the relationships between that heterogeneity and species distribution
patterns and abundances to obtain measures of ecological function (Rollins et al., 2004;
Turner et al., 2003). With new tools and advances in remote sensing we have the opportunity

335
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

to develop sophisticated measures of landscape structure and function which can manage
data at different scale levels and across structural gradients.
This study aims to develop and test metrics of landscape structure and function which
are ecologically relevant to fauna populations in northern Australian and other landscapes
with continuous variation in their structure. It takes advantage of advances in remote sensing
technologies and conceptualises and quantifies landscapes in continuous or gradient-based
terms that are measurable, accurate and of ecological relevance to native fauna. This work is
a first step towards process-oriented measures of ecological function in landscapes with a
continuous rather than discrete structure. We demonstrate the approach and ecological
relevance using a case study landscape with example fauna data. The study builds on a
conceptual model of continuously varying landscapes and develops an integrated spatial
analysis and image processing technique to quantify these gradients from high spatial
resolution (pixels < 5.0m) satellite images. Suitable continuous measures of landscape
structure are identified and then tested as indicators of habitat requirements and movement
behaviour of native fauna. The ecological relevance of these indicators is examined through
analysis of image and faunal survey data.

References
Fahrig, L. (2003). Effects of habitat fragmentation on biodiversity. Annual Review Ecology, Evolution
and Systematics 34, 487-515.
Law, B. S., & Dickman, C. R. (1998). The use of habitat mosaics by terrestrial vertebrate fauna:
implications for conservation and management. Biodiversity and Conservation 7, 323-333.
McArthur, R. H., & Wislon, E. O. (1967). The theory of island biogeography, Princeton University
Press, Princeton, N.J.
Pearson, D. M. (2002). The application of local measures of spatial autocorrelation for describing
pattern in north Australian landscapes. Journal of Environmental Management 64, 85-95.
Rollins, M. G., Keane, R. E., & Parsons, R. A. (2004). Mapping fuels and fire regimes using remote
sensing, ecosystem simulation, and gradient modeling. Ecological Applications 14, 75-95.
Turner, W., Spector, S., Gardiner, N., Fladeland, M., Sterling, E., & Steininger, M. (2003). Remote
sensing for biodiversity science and conservation. Trends in Ecology & Evolution 18, 306-314.
Wiens, J. A. (1997). The emerging role of patchiness in Conservation Biology. (S. T. A. Pickett, R. S.
Ostfeld, M. Shachak & G. E. Likens, eds.),The Ecological Basis of Conservation Vol.
Chapman & Hall, New York, pp. pp93-107.
Woinarski, J. C. Z., Williams, R. J., Price, O., & Rankmore, B. (2005). Landscapes without
boundaries: wildlife and their environments in northern Australia. Wildlife Research 32, 377-
388.

336
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

Testing the performance of landscape structure variables as predictors of

biodiversity: a case study from Dadia NP, Greece

S. Schindler1, V. Kati2, K. Poirazidis3

1
University of Vienna, Department of Conservation Biology, Vegetation and Landscape
Ecology; University of Vienna; Althanstraße 14, 1090 Vienna, Austria.
e-mail: [email protected]
2
University of Ioannina, Department of Environmental and Natural Resources Management,
Seferi 2, GR 30100, Agrinio, Greece
3
WWF Greece, Dadia project, GR-68400, Dadia, Soufli, Greece.

Introduction

Variables quantifying composition and configuration of landscapes are an important tool

for relating aspects of landscape structure with ecological pattern and processes. They are
used for the description of changes of landscapes due to climate change and socioeconomic
drivers, and for the quantification of habitat requirements of species. However, there are few
integrated empirical studied testing the relations between landscape structure and
biodiversity. In this study we evaluated for six taxonomic groups of organisms at five different
scales 1) correlations and correlation patterns between landscape structure and species
richness, and 2) the performance of sets of landscape structure variables as predictors of
species richness, comparing sets of different origin.

Methods

We collected species richness data of six taxa (woody plants, orchids, Orthoptera,
amphibians, reptiles, and small terrestrial birds) from 30 sampling plots at Dadia National
Park, a Mediterranean forest and regional hotspot of biodiversity in north-eastern Greece
(Kati et al. 2004a,b,c, Kati and Sekerlioglu 2006). Using a raster map of 9 landcover
categories and a resolution of 5 m, we clipped in an Arc GIS environment the surrounding
areas of 20, 50, 100, 200 and 500 ha of each sampling plot. For all these areas we computed
53 landscape level variables of landscape structure, using the software FRAGSTATS. In a
first analysis, we grouped the landscape structure variables into the six categories area,
shape, isolation, contrast, texture and diversity and tested for significant relations between
the landscape variables and the species richness of the 6 taxa, calculating partial Spearman
correlation coefficients. Then we evaluated pattern of correlations of the different taxa, metric
groups and scales. In a second analysis we compared different sets of variables as
predictors of species richness. These sets were based on A) correlation analysis of the
landscape variables, B) on expert knowledge and C) on random choice. We applied multiple
linear regressions to examine the performance of the different landscape variable sets as
predictors of the observed species richness. Additionally we performed multiple linear
regression using all variables in a forward procedure to define D) an optimal set of predictors
and to evaluate its performance.

Results

Concerning the correlation matrices, we found out that the scale affected the number of
landscape metrics that had a significant correlation with the species richness of the different
taxa. Woody plants were predicted better by landscape metrics at the scales 50-100 ha,
while reptiles and birds at the scale 500 ha. Other taxa like orthoptera and amphibians
performed best at the lowest scale (20ha), while there was no significant correlation between
the species richness of orchids with any landscape variable at any scale. Finally, the

337
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

correlations of the total species richness (the sum of the number of species of each taxon),
were stable from 20 until 100 ha and declined towards 500 ha (fig. 1).

Figure 1. Relations between landscape structure, organism groups and scale expressed by
the number of variable groups (out of the six groups “area”, “shape”, “contrast”, “isolation”,
“texture”, and “diversity”) containing at least one landscape metric of significant partial
correlation with the species richness.

When comparing the performance of the sets of variables, we could not detect any
significant differences between the sets obtained by correlation analysis, expert knowledge
and random choice (A-C). On the other hand, the optimal set (D) was explaining up to 46.5
% of the remaining variance, performing much better than any of the other sets for each
taxon and scale. These results lead to the conclusion that a priori definition of sets of
variables could be problematic when trying to model the relations of landscape structure and
biodiversity and that a preliminary analysis of the performing of different variables and sets
should be completed.

References
Kati, V.; Devillers, P.; Dufrêne, M.; Legakis, A.; Vokou, D. & Lebrun, Ph. (2004a) Hotspots,
Complementarity or Representativeness? Designing optimal small-scale reserves for biodiversity
conservation. Biological Conservation 120: 475-484.
Kati, V.; Devillers, P.; Dufrêne, M.; Legakis, A.; Vokou, D. & Lebrun, Ph. (2004b) Testing the value
of six taxonomic groups as biodiversity indicators at a local scale. Conservation Biology 18: 667-
675.
Kati, V.; Dufrêne, M.; Legakis, A.; Grill, A. & Lebrun, Ph. (2004c) Conservation management for
Orthoptera in the Dadia reserve, Greece. Biological Conservation 115: 33-44.
Kati, V. & Sekercioglu, C.H. (2006) Diversity, ecological structure, and conservation of the landbird
community of Dadia reserve, Greece. Diversity and Distributions 12: 620–629.

338
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

The interest of modeling the effect of habitat loss and fragmentation on

population dynamics at different nested observation scales.

J.B. Pichancourt1, P. Auger 2, F. Burel 1

1
CAREN, CNRS – Campus de Beaulieu 35042 Rennes cedx, France.
e-mail: [email protected]
2
IRD, GEODES – centre de recherche d'Ile de France 32, avenue Henri varagnat Bondy
cedex 93143 France.

Introduction

Our goal is to propose a framework to study population dynamics at different nested

observation scales in a landscape. We present an application of its interest to understand the
impact of some landscape changes as habitat loss and fragmentation on the population
dynamics of a forest ground beetle (Abax parallelepipedus: Coleoptera, carabidae). This
work is the follow up of two previous papers (Pichancourt et al., 2006a and b).

Material and methods

In a landscape, different nested observation scales can be defined and linked using
aggregation/zoning methods for landscape spatial units (Openshaw and Taylor, 1979). This
method allows us to link four classical nested observation scales used in landscape ecology,
from the most spatially explicit to the most spatially implicit: 1) the micro-habitat scale, 2) the
habitat-patch scale, 3) the habitat-class scale and 4) the landscape scale (McGarigal et al.,
2002). At each scale, one model of population dynamics can be built: 1) a grid cell model, 2)
a multipatch or metapopulation model, 3) a habitat-class model, and 4) a landscape scale
demographic model. To hierarchically link them across scales, we use a method of variables
aggregation (Bravo de la Parra et al., 1995) applied to spatial matrix population models
(Hunter and Caswell, 2005). This avoids aggregation errors between nested variables and
parameters. We propose a synthetic view of the structure and dynamics of A.
parallelepipedus population along gradients of habitat loss and fragmentation at different
scales. Here we present the link between two observation scales: habitat-classes and
landscape scale. In the simulated landscapes, the proportion of woods and crops vary, whilst
it is fixed for hedgerows (5%).

Results

Population analysis at the landscape scale

We show that population viability at the landscape scale is affected by habitat loss and
fragmentation. A critical threshold of population viability appears to depend on the habitat
amount and fragmentation rate. For more than 44% of wood land cover, the population is
viable whatever the fragmentation rate. Below the 44% threshold, extinction depends on
fragmentation rate. A sensitivity analysis of the population viability to demographic
parameters shows that they are all important for habitat loss and fragmentation gradients.
Fecundity is the parameter most affected by habitat loss and fragmentation (f = 2.81 with
100% of wood land cover, f = 2.42 with 44% of wood land cover). Adult survival is less
affected (s = 0.45 to s = 0.435), and survival of Larvae/pupae is not affected, because it is an
immobile stage at this scale.

Analysis at the habitat-class scale

339
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

The demographic parameters (in each class of habitat) and the movement parameters (at
the boundary between two classes of habitats) were fixed at this scale in the model, so they
cannot vary. But we show with a sensitivity analysis on population viability, that the relative
importance of the different habitat-classes and boundaries (via the there associated
parameters) vary according to habitat loss and fragmentation. The more wood land loss or
fragmentation, the more population viability is sensitive to adjacent habitats (crops or
hedgerows), then to inner boundaries (between wood land and hedgerows or crops), and
finally to the non adjacent boundary (between crops and hedgerows). This evolution has
consequences for the definition of the best targets (i.e. habitats and/or boundaries) for
landscape management or for biological data collection (for instance to improve parameter
estimation).

Discussion

The aggregation methods for dynamic systems and for geographic spatial units can be
combined in a general hierarchical theoretical framework to understand population dynamics
at different observation scales in landscape ecology. Each model represents a particular
observation scale and offers a piece of information to fully understand the landscape effects
on population structure and dynamics. In our example, the landscape scale is important in
the understanding the overall demographic sensitivity of this species to global spatial
changes. The habitat class level emphasizes the relative/complementary roles of source and
sink for the different classes of habitats and boundaries. For future work, we propose this
framework to link the four nested scales and the models described above. In this way, we
provide an example of Wu's prospects (Wu, 1999) on the importance of the hierarchical view
of landscape ecology to enhance a full understanding population dynamics and structure.

References
Hunter, C.M. and H. Caswell. 2005. The use of the vec-permutation matrix in spatial matrix
population models. Ecological Modelling 188:15-21.
Bravo de la Parra, R., Auger, P. and Sanchèz, E., 1995. Aggregation methods in discrete models. J.
Biol. Syst. 3: 603-612.
McGarigal K, Cushman SA, Neel MC, Ene E. 2002. FRAGSTATS: spatial pattern analysis program
for categorical maps.
Openshaw, S. and Taylor, P. 1979. A million or so correlation coefficients: three experiments on the
modifiable area unit problem. In Statistical Applications in the spatial sciences, pp. 127- 144.
Edited by Wrighley, N. Pion: London
Pichancourt J-B, Burel F, Auger P. 2006a. Assessing the effect of habitat fragmentation on
population dynamics: An implicit modelling approach. - Ecol. Mod 192: 543-556.
Pichancourt J-B, Burel F, Auger P. 2006b. Assessing the impact of habitat fragmentation on
population dynamics: An elasticity analysis. - C.R. Biologie 329: 31-39
Wu, J. 1999. Hierarchy and scalling: extrapolating information along scaling ladder. Canadian Journal
of Remote sensing 25: 367-380.

340
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

Patchy responses to a patchy landscape: the spatial structure of habitat selection

S.J. Mayor1,2, D.C. Schneider1, J.A. Schaefer1,3, S.P. Mahoney2

1
Memorial University of Newfoundland, Department of Biology, St. John’s, NL, Canada.
E-mail: [email protected].
2
Institute for Biodiversity and Ecosystem Science and Sustainability, Government of
Newfoundland and Labrador, P.O. Box 8700, St. John’s, NL, A1B 4J6, Canada.
3
Trent University, Biology Department, 1600 West Bank Dr., Peterborough, ON, K9J 7B8,
Canada.

The caribou’s-eye-view

Habitat selection, the disproportionate use of available resources, is a multi-scaled

phenomenon. Selection depends on the scales perceived by organisms, and our ability to
detect selection depends on analytical scale. However, conventional multi-scaled studies of
selection have been limited by the use of discrete, arbitrary scales because a quantitative
basis has not existed for evaluating how animals perceive the availability of habitat. We
examine selection from the perspective of woodland caribou (Rangifer tarandus caribou) in
the maritime barrens of Newfoundland, Canada, by allowing their response to winter habitat
define the scale domains of selection.

Spectrum of selection

We unite habitat selection research with spatial and geostatistical analyses to

investigate the spatially structured response of mobile animals to a heterogeneous
landscape by caribou. When habitat components are correlated across scales
interpretations of correlations between organisms and habitat can be impeded, so we
developed new two new approaches to quantify habitat selection as a reduction in variance.
We employ spatial continua to represent patterns of heterogeneity across scales by
developing two new analytical approaches – one based on lagging, and a second based on
coarse-graining (Fig. 1). Variogram analysis operates in the distance domain and compares
variability between pairs of samples at given separation distances (or lags) (Matheron, 1960).
Blocked quadrat variance, a form of coarse-graining, operates in the frequency domain and
compares variability among contiguous blocks in different sized grids, in relation to the size
of those blocks (Grieg-Smith, 1952). Employing spatial continua avoided assumptions of the
spatial extent of behavioural levels, thereby enabling these levels to be customized to the
specific ecology of the taxon under investigation. We applied these techniques at four levels
of habitat use (population winter range, travel routes, feeding areas, and feeding microsites),
each nested within available habitat at coarser levels.

Spatially structured habitat selection

By comparing the spatial structure of habitat components in used and available sites
(at several levels of behaviour) we went beyond describing patterns of spatial variability in
the environment to evaluating the behavioural processes resulting from those patterns. We
stepped from quantifying the spatial structure of habitat to quantifying the spatial structure of
habitat selection.
Caribou consistently selected favourable (and avoided poor) habitat such that
preferred habitat components were less variable in selected sites than in the available
environment. Selection was strongest for Cladina lichens (the herd’s primary winter food
resource) and snow depth (an indicator of the energetic costs of foraging and moving) at lag
distances up to 13 km (Fig. 2a). Within this range, selection decreased with lag, but
avoidance of deep snow occurred at all lags and was accomplished at several behavioural
levels (Fig. 2b).

341
Theme 3. Ecological Networks, fragmentation and connectivity
3.5 Open Session 10: Landscape modelling and populations ecology

Caribou responded to habitat heterogeneity at all scales, and this response was
greatest at the scales of highest patchiness (Fig. 3). Our results implicate habitat
heterogeneity as an underlying cause of multi-scaled habitat selection.

Implications for limiting factors

When habitat components are selected at different spatial scales, the scales of
selection for each resource may reflect the relative importance of escaping the effects of
limiting factors (Rettie and Messier 2000). However, caribou selected habitat variables
across overlapping scaling domains, suggesting that caribou selected for Cladina cover and
soft, shallow snow to make a trade-off between the potentially limiting effects of forage
abundance and accessibility.

Figure 1. Comparison of Figure 2. Variograms, with Figure 3. Blocked quadrat

analytical approaches for habitat selection variance, with habitat
multiscale habitat selection represented by the selection represented by the
research. Points represent discrepancy between discrepancy between
sampling locations. Top row semivariance at crater (♦), semivariance at crater (♦),
shows conventional feeding area (■), and travel feeding area (■), and travel
hierarchical approach. route (▼) levels relative to route (▼) levels relative to
the winter range (●) level. the winter range (●) level.

References
Greig-Smith P (1952) The use of random and contiguous quadrats in the study of the structure of
plant communities. Annals of Botany, New Series 16: 293-316.
Matheron G (1960) Principles of geostatistics. Economic Geology 58: 1246-1266.
Rettie WJ, Messier F (2000) Hierarchical habitat selection by woodland caribou: its relationship to
limiting factors. Ecography 23: 466-478.

342
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

3.6 Open Session 11: Landscape modelling and bird populations

Conservation of understory birds in fragmented landscapes: relative importance

of habitat cover and configuration

A. C. Martensen & J. P. Metzger

Depto. of Ecology, Bioscience Institute, University of São Paulo, Rua do Matão, 321,
travessa 14, 05509-900, São Paulo, SP, Brazil.
e-mail: [email protected]

Introduction
Habitat loss and fragmentation are considered key factors currently threatening
biodiversity (Wilcox and Murphy, 1985, Fahrig, 2003). There is a current debate in the
literature (see revision in Fahrig 2003), where the effects of habitat cover and configuration
are discussed, and there is no agreement at present. Some authors have suggested a major
importance in habitat amount variables (McGarigal & McComb 1995, Drolet et. al. 1999,
Trzcinski et. al. 1999), whilst others have suggested a larger importance of habitat
configuration (Andrén 1994, Fahrig 1997, Villard et. al. 1999, Develey & Metzger 2006).
There is a further hypothesis, where the relative importance of these variables may change
along the habitat conversion gradient. Some authors even go further and have suggested a
threshold point where configuration aspects could have a larger importance (around 20%,
Andrén 1994, Fahrig 1997, 2001 e 2003, Flather & Bevers 2002).

Objective & Methods

In this context, our objective was to analyze the influence of the extent habitat and its
configuration in the bird community, modeling the influence of the percentage of forest in
three different radii (300, 500 and 800 m) as a measure of habitat amount; and fragment size
and connectivity in different spatial scales, as configuration metrics. We sampled 34
fragments with different sizes and connectivity values, split equally in two regions of 10000
ha each, with different proportions of habitat (10 and 30%, that was also used in the habitat
amount analyzes as two categories, and were chosen because of the threshold point
suggested above) in the Atlantic Plateau of São Paulo, with the same forest quality and
structure conditions. Connectivity variables taken in consideration were different scales of
perceiving connectivity, diverse distances through open habitats (20, 40, 60 and 80 m) and
the amount of habitat linked by a corridor of < 100m of width. As bird metrics we used
community richness and abundance and species abundance. The sizes of the fragments
were given by the area of the fragment that has width > 100 m, in addition to the areas with
width < 100 m that do not connect to other forests areas. The areas < 100 m of width that
connect to other forest areas were considered in the connectivity metric that accounts for the
amount of habitat linked by corridors. The other connectivity metrics were given by the
amount of habitat accessible for an individual of a species that is capable of crossing 20, 40,
60 and 80 m of open habitat. Matrix heterogeneity was not considered, since it was very
similar. Comparisons within regions were made with the observed values of richness and
abundance, since effort in all sampling spots from the same region was very similar.
However, for comparisons between regions, we did 9908 bootstraps of 1000 individuals in
each case, and then compared the obtained simulated results. We used AICc obtained from
likelihood regressions with Poisson error distribution and the evidence ration, following
Burnham and Anderson (2002) to compare the models.

Results
Our results showed the frequent presence of the variable representing the two different
regions (10 and 30% different proportions of forest), both in their influence as a covariable
and as an interaction, showing that bird community behavior differs from region to region,

343
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

and that landscape parameters also vary from region to region. We observed the whole
gradient from the positive affected by a smaller amount of habitat in the landscape, such as
Conopophaga lineata and Basileuterus leucoblepharus. These are edge species which are
not severely affected whereas the drastically negatively affected species are mainly
terrestrial insectivores, trunk insectivorous, frugivorous/omnivorous and small insectivorous
species. These species performed consistent through the different metrics, responding
positively in relation to the extent of habitat and in relation to fragment size and connectivity,
in an opposite way to the edge species. Some particularly affected species were principally
affected by corridor connectivity, especially Sclerurus scansor, a terrestrial insectivore and
Lepidocolaptes fuscus and Sittasomus griseicapillus¸ two trunk insectivorous species.

Discussion
Our results showed that in regions where a large proportion of forest is still present, bird
community composition and distributions are mainly related to connectivity variables and also
to habitat amount, whilst in areas where a lower amount of forest is present, fragment size
was more important.
Management policies should therefore be different in areas with different amounts of
forest, in some cases favoring the enhancement of fragment size and in others connectivity.
Nevertheless, while increasing fragment size or connectivity, the amount of habitat
automatically increases. Therefore, depending on the extent of the restoration, the focus
should be on connectivity.
Nevertheless, some species seem to respond differently according to the analyzed scale
for all variables. For example L. fuscus responds well to the extent of habitat at the smaller
scales, but poorly at all other scales. This suggests that for different analyzed scales, we can
obtain different relative influences, varying according to the species perception of
landscapes.

References
Andrén, H. 1994. Effects of habitat fragmentation on birds and mammals in landscapes with different
proportions of suitable habitat - A Review. Oikos 71(3): 355-366.
Burnham, K. P., & Anderson, D. R. 2002. Model selection and multimodel inference: A practical
information-theoretic approach. New York: Springer-Verlag.
Develey, P. F. & Metzger, J. P. 2006. Birds in Atlantic forest landscapes: effects of forest cover and
configuration In: Emerging Threats to Tropical Forests ed. W. F. Laurance. Chicago : University of
Chicago Press.
Drolet, B.; Desrochers, A. & Fortin, M. J. 1999. Effects of landscape structure on nesting songbird
distribution in a harvested boreal forest. Condor 101: 699-704.
Fahrig, L. 1997. Relative effects of habitat loss and fragmentation on population extinction. Journal of
Wildlife Management 61: 603-610.
Fahrig, L. 2001. How much habitat is enough? Biological Conservation 100: 65-74.
Fahrig, L., 2003. Effects of habitat fragmentation on biodiversity. Annual Review of Ecology, Evolution
and Systematics 34, 487-515.
Flather CH, Bevers M. 2002. Patchy reaction diffusion and population abundance: the relative
importance of habitat amount and arrangement. American Naturalist 159:40–56
McGarigal, K. & McComb, W. C. 1995. Relationship between landscape structure and breeding birds
in the Oregon Coast Range. Ecological Monographs 65: 235-260.
Trzcinski, M.K., Fahrig, L., Merriam, G. 1999. Independent effects of forest cover and fragmentation
on the distributionof forest breeding birds. Ecological Applications 9: 586-593.
Villard, M. A., K. Trzcinski, and G. Merriam. 1999. Fragmentation effects on forest birds: relative
influence of woodland cover and configuration on landscape occupancy. Conservation Biology 13:
774-783.
Wilcox, B.A., Murphy, D.D., 1985 Conservation strategy: the effects of fragmentation on extinction.
American Naturalist 125 (6), 879-887.

344
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Understory birds and forest cover: a strong association in a highly forested

region in the Atlantic forest, Brazil

C. Banks, J.P. Metzger

Universidade de São Paulo, Instituto de Biociências, Departamento de Ecologia, Rua do

Matão 321, trv 14, postal code 05508-900, Cidade Universitária, São Paulo – SP, Brazil.
e-mail: [email protected]

Introduction

Habitat loss has been indicated as the major factor contributing to species extinction, while
the influence of fragmentation is not as well established since both positive and negative
outcomes have been reached (Fahrig, 2003). These effects may depend on several factors,
such as the amount of habitat cover in the landscape and species behavior. It has been
proposed that over 20-30% of forest cover the effects of fragmentation are reduced or
inexistent (Fahrig, 2003). In highly forested areas it is expected to find more sensitive
species, with greater area requirements and less ability to cross open areas, while in
intensively fragmented landscape (Stotz et al., 1996). This project aimed to study how habitat
loss and fragmentation affect understory birds in an Atlantic Forest region located in the
Plateau of São Paulo, Brazil.

Methods

We studied two landscapes of 10.000 ha each, one with 45% of overall forest cover
(fragmented landscape) and one with > 90% of forest cover (forested landscape). In the
fragmented landscape, we chose 19 sampling points ranging from 20 to 80% of forest cover
within a 800 m radius (approximately 200 ha), considering different fragmentation levels,
measured by the Matheron Index. Understory species usually have area requirements of
approximately 10 ha (Stotz et al. 1996), so this area should comprise home range of several
individuals. We also sampled four sites in a 100% forest cover context in the forested
landscape. Each site was located at least 50 m from forest edges and distant to each other
by a minimum of 1200 m. The understory bird community was sampled with the use of mist
nets for 510 net-hours in each site. We used linear regressions to analyze how forest cover
percentage and fragmentation affected species richness and abundance of the overall
assemblage of birds, insectivorous, frugivorous, nectarivorous and forest interior bird
species.

Results

In total 1613 birds of 91 species were captured, including 1222 individuals of 72 species at
the fragmented landscape and 367 individuals of 61 species at the forested landscape.
Among patches, forest cover positively affected the richness of the overall bird assemblage,
frugivorous and forest interior birds, and the abundance of the forest interior birds. When
continuous forest sites were included in the regression models, forest cover continued to
positively affect the same variables, except for forest interior species abundance.
Fragmentation measured by the the Matheron Index positively affected the overall bird
species richness. However, when continuous sites were included, fragmentation instead
negatively affected the abundance of nectarivorous birds (Table 1).

Discussion
As predicted (Fahrig 2003, Develey & Metzger, 2006), forest cover consistently affected
species richness and abundance. In addition, the squared multiple R shows that a great part
of the variability in species richness and abundance is explained by the forest cover
percentage in an 800 m radius neighborhood. This shows that in highly forested regions, the

345
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

estimation of habitat cover in an area of 800 m radius could be a good predictor of the overall
species richness and of forest interior species richness, those of greatest value for
conservation. Conversely, fragmentation yielded weaker and contrasting results. The overall
species richness among forest patches was positively affected by fragmentation, but due to
an interaction with forest cover. This result indicates that more forested and fragmented sites
have higher richness since they are able to comprise forest interior species and edge habitat
species in the same place. Hummingbirds, which are thought to be benefited by habitat loss
and fragmentation (Stouffer & Bierregaard, 1995), in this study, had its abundance negatively
affected by the Matheron Index. It is possible that these birds are more sensitive than
previous studies indicate. Therefore, in highly forested regions such as this, fragmentation is
in fact less important to predict species richness and abundance than the amount of forest
cover. Although it benefits edge habitat species, the effects of fragmentation, in this region,
are not strong enough to endanger more sensitive species.

Table 1. Linear regression results for the overall bird assemblage, nectarivorous,
insectivorous, frugivorous and forest interior bird species, in relation to forest cover
percentage (% cover) and the Matheron Index (Frag). Squared multiple R and probability of
type I error (P) are separated for forest fragments only (19 sampling areas) and for forest
fragments and continuous forest sites (23 areas). When independent variable was
significant, it is indicated inside parenthesis whether effects were positive or negative. Inside
brackets it is indicated the total richness and abundance of each bird assemblage.
Continuous forest and
Forest fragments
fragments
r2 % cover Frag r2 % cover Frag
Overall bird species
Richness [91] 0.288 0.027 (+) 0.043 (+) 0.414 0.04 (+) 0.437
Abundance [1613] 0.004 0.911 0.807 0.257 0.973 0.319
Nectativorous birds
Richness [9] 0.241 0.592 0.238 0.283 0.556 0.115
Abundance [124] 0.173 0.157 0.821 0.321 0.144 0.023 (-)
Insectivorous birds
Richness [58] 0.157 0.118 0.156 0.231 0.127 0.486
Abundance [987] 0.009 0.712 0.775 0.111 0.714 0.786
Frugivorous birds
Richness [19] 0.296 0.02 (+) 0.086 0.415 0.015 (+) 0.187
Abundance [312] 0.002 0.947 0.951 0.24 0.898 0.274
Forest interior species
Richness [45] 0.493 0.003 (+) 0.277 0.796 0.002 (+) 0.97
Abundance [365] 0.39 0.024 (+) 0.803 0.775 0.149 0.071

Financial support was provided by CNPq («No_Proc_Inst») for field work and a PhD scholarship.

References
Develey, P.F. & Metzger, J.P. (2006) Extinction thresholds in Atlantic forest birds: effects of
landscape cover and configuration. In: Laurance, W. & Peres, C. (Eds.). Emerging Threats to
Tropical Forests. University of Chicago Press, Chicago.
Fahrig, L. (2003) Effects of habitat fragmentation on biodiversity. Annual Review of Ecology, Evolution
and Systematics, 34:487-515.
Stotz, D. F.; J. W. Fitzpatrick; T. A. Parker III and D. K. Moskovits, (1996) Neotropical Birds:
Ecology and Conservation. Chicago Press
Stouffer, P. C. & Bierregaard, R.O. (1995) Effects of forest fragmentation on understory
hummingbirds in Amazonian Brazil. Conservation Biology 9(5):1085-1094.

346
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Modelling the influence of landscape structure on the incidence pattern of bird

species living in fragmented Brazilian Atlantic forest

D. Boscolo1, J.P. Metzger 2, K. Frank 3.

1,2
Departamento de Ecologia, Instituto de Biociências, Universidade de São Paulo - USP,
Rua do Matão, trav. 14, nº 321, Cid. Universitaria, São Paulo, Brazil, zip: 05508-900.
1
e-mail: [email protected]
3
Helmholzt zentrum für Umweltforschung – UFZ, Ökologische Systemanalyse-OESA, PF
500135, 04301 Leipzig, Germany.

Introduction

Habitat loss and fragmentation are among the most important threats to biodiversity
conservation. They reduce habitat availability, increase isolation and generate patchy
environments. At highly fragmented landscapes, changes in habitat spatial structure may
profoundly affect species survival, leading several populations to higher chances of local
extinction (Wiens, 1995; Lindenmayer et al., 1999; Meyer and Cameron, 2003). Proposing
efficient conservation strategies for such species relies initially on understanding how the
structure of fragmented landscapes may influence their incidence pattern (Cushman and
McGarigal, 2004). The objective of the current study was to understand the influence of
forest spatial distribution on the occurrence of three Brazilian Atlantic Forest bird species
(Chiroxiphia caudata, Xiphorhynchus fuscus and Pyriglena leucoptera) found in fragmented
environments in order to model their incidence pattern based on patch size, connectivity and
the surrounding landscape structure.

Methods

Presence/absence data
Four fragmented landscapes from the Atlantic plateau of São Paulo, Brazil, were used for
the current study. Data on the birds’ presence/absence pattern were collected at the central
point of 80 forest patches (20 at each landscape) with the use of playback techniques. The
methodology used enabled a large number of sites to be surveyed with great precision, while
reducing the risk of false absence records (Boscolo et al., 2006).

Incidence models
Patch size and other eight indices describing the landscape structure inside an 800 m
radius circle surrounding each sampled point were estimated using FRAGSTATSTM
(McGarigal et al., 2002). Multivariate logistic models were used to analyse the relationship
between the incidence pattern of the birds and the landscape structure. Such models were
generated using backward stepwise regressions and the minimal adequate ones were
selected by the lowest Akaike’s Information Criterion (AIC).

Results and Discussion

The minimum adequate models of each species contained only two (C. caudata and P.
leucoptera) or three (X. fuscus) of the original explanatory variables (Tab. 1). Bird incidence
was in general positively affected by higher habitat cover and functional connectivity and by
smaller inter-patch distances. These results indicate that the studied species are able to
overcome short distances through the matrix. This allows the birds to supplement their
habitat needs in highly fragmented landscapes by including several nearby patches in their
home-ranges, a process called landscape supplementation (Dunning et al., 1992). Broad
conservation strategies concerning small passerine birds living in fragmented Atlantic forest
should focus firstly, but not exclusively, on increasing patch density and connectivity, in order
to reduce the average distance between forest remnants, essential for the persistence of a

347
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

wide range of species. This is especially important for places where human activity is intense
and the implementation of large reserves is not possible. In such cases, the promotion of
forest networks and stepping-stones between patches might be the most effective restoration
strategy. The generated models can also be used to produce habitat maps based on the
birds’ perception of the landscape, which are useful to be applied in spatially explicit
population analysis.

Table 1. Final multivariate models for all species. β: regression coefficient for each selected
variable; χ2: test values with significance in brackets; AIC: Akaike’s information
criterion; R2: model’s predictive power. See McGarigal et al. (2002) for detailed
variable descriptions.
Species Explanatory variables β χ2(p) AIC R2

C. caudata •
Patch size 0.1597 32.32 (<0.001) 46.289 0.563
•
Mean Euclidian distance to the -0.0325
nearest patch

X. fuscus •
Patch density 0.8086 30.18 (<0.001) 64.984 0.475
•
Percentage of patch connections 0.1908
considering a functional linkage
distance of 150 m
•
Mean Euclidian distance to the 0.0211
nearest patch

P. leucoptera •
Patch density 0.1212 27.09 (<0.001) 89.365 0.384
•
Proportion of forest 0.0845

Financial support was provided by FAPESP (n. 05123-4) and CNPq (n. «No_Proc_Inst» and
200848/2005-4).

References
Boscolo, D; Metzger, J.P. & Vielliard, J.M.E. (2006) Efficiency of playback for assessing the
occurrence of five bird species in Brazilian Atlantic Forest fragments. Anais da academia brasileira
de ciências 78: 629-644.
Cushman, S.A. & McGarigal, K. (2004) Patterns in the species-environment relationship depend on
both scale and choice of response variables. Oikos 105: 117-124.
Dunning, J.B; Danielson, B.J. & Pulliam, H.R. (1992) Ecological processes that affect populations in
complex landscapes. Oikos 65: 169-175.
Lindenmayer, D.B; McCarthy, M.A. & Pope, M.L. (1999) Arboreal marsupial incidence in eucalypt
patches in south-eastern Australia: a test of Hanski’s incidence function metapopulation for pacth
occupancy. Oikos 84: 99-109.
Meyer, A.L. & Cameron, G.N. (2003) Landscape characteristics, spatial extent, and breeding bird
diversity in Ohio, USA. Diversity and Distributions 9: 297-311.
Wiens, J.A. (1995) Habitat fragmentation: island vs landscape perspectives on bird conservation. Ibis
137: 97-104.

348
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Airborne Laser Scanning data provide three-dimensional information for analyses

of woodland bird habitats

H. Hashimoto1, J. Imanishi2 and Y. Morimoto2

1
Faculty of Agriculture, Meijo University, 501 Shiogamaguchi 1, Tenpaku-ku, Nagoya City,
468-8502, Japan.
e-mail: [email protected]
2
Graduate School of Global Environmental Studies, Kyoto University, Kyoto City, 606-8502,
Japan

Introduction

Woodland bird species habitats are affected by the vertical forest structure and by
topography. We analyzed woodland birds - habitat relationship in a hillside forest in Kyoto
City, central Japan, from a landscape ecological perspective using a multilayer GIS map
created from two types of remote sensing data.

Method

Study area
The study area is the Kamigamo Experimental Station of Field Science Education and
Research Centre Kyoto University, a hillside secondary forest in Kyoto City, central Japan.
This forest consisted from mainly evergreen coniferous trees and deciduous broad-leave
trees.

Bird data
Bird data were collected using an hour line census during the breeding seasons and the
wintering seasons of 2002-2004. Bird survey was conducted twelve mornings without rain.

Environmental variables
Remote sensing data we used were airborne laser scanning data (RAMS library data)
obtained in the spring of 2001 and a high spatial resolution satellite image by Quick Bird
obtained in the autumn of 2003. Density of RAMS data is approximately a laser flux per a 2
m x 2 m space, and it records maximally five pulses per a laser flux.
We created maps of maximum tree height, mean vegetation height, slope gradient and
slope aspect from the airborne laser scanning data, and the percentages of vegetation cover
maps of evergreen trees, deciduous trees and shrub (under 5 m tall) from the Quick Bird
satellite image and laser scanning data. These indices were calculated in each 30 m x 30 m
grid. Therefore, a GIS based vegetation and topography map consisted from seven raster
layers was derived. The procedure for creating DEM from laser scanning data was followed
Yonedu and Hasegawa (2004). Tree height was calculated from 5 m x 5 m grid of DSM and
DEM.
Maximum tree height, mean vegetation height, and the percentages of evergreen trees,
deciduous trees and shrub in each 90 m x 90 m grid were also calculated from the GIS map
as environmental variables.

Data analysis
A GIS based vegetation and topography map and bird data were overlaid (Figure 1), and
habitat preferences of each species were analyzed.

349
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Figure 1. A GIS based vegetation map and bird data. A grid size is 30 m x 30 m.

Results & Discussion

The results of overlaying vegetation and topographic map and bird data, habitat
preferences of some species were quantified. For instance, Jungle Crows Corvus
macrorhynchos preferred steep slope, Black-faced Buntings Emberiza spodocephala in
wintering season and Bush Warblers Cettia diphone in breeding season need some area of
shrub vegetation.
Application of laser scanning data for bird habitat evaluation was proposed by Hinsley et
al. (2002). Laser scanning data provide us three-dimensional information such as vegetation
height and topographic information also. Especially vegetation height is very important for
evaluating woodland bird habitat. Although field measurement of tree height in broad area is
difficult, airborne laser scanners can easily measure tree height distribution in broad area.
Although obtaining high resolution of laser scanning data is very expensive, library data
have already prepared around urban area in Japan, and we can easily purchase them.
Although there are several limitations in use of library data (i.e. we cannot design scanning
year and season), laser scanning data provide us useful three-dimensional information of
forest vegetation for evaluating woodland bird habitats in landscape level.

References
Hinsley, S.A.; Hill, R.A.; Gaveau, D.L.A. & Bellamy, P.E. (2002) Quantifying woodland structure and
habitat quality for birds using airborne laser scanning. Functional Ecology 16: 851-857.
Yonedu, K & Hasegawa, N. (2004) A filtering method for creating DEM from airborne LiDAR in
mountainous forests. Abstracts for 115th meeting of the Japanese Forestry Society: 476. (in
Japanese)

350
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Effects of landscape patterns and resource distribution on frugivorous birds in

the Central Andes of Colombia: tracking ecological thresholds

J. E. Mendoza1, G. H. Kattan2, J. de Jong3

1
Swedish Biodiversity Centre – Swedish University of Agricultural Sciences, Uppsala,
Sweden and Alexander von Humboldt Institute, Diag. 27 #15-09, Bogotá, Colombia.
e-mail: [email protected]
2
Fundación EcoAndina/Wildlife Conservation Society Colombia Program, Av. 2 oeste #10-
54. Cali, Colombia.
3
Swedish Biodiversity Centre – Swedish University of Agricultural Sciences, Bäcklösavägen
10, Box 7007 S-750 07 Uppsala, Sweden

Introduction

Understanding wildlife distributional patterns and their relationship with landscape

characteristics and resource distribution is key for their conservation, especially in rural
landscapes. Theoretically, there are threshold values in habitat amount and configuration,
below which species persistence is compromised. Searching for ecological thresholds is one
of the most promising avenues to ensure the conservation of many species (Hugget, 2005).
In this sense, the recognition of empirical thresholds existing in landscapes is a major step
forward in understanding community level responses to landscape changes (Radford et al.
2005). We analyzed the response of forest frugivorous bird species to the distribution of key
fruiting trees (Cecropia, Ficus and Miconia) and landscape characteristics of forest patches
(habitat amount, edge length, number of land covers, number of patches) at three different
spatial scales, local (3 ha), middle (312 ha) and landscape (2500 ha), in the Central Andes of
Colombia to look for evidence of statistically significant ecological thresholds.

Methods

We used data from four 2,500 ha rural landscapes with different fragmentation levels
(80%, 46%, 35% and 25% forest cover) between 1,700 and 2,100 m on the western slope of
the Central Andes. Data on species richness and abundance of frugivorous birds were
collected in point counts (50m radius) in forest patches, and abundance of key fruiting trees
were collected in 4 adjacent 50x4m transects. To evaluate relationships between bird
species abundance and distribution, and potential explanatory variables (landscape
characteristics, and fruiting tree abundance and richness) at the three ecological scales, we
ran canonical correspondence analyses (CCA). We combined stepwise logistic regressions
and receiver operation characteristic (ROC) curves to detect thresholds in species
occurrence (Guénette & Villard, 2004, 2005). For this, we used single specific explanatory
variables to test the bird responses. Because we found significant spatial autocorrelation in
our variables (Moran’s Index between 0.16 to -0.22), and to reduce the likelihood of
committing Type I errors, we used a conservative level of significance in our analysis (α =
0.01). However, for threshold identification, marginally significant values (α = 0.02) were also
considered.

Results

We found a consistent response of the explanatory variables across scales. Axis 1 in the
CCAs grouped landscape variables, which explained 28-40% of the variance in bird species
richness. Amount of suitable habitat was always opposite to edge length, number of land
covers and number of patches in this axis. All forest interior bird species were positively
correlated with this variable. Axis 2 in CCAs represented fruiting resources distribution at all
scales. The three first eigenvalues explained 43% of total variability at local scale, 59% at
middle scale and 79% at the landscape scale.

351
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

At the local scale (3 ha), out of 23 bird species abundant enough to be included in the
analysis, 14 sensitive species showed a significant relationship to at least one variable. Nine
species showed a negative relationship to edge length and number of land covers, and 13
were positive to variables like patch area, amount of suitable habitat amount and abundance
of Ficus. Curves of the expected number of sensitive species (based on their ROC-derived
thresholds) showed that forest patches of 98 ha are necessary to conserve species
negatively affected by habitat alteration such as Penelope perspicax, Mionectes olivaceus,
and Aratinga wagleri. Habitat heterogeneity was detrimental for species like Aulacorhynchus
prasinus, Penelope perspicax, Pharomachrus auriceps which need less than 1.5 kinds of
land covers within 3 ha to ensure their presence. Furthermore, the presence of these forest
interior species seems to depend on edge lengths being shorter than 127.55 m.

At the middle scale (312 ha), 6 sensitive species showed a positive relationship to amount
of suitable habitat and Ficus abundance, whereas 2 species showed a negative relationship
to number of land covers. Arrangements of minimum 84 ha of forest, and fewer than 3.5 land
covers will ensure the presence of Euphonia xanthogaster. For Penelope perspicax and
Pyroderus scutatus, forest patches of 212 and 207 ha respectively are needed to ensure
their presence. At the landscape scale no significant relationships were found because of the
small number of replicates.

Conclusions

The community of frugivorous bird species responded more strongly to landscape

characteristics than to Cecropia, Ficus and Miconia abundance and richness. large-bodied
bird species, which are highly endangered, require large forest patches with few edges to be
conserved. Our study shows the relevance of local and middle scales in bird species
responses of tropical mountain ecosystems to landscape changes. The conservation of
species negatively affected by habitat alteration is a great challenge in these ecosystems,
and frugivorous birds are one of the most vulnerable guilds to landscape transformation
processes. Our work is one of the first initiatives to identify ecological thresholds related with
habitat requirements for tropical species and provide tools for conservation strategies,
landscape planning and ecological restoration initiatives which have sensitive bird species as
targets or phase indicators.

References
Guénette, J. S. and Villard, M. A. (2004). Do empirical thresholds truly reflect species tolerance to
habitat alteration? Ecological Bulletins 51:163-171.
Guénette, J. S. and Villard, M. A. (2005). Thresholds in forest bird response to habitat alteration as
quantitative targets for conservation. Conservation Biology 19:1168-1180.
Hugget, A.J., (2005). The concept and utility of ecological thresholds in biodiversity conservation,
Biological Conservation 124: 301–310.
Radford, J.Q., Andrew, A.F. & Cheers, G.J. (2005). Landscape thresholds of habitat cover for
woodland – dependent birds. Biological Conservation 124:317-337.

352
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Continent-scale patterns in temporal dynamics of avian assemblages

M. Maron1,2, P.K. Dunn 1,2 and A. Apan1,3

1
Australian Centre for Sustainable Catchments, University of Southern Queensland,
Toowoomba, Queensland 4350, Australia.
e-mail: [email protected]
2
Faculty of Sciences, University of Southern Queensland, Toowoomba, Queensland 4350,
Australia.
3
Faculty of Engineering & Surveying, University of Southern Queensland, Toowoomba,
Queensland 4350, Australia.

Introduction

A common method of characterizing the bird assemblages of sites in landscape

ecological research is to conduct a series of snapshot-type surveys, often over less than one
year (Maron et al. 2005). However, highly mobile taxa such as birds show substantial inter-
annual variability in species richness and community composition (e.g. Holmes et al. 1986),
and this may have implications for the veracity of snapshot-derived patterns. The
environmental correlates of inter-annual variability in bird assemblages are not well known,
although it seems likely that more climatically stable regions support less dynamic avian
assemblages (Jarvinen 1979). We conducted a continent-scale study to investigate
associations between environmental factors and annual bird assemblage turnover rates in
Australia. Correlates of inter-annual bird assemblage variability have rarely been investigated
at this scale as long-term data from across a broad area are required. However, the
availability of long-term, large-scale databases from volunteer-based monitoring initiatives
provides opportunities for comparisons across large areas. For example, data from the North
American Breeding Bird Survey revealed increased annual turnover in landscapes with
smaller forest patches (Boulinier et al. 2001).

Methods

We used Birds Australia Atlas data from the period 1998-2005 to examine bird
assemblage variability in regularly surveyed 2-ha sites across Australia. Only sites which had
been surveyed for a minimum of three consecutive years and in all seasons per year
(autumn, winter, spring, summer) were included in the analysis. For sites with multiple
surveys per season, data for one survey per season in each year were randomly selected for
inclusion. We excluded sites in close proximity so that all were at least 1km distant from one
another. Apparent annual turnover (T) for each pair of consecutive years was calculated by
expressing the number of compositional changes (new presences and new absences) as a
proportion of the total ‘species pool’ as follows (after Maron et al. 2005):
E +C
T= × 100
S tot
where E = number of species occurring in the first but not the second year, C = number of
species occurring in the second but not the first year, and Stot = total number of species
recorded across both years. The mean of T for each site is hereafter referred to as ‘turnover’.

Data collected by the volunteer observers about the site included the area of the vegetation
patch in which the 2-ha site was situated, grazing and logging history, distance to permanent
water, conservation status, degree of isolation, latitude, and longitude. We used data from
the Australian Bureau of Meteorology to add information on mean annual rainfall and the
annual rainfall variability index. Stepwise multiple regression using Akaike’s Information

353
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Criterion and weighted by the number of year-pairs available for a site was used to develop
predictive models for both turnover and mean annual species richness.

Results & discussion

The final multiple regression model for turnover included the variables rainfall variability
and distance to permanent water, with mean annual turnover higher in areas of higher rainfall
variability and further from water. However, there was a large number of missing values for
distance to water, and the overall explanatory ability of the model was low (R2 = 0.06). More
variable rainfall patterns result in increased year-to-year vegetation change, particularly in
the arid zone which comprises much of Australia. Nomadic behaviour in birds is likely to be
more prevalent in areas of higher rainfall variability (Wiens 1991) and such behaviour was
probably a major contributor to the pattern of higher bird community variability in sites with
higher rainfall variability.

The final model for mean annual species richness was able to account for 26% of the
variation in species richness and suggested higher species richness for sites with higher
average annual rainfall, lower annual rainfall variability and at lower latitudes. Sites with
higher mean species richness were more likely to have more stable bird communities.

Mean apparent annual turnover across all sites was 55%. This is broadly comparable with
the 63% apparent turnover recorded for birds in a series of woodland sites in southeastern
Australia (Maron et al. 2005) and the 52-60% observed by Collins et al. (2000) for breeding
birds in tallgrass prairie in the USA. The substantial interannual variability recorded suggests
that bird assemblages in Australia should not be viewed as stable or equilibrial but as
fluctuating substantially, particularly in areas of high climatic variability and low species
richness. The generally high turnover recorded even in fairly climatically stable regions
suggests that pictures of avian assemblages based on a few surveys over one year are
substantially incomplete.

References
Boulinier, T; Nichols, J. D; Hines, J. E; Sauer, J. R; Flather, C. H. & Pollock K. H. (2001) Forest
fragmentation and bird community dynamics: Inference at regional scales. Ecology 82: 1159-
1169.
Collins, S. L. (2000) Disturbance frequency and community stability in native tallgrass prairie.
American Naturalist 155: 311-325.
Holmes, R. T; Sherry, T. W. & Sturges, F. W. (1986) Bird Community Dynamics in a Temperate
Deciduous Forest: Long-Term Trends at Hubbard Brook. Ecological Monographs 56: 201-220.
Jarvinen, O. 1979. Geographical gradients of stability in European land bird communities. Oecologia
38: 51-69.
Maron, M; Lill, A; Watson, D. M. & Mac Nally, R. (2005) Temporal variation in bird assemblages:
how representative is a one-year snapshot? Austral Ecology 30: 383-394.
Wiens, J. A. (1991) Ecological Similarity of Shrub-Desert Avifaunas of Australia and North America.
Ecology 72: 479-495.

354
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Response of two diurnal raptors, the Common Buzzard (Buteo buteo) and the
Eurasian Kestrel (Falco tinnunculus), to agricultural intensity in three landscape
units of Western France

N. Michel1, V. Comor2, Y. Rantier2, Y. Delettre2, F. Burel2, A. Butet2

1
Lab Ecology, Systematics and Evolution. Orsay University. CNRS. Bat. 360. Orsay France
e-mail: [email protected]
2
Centre Armoricain de Recherche en Environnement, UMR CNRS 6553 « Ecobio »,
Université de Rennes 1, 35042 Rennes France

Introduction

In front of agricultural intensification, there has been a wide decline in biodiversity.

Several floral and faunal groups have been concerned by a reduction of abundance in
farming landscapes. In particular, the period of intensification of farm management coincides
with the decline of many farmland bird species (Fuller et al., 1995). In this study, we focused
on two diurnal raptor species, the Common buzzard (Buteo buteo) and the Eurasian kestrel
(Falco tinnunculus). The Common Buzzard (Buteo buteo) seem to be recovering in Europe
from low populations due to past persecution and pesticide effects, while the Eurasian
Kestrel (Falco tinnunculus) do not show similar positive trends (Hagemeijer and Blair, 1997).
The populations of those two species in intensively-cultivated areas may be regarded as
important ecological indicators and the monitoring of their trends as an important
conservation task (Boano and Toffoli, 2002).

Methods

We carried out a one year survey of the abundance of these two species in several
observation points in three agricultural landscapes of Western France (BOC1, BOC2 and
POL) differing by their level of agricultural intensification The three sites are known to be on a
gradient of land-use intensity (BOC1<BOC2<POL) and hedgerow network density
(BOC1>BOC2>POL). We also made an estimation of the prey availability by trapping small
mammals in hedgerows of the same sites during one year (7 trapping sessions).

Results & Discussion

The local abundance and biomass of small mammals differed between the three
sites: they were significantly more numerous in the hedges of the most intensified site than in
hedges of the two others. On the contrary, at the landscape scale, small mammal biomass
was the lowest in the most intensified site because of the fragmentation of the hedgerow
network, then in the medium site and it was the highest in the more preserved site
(BOC1>BOC2>POL)(Figure 1).
It appeared that the two raptor species responded differently to agricultural
intensification (Figure 1 & 2): the abundance of the Common Buzzard significantly decreased
with the reduction of semi-natural elements such as hedgerows, woodlots and grasslands, as
well as with the decrease of small mammal availability at the landscape scale (Suetens
1989). The abundance of the Eurasian kestrel showed the same tendency, but the
correlation was not significant. We also showed that the Common Buzzard was very
dependant to woody habitats, whereas the Eurasian kestrel was much more associated with
grasslands (Suetens 1989) (figure 1).

355
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

A
B

availability (g/ha)
Mean raptor abundance

20

Small mammal
900,00
800,00
15 BOC1 700,00
600,00
10 BOC2
500,00
POL 400,00
5 300,00
200,00
100,00
0
0,00
Buzzard Falcon Total BOC1 BOC2 POL

Figure 1. (A) Mean abundances of buzzards and falcons (± SD) (separated and combined),
and (B) small mammal availability in hedgerows (g/ha) in the three landscape units.

1
-1 1
Hedges -1

Crop
Buzzard
Wood

Build

Grass

Kestrel

Figure 2. Projection of raptor species and landscape variables on the co-inertia plane (co-
inertia analysis of landscape and faunal data) (F1= 98%; F2= 2%).

References
Fuller, R.J., Gregory, R.D., Gibbons, D.W., Marchant, J.H.W., Baillie, S.R. & Carter, N., (1995).
Population declines and range contractions among lowland farmland birds in Britain. Conservation
Biology, 9: 1425-1441.
Hagemeijer, W.J.M. & Blair, M.J., (1997). The EBCC atlas of European breeding birds. T & A.D.
Poyser Ed., London, U.K.
Boano, G. & Toffoli, R., (2002). A line transect survey of wintering raptors in the western PO Plain of
northern Italy. Journal of Raptor Research 36: 128-135.
Suetens, W., (1989). Les Rapaces d'Europe. Perron Editions

356
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Effects of biogeographical factors on bird species sensibility to habitat

fragmentation in the Atlantic Rainforest

R.G. Pimentel, E. Hasui, M.C. Ribeiro, J.P. Metzger

Landscape Ecology and Conservation Laboratory – LEPAC, Department of Ecology, Institute

of Biosciences, University of São Paulo, Rua do Matão - travessa 14 no 321 CEP 05508-900,
São Paulo, Brazil.
e-mail: [email protected]

Introduction

Habitat loss and fragmentation are the main factors acting on species decline and local
extinctions. Nowadays, there is a vast literature concerning this issue, what allows some
factors that exert influence on organisms during the fragmentation process to be identified
(Laurance et al., 2002; Henle et al., 2005). Size and connectivity of the remaining patches
are two factors that have been already discussed on their influence on organisms. On the
other hand, there are some other factors that are still poorly known. Among these factors are
the influences of biogeographical effects on species vulnerability to fragmentation. Some
authors suggest that populations situated in the periphery of the total distribution of species
are more sensitive than those situated in the centre (Gaston, 1990; Kattan et al., 1994).
However, there is a lack of empirical evidences that support this model.

Objective

The present study aims to verify which factors best explain the sensibility of populations in
a fragmented landscape, considering the habitat availability, connectivity of the patches and
the localization in relation to total geographic distribution

Materials and Methods

Study area

The study area comprises a vast territorial extension of ca. 1.2 million of km2 in the
southeastern part of Brazil (14o17´-29o25´S; 39o40´-54o37´W). The predominant biome of the
study area is the endangered Atlantic Rainforest. The selection of this area is based on a
center of bird endemism, proposed by Silva et al. (2004). The vastness of this study area
tends to comprise a high biogeographical variation.

Biological data

Data of species occurrence and abundance have been compiled in a database. These
data have been obtained from the literature (i.e. papers and thesis) and museum records.
This database comprises ~26,000 points of occurrence for 160 bird species.

Species distribution model

As part of this project will be generated potential distribution models for the bird species.
For this purpose will be utilized only species occurrence data from the database. These
models will be made for species which present more than 30 records. The potential
distributions of birds will be modeled utilizing the GARP software. GARP utilizes genetic
algorithm to select a set of association rules between species occurrence and environmental
variables, and thus predict the species distribution (Stockwell and Peters, 1999). Thematic
maps of the whole study area in a scale of resolution of 1:250,000 will be used as

357
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

environmental variables. The probability of occurrence, obtained from the modeling

procedure will be the biogeographical variable utilized in the posterior analyses.

Analysis of the effects of area, connectivity and biogeographical factor

To analyze the influence of biogeographical factors, area and connectivity on species

abundance, multiple regression models will be built, having species abundance as
dependent variable and as independent variables the area and connectivity of the patches
and biogeographical factor. Species abundance data will be taken from the database. For the
purpose of compare and estimate the best models generated by the multiple regressions, we
will utilized the Akaike´s Information Criteria (AIC).
We will also conduct a landscape-level analysis. Landscape will be defined as an area
with ca. 10,000 ha, containing at least 10 sampling points. In this landscape-level analysis a
vulnerability index will be created as a function of the abundance of species in the sampled
patches over the area of these patches. Then, this vulnerability index will be included as
dependent variable in a multiple regression that will have the biogeographical factor,
percentage of habitat and an aggregation index of remaining patches in the landscape as
independent variable. AIC will also be used to estimate the best models.

Final Considerations

With this research we pretend to improve the comprehension of the factors that lead to a
differential sensibility of species after the habitat fragmentation process. Besides,
understanding differential sensibility of species mediated by biogeographical factors may
help the selection of priorities areas for conservation and restoration in one of the most
threatened biomes in the planet, the Atlantic Rainforest.

References
Gaston, K.J. (1990) Patterns in the geographical ranges of species. Biological Reviews 65: 105-129.
Henle, K.; Davies, K.F.; Kleyer, M.; Margules, C. & Settele, J. (2004) Predictors of species
sensitivity to fragmentation. Biodiversity and Conservation 13: 207-251.
Kattan, G.H.; Alvarez-López, H. & Giraldo, M. (1994) Forest fragmentation and bird extinctions: San
Antonio eighty years later. Conservation Biology 8(1): 138-146.
Laurance, W.F.; Lovejoy, T.E.; Vasconcelos, H.L.; Bruna, H.E.M.; Didham, R.K.; Stouffer, P.C.;
Gascon, C.; Bierregard, R.O.; Laurance, S.G. & Sampaio E. (2002) Ecosystem decay of
Amazonian forest fragments: a 22-year investigation. Conservation Biology 16(3): 605-618.
Silva, J.M.C.; Sousa, M.C. & Castelletti C.H.M. (2004) Areas of endemism for passerine birds in the
Atlantic Forest, South America. Global Ecology and Biogeography 13: 85-92.
Stockwell D. & Peters D. (1999) The GARP modeling system: problems and solution to automated
spatial prediction.

358
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Effects of forest landscape composition and structure on forest bird species

richness in the Mediterranean at two different spatial scales

A. Gil-Tena, O. Torras, S. Saura

Departament d’Enginyeria Agroforestal. ETSEA. Universitat de Lleida. Av. Alcalde Rovira

Roure, 191. 25198. Lleida. Spain.
e-mail: [email protected]

Introduction

In the Mediterranean region there is a lack of knowledge on how forest landscape

characteristics are related to different components of biodiversity. The required research in
this topic should be tackled at different landscape scales, since ecological processes do not
occur at a single scale, and forest management affects biodiversity in a wide range of scales,
not only at the stand level. In this study we analyzed the influence of forest landscape
characteristics on forest bird species richness at two different spatial scales and determined
the most relevant factors for explaining forest bird species distribution.

Material and methods

The study area was Catalonia (a region of about 32,000 km2 in NE Spain). Bird species
richness was obtained at two spatial scales (1 km2 and 100 km2) from census carried out by
volunteers within the Catalan Breeding Bird Atlas (1999-2002) and the Spanish Breeding
Bird Atlas, and the forest landscape characteristics were extracted from the recent Spanish
Forest Map (scale 1:50,000), developed within the Third Spanish National Forest Inventory.
We considered a sample of 2,923 1 x 1 km UTM cells scattered throughout Catalonia and
the whole lattice of 283 10x10 km UTM cells. The forest landscape variables were related to
forest composition and structure (forest area, development stage, canopy cover, forest tree
species diversity, etc.) and to forest configuration (fragmentation and shape indices).
Analyses were performed separately for specialist, generalist and total species richness at
the two scales of analysis.

Results and discussion

Too high forest canopy cover (above a threshold of 70 %) did not favour (p<0.01) species
richness (see Figure 1), since too closed canopies may not allow the development of shrub
strata which provide relevant forage and nest sites. Landscape characteristics explained
more variation at broader scales (see Table 1) and this could be due to the greater data
variability at 1 km2 or to a better match with the home range of some forest bird species at
the scale of 100 km2. Agreeing with Mitchell et al. (2001), we found that specialists were
more associated than generalists to forest landscape characteristics since the former use or
appear in a greater extent in forest landscapes. As in previous researches forest area was
the variable with a higher explanatory power (McGarigal and McComb, 1995; Trzcinski et al.,
1999; Radford et al., 2005). Moreover, forest configuration had a minor effect on bird species
richness compared to other composition variables like tree species diversity, which favoured
bird species at both spatial scales. The relevance of some variables at one scale but not at
the other (such as the development stage and the percentage of coniferous species)
highlighted the need to consider different spatial scales for an adequate and integrated
analysis and forest landscape management in the Mediterranean.

Table 1. Stepwise multiple regression analysis for total forest bird species richness and
forest composition and structure variables. β: standardized coefficient; p:
significance.

359
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

1 km2 Partial R2 β p
Forest area 0.399 0.435 <0.0005
Forest development stage 0.064 0.240 <0.0005
Tree species diversity 0.024 0.156 <0.0005
Canopy closure diversity 0.001 0.043 0.005
Coniferous species percentage 0.001 -0.042 0.013

10 km2 Partial R2 β p
Forest area 0.444 0.529 < 0.0005
Tree species diversity 0.103 0.473 < 0.0005
Coniferous species percentage 0.072 0.298 < 0.0005

Figure 1. Correlations between forest bird species richness and forest area (m2), defined
as the area of land with a forest canopy cover (FCC) above a certain FCC
threshold for the 1 x 1 km and the 10 x 10 km UTM cells.

References
Mitchell, M.S; Lancia, R.A. & Gerwin, J.A. (2001) Using landscape-level data to predict the
distribution of birds on a managed forest: Effects of scale. Ecological Applications 11: 1692-1708.
McGarigal, K. & McComb, W.C. (1995) Relationships between landscape structure and breeding
birds in the Oregon Coast Range. Ecological Monographs 65: 235-260.
Radford, J.Q; Bennett, A.F. & Cheers, G.J. (2005) Landscape-level thresholds of habitat cover for
woodland-dependent birds. Biological Conservation 124: 317-337.
Trzcinski, M.K; Fahrig, L. & Merriam, G. (1999) Independent effects of forest cover and
fragmentation on the distribution of forest breeding birds. Ecological Applications 9: 586-593.

360
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Relative and independent effects of habitat amount, fragmentation, matrix quality,

and road density on forest bird diversity

A.C. Smith1, L. Fahrig 1, C. Francis 2

1
Geomatics and Landscape Ecology Research Laboratory, Carleton University,
Ottawa, Ontario, Canada, K1S 5B6.
e-mail: [email protected]
2
Environment Canada, National Wildlife Research Centre, Carleton University
Ottawa, Ontario, Canada, K1A 0H3.

Introduction

The limited resources available for biodiversity conservation must be allocated towards
conservation measures that will have the greatest impact. Habitat loss has a consistently
strong, negative effect on biodiversity but in contrast, the effects of habitat fragmentation are
about as likely to be positive as they are negative (Fahrig, 2003). In addition, multiple
ecological processes responsible for configuration effects may have different effects on
individual species within disparate ecological neighbourhoods and therefore fragmentation
effects may be particularly scale dependent (Krawchuck and Taylor, 2003).

Methods

In order to determine what landscape factors most strongly influence forest bird diversity,
we assessed the relative strength of the effects of amount of forest, habitat fragmentation,
matrix composition, and roads on the species richness of forest birds in North American Bird
Conservation Region 13 (Southern Ontario, Canada). To investigate the scale dependency
of these relationships we compared the relative effects of the predictor variables measured
within landscapes of different sizes (1-20 km radii) after controlling for local effects. We used
bird data collected during the recent, Ontario Breeding Bird Atlas project and province-wide
landcover data.
We used hierarchical variance partitioning to estimate the total independent influence of
the five correlated predictor variables (Chevan and Sutherland, 1991) on the species
richness of all forest dependent bird species and on a subset of species that in previous
studies have shown significantly negative effects of fragmentation.

Results and Implications

Landscape Effects at 20 km

Landscape context explained an increasing proportion of the variation in forest bird

species richness up to distances of 20km (Figure 1). To our knowledge, no previous study
has shown effects of landscape context for any species at such a broad scale. These results
suggest that forest birds in this region are sensitive to landscape features at distances much
greater than are generally considered. It also highlights the fact that comparisons of the
relative effects of local and landscape variables are scale dependent and may be highly
biased.

Habitat Amount Stronger than Fragmentation or Matrix

Within landscapes of less than 30% forest cover, the amount of forest in the landscape
explained significantly more variation in species richness of fragmentation sensitive birds
than did either fragmentation measure or the amount of hostile matrix in the landscape at a
20km scale. Surprisingly, this group of presumably fragmentation sensitive species were

361
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

much more strongly affected by the amount of habitat in the surrounding landscape than by
either measure of fragmentation. The amount of habitat also had a consistently positive
influence on species richness at all scales (85 – 99% of partial regression coefficients were
positive). In addition, this subgroup did not respond consistently to either fragmentation
metric with the exception of forest edge measured in relatively small landscapes (25 – 50%
of partial regression coefficients were positive).
From these results we conclude that there should be one major conservation priority for
forest birds in this region: the protection and/or restoration of forests. Managing the
configuration of forest habitat to conserve forest bird diversity is likely a waste of precious
resources.

Figure 1. Average percent of explained variation within poisson regression models (n =

500 bootstrapped samples) that can be independently allocated to one of five correlated
landscape predictor variables using hierarchical variance partitioning, within landscapes of
increasing radius. Error bars indicate 90% confidence limits and printed percent values
within the plot indicated the proportion of total variation in the response explained by the
model containing all five predictors.

References
Chevan, A; & Sutherland, M. (1991) Hierarchical partitioning. The American Statistician 45: 90-96.
Fahrig, L. (2003) Effects of habitat fragmentation on biodiversity. Annual Review of Ecology, Evolution
and Systematics 159: 40-56.
Krawchuck, M; & Taylor, P. (2003) Changing importance of habitat structure across multiple spatial
scales for three species of insects. Oikos 103: 153-161.

362
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

Birds respond to large landscapes more than to “islands”

M.A. Cunningham1,2, D.H. Johnson2

1
Vassar College, 124 Raymond Ave., Poughkeepsie NY 12604, USA.
e-mail: [email protected]
2
Northern Prairie Wildlife Research Center, United States Geological Survey, Biological
Resources Division, Jamestown, ND, 58401 USA.

Introduction

The “island-matrix” model has long been used to describe fragmented habitat areas. This
model, or metaphor, has been useful but also has inherent flaws, including the assumption
that researchers can accurately characterize habitat patches and boundaries, that species
use only one habitat type, and that matrix is a dangerous, or at best neutral, component of a
landscape. Recently, landscape ecologists have sought alternatives to the island-matrix
framework in modeling fragmented landscapes (Villard 1998, Haila 2002, Kupfer et al. 2006).
As an alternative, we find that a useful, parsimonious method of characterizing landscapes is
to calculate percentage of habitat within buffers around sample points. While this method still
assumes that the researcher can meaningfully differentiate habitat types, if avoids critical
assumptions that multiple patches are functionally separate and that a single patch is
internally similar. This approach is simultaneously sensitive to the extensiveness of individual
patches and to a patch’s surrounding environment. It can easily be calculated at a range of
scales, and for most species in our study area, habitat responses changed little as scales of
analysis increased.

Habitat responses in an oak savanna landscape

We examined responses of woodland and grassland birds to percentage tree cover on

>3,000 transect segments in a mixed grassland-woodland landscape in eastern North
Dakota, USA (Cunningham and Johnson 2006). We evaluated probability of occurrence in
response to percentage tree cover in the landscape. A majority of species responded
similarly at a range of scales (figure 1). Even when proximate landscape conditions were
held constant, most of these species selected disproportionately wooded or open conditions
in large landscapes. Thus large landscapes had influence independent of proximate
landscape conditions. At the same time, there was interaction between scales. Many
woodland birds, for example, selected for more wooded proximate conditions when in open
landscapes than when in wooded landscapes. In contrast, species had little response to
patch size (figure 2).
In addition to the advantages noted above for this approach, our method allows
consideration of species that rely on more than one habitat type. In a study of 40 woodland
species occupying an oak savanna habitat, 17 species preferentially occupied mixed habitat
and were most abundant at 30-50 percent tree cover. Many species also occupied more than
one habitat type, although there was a spectrum of preferences from strong interior to strong
edge-association.

Conclusions

These results suggest that percentage cover is a useful alternative to the “island-matrix”
approach in modeling fragmented landscapes, especially where (1) the “matrix” is not
absolutely unsuitable, (2) species move readily among fragments, or (3) where species use
multiple habitat types, either opportunistically or as obligates.
For presence/absence data, our observations indicate that landscape composition
provides a parsimonious, easily reproduced approach. Patch size, on the other hand,

363
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 11: Landscape modeling and bird populations

predicts occurrence much more weakly, a finding consistent with ambiguous results in
previous studies.

Figure 1. An edge species’ response to increasing percentage tree cover, at three scales
(200, 400, and 1200 m radius around sampled transects). Of 40 species, 28 had similar
responses at increasing scales, though thresholds, peaks, asymptotes, and slopes might
shift.

Figure 2. Strength of percentage tree cover and patch size in predicting presence/absence
for interior woodland species.

References
Cunningham, M.A. & Johnson, D.H. (2006) Proximate and landscape factors influence grassland
bird distributions. Ecological Applications 16: 1062-1075.
Haila, Y. (2002) A conceptual genealogy of fragmentation research: From island biogeography to
landscape ecology. Ecological Applications 12: 321-334.
Kupfer, J.A.; Malanson, G.P.; & Franklin, S.B. (2006) Not seeing the ocean for the islands: The
moderating influence of matrix-based processes on forest fragmentation effects. Global Ecology
and Biogeography 15: 8-20.
Villard, M.A. (1998) On forest-interior species, edge avoidance, area sensitivity, and dogmas in avian
conservation. Auk 115: 801-805.

364
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect
populations

Landscape change and extinction debt: measuring long-term change in the

mammal fauna of forest fragments

A.F. Bennett1, A. Haslem1 & R. Mac Nally2

1
Landscape Ecology Research Group,
School of Life and Environmental Sciences, Deakin University,
Burwood, Victoria 3125, Australia.
e-mail: [email protected]
2
Australian Centre for Biodiversity: Analysis, Policy and Management
School of Biological Sciences, PO Box 18, Monash University, Victoria 3800, Australia

Introduction

The consequences of landscape change are not always immediately apparent. In rural
environments, patches of native vegetation have a critical role in the conservation of
biodiversity but a major gap in understanding is the way in which faunal assemblages in such
patches change through time. Theory suggests that following isolation, remnant patches will
experience a gradual loss of species through time, but that there will be a time-lag before the
full consequences of isolation and landscape change are realized (i.e. there is an extinction
debt, sensu Tilman et al. 1994). The potential for an extinction debt in rural landscapes has
profound implications for nature conservation; it implies that species currently present will not
necessarily persist in the longer term.

In 1976–1980, mammal surveys were undertaken in two large sets of forest patches in
study areas in Gippsland (Suckling 1980, 1982) and Western Victoria (Bennett 1987, 1990),
both in temperate south-eastern Australia. In this study, we re-surveyed the terrestrial
mammals in these patches to directly measure potential change in the fauna over a 20+ year
period, and to test predictions concerning the types of species most vulnerable to decline.

Methods and Results

Field surveys were carried out in 36 forest patches in W Victoria (0.3 to 85 ha in size) and
in 32 patches in Gippsland (0.5 to 57 ha). Survey techniques and survey effort were closely
matched to that documented for the historic surveys, and involved trapping, spotlight
transects, hair-sampling tubes, direct observation, and records of tracks and signs.
Substantial change has occurred to forest vegetation in both regions. In W Vic, the overall
extent of tree cover has not changed but a wildfire in 1983 burned 87% of the area. In
Gippsland, establishment of plantation forestry is a major change in land use. In both study
areas, a number of the original forest patches have been cleared or further fragmented.

Species richness of native mammals in forest patches showed little change over the 20-
year period, but in both study areas there were regional changes in the overall status of the
mammal fauna. A marked increase in the frequency of occurrence of the Common Wombat
Vombatus ursinus in Gippsland; and the Common Brushtail Possum Trichosurus vulpecula
and Koala Phascolarctos cinereus in W Victoria, masked a regional decline in occurrence of
several rarer species (e.g. Southern Brown Bandicoot Isoodon obesulus, Long-nosed
Potoroo Potorous tridactylus).

Predictions were made (before commencing the study) of the vulnerability of each species
to long-term decline, based on relative abundance, capacity to move through the rural

365
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

landscape and sensitivity to disturbance (feral predators, grazing by stock). For each study
area, there was a significant correlation between predicted vulnerability and actual change in
the status of native mammal species. The most vulnerable species were those present
historically in low abundance, dependent on native forest vegetation, and sensitive to
predation and loss of understorey cover. Turnover in the identity of species in forest patches
over the 20-year period was greatest in small (<5 ha) patches.

Discussion and Conclusions

Understanding the long-term dynamics of faunal populations in rural environments is

complex because landscape structure and patterns of human land-use are not static. In
addition to faunal changes that may be due to time-lag effects from biogeographic change
(e.g. overall forest extent, patch size, isolation), the effect of ongoing patterns of land use
and environmental disturbance must also be recognised. We highlight four conclusions from
this study concerning the conservation of native fauna in rural landscapes.

1. Networks of natural vegetation have a key role in the long-term persistence of wildlife in
farmland environments. Here, small forest patches (most < 20 ha) have maintained
populations of at least 10 species of native mammal over a 20-year period.

2. The status of native fauna in remnant habitats in rural environments is dynamic. In this
study, several species experienced a long-term increase in frequency of occurrence in forest
patches, some showed little change, while others experienced decline and possible local
extinction.

3. Patch-level and landscape-level assessments of the fauna provide complementary

insights. In both study areas there was little change over the 20-year period in species
richness at the patch level, but at the landscape level long-term change in the fauna was
evident. The overall status of the mammal fauna in each area has deteriorated: at least four
species in W Victoria and six species in Gippsland have declined or are now scarce, such
that their future persistence for the next 20 years is not assured.

4. The differential vulnerability of species to long-term change is influenced by their

ecological characteristics. Here, native mammals most vulnerable to decline were those that
initially occurred in low abundance, are dependent on forest understorey vegetation, and are
sensitive to predation by feral predators (i.e. Red Fox Vulpes vulpes).

References
Bennett, A.F. (1987) Conservation of mammals within a fragmented forest environment: the
contributions of insular biogeography and autecology. D.A. Saunders, G.W. Arnold, A.A.
Burbidge & A.J.M. Hopkins (Eds.). Nature Conservation: The Role of Remnants of Native
Vegetation, Surrey Beatty & Sons, Chipping Norton, pp. 41-52.
Bennett, A.F. (1990) Land use, forest fragmentation and the mammalian fauna at Naringal, south-
western Victoria. Australian Wildlife Research 17: 325-347.
Suckling, G. C. (1980) The effects of fragmentation and disturbance of forest on mammals in a region
of Gippsland, Victoria. PhD thesis, Monash University, Victoria.
Suckling, G. C. (1982) Value of preserved habitat for mammal conservation in plantations. Australian
Forestry 45: 19-27.
Tilman, D., May, R.M., Lehman, C.L. & Nowak, M.A. (1994) Habitat destruction and the extinction
debt. Nature 371: 65-66.

366
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

Pond or landscape characteristics – which is more important for common toads

(Bufo bufo)? A case study from central Romania

T. Hartel1, S. Nemes2, L. Demeter3, K. Öllerer1

1
Institute of Biology – Romanian Academy, Splaiul Independenţei 296, Bucharest, Romania.
e-mail: [email protected]
2
School of Mathematical Sciences, University of Gothenburg, Göteborg 41296, Sweden
3
Sapientia Hungarian University of Transylvania, Department of Technical and Nature
Sciences, P-ţa. Libertăţii 1, 530104 M-Ciuc, Romania.

Introduction

The primary anthropogenic factor causing amphibian decline in Europe is habitat loss and
fragmentation (Stuart et al., 2004). In order to implement efficient conservation measures on
a regional level, baseline data are needed regarding local distributions and the factors
influencing habitat use of different amphibian species. Here we explore the effects of aquatic
and terrestrial habitat variables on the abundance (adult counts) of the common toad (Bufo
bufo) in 43 permanent ponds in central Romania surveyed between the years 2000 and
2005.

Materials and Methods

Study area

The Târnava Mare Valley is located in central Romania. The area selected for this study
covers approximately 2600 km² and is situated in the middle section of the valley. Important
land use types in the area include: deciduous woodland (33%), shrubland (5%), pastures and
grassland (41%), orchard (2%), vineyard (1%), marsh (1%) and urban area (1%). Two large
roads and one railway run through the valley. An other highway is planned to be constructed
through this valley in the near future.

Data collection

The surveys were made between the years 2000-2005. Ponds were located using 1:25
000 scale topographic maps. Adult toad counts were gathered for 43 populations (here we
considered adults breeding in one pond as being a population) in the afternoon and at night
from the end of March till the second half of April.
We measured four aquatic habitat variables and seven landscape related variables at
each sampled site. The aquatic variables were: area (m²), percentage of emergent aquatic
vegetation cover (Phragmites sp., and Typha sp.), percentage of shallow water (< 50 cm
depth) and presence / absence of non-predatory and predatory fish. Landscape variables
included: distance of pond from forest (m), the percentage of forest cover, the
presence/absence of green connecting corridors, pastures and grassland cover (%), arable
land cover (%), and the presence/absence of high traffic roads.

Data analysis

The predictor variables are continuous and discrete binary variables. The nonlinear
iterative NIPALS–PCA (Nonlinear Iterative Partial Least Squares-Principal Component
Analysis) algorithm was used to explore the relationship between the environmental factors
and toad counts. Statistical analyses were performed with Matlab 7.01.

Results

367
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

Common toads were present in 81.4% of the surveyed sites. Out of the 43 ponds
sampled, 18.6% lacked fish, 30.2% had only non-predatory fish, while 51.1% contained both
predatory and non-predatory fish. 74.4% of the ponds are connected to forest through a
green corridor, 32% are next to pastures and 37.2% are close to roads with heavy traffic.

Table 2. Descriptive statistics for eight pond and landscape related variables. (SD = standard
deviation; CV = coefficient of variation)

Variable Mean SD CV
Area (m²) 99,2 319,3 3.22
Shallow water (%) 34.5 27.7 0.80
Emergent aquatic vegetation cover (%) 32.2 24.4 0.76
Elevation (m) 395.4 72.1 0.18
Grass/pasture cover (%) 32 20.0 0.71
Forest cover (%) 34 22.0 0.73
Distance to forest (m) 298.6 339.1 1.14
Arable land cover (%) 15.8 20.2 1.27

Toad counts can be related to three landscape variables, namely the percentage of forest
cover (positive association), presence/absence of roads (negative association between the
high volume traffic roads and toad counts) and habitat corridors (positive relationship). These
variables together with the presence/absence of fish accounted for 49% of the total variance.

Conclusions

The common toad counts in this area are significantly associated only with landscape
characteristics. Our study highlights the role of landscape composition and configuration in
maintaining common toad populations in this area, and confirms the negative effect of
landscape fragmentation. To efficiently protect amphibians in Romania, appropriate
legislation and a strong collaboration between landowners, landscape planners and
herpetologists is needed.

References
Stuart, S.N; Chanson, I.S., Cox, N.A., Young, B.E., Rodrigues, A.S.L., Fishman, D.L., Waller,
L.W. (2004) Status and trends of amphibian declines worldwide. Science 306: 1783 - 1786
Hartel, T. et al (In press 2007). Article in Applied Herpetology, 2007 Vol. 4(1).

368
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

Changes in Female Grizzly Bear (Ursus Arctos) Homerange Configuration in the

Multi-use Albertan Rocky Mountain Foothills, Canada.

J. Linke1, A. Pape2, J. Cranston3, G. McDermid1, M. Hall-Beyer1, S.E. Franklin2, G.

B. Stenhouse3,4.
1
Department of Geography, University of Calgary, 2500 University Drive N.W. Calgary, AB
T2N 1N4, Canada.
e-mail: [email protected];
2
Department of Geography, University of Saskatchewan, Saskatoon, SK, S7N 4P3 Canada;
3
Foothills Model Forest, Box 6330, Hinton, A.B T7V 1X6, Canada, 4Alberta Sustainable
Resource Department, Fish and Wildlife Division, Edmonton, A.B. Canada.

Introduction
Grizzly bear populations in Alberta, Canada are under tremendous pressure from land
use changes related to development of the province’s vast forest, agriculture, and petroleum
resources. For example, Alberta contains roughly 70% of Canada’s oil and gas reserves,
and is forecasting annual production increases of 5% over the next decade (CAPP 2005).
Understanding the impacts of these changes, and providing resource managers with the
knowledge and planning tools necessary to ensure the long-term conservation of Alberta’s
grizzly bears has been the goal of The Foothills Model Forest Grizzly Bear Research
Program (FMFBGRP) since 1999. In this paper, we report on a research project designed to
quantify the impacts of human-induced landscape change on grizzly bear home ranges using
parsimonious metrics of landscape structure (Linke and Franklin 2006), as observed through
annual satellite imagery covering the years 1999 through 2003. The work is part of a
broader initiative designed to understand the impacts of resource development on grizzly
bear health and habitat selection.

Methods
A series of six annual Landsat 5 TM and Landsat 7 ETM+ images covering the time frame
1998-2003 were orthorectified and radiometrically normalized over a 7450 km2 study area
southeast of Hinton, Alberta, Canada. The enhanced wetness differencing index (EWDI) of
Franklin et al. (2001) was used to produce 5 layers of unlabelled annual change. We then
used an object-oriented image processing package (Definiens Professional 5.0), to segment
the change pixels into identifiable objects, and applied a series of logical decision rules to
classify change objects into landuse/disturbance categories, including wellsites, cutblocks,
natural burns, and mine sites. Additional disturbance layers that were not captured by the
automated change detection process, such as roads and small disturbance features, were
digitized manually with the help of supplemental imagery such as IRS, SPOT and aerial
orthophotos. Once assembled, we used these layers of labelled annual change to update a
10-class landcover map (circa 2003) of the study area produced by McDermid et al. (2007).
A second set of decision rules were used to reclassify the change features to the appropriate
landcover class, and backdate the 10-class map to each of the years 1998 through 2003.
We used the updated annual landcover maps to calculate four landscape-level metrics with
Fragstats 3.3 in six female minimum convex polygon (MCP) home ranges with low (G004,
G016), medium (G011, G027), and high (G020, G023) exposure to human use (Linke et al.
2005).

Results/Conclusions
The overall study area experienced land use changes every year, with forestry, road
construction, and well sites contributing the most dominant disturbance features (Table 1).
However, changes varied spatially across the study area, affecting the landcover structure of
the six home ranges to varying degrees, with two MCPs (G004, G016) remaining
unchanged. Annual changes in metrics were observed for the remaining four MCPs, with the

369
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

overall largest magnitudes being recorded in home ranges with originally medium exposure
to human use (Figure 1). In conclusion, landscape metrics derived from annual maps of
landcover from remote sensing are shown to be effective tools for capturing and
summarizing changes in landscape structure caused by human development. The
techniques presented here have been used to quantify a broad spectrum of human-induced
changes to grizzly bear home ranges on multi-use lands in Alberta, Canada. The work
establishes a strong foundation for on-going monitoring activities, and further investigations
into wildlife habitat inferences frequently drawn from multi-temporal wildlife data sets.

Table 1. Annual extents of land uses/disturbances as derived from Landsat image change
detection in the 7450 km2 grizzly bear foothills study area between 1999 and 2003.
Disturbance Feature 1999 2000 2001 2002 2003
2
Forestry Cutblocks (km ) 32.5 34.0 34.1 42.7 44.1
Mining (km2) 2.1 0.0 2.2 2.2 0.0
Burns/Forest Fires (km2) 0.4 0.0 0.0 0.0 0.2
Wellsites (#, km2) 117,1.9 113, 2.1 96, 1.6 65, 1.3 68, 1.1
Roads (km) 94.4 152.5 136.2 87.8 82.3

Figure 1. Percent annual cumulative changes in four female grizzly bear homeranges
with medium and high exposure to human use between 1999 and 2003 as measured in
relation to landcover structure in 1998 with A) mean shape index, B) mean patch size, C)
largest patch index, and D) contrast weighted edge density.

References
Canadian Association of Petroleum Producers (CAPP). 2007. Canadian crude oil production and
supply forecast 2005-2015. Calgary, Alberta, Canada. Retrieved on Jan. 29, 2007, from
http://www.capp.ca/default.asp?V_DOC_ID=6
Franklin, S.E., Lavigne, M.B., Moskal, L.M., Wulder, M.A. and T.M. McCaffery. 2001. Interpretation
of Forest Harvest Conditions in New Brunswick Using Landsat TM Enhanced Wetness Difference
Imagery (EWDI). Canadian Journal of Remote Sensing, 27, 118-128.
Linke, J. and S.E. Franklin. 2006. Interpretation of landscape structure gradients based on satellite
image classification of land cover. Can. Journal of Remote Sensing 32(6): in press
Linke, J., S.E. Franklin, F. Huettmann and G.B. Stenhouse. 2005. Seismic cutlines, changing
landscape metrics and grizzly bear landscape use in Alberta. Landscape Ecology 20: 811-826.

370
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

Space use in desert areas by guanaco (Lama guanicoe) and its seasonal
dependence on the most productive patches.

P. Acebes1, J.E. Malo1, F. Suarez1, C.E. Borghi2, S.M. Giannoni2 & J. Traba1
1
Departamento de Ecología. Universidad Autónoma de Madrid. E-28049 Madrid. Spain.
e-mail: [email protected]
2
Museo de Ciencias Naturales. Universidad Nacional de San Juan. San Juan. Argentina.

Local habitat selection and seasonal movements are key elements in life strategies of
many animal species as means to cover their vital needs. This is specially true for
populations living in areas close to species' distribution extremes as their proximity to the
realizable niche limit makes survival more dependent on an efficient use of available
resources. In this context, we describe seasonal variability in space and habitat use by a
small population of guanaco (Lama guanicoe Müller 1776) living at the dryest limit of its
Argentinean distribution.
During the wet and dry seasons of 2005 and 2006 we GPS-located all guanaco
observations obtained during systematic surveys through the Ischigualasto-Talampaya World
Heritage Site (Argentina), located at the phytogeographic region of the “Monte árido”,
dominated by shrubby plants such as Larrea spp, Zuccagnia punctata, Atriplex spp. and
Prosopis spp. in addition to several cacti species (Trichocereus, Cereus, Tephrocactus,
Opuntia spp). Temperatures vary from -10º to 50ºC and annual precipitation is about 100
mm. Guanaco observations were obtained form daily surveys of the area totalling
approximately 150 hours by season. Variables of the environment and the group of animals
were also collected for each observation. We defined 6 different plant communities to
characterize the habitat where guanacos were found: dense scrubland (dominated by shrubs
such as Zuccagnia punctata, Larrea cuneifolia and Geoffrea decorticans with >20% plant
cover), open scrubland (with shrubs like Larrea spp. and Plectocarpa tetracantha, always
with <20% plant cover, and sometimes lower than 5%), treelike cacti-dominated slopes (hills
covered by mixed shrubland with Trichocereus terscheckii), algarrobal (Prosopis spp
associated to dry rivers), zampal (sandy areas with a low cover of Atriplex spp.), and peladal
(barren areas).
In all we carried out 209 observations of guanaco herds with a total number of 1343
individuals (mean ± standard error; 6.43 ± 0.29 animals by observation). Mean group size
varied significantly between years (ANOVA test: F = 7.05; df = 1; p=0.008), but not between
seasons (ANOVA test, F = 8.72; df = 1; p=0.706). The estimation of the whole area occupied
by herds (Minimum Convex Polygons) showed a high seasonal overlap between wet and dry
seasons (average for two years; Cole Concordance Index >66%). Nevertheless, the analysis
of the core areas (estimated using a density estimator kernel at 50%) showed a much lower
seasonal overlap (average for two years; CCI <36%).
Guanacos habitat selection was significantly different than expected from availability (chi-
squared= 71.41; df = 15; p<0.0001). Ivlev habitat selection index showed a somewhat
relaxed selection pattern during wet seasons and a more neat one during dry seasons.
Accordingly to this, guanacos showed preference for zampal and algarrobal during 2005 and
2006 wet seasons and avoidance of treelike cacti-dominated slopes and dense shrublands in
2005 and of zampal in 2006. However, in both dry seasons guanacos avoided peladal, and
zampal in 2005 and algarrobal in 2006. No positive selection was recorded for dry seasons.
These fuzzy seasonal patterns were further analysed using a productivity approach. We
used satellite imagery (LANDSAT 7 ETM+) to analyse guanacos habitat preferences, from a
regional approach, based on a Vegetation Index (NDVI) of the area. NDVI for each guanaco
location was calculated as the average in a 100 m buffer around it. NDVI estimated for dry
season locations were significantly higher than for wet seasons (ANCOVA test; F=35.09;
p<0.0001), apparently showing a preference towards the most productive patches of the area
during dry seasons. The covariate number of guanacos per observation showed a tendency
for larger herds to occupy locations of higher plant productivity.

371
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

The combined results show guanaco's flexibility in space use and habitat selection under
extremely arid conditions, and rise two main points with potential conservation implications.
First we detect a seasonal dependence of animals on the habitats with somewhat higher
productivity, which occupy a small fraction of the area. Secondly, the smaller groups seem to
be displaced to suboptimal habitats throughout the year, a behaviour coherent with a
resource-defense polygyny mating system, where the number of females attracted by a male
is related to the quality (productivity) of resources in the territory. In addition, they highlight
the relevance of NDVI-productivity analyses to detect changes in habitat use, as a
complement to standard fine scale habitat assessments.
Finally, the low density populations of the World Heritage Site seem to be really close to
the arid limit of their niche and they could be easily threatened by competition with feral
livestock that rely heavily on the few most productive locations and water sources of the
area.

372
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

Small mammals in fragmented Atlantic forest landscapes - importance of patch

size in landscapes with different amounts of remaining habitat

A.A. Buen1, J.P. Metzger 2, R. Pardini 1

1 Departamento de Zoologia and 2 Departamento de Ecologia, Instituto de Biociências,

Universidade de São Paulo, Rua do Matão, travessa 14, 101, São Paulo, SP, Brazil.
e-mail: [email protected]

Introduction and Methods

Habitat fragmentation implies habitat loss, reducing patch size and increasing distance
among patches. A fragmentation threshold, defined as the amount of habitat below which
patch size and isolation synergistically interact with habitat loss increasing extinction rates in
fragmented landscapes, was proposed based on the non-linear relationships between habitat
amount and the number, size and isolation of remnants, and the review of bird and small
mammal empirical studies (Andrén, 1994). However, interactions between habitat loss at the
landscape scale and patch characteristics were rarely investigated, especially in tropical
forests. Here, we aim to investigate if the influence of patch size on small mammal
assemblages varies among landscapes with different amounts of remaining Atlantic Forest.
Three 10,000-ha landscapes with similar abiotic conditions were chosen in the Atlantic
plateau of São Paulo State, Brazil, retaining 45, 30 and 15% of forest. Small mammals were
sampled with a standardized protocol with pitfall traps at 50 forest patches, 15 - 20 in each
landscape, randomly chosen in classes of size. We used analysis of variance and Tukey test
to investigate if richness and abundance varied among landscapes, and linear regression to
investigate the influence of patch size on small mammals in each of the three landscapes.

Results and Discussion

We captured 1,295 individuals of 27 small mammal species, 17 of which are endemic to

the Atlantic forest and 10 are also found in open Brazilian biomes. Richness differed among
the three landscapes considering the endemic (F= 28.971, P< 0.001) but not the non-
endemic species. Richness of endemic species was lower in the 15% landscape compared
to the more forested landscapes (Figure 1). Abundance differed among the three landscapes
considering both endemic (F= 14.516, P< 0.001) and non-endemic species (F= 6.398, P=
0.003). However, while the abundance of endemic species was lower in the 15% landscape,
the abundance of non-endemic species was lower in the 45% landscape (Figure 1).
Richness and abundance were not influenced by patch size in the 15% landscape, whereas
both richness and abundance of endemic species increased with patch size in the 30%
landscape (R²= 0.384, P= 0.002; R²= 0.283, P< 0.009, respectively) and richness of non-
endemic species decreased with patch size in the most forested landscape (R²= 0.241, P=
0.036) (Figure 2). Five endemic species were present only in the two more forested
landscapes. Among those, Oryzomys russatus and Thaptomys nigrita were rare in both
landscapes, Brucepattersonius soricinus was more common in the 45% compared to the
30% landscape (F= 5.058, P< 0.031), and the abundance of Delomys sublineatus and
Marmosops incanus did not vary among landscapes. The abundance of the last three
species increased with patch size in the 30% landscape (R²= 0.192, P= 0.030; R²= 0.133, P=
0.064; R²= 0.253, P= 0.014, respectively), but not in the most forested one. The abundance
of the commonest non-endemic species (Oligoryzomys nigripes) differed among landscapes
(F= 10.456, P< 0.001), was lower in the 45% landscape compared to the other two, and was
not influenced by patch size in none of the three landscapes.
Our results highlight the importance of habitat amount for determining small mammal
diversity in fragmented tropical landscapes. Similar-sized remnants harboured distinct
assemblages in different landscape contexts, corroborating the existence of a fragmentation
threshold below which patch characteristics exacerbate habitat loss effects (Andrén, 1994).

373
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

As expected, this threshold is evident only for endemic species, which are supposed to
present a stronger association to forested habitats. For those, patch size was important in
the 30% landscape, whereas the degree of isolation and/or average size of remnants may
have, on the one hand, strengthened habitat loss effects causing the extinction of several
endemic species and thus reducing the influence of patch size in the 15% landscape; and,
on the other hand, allowed dispersion and connectivity among sub-populations and promoted
the presence of endemic species regardless of patch size in the 45% landscape. For non-
endemic species, on the contrary, richness decreased with patch size in the 45% landscape,
indicating that open habitats may act as a source for these species. However, the threshold
seems to vary among species, as proposed by Metzger and Décamps (1997). Two of them
were uncommon even in the 45% landscape and three clearly benefited from an increase in
habitat amount, being able to occupy small patches only in the landscape with the highest
habitat amount, while a generalist non-endemic species dominated the 15% landscape.

Figure 1. Mean and standard deviation of small mammal richness and abundance in forest
patches of landscapes with different proportions of remaining Atlantic Forest in Brazil.

Figure 2. Variation of small mammal richness and abundance as a function of forest patch
size in landscapes with different proportions of remaining Atlantic Forest in Brazil.

Acknowledgements

To FAPESP (99/05123-4, 05/56555-4), CNPq (690144/01-6) and CAPES, for support.

References
Andrén, H. (1994) Effects of habitat fragmentation on birds and mammals in landscapes with different
proportions of suitable habitat: a review. Oikos 71: 355-366.
Metzger, J.P. & Décamps, H. (1997) The structural connectivity threshold: an hypothesis in
conservation biology at the landscape level. Acta Oecologica 18: 1-12.

374
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

European rabbit (Oryctolagus cuniculus L.) abundance at a regional scale:

controlling factors

A. Saldaña, G. García-Salgado and S. Rebollo

Department of Ecology, Faculty of Sciences, University of Alcalá, Ctra. Madrid-Barcelona Km

33.6, 28871 Alcalá de Henares (Madrid), Spain.
e-mail: [email protected]

Introduction

The European rabbit (Oryctolagus cuniculus) is the most important small game species in
the Iberian Peninsula, and a major prey for almost 30 raptors and mammalian carnivores,
including several endangered species such as the Spanish imperial eagle and the Iberian
lynx (Fernández 2005). For this reason, the decline of their populations in the last decades
has triggered important socio-economic impacts and conservation concerns in
Mediterranean ecosystems. The need to recover rabbit populations demands appropriate
knowledge about factors controlling their distribution at wide scales.
The main objective of this investigation was to analyse the major environmental and
human factors affecting the current distribution and abundance of European rabbits in the
Central Iberian Peninsula, an area which hosts some of the most abundant native rabbit
populations. This study may provide useful guidelines for developing landscape-oriented
management strategies for the recovery of rabbits over broader areas.

Material and methods

The study area (Madrid province, between the coordinates 39º 53’ N and 41º 9’ N and 3º
3’ W and 4º 34’ W), covering about 8,000 km2, exhibits a high environmental heterogeneity.
Altitude ranges from 430 m to 2400 m. The climate is continental Mediterranean, with hot dry
summers and cold winters. Average annual temperature and rainfall range from 4ºC to 16ºC
and 400 mm to 1600 mm, respectively. The predominant soil types include Entisols, Inceptisols
and Alfisols according to USDA Soil Taxonomy. Land use is mainly forestry and cattle breeding
in the north, while agriculture and urban uses are dominant in the south. Natural vegetation
types are mainly mountain grasslands, shrublands, and forests of Quercus rotundifolia, Q.
pyrenaica and Pinus sylvestris.
Rabbit abundance was estimated through interviews with forest rangers during 2002 and
2003. Seven classes of rabbit abundance were defined on 1:25,000 maps. These data were
then re-classified into four rabbit abundance classes for statistical analyses. Associations
with environmental variables such as altitude, slope, mean annual temperature, annual
rainfall, soil types, and land use were analysed using GIS. We explored the relationships
between rabbit abundance and environmental factors by using multivariate ordination
analyses (correspondence analysis) and bi-variate analyses (corrected frequencies).

Results

Rabbits were absent or had low abundances in 65% of the area. Rabbit abundance was
strongly influenced by the environmental gradients of altitude, precipitation and temperature,
and by the different types of soils and land uses (Figure 1). The lowest rabbit abundances
were found in areas higher than 1000 m a.s.l., with slope greater than 30%, of annual
precipitation above 800 mm, and mean annual temperature below 11ºC. In addition, low
abundances were usually found where outcrops and Alfisols occurred, in pine forests, and in
urban areas. In contrast, the highest rabbit abundances were found in areas lower than 800
m a.s.l., with slopes less than 30%, annual precipitation below 650 mm, and mean annual
temperature above 13ºC. These high abundances occurred mainly on Inceptisols and

375
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

Entisols, and on land uses such as agricultural areas (rainfed and irrigated land), evergreen
forests (mainly Quercus rotundifolia), shrublands or mosaics of these types of land uses.

Figure 1. Correspondence
analysis of rabbit abundance
and environmental variables
(black circles for variables with
cosine greater than 0.9) R:
rabbit abundance (R0-R3:
rabbit absence to high rabbit
abundance). A: Altitude, m (A1-
A4: >1000; 1000-800; 800-600;
<600). PE: slope, % (PE1-PE4:
>30; 30-10; 10-3; <3). S: soil
type (S1-S6: outcrops; Entisols;
Inceptisols; Mollisols; Alfisols;
urban areas). T: temperature,
ºC (T1-T5: >13; 13-11; 11-9; 9-
7; <7). P: precipitation, mm (P1-
P5: >1200; 1200-800; 800-650;
650-500; <500). U: land use (U1: mountain grassland; U2: irrigated land; U3: rainfed; U4:
mosaic; U5: shrubland; U6: juniper; U7: evergreen forest; U8: deciduous forest; U9: pine
forest; U10: water body; U11: scarpment; U12: urban)

Discussion and conclusions

Our findings show a strong association between rabbit abundance and landscape
structure, and also reflect the high environmental heterogeneity of the Madrid province. The
highest rabbit abundance was found in mosaics, which provide abundant food and
availability of shelter (Lombardi et al. 2003; Fernández 2005). Rabbits rejected areas with
substrates where excavating warrens is difficult, such as outcrops and soils with the argillic
horizon near the soil surface (Alfisols). Our results reflect the ability of rabbits to colonize
different habitats and highlight the ecological plasticity of this species. However, rabbits were
absent or had very low abundance in two zones of the study area: (1) the mountainous
areas, due to limiting physical conditions -low temperature, high precipitation and slope, and
areas with abundant outcrops-; (2) areas of urban expansion which would otherwise be
suitable for rabbits. The results highlight that European rabbit populations in the Madrid
province are very fragmented due to urbanization and there is an urgent need for the
establishment of corridors of natural vegetation connecting the different fragments in order to
increase the stability of populations. Riverine woody vegetation along rivers flowing through
the region in a southerly direction may be a key element in this conservation strategy. Such
zones are not suitable for urbanization because of flooding risk, but they can support high
rabbit densities and would allow the exchange of individuals between different populations.

References
Fernández, N. (2005) Spatial patterns in European rabbit abundance after a population collapse.
Landscape Ecology 20: 897-910.
Lombardi, L., Fernández, N. Moreno, S. & Villafuerte, R. (2003) Habitat-related differences in rabbit
(Oryctolagus cuniculus) abundance, distribution, and activity. Journal of Mammalogy 84: 26-36.

376
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

Arthropod assemblage responses to agricultural intensification in heterogeneous

landscapes – local testing of global patterns

S.J. Attwood¹, M. Maron¹, A.P.N. House², C. Zammit³

¹Australian Centre for Sustainable Catchments and Faculty of Sciences, University of

Southern Queensland, Toowoomba, Queensland, 4350 Australia.
e-mail: [email protected]
²CSIRO Sustainable Ecosystems, Level 3, 306 Carmody Road, St. Lucia, Queensland, 4067,
Australia
³Natural Resource Management Policy Branch, Department of the Environment and
Heritage, King Edward Terrace, Parkes, ACT, 2600, Australia

Introduction

Agricultural expansion and intensification are major threats to global biodiversity, with
associated habitat loss, reduced habitat heterogeneity and other threatening processes
causing biodiversity decline (Stoate et al. 2001; Benton et al. 2003). Despite this, agricultural
landscapes often support complex and dynamic biological communities within a diverse array
of land uses (Benton et al. 2003), ranging from highly modified cropping systems to native
vegetation remnants. The responses of arthropods to agricultural intensification are of
particular interest, as they are known to contribute to vital ecosystem processes (Tscharntke
et al. 2005) and are considered responsive to environmental changes (Andersen and Majer,
2004). Whilst many individual studies indicate that arthropod diversity declines with
increasing agricultural intensification, we investigated whether this response was consistent
among habitats, taxa, management systems and global regions, and subsequently tested
whether similar patterns were evident in an eastern Australian agricultural landscape.

Method

To compare the biodiversity profiles of land uses ranging from low to high intensification,
we performed meta-analyses of arthropod richness and abundance responses to several
agricultural intensification scenarios reported in the global scientific literature. These findings
were then field-tested at a local scale, using a subset of the land uses investigated in the
meta-analyses, with ants as the focal taxon. Nine gradients of increasing land-use
disturbance, incorporating native woodland remnants, grazed grassland/pasture and cereal
crops, were examined in southern Queensland. Ants were sampled using pitfall traps in the
core of each land use, and at the interfaces between woodland/pasture and
pasture/cropping.

Results

Globally, arthropod abundance displayed no clear trend with intensifying land use, but
richness declined consistently and significantly as land use intensified. Richness was greater
in native vegetation than agricultural land uses, greater in woodland than pasture, greater in
pasture than cropping and greater in reduced-input cropping than conventional cropping. In
the local-scale field-testing, ant abundance responses were variable, but ant richness
declined significantly as land use intensified, consistent with the global trend (Fig.1).

377
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

c c
20
18
16
b

14 b
12
10
8
a
6
4
2
0
CC CE PC WE WC

Figure 1. Mean count ant morphospecies richness in core land uses and interfaces. CC =
cropping; CE = cropping/pasture interface; PC = pasture; WE = pasture/woodland interface;
WC = woodland. Different letter denotes statistically significant difference.

Multivariate analyses revealed that ant assemblages within each core land use were
distinct from one another. No differences were found between the woodland/pasture
interface and its adjacent habitats. The pasture/cropping interface exhibited a different
assemblage to cropping but not pasture.

Discussion

The observed decline in taxonomic richness may be due to frequent ‘resetting of the
successional clock’, with high disturbance areas only supporting early-successional taxa.
This was illustrated in our local-scale study, with opportunistic taxa being more prevalent in
the cropping and crop/pasture interface than in other land uses. The marked richness decline
in the cropping core indicates that this land use may act as an ecological filter for many taxa.
Land use type exerted a greater influence over assemblage composition than the spatial
proximity of treatments or other local factors. This may be due to the limited dispersal
abilities of minor worker ants, where seemingly modest areas of habitat may effectively
represent the operational landscape for many taxa.
It appears that even fragmented and highly degraded woodland remnants, such as those
studied, and less intensive agricultural land uses such as pasture, may play an important role
in maintaining arthropod biodiversity in agricultural landscapes.

References
Andersen, A. N. & Majer, J.D. (2004) Ants show the way Down Under: invertebrates as bioindicators
in land management. Frontiers in Ecology and the Environment 2: 291-298.
Benton, T. G., Vickery, J. A. & Wilson, J.D. (2003) Farmland biodiversity: is habitat heterogeneity
the key? Trends in Ecology & Evolution 18: 182-188.
Stoate, C., Boatman, N. D., Borralho, R.J., Carvalho, C.R., de Snoo, G.R. & Eden, P. (2001)
Ecological impacts of arable intensification in Europe. Journal of Environmental Management 63:
337-365.
Tscharntke, T., Klein, A. M., Kruess, A., Steffan-Dewenter, I., & Thies, C. (2005) Landscape
perspectives on agricultural intensification and biodiversity - ecosystem service management.
Ecology Letters 8: 857-874.

378
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

Influence of habitat quality and landscape structure on the distribution of

Maculinea teleius and Maculinea nausithous

N. Örvössy, Á. Kőrösi, P. Batáry, L. Peregovits

Department of Zoology, Hungarian Natural History Museum Baross u. 13, H-1088 Budapest,
Hungary
e-mail: [email protected]

Maculinea butterflies are considered to be good indicator species, because of their

special life cycle – obligate parasitism on Myrmica ants and larval host plant use. Maculinea
teleius and Maculinea nausithous occur sympatricaly in wet meadows (Nowicki et al. 2005)
and use the same food plant, Sanguisorba officinalis, but different host ant species (Thomas
et al. 1998). Niche segregation and habitat requirements have been examined between the
two species from the 80’s (Thomas 1984), but are still a hotspot in Maculinea research.
Thomas and Elmes (2001) found that M. teleius and M nausithous differ in food plant
selection, Johst et al. (2006) showed difference in the effects of mowing regime on these
butterflies.
M. teleius and M. nausithous are widespread in western Hungary, where mowing regimes
maintain suitable wet meadow habitats. This area is a very mosaic landscape due to
undulating surface, land ownership and management; therefore numerous different habitat
patches are present. Transformation of mowing regime or abandonment cause changes in
the quality of the habitats and consequently in their suitability for butterflies. Johst et al.
(2006) simulated the effects of different mowing regimes on M. teleius and M. nausithous
populations and found that traditional twice a year mowing is detrimental. Our aims were to
determine how M. teleius and M. nausithous abundance depend on habitat quality and what
the effects of mowing, abandonment and landscape parameters are on butterfly density.
84 habitat patches were chosen for sampling in three stream valleys (region), relying
upon the presence of host plant. The habitat patches were outlined based on presumable
barriers (forest, road) and difference in habitat management. Transects were used to
estimate the density of the butterflies and the quality of the habitat patches. The density was
calculated based on two visits per patches. Habitat quality was described by host plant
density, host plant and vegetation height and coverage of shrubs. Transect lengths
depended on the area of the patch; the width was 3 meters on both sides. Area and
perimeter of patches were measured by GPS, year of last mowing or presence of previous
mowing in the study year were also noticed. Linear regression analyses were used to
determine the effects of described parameters on butterfly density, the best models were
chosen by backward selection.

Table 1. Results of linear regression analyses on M. teleius and M. nausithous.

Species model F df p parameters Beta t value p
M. teleius 62.15 59 <0.01 Perimeter/Area ratio 0.670 3.470 0.001
Mowing before the survey 0.088 2.286 0.026
Year of last mowing -0.007 -2.545 0.014
Height of food plant 0.002 2.222 0.030
Region1 14.186 2.536 0.014
Region2 14.267 2.550 0.013
Region3 14.197 2.535 0.014
M. nausithous 15.46 48 <0.01 Perimeter/Area ratio 0.761 6.388 <0.01
Year of last mowing -0.005 -2.501 0.016
Mowing before the survey 0.031 1.441 0.156

M. teleius in all, M. nausithous in most of the patches were present. In the case of M. teleius
the best linear regression model included habitat quality factors like the host plant height and
mowing regime and landscape factors, like region and perimeter/area ratio as well (Table 1).

379
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

In spite of that model M. nausithous density is best described by perimeter/area ratio and the
year of last mowing (Table 1). Patch sizes and perimeter area ratios are different in the three
regions (Kruskal-Wallis test: patch size chi-square = 16.268 df = 2, p < 0.001; perimeter area
ratio chi-square = 13.65, df = 2, p = 0.001), and this difference has an effect on the density of
the species.
In the study area the mowing regime has been changed in the last decade, the traditional
twice a year cut does not exist anymore, the majority of the patches are mown once a year
with no time considerations and several patches have been abandoned. We found that
abandonment has a positive effect on both species, although this survey does not contain
patches abandoned more than 15 years ago. In addition preserving wet meadow habitats
probably require management to prevent succession. Further, mowing the meadows just
before the flight period of Maculinea species has a negative effect on density. These results
are similar with the simulation models of Johst et al. (2006), that less frequent mowing
frequency than in traditional regime increases the survival of both species. Our result support
the simulation model that the cut should be at least some weeks before the flight period, in
front of the traditional mid flight period mowing.
Even though niche segregation exists in the use of food plant between the two Maculinea
species (Thomas and Elmes 2001), density of food plant had no significant effect on the
butterflies (Table 1). On the other hand growing host plant height increased the density of M.
teleius, but not M. nausithous, which can support the niche segregation by host plant use
hypothesis (Thomas and Elmes 2001).
Perimeter area ratio had the strongest impact on the density of the butterflies, particularly
M. nausithous preferred patches with more edge, than core area (Table 1), likewise earlier
MRR studies refer to it in the study area (Kőrösi 2005) and in Poland (Nowicki et al. 2005).
Our results suggest that for preserving these species both required mowing regime that
differs from traditional mowing and landscape structure, especially the enormous effect of
perimeter area ratio should be taken into account during management. Based on our results,
on this area mowing should be more frequent in smaller patches with large edges, which are
more exposed to succession and less frequent in large patches to increase butterfly density.

Research was funded by NKTH Faunagenesis project (3B023-04).

References
Johst, K; Drechsler, M; Thomas, J. & Settele, J. (2006) Influence of mowing on the persistence of
two endangered large blue butterfly species. Journal of Applied Ecology 43: 333–342.
Kőrösi, Á. (2005) Habitat-use of wetland Maculinea species – a case study. J. Settele, E. Kühn & J.
Thomas (Eds). Studies on the Ecology and Conservation of Butterflies in Europe. Pensoft, Sofia-
Moscow, pp. 132.
Nowicki, P; Witek, M; Skórka, P; Settele, J. & Woyciechowski, M. (2005) Population ecology of the
endangered butterflies Maculinea teleius and M. nausithous and the implications for conservation.
Population Biology 47: 193-202.
Thomas, J. A. (1984) The behaviour and habitat requirements of Maculinea nausithous (the Dusky
Large Blue Butterfly) and M. teleius (the Scarce Large Blue) in France. Biological Conservation
28: 325-347.
Thomas, J.A; Elmes, G.W. & Wardlaw, J.C. (1998) Polymorphic growth in larvae of the butterfly
Maculinea rebeli, a social parasite of Myrmica ant colonies. Proceedings of the Royal Society of
London B 265: 1895–1901.
Thomas, J.A. & Elmes, G.W. (2001) Food-plant niche selection rather than the presence of ant nests
explains oviposition patterns in the myrmecophilous butterfly genus Maculinea. Proceedings of the
Royal Society of London B 268: 471–477.

380
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

Effect of landscape structure on Collembola communities in 30 oilseed rape fields

P. Querner, A. Bruckner, T. Drapela, D. Moser, J.G. Zaller, T. Frank

Institute of Zoology , Department of Integrative Biology and Biodiversity Research, University

of Natural Resources and Applied Life Sciences, Gregor-Mendel-Strasse 33
A-1180 Vienna, Austria.
e-mail: [email protected]

Soil animals are seldom investigated in landscape ecology although they contribute with a
very high local diversity and abundance to the important nutrition cycles and the soil
development processes. Are surface active Collembola assemblages more affected by the
surrounding landscape than the soil living species? Do species migrate from the adjacent
forest or grassland into the crop fields and from what distance? To study these questions, we
sampled the surface active and the soil living springtail fauna of 30 winter oilseed rape fields
in a 160 km2 study area in eastern Austria. The proportion of non-crop area around the sites
varied from 10% to 70%. Surface active Collembola were sampled with pitfall traps and the
endogeic species with soil cores. All specimens were counted, identified to species level and
classified as grassland, woodland or ubiquistic species. Community structure and
composition were correlated with (1) local variables (soil type, soil index, cultivation intensity)
and landscape variables as the (2) proportional area of crop, vineyards and semi natural
habitats, (3) minimum distance to semi-natural habitats (fallow, grasslands and woodlands)
and (4) landscape diversity indices at scales of 250 to 2000 m radius around the sites.
Surface active Collembola were found to be influenced by the regional landscape
composition more than the local soil properties. This was expected as they are larger and
more mobile and can migrate from one patch to another. Passive wind dispersal is also a
possible explanation. The small endogeic Collembola living in deep soil layers are less
mobile and depend more on the local soil properties, than on the surrounding landscape
structure.

381
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

Landscape structure and small mammal distribution – the importance of matrix

quality for the connectivity of a fragmented Atlantic forest landscape

F. Umetsu 1, R. Pardini 1, J.P. Metzger 2

1
Departamento de Zoologia and 2 Departamento de Ecologia, Instituto de Biociências,
Universidade de São Paulo, Rua do Matão, travessa 14, 101, CEP 05508-900, São Paulo,
SP, Brazil.
e-mail: [email protected]

Introduction
If, on the one hand, theoretical metapopulation and landscape ecology models have
suggested that taking into account the suitability of altered habitats for the occurrence/
dispersion of organisms is essential to understand processes in fragmented landscapes; on
the other hand, this often requires detailed and long term data on the biology of species,
which, particularly in the tropics, are not available and are time-consuming to gather. Using
the distribution of small mammals in forest remnants and in the main types of altered habitats
in an Atlantic forest landscape, we investigated how the explanatory power varies between
landscape variables that incorporate or not matrix quality and the importance of spatial scale
for analyzing the influence of landscape structure. We assumed that information on the
occurrence of species gathered through standardized sampling in different habitat types of a
heterogeneous landscape is useful as indices of matrix quality.

Methods
The fragmented landscape under study, located in Caucaia do Alto, State of São Paulo,
Brazil, comprises 10,000 ha and harbors 31% of forest remnants surrounded by areas of
agriculture (38%), rural/ urban areas with buildings (14%), native vegetation in initial stages
of regeneration (7%), and homogeneous pine/ eucalyptus plantations (7%). The small
mammal community was sampled using a standardized protocol with Sherman live-traps at
36 sites, 20 in forest fragments (2 – 275 ha) and 16 in the matrix (four in each of the four
predominant altered habitats). We calculated two indices of quality for each sampled habitat,
one represented by the relative abundance and one by the occurrence of each species. For
each forest fragment, we calculated two landscape variables - habitat quantity and habitat
connectivity based on the distance and size of surrounding habitat patches - in different
spatial scales (50 – 800 m). For each variable, we calculated one metric that considers forest
fragments as equally suitable and all altered habitats in the matrix as unsuitable for all small
mammal species, and one that takes into account the variation in quality among habitats for
each species. We compared the explanatory power of logistic regression models describing
the chance of occurrence of species in the 20 fragments among metrics that consider and
those that do not consider matrix quality.

Results
We analyzed the distribution of 8 small mammals (Akodon montensis, Delomys
sublineatus, Didelphis aurita, Marmosops incanus, Micoureus paraguayanus, Oligoryzomys
nigripes, Oryzomys angouya and Oryzomys russatus). Only D. aurita and O. russatus did not
occur in any of the matrix habitats and only A. montensis and O. nigripes occurred in all.
Open anthropogenic habitats harbored the smallest (2), and native vegetation in initial stages
of regeneration the highest (5), number of species. While a consistent increase in the
explanatory power when considering matrix quality in relation to considering the matrix as
inhospitable was observed for just one species (D. sublineatus) among models of habitat
quantity, it was observed for all six species that occurred in the matrix among models of
habitat connectivity. Both for habitat quantity and connectivity, models presented higher or
similar explanatory power using the index based on species occurrence in comparison to that
based on species abundance. The models of habitat quantity significantly explained the
chance of occurrence of just the two species that did not occur in the matrix and for which

382
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

there was no difference between metrics that considered or not matrix quality (D. aurita and
O. russatus). The occurrence of O. russatus was also significantly explained by models of
habitat connectivity. However, these models, when considering matrix quality, also explained
the occurrence of M. paraguayanus, a species that occurred in only one type of matrix
habitat. Both models of habitat quantity and connectivity for the two species that did not
occur in the matrix varied with spatial scale. Among the remaining species, only the models
for D. sublineatus did not vary with spatial scale for all variables and metrics considered,
while the models of the others varied depending not only on the type of landscape variable
and on considering or not matrix quality, but also on spatial scale. For them, models were
clearly more independent of spatial scale for habitat connectivity compared to habitat
quantity models.

Discussion
For many species and spatial scales, metrics that considered the heterogeneous quality
of the matrix presented a higher explanatory power for the distribution of small mammals in
forest fragments. This is in accordance with the results of the few published studies that
compared the performance of models considering or not matrix heterogeneity, all carried out
in temperate regions (e.g., Sutcliffe et al. 2003; Verbeylen et al. 2003; Revilla et al. 2004).
However, in those studies matrix heterogeneity was quantified based on the researcher
experience or inferred from auto-ecological studies. Our study shows that it is feasible to
obtain indices of matrix quality through standardized sampling in different types of matrix,
which may increase the predictive power of landscape structure variables. Thus, the matrix
of altered habitats not only plays a fundamental role in ecological processes, but also could
be taken into account for modeling and managing human dominated landscapes, even if
detailed information on the biology of species is not available. However, the increase in
explanatory power between metrics that considered or not matrix heterogeneity is more
consistent across species for habitat connectivity compared to habitat quantity models. This
clearly indicates the importance of taking into account the distance among patches, which is
a simple way of representing the chance of dispersion. It is probable that some types of
matrix that do not provide a suitable habitat for small mammals allow periodical dispersion to
take place. Besides the consistency across species, habitat connectivity models were also
less variable across spatial scales, corroborating the findings of simulation models (Bender
et al. 2003) and favoring the use of these metrics, since establishing appropriate scales is a
difficult task. Finally, all three species with distribution in fragments significantly explained by
landscape variables did not occur in matrix habitats or in just one, indicating that the
importance of landscape structure is higher for species that are not widespread in the matrix.

References
Bender, D.J; Tischendorf, L. & Fahrig, L. (2003) Using patch isolation metrics to predict animal
movement in binary landscapes. Landscape Ecology 18: 17-39.
Revilla, E; Wiegand, T; Palomares, F; Ferreras, P. & Delibes, M. (2004) Effects of matrix
heterogeneity on animal dispersal: from individual behavior to metapopulation-level parameters.
The American Naturalist 164: 130-153.
Sutcliffe, O.L; Bakkestuen, V; Fry, G. & Stabbetorp, O.E. (2003) Modelling the benefits of farmland
restoration: methodology and application to butterfly movement. Landscape and Urban Planning
63: 15-31.
Verbeylen, G; De Bruyn, L; Adriaensen, F. & Matthysen, E. (2003) Does matrix resistance
influence Red squirrel (Sciurus vulgaris L. 1758) distribution in an urban landscape? Landscape
Ecology 18: 791-805.

383
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

A. Janin1, J.P. Léna 1, C. Delacourt 2, P. Allemand 2, P. Joly 1

1
Ecology of Fluvial Hydrosystems and 2 Earth Sciences – University Lyon 1, 43 boulevard du
11 Novembre 1918 F-69622 Villeurbanne Cedex, France.
e-mail: [email protected].

Introduction

Human activities generate a land conversion process resulting in landscapes composed

of multiple habitats of varying quality according to the species’ perception. This phenomenon,
which generally reduces and fragments suitable habitats, is widely recognised as a major
threat to the long-term persistence of populations. Hence, there is a growing demand for
tools to predict and evaluate the impact of landuse changes. Traditional studies simplify the
complexity of real landscapes by assuming a binary landscape structure which consists of a
fragmented suitable habitat (the patches) embedded in a uniform non-habitat (the matrix).
These studies have mainly focused on the patches, whereas considerably less attention has
been devolved to the matrix. Yet, it is increasingly evident that the matrix is composed of a
mosaic of habitats differing in their resistance to animal movement and thus affects
landscape functional connectivity by enhancing or impeding movement (Ricketts, 2001).
In this context, we argue in favor of new approaches which integrate both the composition
and the configuration of the whole landscape including the matrix. Percolation models such
as Cost Distance models (Adriaensen et al., 2003) are relevant tools to capture the
landscape complexity and assess connectivity in a functional way. However, the accuracy of
these models is strongly dependent on pre-existent knowledge of movement characteristics
of the species considered. Unfortunately, this type of biological knowledge remains scarce. In
the present work, we propose to adapt a cross-validation method to circumvent this problem:
1) we estimate the habitat resistance values required in Cost Distance models through a
calibration procedure with no a priori knowledge; 2) we test the model’s validity and assess
the relative influence of composition and configuration of the complete landscape. We
illustrate this approach through a case study conducted on the common toad (Bufo bufo)
which is expected to harshly undergo the landscape impact because of its ground-dwelling
habits and its obligatory seasonal migrations between forests and breeding ponds.

Methods

Cost Distance modelling (a toolbox in GIS-systems) provides a measurement of distance

weighed by the species-specific resistance of each habitat constituting the landscape
(Adriaensen et al., 2003). Using this method, we built an integrative variable of landscape
composition and configuration corresponding to the migration area, i.e. the area in which
toads can potentially move from their pond to complete their biological cycle (Ray et al.,
2002, Joly et al., 2003). We calibrated the resistance values of four habitat types (forest,
crops, meadow and urban area) on a first dataset (Rhone-Alps, France) without a priori
knowledge. For this purpose, we explored the relationship between migration areas,
computed with a systematically explored range of resistance values, and toad occurrence.
Next, we checked the validity of the selected resistance values with another dataset in the
same region. We also verified that the model remained valid in a different landscape
configuration (Geneva, Switzerland). To segregate the relative role of functional connectivity,
we compared the predictive power of the migration area, which synthetizes composition and
configuration, with that of landscape composition alone, which was assessed by the amount
of each habitat surrounding the pond. All analyses were performed using logistic regressions.

Results

384
Theme 3. Ecological Networks, fragmentation and connectivity
3.7 Open Session 12: Landscape modelling and mammal, amphibian and insect populations

The calibration selected high resistance values for crops and low resistance values for
meadow and forest. The validation step showed that the migration area computed using the
selected resistance values had a significant impact on occurrence (p<10^-5 for Rhone-Alps
and p<10^-4 for the Swiss region): the probability of presence increased with the size of the
migration area (Fig.1). Moreover, for both validation datasets, our integrative variable had a
significantly better predictive power than landscape composition. Indeed, even after
removing the effects of the amount of each habitat type, the influence of the migration area
on occurrence remained significant: p<10^-5 for Rhone-Alps and p<10^-2 for the Swiss region.
The explanatory power of landscape composition was surprisingly low, particularly for the
Rhone-Alps dataset in which the amount of forest had no effect at all on occurrence (p>0.20).

Logistic regression Simulated migration

1.0

Ly
0.8
0.6
Occurrence

The Rhone
0.4
0.2
0.0

0 1 2 3 4 5 Po
Size of migration area in
Figure 1. Toad occurrence related to migration areas for the Rhone-Alps validation dataset.

Discussion and Perspectives

The resistance values determined by calibration fully agree with current knowledge of the
behavior of the common toad that avoids crops whilst it crosses large meadow and settles in
forest (Ray et al., 2002). Moreover, the migration areas computed with these resistance
values are accurate predictors of toad occurrence since they explain significantly more
variance than landscape composition alone. These results highlight the advantage, in order
to reliably predict species occurrence, of capturing both the composition and the
configuration of the whole landscape through a unique variable. Used here to assess the
possibility for animals to complete their biological cycle, this tool can easily be adapted to
design connections necessary to maintain the functioning of metapopulations.

References

Adriaensen, F; Chardon, J.P; De Blust, G; Swinnen, E; Villalba, S; Gulinck, H. & Matthyssen, E.

(2003) The application of ‘least cost’ modelling as a functional landscape model Landscape and
Urban Planning 64: 233-247.
Joly, P; Morand, C. & Cohas, A. (2003) Habitat fragmentation and amphibian conservation: building
a tool for assessing landscape matrix connectivity C. R. Biologies 326: S132-S139.
Ray, N; Lehmann, A. & Joly, P. (2002) Modeling spatial distribution of amphibian populations: a GIS
approach based on habitat matrix permeability Biodiversity and Conservation 11: 2143-2165.
Ricketts, T.H. (2001) The matrix matters: effective isolation in fragmented landscapes The American
Naturalist 158(1): 87-99

385
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

3.8 Open Session 13: Habitat Fragmentation and Mitigation

Landscape dynamic effecting bird diversity in an Atlantic Forest fragmented

region: an evidence for time lag response.

A. Uezu1 and J.P. Metzger2

1
– IPÊ - Instituto de Pesquisas Ecológicas, Rodovia Dom Pedro I, Km 47 Caixa Postal
4712960-000, Nazaré Paulista, SP, Brazil. www.ipe.org.br. E-mail: [email protected]
1,2
– Departamento de Ecologia, Instituto de Biociências, Universidade de São Paulo, Rua do
Matão, 321, travessa 14, 05508-900, São Paulo, SP, Brazil.

Introduction

In several cases, species extinction does not occur immediately after habitat fragmentation,
what creates a time lag between these two processes (Tilman et al., 1994; Brooks et al.,
1999; Hanski & Ovaskainen, 2002), making landscape dynamic an important factor to assess
species threaten. An approach to evaluate such effects is to relate actual biological pattern
with past landscape structures (Lindborg & Eriksson, 2004). The aim of this study was to
verify the effects of historical processes of fragmentation on current richness patterns of
forest birds in an Atlantic Forest region (Pontal do Paranapanema) and to estimate the time
lag of species responses to structural landscape changes.

Methods

Historical survey and landscape analysis

We mapped the region using different types of information. For the most recent years we
used Satellite images (Landsat 5) and for older ones topographic maps and aerial
photographies. The maps were produced for each decade: 1957, 1965, 1978, 1984, 1993
and 2003. For each period we calculated landscape indices which represent patch size
(logArea) and isolation (logProx).
Bird survey
Birds were surveyed using point count methodology in 21 patches plus seven sites in a
control reserve (The State Park of Morro do Diabo, with 36,000 ha). Patches varied in size:
seven large (>400 ha), seven medium (100 - 200 ha) and seven small (30 – 80 ha). Species
richness was calculated for forest and generalist species and for different functional groups.
Data analysis
Model selection, based on Akaike’s Information Criterion, was performed to distinguish
which landscape indices, older or recent ones, explain better species richness. Nine
regression models were proposed combining landscape parameters from the years of 1965,
1978 and 2003 (Figure 1), which presented the lowest correlation, and relating them with bird
community parameters.

Results

Patch size of 1978 was the most important variable to explain the richness of the most
sensitive groups (Table 1), what provides an evidence for a time lag of bird response to
fragmentation of about 25 years. This period also corresponds to a moment of intense
fragmentation process (Figure 1), when, probably, the landscape passed through the
fragmentation threshold. For non-sensitive species, there is a high uncertainty, although the
proximity index of 2003 was more often selected as the most important variable.

Discussion

386
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Our results indicate that the sensitive species did not get yet into a steady state condition.
More losses might occur from these groups in a near future if no landscape management is
carried out to revert the negative consequences of forest fragmentation in the region.
Probably, other landscapes in the Atlantic Forest are in the same or even worst condition as
in many cases, human occupation started earlier than in the studied region and theoretical
fragmentation threshold was reached more than 50 years ago. The present study highlight
the necessity to consider the historical process of fragmentation to evaluate the real status of
species threaten and to preview where the futures extinctions will occur, making possible for
us to anticipate the conservation actions to avoid such losses.

Figure 1 – Fragmentation process in the Pontal do Paranapanema region, southeast Brazil.

The rivers are represented in dark gray, the forest in light gray and the black lines delineate
the studied forest patches in 2003.

Table 1 – Relative importance of independent variable: sum of w AIC (chance of the model
to be selected) of all models (nine), where a particular variable was present. Higher values
represent more important variable. Pontal do Paranapanema, southeast Brazil.
Functional groups (sensitive) Area65 Prox65 Area78 Prox78 Area03 Prox03
1-2 types of Forest 0,00 0,00 0,90 0,18 0,10 0,03
High sensitivity 0,03 0,01 0,56 0,11 0,40 0,13
Medium sensitivity 0,00 0,00 1,00 0,24 0,00 0,00
Limit of distribution (<100 km) 0,01 0,00 0,69 0,20 0,29 0,09
Limit of distribution (100-200 km) 0,00 0,00 0,90 0,19 0,10 0,02
Low abundance 0,00 0,00 0,96 0,22 0,04 0,01
Endemic in the Atlantic Forest 0,00 0,00 0,91 0,32 0,09 0,02
Terrestrial Frugivorous and insectivorous 0,06 0,02 0,80 0,18 0,14 0,03
Frugivorous and omnivorous from canopy 0,05 0,04 0,78 0,15 0,17 0,08
Understore insectivorous and omnivorous 0,00 0,00 0,99 0,58 0,01 0,00
Large understore Insectivorous 0,00 0,00 0,80 0,41 0,18 0,06

References
Brooks, T.; Pimm, S.L.; Oyugi, J.O., 1999. Time lag between deforestation and bird extinction in
tropical forest fragments. Conservation Biology 13, 1140-1150.
Lindborg, R. & Eriksson, O., 2004. Historical landscape connectivity affects present plant species
diversity. Ecology 85, 1840-1845.
Hanski, I. & Ovaskainen, O., 2002. Extinction debt at extinction threshould. Conservation Biology 16,
666-673.
Tilman, D.; May, R.M.; Lehman, C.L. & Nowak, M.A., 1994. Habitat destruction and the extinction
debt. Nature 371, 65-66.

387
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Using a consistent habitat network approach to reduce fragmentation at site to

national scales.

A.E. Eycott, K. Watts

Forest Research, Alice Holt Lodge, Farnham, Surrey, GU10 4LH, UK.
e-mail: [email protected]

Introduction

Improving the functional connectivity of landscapes is important for biodiversity

conservation. Movement between patches is beneficial for the maintenance of structured
populations and may also allow species to adjust their range in response to environmental
change.
There are obligations for countries to reduce fragmentation at local, national and
international scales (e.g. European Union Habitats Directive Article 10, UK Biodiversity
Action Plan – EEC, 1992; UK Biodiversity Steering Group, 1995). At the local scale, there is
a need to prioritise individual sites to ensure the effectiveness of site-based action. At more
regional and national scales, finding areas with the greatest potential for improvement or a
key role in large-scale connectivity will also focus action in a more effective manner.

Method

Habitat networks based on functional connectivity have been used to examine

fragmentation (e.g. Bani et al., 2002; Vuilleumier and Prelaz-Droux 2006) but how such
networks should be used to target action is less frequently demonstrated. There has not
been a consistent, integrated approach to reducing fragmentation that can be applied across
the various scales and link them up.
We use a functional approach to define habitat networks based on generic focal species
and least-cost techniques (Adriaensen et al., 2003) to account for matrix permeability.
Functional habitat networks were generated for broadleaved woodland, an important habitat
suffering fragmentation in temperate Europe. The ecological permeability of the matrix was
set by the degree of naturalness and complexity of vertical structure (Watts et al., 2005).

Locating latent networks for site based and national targeting

Once the initial functional networks had been mapped, the maximum dispersal distance
was gradually increased by small increments relative to the initial dispersal distance. This
allowed the identification of the functionally closest networks (Figure 1). These were termed
‘latent networks’ as they would potentially become single networks with only a small
adjustment to the landscape, e.g. woodland planting or farming extensification.
Beyond local scale action, it is possible to identify particular regions which have a
disproportionately high number of latent networks, thus offering a high potential to reduce
fragmentation. Such areas could be the target of strategic action and incentive schemes.

A hierarchy of networks for different scales of planning

To complement the above method, instead of small increments in the dispersal distance,
larger intervals were used. These provide a framework for action at increasingly large
scales, appropriate to different elements of biodiversity or more infrequent long-distance
dispersal events. Short distances identify existing functionally connected ‘core’ networks,
moderate distances provide a framework for consolidating and expanding core networks into,
and very high maximum dispersal distances form networks large enough to span near-
national scales allowing the identification of potential linkages and barriers to large-scale

388
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

movement. These scales represent a gradient of aspiration, starting with a need to protect
existing networks and ending when species are able to adapt to climate change and shifting
range margins by migration.

Conclusions

Our work addresses the need for consistency over different scales. It also addresses the
need for improving connectivity while acknowledging multi-use landscapes, as it allows for
guidance of solutions as well as targeting specific locations.
The issue of scale has dominated landscape research since the inception of IALE. While
it is important to consolidate the existing resource (i.e. functioning habitat networks) in the
face of continuing biodiversity loss, in the near future species range margins may be forced
to change dramatically. However, if on the other hand we act to close only the gaps on a
national scale, will the underlying networks be robust enough for species to respond?

Figure 1. Increasing the maximum dispersal distance from the functional networks mapped
allows the detection of functionally close networks.

References
Adriaensen, F.; Chardon, J.P.; De Blust, G.; Swinnen, E.; Villalba, S.; Gulinck, H. & Matthysen,
E. (2003) The application of 'least-cost' modelling as a functional landscape model. Landscape
and Urban Planning 64:233-247.
Bani, L.; Baietto, M.; Bottoni, L. & Massa, R. (2002) The use of focal species in designing a habitat
network for a lowland area of Lombardy, Italy. Conservation Biology 16:826-831.
EEC (1992) Council Directive 92/43/EEC of 21st May 1992 on the conservation of natural habitats and
of wild flora and fauna. HMSO, London
UK Biodiversity Steering Group (1995) Biodiversity: The UK Steering Group Report - Volume II:
Action Plans (Annex G - species action plans, habitat action plans and habitat statements).
HMSO, London.
Vuilleumier, S. & Prelaz-Droux, R. (2002) Map of Ecological Networks for landscape planning.
Landscape and Urban Planning 58:157-170.
Watts, K.; Griffiths, M.; Quine, C.; Ray, D. & Humphrey, J. W. (2005) Towards a Woodland Habitat
Network for Wales. Countryside Council for Wales, Bangor, UK.

389
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Combining ecological networks on different spatial scales leads to synergy

C.J. Grashof-Bokdam, M. van Adrichem, J.M. Baveco, M. van de Berg, P. Chardon,

R. Jochem, H.A.M. Meeuwsen, P. Schippers, M. van der Veen, J. Verboom, C. Vos

Alterra, Landscape Centre, p.o box 47, 6700 AA Wageningen, the Netherlands.
e-mail: [email protected]

Introduction

In the Netherlands, large networks of nature areas, like the National Ecological Network
or Natura 2000 areas, are often under national protection. On the other hand, small networks
of (semi-) natural landscape elements in agricultural or urban areas, also called green (or
blue) veining, are often a responsibility of regional or local authorities. As many species use
both larger nature areas and green veining as (partial) habitat, combining both networks in
landscape design could lead to synergy: the amount of total habitat needed for sustainable
(meta) populations could be optimized, and, if species can use both networks in an optimal
way, populations may have higher sustainability or resilience in changing landscapes and in
a changing climate.

In this study we focussed on networks of woody habitat on the higher sandy soils of the
Netherlands. Results of field studies show what species benefit from combining networks
and under what spatial conditions. Simulation modelling provides more insight in the
underlying (meta) population processes of synergy between large and small networks.
Finally, the found spatial conditions will be assessed in a multi-criteria analysis in which
green networks are also used for non-ecological functions like recreation.

Field studies

On a regional scale, we related the occurrence in km grids of 40 plant, butterfly and bird
species of forest habitat to spatial connectivity of large and of small networks in the
surroundings within the dispersal range of the studied species. Many species, e.g. the
butterfly speckled wood (Pararge aegeria), appeared to occur more often in landscapes were
both networks are spatially combined than in landscapes dominated by either large forest
areas or by small forest patches and wooded banks (figure 1). We found a clear relation
between synergy and the spatial character traits and habitat preferences of species.

On a local scale, we compared the occurrence of two butterfly species, speckled wood
and ringlet (Aphantopus hyperantus), in isolated large networks, in combinations of large and
small networks and in isolated small networks. On this scale, only the less mobile ringlet
benefited from combining networks. Apparently, combination of networks is only beneficial if
it happens on the correct spatial scale for the species under concern.

Model studies

We developed several artificial landscapes that all consist of a fixed percentage of woody
habitat, but of a different proportion of large and small networks (patchy and linear habitat).
We simulated population performance of several species differing in habitat preferences and
spatial character traits. For this purpose we combined METPHOR, an individual-based
(meta)population model (Verboom et al., 2001), with SmallSteps (Baveco, 2002) a
movement model. For each of the species types, we searched for the minimum spatial
conditions for synergy in combined networks.
a) b)
p(occurrence) speckled wood p(occurrence) speckled wood

390
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

SN max LN max
SN average LN average
SN min LN min

connectivity large networks connectivity small networks

Figure 1. Probability of occurrence of the butterfly species speckled wood versus spatial
connectivity of large networks (LN) and small networks (SN). Increase in occurrence versus
large networks is higher if connectivity of small networks is also higher (a). Increase in
occurrence versus small networks is higher if connectivity of large networks is also higher (b)
(Grashof et al., in prep).

References
Baveco, J.M (2002) The SmallSteps movement model. http://purl.oclc.org/NET/alterra/movement.
Grashof, C.J. et al. (in prep). Synergy between large and small networks on the higher sandy soils of
the Netherlands.
Verboom, J. et al. (2001). Introducing the key patch approach for habitat networks with persistent
populations: an example for marshland birds. Biological Conservation 100(1): 89-101.

391
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Habitat availability and connectivity for a focal species (jaguar) in the Yungas
region, Argentina

D. Somma1, P. Perovic2, R.H.G. Jongman3, H. van Lier3, A. Cerezo4 , R.J.A. van

Lammeren3
1
National Parks Administration of Argentina, Av. Santa Fe 690, 1059, Capital Federal,
Argentina.
e-mail: [email protected]
2
Instituto de Biología de la Altura, Universidad Nacional de Jujuy. Av. Bolivia 1711, 4600,
San Salvador de Jujuy, Jujuy, Argentina.
3
ALTERRA, Po Box 47, 6700 AA, Wageningen, The Netherlands.
4
Facultad de Agronomía – Universidad de Buenos Aires, Av. San Martín 4453, 1417 Buenos
Aires - Argentina.

Introduction

Our aim was to evaluate wildlife habitat availability and landscape connectivity in the
Argentine Yungas at a regional scale for a selected feline taken as focal species: the jaguar
(Panthera onca).

Jaguars have been proposed as landscape detective species in Brazil (Cullen, 2004).
Landscape detectives are thus defined as organisms which can show how to plan and
manage reserves and large interconnected eco-regions, because their requirements for
survival reveal factors important to maintaining ecologically healthy conditions.
The central hypothesis is that by using jaguars as a landscape detective it is possible to
identify and assess three important and independent features that characterize large
carnivores and large scale conservation planning: (1) prey diversity and density, (2) large
core areas and important habitat patches for biodiversity conservation, and (3) biological
corridors and landscape connectivity. In our research we developed the items (2) and (3).

In particular, we wanted to establish a set of working hypotheses about the spatial

distribution of high quality habitat for jaguars. We aim to identify areas of high conservation
value, which of these are in potential conflict with human presence or uses, and possible
habitat strips which can act linking corridors between core areas (e.g. Baritu and Calilegua
National Parks, and Pintascayo Provincial Park). The jaguar is used as “primer” in the
identification of key sites for biodiversity conservation across the Yungas region, an
approach similar to the case of bears in northern Spain (Naves et al., 2003).

Methods

A number of natural and human variables were considered, a statistical validation of the
analysis and a regional model of habitat availability are presented (Austin, 2002). In this
research, “habitat” is considered a species-specific concept (Garshelis, 2000).

Published and new data were combined for the analysis. The new data was obtained through
fieldwork transects, meetings with park rangers, scientist, teachers, farmers, ranchers and
special interviews to a singular and species-specific local group: the trackers (“tigreros”). A
logistic regression analysis was carried out using presence data and a group of possible
explanatory variables. This resulted in a habitat quality model for the jaguar. The problem of
autocorrelation was resolved by the application of a scale dependent - random labeling
methodology for point pattern analysis (Wiegand and Moloney, 2004). The performance of
the jaguar habitat model was evaluated with ROC analysis.

392
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Connectivity was managed as a species and landscape specific parameter. The jaguar was
used as the focal species to perform the assessment of connectivity among habitat patches
at regional scale. We applied percolation theory, graph theory and a range of analytical tools
based on this conceptual framework. These tools were a landscape ecological metric:
correlation length -C- and two derived indexes for patch evaluation applied to the analysis of
individual patch relevance.

Results

Figure 1. Jaguar habitat probability model.

The ROC analysis qualifies the model as reasonably good (AUC: 0.701). The connectivity
analysis outlined different possible configurations of high quality habitat patches connecting
the core areas. It would give options to the planner that could be managed as negotiation
alternatives in a land use planning participatory process.

References
Austin, M.P. (2002) Spatial prediction of species distribution: an interface between ecological theory
and statistical modeling. Ecological Modeling 157: 101-118.
Cullen, L. (2004) Jaguar as Landscape detectives for the Atlantic Forest. Brazil. 6th International
congress on wildlife management in Amazonas and Latin America. September, 5th - 10th, 2004,
Iquitos, Perú.
Garshelis, D.L. (2000) Delusions in habitat evaluation: measuring use, selection, and importance. L.
Boitani & T. K. Fuller (eds.) Research techniques in animal ecology: controversies and
consequences. Columbia University Press, New York, USA. pp. 111-164.
Naves, J., Wiegand, T., Revilla, E. and Delibes, M. (2003) Endangered species constrained by
natural and human factors: the case of brown bears in Northern Spain. Conservation Biology 17:
1276-1289.
Wiegand, T., and Moloney, K.A. (2004) Rings, circles, and null-models for point pattern analysis in
ecology. Oikos 104: 209-229.

393
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Harvesting: a keystone process in cover type evolution at landscape level?

Fourteen landscape histories in Quebec (Canada)

E. Alvarez1, L. Bélanger2, L. Archambault3 and F. Raulier2

Wood and forest sciences department, Forestry and geomatic faculty, Laval University, G1K
7P4, Quebec, Canada.
e-mail: [email protected]
2
Forestry and geomatic faculty, Laval University, Quebec, Canada.
3
Natural Ressources Canada, Canadian Forest Service, Quebec, Canada.

Introduction

The natural range of variability (NRV) concept is a useful tool in forest management
(Landres et al. 1999). However, despite a large consensus on the usefulness of this concept,
many authors in recent years have noted its limits.
The first limit is expressed by the key question "Which period of time should we use?”
Sprugel (1991) mentioned that taking the pre-colonial period may be misleading because of
climate changes. An idea defended by Millar and Woolfenden (1999).
Second, this method is mostly valid in an "equilibrium" or "shifting-mosaic steady
state" context such as a long fire-free period (Baker 1989; Spies and Turner 1999). In the
presence of catastrophic fires, this equilibrium may no longer exist (Baker 1989, Shinneman
and Baker 1997). As our study area, a temperate mixedwood forest in central Quebec
(Canada), has a long history of major natural disturbances, we have to be cautious because
we probably are in a non-equilibrium situation.
A useful complement to the NRV concept would be to keep the integrity of ecological
processes. Marcucci (2000) talked about "keystone process". He defined it as "(…) those
formatives processes that influence the trajectory of landscape change(…)”.
The objective of the study is to validate if harvesting has been a keystone process at
landscape level.

Methods

We monitored the evolution of forest cover for four periods of time (1946, 1957, 1976
and 1996) for fourteen landscapes that average 300 km2. For each landscape, we also
analyzed archival data to detail past natural disturbances and harvesting. Analysis was
based on the combination of three approaches. 1) We evaluated if 1996 landscapes were
within their natural range of variability. 2) We verified hypotheses on the evolution of forest
cover following harvesting. 3) We grouped landscapes based on three cluster analyses.
Cover type appellations were standardized as appellations changed over the years.
The area of each cover type was extracted from photo-interpretation databases. As we have
not done photo-interpretation ourselves, we had to consider an error in the percent area of
each cover type for each period of time considered. Unfortunately, we were unable to specify
a level of error as very few studies have been done on the subject. (Naesset 1999). Also,
natural disturbance areas were underestimated due to limitations of archival data.

Results

Natural disturbances have been so important that between the big fires of 1922/23
and 1996, only three landscapes have had much more harvesting than natural disturbances.
There were no cases for which all our hypotheses about the evolution of forest cover
following harvesting matched. As expected however, maximum concordance occurred for
two of the most harvested landscapes in the last forty years. The least respected hypothesis
was about intolerant hardwoods stands. They decreased instead of increasing.

394
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Hard(intolerant)-soft and jack pine stands were the most often outside the NRV.
In 1996, based on cluster analysis, up to 12 of the14 landscapes were in the same
group despite very different disturbance histories.

Discussion

We could not conclude that harvesting has been, globally, a keystone process at
landscape level for the period considered. Only jack pine stands responded fully to our
criteria. For hard(intolerant)-soft stands, harvesting impacts were additive to what was
already a natural tendency.
Based particularly on cluster analysis, it appeared that harvesting may have
homogenized the landscapes. At landscape level, some studies confirmed the idea that the
main effect of harvesting would be to modify the magnitude of the effects when compared to
natural disturbances (Weir and Johnson 1998, Boncina et al. 2003).
Based on hierarchy theory, we hypothesized that harvesting is not a constraint but rather
limits the displacement of landscapes inside their “constraint envelope” (O'Neill et al. 1989).

Management Implications

The main recommendation is to ensure that pure softwood stand (black spruce and
jack pine) proportions reach historical levels.

References
Baker W L (1989) Landscape ecology and Nature reserve design in the boundary Waters Canoe
Area, Minnesota. American Scientist 70(1): 23-35
Boncina A, Gaspersic F, Diacj J (2003) Long-term changes in tree species composition in the
Dinaric mountain forests of Slovenia. The Forestry Chronicle 79(2): 227-232
Landres P B, Morgan P, Swanson F J (1999) Overview of the use of natural variability concepts in
managing ecological systems. Ecological Applications 9(4): 1179-1188
Marcucci D J (2000) Landscape history as a planning tool. Landscape and Urban Planning 49: 67-81
Millar C I, Woolfenden W B (1999) The role of climate change in interpreting historical variability.
Ecological Application 9(4): 1207-1216
Naesset E (1999) Assessing the effect of erroneous placement of forest stand boundaries on the
estimated area of individual stands. Scandinavian Journal of Forest Research 14(2): 175-181
O'Neill R V, Johnson A R, King A W (1989) A hierarchical framework for the analysis of scale.
Landscape ecology 3(3/4): 193-205
Shinneman D J, Baker W L (1997) Nonequilibrium dynamics between catastrophic disturbances and
old-growth forests in Ponderosa Pine Landscapes of the Black Hills. Conservation biology 11(6):
1276-1288
Spies T A, Turner M G (1999) Dynamic forest mosaïcs. In: Hunter Jr M L (ed.), Maintaining
biodiversity in forest ecosystems, Cambridge University Press, Cambridge, pp 95-160
Sprugel D G (1991) Disturbance, equilibrium, and environmental variability: What is "Natural"
vegetation in a changing environment? Biological conservation 58: 1-18
Weir J M H, Johnson E A (1998) Effects of escaped settlement fires and logging of forest
composition in the mixedwood boreal forest. Canadian Journal of Forest Research 28: 459-467

395
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

At which spatiotemporal landscape scales does habitat amount explain the local
abundance of species?

H. Paltto1, A. Moilanen2, B. Nordén1

1
Department of Plant and Environmental Science, Göteborg University – Box 461, 405 30
Göteborg, Sweden.
e-mail: [email protected]
2
Metapopulation Research Group, Department of Biological and Environmental Sciences –
PO Box 65, 00014 Helsinki, Finland

Background

Empirical studies show that different habitat specialist species and organism groups
(taxa/guilds) respond differently to the amount of suitable habitat in the surrounding
landscape (Paltto et al., 2006). Studies including (1) many study sites, (2) habitat availability
within several spatial scales (e.g. formed as concentric rings around a focal patch) and
(3) more than one time scale (historical maps also considered) are scarce, but point to the
possibility that local species responses for a certain spatial scale change with changing
temporal scale. Since species dispersal is a distance per time-process, the strongest effect
of the amount of habitat in the surrounding landscape may be found at a scale corresponding
to the dispersal capacity of a species. Using a metapopulation modeling approach we
investigated how species with different life-history strategies respond to the amount of habitat
in the current and historic landscape under a scenario of ongoing habitat fragmentation.

Methods

A grid-based stochastic patch-occupancy model simulating colonisations and extinctions

of a species in a landscape was constructed. The model included a reduction of suitable
habitat during simulations. We run the model for a number of hypothetical species with
different dispersal and extinction characteristics (n=99 replicates). The species abundance in
the focal habitat patch (25 grid cells) at a certain fragmentation degree (mean 2%, 5% and
20%) was related to the amount of habitat in the current and past landscape at various points
in time (with 10 years interval) and at different spatial scales (circular areas around the focal
patch within 0-1 km, 1-5 km and 5-20 km radii).

Results

A hypothetical species with slow dispersal and extinction rates was strongly related to the
current amount of habitat at the smallest spatial scale (R=0.81; Figure 1a), while the
relationship to the largest scale was weak and strongly delayed in time (R=0.27; 240 years).
For another species (with higher dispersal rate but the same extinction rate) the intermediate
spatial scale (R=0,40; Figure 1b) became relatively more important and slightly delayed in
time (30-70 years) in comparison with the smallest scale. The relationship to the largest
scale was weak and strongest for the current landscape (R=0,14). A third species (good
dispersal ability and high extinction rate) was related to the smallest and intermediate spatial
scales (R=0.32-0.41 during 130 years BP) and very weakly to the largest scale.
Analyses of species in less fragmented landscapes (mean 20 %) showed roughly the
same pattern as mentioned in the paragraph above (mean 5 %), but with a slightly stronger
relationship to the intermediate spatial scale, a weaker relationship to the largest scale and
fewer time delays in species responses. Analyses of slow species in highly fragmented
landscapes (mean 2 %) revealed clear time delays in their responses (Figure 1d; 100 years
for the smallest, 50 years for the intermediate and 300 years for the largest scale).

396
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Figure 1. Spearman Rank Correlation coefficients for tests between habitat amount in
the landscape (circle radius 0-1 km ▲, 1-5 km □, 5-20 km ■; n=99) and species abundance
in a focal patch in the end of a run: (a) a species with slow dispersal and extinction rates, (b)
a species with slow dispersal and high extinction rate, (c) a species with high dispersal and
high extinction rate. (a, b, c: mean 5 % habitat in the end of the run), (d) same as (a) but in a
more fragmented landscape (mean 2 % habitat). The lines show the habitat amount as
proportion of the landscape: maximum, mean and minimum.

Conclusions and implications

Many empirical studies have explored the relationship between species occurrences/
abundances and habitat at landscape scale. Most of these studies include only a few
spatiotemporal scales and thus there is a risk – if wrong scales are studied – that an effect of
landscape connectivity is not detected. Our study shows that the relationship between the
local abundance of a species and the landscape depends on the scales included, and that
the response is dependent on the species dynamics. For the future it is of outmost
importance to conduct empirical studies at multi-spatiotemporal scales to complement our
results. It is also important that potential effects of scales not included in current projects are
discussed; otherwise the conclusions may be misleading.

References
Paltto, H.; Nordén, B.; Götmark, F. & Franc, N. (2006) At which spatial and temporal scales does
landscape context affect local density of Red Data Book and Indicator species? Biological
Conservation 133: 442-454.

397
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Modelling the effects of elevation and topographical-based behaviour on inter-

patch dispersal

J. Alderman1, S.A. Hinsley 2

1
Landscape and Biodiversity Research Group, School of Applied Sciences, University of
Northampton, Park Campus, Northampton, NN2 7AL, UK. email:
[email protected]
2
Centre for Ecology and Hydrology, Monks Wood, Abbots Ripton, Huntingdon,
Cambridgeshire, PE28 2LS, UK.

Introduction

One of the effects of fragmentation is to increase inter-patch dispersal distance,

increasing the risk of predation en-route. Perceptual range (the ability to detect habitat
patches at a distance) is suggested as one characteristic that would be advantageous to a
dispersing individual by providing a greater choice of patches, resulting in more direct
dispersal paths. Most real landscapes have topographical elements, such as hills, valleys
and urban developments, which can all act to modify a species’ perceptual range and directly
influence movement behaviour. However, the majority of individual-based spatially explicit
population models (SEPM) employ two-dimensional landscapes, which fail to model the
effects of topography. In two dimensions, all habitat patches within a species’ perceptual
range are generally detectable, but topographical features may substantially alter
‘detectability’, with the associated likelihood of erroneous and misleading predictions.

Using an individual-based Spatially Explicit Population Model (SEPM), the aim of this
study was to explore the effects of elevation and topography on inter-patch dispersal flow
paths. Previous work showed that spatial configuration of the landscape influenced
(modelled) flow paths (Alderman et al. 2005), thus it was thought likely that elevation would
also affect dispersal.

Methods

This study was based on an intensively farmed 40,000 ha (20 × 20 km) landscape in
eastern England, centred on Monks Wood (52°24’N, 0°14’W) in Cambridgeshire.
Approximately 4% (1,660 ha) of the area consisted of fragmented woodland. Ranging in
elevation from 0-70 m above sea level (ASL), several valleys run largely north-west to south-
east, with Monks Wood itself lying on a shallow north-facing slope. The nuthatch (Sitta
europaea) was used as a test species. Resident in the UK, the nuthatch is a small (c. 22-26
g) cavity-nesting, woodland passerine, which has been shown to be sensitive to forest
fragmentation. Dispersal from the natal patch is common in fragmented habitat (Enoksson et
al. 1995).

The model, called PatchMapper (Alderman et al. 2005), combines an individual-based

nuthatch population simulator with a grid-based representation of the landscape. Perceptual
range mechanisms were integrated into the inter-patch dispersal rules, allowing the disperser
to move directly to the nearest wood within the specified range and field-of-view.

To meet the aims of this study, three tests were devised. As a reference, the first test
modelled dispersal in a two-dimensional landscape, with no topographical data. To gauge
the effects of elevation on dispersal, the second test incorporated topography together with a
topographical-based behavioural response in the form of a valley-seeking algorithm. The
third test added an urban avoidance element to the dispersal rules of the second test. The
overall consequences of dispersal under each test were measured by recording their

398
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

respective effects on dispersal paths through a landscape of fragmented habitat patches and
an urban development, by recording population size in Monks Wood (located in the centre of
the landscape). All tests utilised 5,000 yearly cycles, with an immigration rate of 15 individual
birds per cycle, from each of the cardinal directions in turn. The tests were repeated for four
(modelled) perceptual ranges of 0, 0.2, 2 and 5 km.

Results and Discussion

Compared to the two dimensional version of the landscape, the inclusion of elevation and
topographical-based behaviour resulted in predicted population sizes that were strongly
dependant on landscape topography at lower perceptual ranges. This dependence
diminished as perceptual range increased, indicating that the scale at which an organism
interacts with its landscape is important. The three modelling scenarios gave significantly
different results, suggesting the existence of species and landscape dependant optimal
perceptual ranges. Thus, the perceptual behaviour of the same species may differ in different
landscapes. The urban development acted in a similar manner to a rock in a stream.
Dispersal flow densities increased ‘upstream’, with the flow being deflected to either side of
the development, meeting up again some distance ‘downstream’. Flow density was much
reduced in an area immediately behind the development.

During dispersal, the areal extent of the viewable landscape is continually updated. Thus,
a disperser can continually revise its movement decisions as previously unseen topography
comes into view. Visually detectable topography will also have a large influence on initial
choice of dispersal direction. It is likely, therefore, that the detectable topographical features
and other landscape elements of a landscape will all play some part in determining inter-
patch dispersal paths. Also, when presented with a choice of topographical elements, a
disperser may, for example, select the path of least energy consumption, such as a
reluctance to ‘hill-climb’ demonstrated by some species. Alternatively, a disperser may take
the path of least perceived risk, such as utilising the cover afforded by a hedgerow network
and the avoidance of unfamiliar habitat such as urban areas.

Conclusions

At the higher perceptual ranges, the traditional two-dimensional modelling method

consistently under-estimated population sizes, compared with the three-dimensional
approach. There were also significant differences between the three sets of inter-patch
dispersal paths. These findings suggest that using population predictions based on two-
dimensional landscapes to aid landscape management planning decisions may be
problematic. One example of this may be the construction of a wind farm. Such a
development may, depending on its location, deflect dispersal flow towards a particular patch
and increase its population or shield it from dispersal, and reduce the population. Thus, it is
recommended that both elevation and topographical-based behaviour, commensurate with
the studied species, be applied to the movement rule-base of dispersal simulation models.

References
Alderman, J; McCollin, D; Hinsley, S.A; Bellamy, P; Picton, P. & Crockett, R. (2005) Modelling the
effects of dispersal and landscape configuration on population distribution and viability in
fragmented habitat. Landscape Ecology 20: 857-870.
Enoksson, B; Angelstam, P. & Larsson, K. (1995) Deciduous forest and resident birds: the problem
of fragmentation in a coniferous forest landscape. Landscape Ecology 10: 267-275.

399
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Importance of riparian corridors for native fauna – small mammals in forest

patches, matrix habitats and linear structures in an Atlantic forest fragmented
landscape

L. Naxara 1, R. Pardini 2
1
Departamento de Ecologia and 2 Departamento de Zoologia, Instituto de Biociências,
Universidade de São Paulo, Rua do Matão – travessa 14, 101, São Paulo, SP, Brazil.
e-mail: [email protected]

Introduction and Methods

Landscape connectivity, defined as the amount of flow of organisms or genes in a

landscape, is influenced by the spatial configuration of remnants, matrix permeability and the
presence of stepping stones and corridors. Although the maintenance and restoration of
riparian corridors have frequently been the focus of conservation strategies and
environmental legislation in the tropics, and in the Atlantic forest in particular, studies on the
importance of these linear structures to native species are rare. Studies performed in other
regions of the world point not only to positive but also to negative ecological effects of
corridors, which can potentially be stronger in tropical forests given the clear edge-induced
habitat changes observed in this type of forest (Laurance et al., 2002). Here, we aim to
evaluate the importance of riparian corridors to endemic and invasive small mammals in an
Atlantic forest fragmented landscape.
Four similar systems were selected in a fragmented landscape in Tapiraí and Piedade,
São Paulo State, Brazil. Each of those encompasses two forest patches (20-36 ha)
connected by one riparian corridor (420-613 m in length and 50-90 m in width) and
surrounded by open areas of agriculture or pasture. In each system, we set four 50-m long
and 20-m wide grids of Sherman and pitfall traps, a 100 m from each other: one in the
interior of the forest patch, one at the edge of the forest patch, one in the corridor and one in
the open matrix. We carried out a 6-day sampling session every two months, totaling five
sessions and 30 sampling days in each of the16 grids in 2006. We compared small mammal
richness and abundance among habitat types using Repeated Measures ANOVA and Tukey
test. When necessary, data was ranked-transformed to attain homocedasticity.

Results and Discussion

We captured 555 individuals of 16 small mammal species, seven endemic to the Atlantic
forest and nine also found in open biomes. While the two most abundant species - Akodon
montensis and Oligoryzomys nigripes, which are not restricted to the Atlantic forest - were
captured in all sampled habitats, nine species - seven of which endemic to the Atlantic forest
- were captured only in forested habitats, three non-endemic species only in the matrix and
two non-endemic species in the matrix, corridors and edges. Although richness did not vary
among habitat types (F= 0.163, P= 0.918), total abundance showed a significant variation
(F= 4.000, P= 0.034), decreasing from the matrix through corridors, edges and interiors
(Figure 1). Nine species had more than 20 individuals captured and were analyzed
individually. The abundance of the two non-endemic, most abundant and widespread
species, A. montensis (F= 4.107, P= 0.032) and O. nigripes (F= 7.278, P= 0.004), varied
among habitats: the abundance of O. nigripes was lower in the matrix and interiors compared
to corridors and edges, and A. montesis was less abundant in interiors compared to the other
habitats (Figure 1). Whereas the abundance of the most common endemic species -
Brucepattersonius soricinus, which was not present in the matrix - varied significantly among
habitat types (F= 19.903, P< 0.001), being more abundant in corridors, followed by interiors
and edges, the abundance of the second most abundant endemic species - Marmosops
incanus, which is also absent from the matrix - was not significantly different among forested
habitats (F= 2.464, P= 0.112) (Figure 1). The two most abundant species in the matrix,

400
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Necromys lasiurus and Calomys tener, are savanna species and were exclusively captured
in this habitat. However, the other non-endemic species, Oligoryzomys flavescens, although
absent from interiors, varied significantly in abundance among habitats (F= 4.150, P= 0.031),
being more abundant in the matrix and corridors compared to edges (Figure 1).
Results point to a strong segregation in small mammal distribution between open areas
and forest remnants - including interiors and edges of patches and corridors, with only two
generalist species not restricted to the Atlantic forest being able to occupy all habitats types.
Open areas are dominated by invasive species from savanna biomes, which can attain
extremely high density increasing total small mammal abundance in those areas, whereas
riparian corridors and forest patches harbor endemic species. Corridors represent an
extension of forest patches for endemic species, since the abundance of none of them
declined from interiors or edges of patches to corridors, indicating the potential of corridors
for mitigating fragmentation effects for native small mammals. On the other hand, corridors
may function as a barrier to the expansion of invasive species, since most of those did not
occupy corridors. Our results indicate that riparian corridors are important not only for
increasing connectivity by providing high quality habitat for native species, as already
observed in a more forested landscape in Amazonia (Lima & Gascon 1999), but also for
preventing the expansion of invasive species in highly fragmented landscapes.

Figure 1. Small mammal richness and abundance in open matrix (M), riparian corridors (C)
and edge (E) and interior (I) of forest patches in four systems (different symbols) in a
fragmented Atlantic forest landscape in Brazil.

Acknowledgements

To FAPESP (05/56555-4, 05/57307-4), for financial support.

References
Laurance, W. F; Lovejoy, T.E; Vasconcelos, H L; Bruna, E. M; Didham, E. K; Stouffer, P. C;
Gascon, C; Bierregaard, R. O; Laurance, S. G & Sampaio, E. (2002) Ecosystem decay of
Amazonian forest fragments: A 22-year investigation. Conservation Biology 16: 605-618.
Lima, M.G. & Gascon, C. (1999) The conservation value of linear forest remnants in Central
Amazonia. Biological Conservation, 91: 241-247.

401
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Ecological Infrastructure: The Concept, Development and Application

H. L. Liu1 2, D. H. Li1 and X.L.Han1

1
Graduate School of Landscape Architecture, Peking University, Beijing, China 100871 e-
mail: [email protected].
2
Department of Landscape Architecture, Tsinghua University, Beijing, China 100084.

The concept of ecological infrastructure originally emerged in the 1980s, and several
theories and ideas facilitated its later development. Basing on available literatures, a
systematic review of five aspects can be made.
Ecological infrastructure in Conservation Biology
Mander (1988) and Selm (1988) have laid the foundation of ecological infrastructure
concept simultaneously. Obviously, since late 1960s when Landscape Ecology and Island
Geography theories grew up, the recognition of ecological corridors and other key landscape
elements in bio-diversity conservation boosted the formation of this idea. Actually,
considerable similar ideas (Ahern, 1995), such as environmental corridors (Lewis, 1964),
Habitat Networks (Noss & Harris, 1986), landscape linkages (Harris & Gallagher, 1989),
Dutch Ecological Main Infrastructure (Bohemen, 2002), National Ecological Network of
Netherlands (Ahern, 1995) and EECONET (Jongman, 1995; Bennett, 1994) etc, have
indicated that the nature conservation is changing from specie-centered and site protection
approaches to ecosystem-oriented ones focusing on the integral conservation infrastructures
(Jongman, 1995).
Ecological infrastructure in ecosystem service study
As Constanza mentioned the minimum level of ecosystem infrastructure and a certain
critical ecosystem structure (Constanza et al. 1997, from Fromm, 1999), it is obvious that the
original intention of eco-economist’s studies is to explore the value of ecosystem services
provided by the earth’s ecological infrastructure (World Resources Report, 1998-1999).
Accordingly, it is an important progress to combine ecosystem service with basic ecological
structure (Yu & Li, 2002, 2003). Ecological infrastructure serves the bridge connecting
ecosystem service with landscape planning, so people can take operable measures to
safeguard and conserve the sustainability both of nature and people.
Ecological man-made infrastructure
Flink (1997) has classified infrastructure into three types: man-made physical, natural
and social ones and man-made physical infrastructures have caused enormous impacts to
natural system and ecological processes. According to Williamson (2003), infrastructure
reflects the social priorities of diverse cultures, so the solution for today is green/ecological
infrastructure. Scientists and engineers try to solve problems including fragmented habitats,
constrained floodplains and impermeable surface by making compensation and restoration of
ecosystem service and enhancing the landscape continuity (Cuerus et al. 1993, from Seiler &
Eriksson, 1995; Bohemen, 1998). A number of cities in North America and Europe are active
to implement such kind of programs (Benedict & McMahon, 2002)).
Ecological infrastructure and open space planning
Parks and green spaces have served as fundamental infrastructure to solve urban
problems such as congestion and filth. Based on Randolph (2004), concepts such as Park,
Park System, Greenbelt, Green Heart, Parkway, Open space and Greenway, evolve
gradually and their objectives varied in different stages. Ecological infrastructure, being
continuous green space network, is increasingly recognized as a new strategy of open space
planning and land conservation Schneekloth, 2003; Randolph, 2004; Yu & Li, 2001,
2002 and has been advocated in 1999 Report of the American President’s Council on
Sustainable Development as nation’s natural life support system etc (Williamson, 2003).
Ecological infrastructure and the planning methodology reformation
Ecological infrastructure concept has driven the reformation of planning methodology.
Honachefsky (1999) stressed on the value of making preservation of the ecological

402
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

infrastructure the priority of the municipal master plan and Benedict et al. (2002) advanced a
framework for smart conservation and smart growth that indicates land conservation should
be concert with rather than isolated or opposite to development. New ideas, such as Smart
Growth and Regional City etc, partially hold such points (Benedict et al. 2002).Yu and Li
(2001, 2002) have explored comprehensive urban ecological infrastructure strategy and the
Negative Planning approach to guarantee urban sustainability in China.
To sum up, ecological (green) infrastructure is the critical natural system on which the
sustainability is based and from which residents, not only the wildlife but also the people, can
continuously enjoy the benefits of ecosystem services.

References
Mander, U. & Jagonaegi, J. et al. (1988) Network of compensative areas as an ecological infrastructure
of territories. Connectivity in Landscape Ecology, Proceedings of the 2nd International Seminar of
the International Association for Landscape Ecology, Ferdinand Schoningh. Paderborn, 35-38.
Selm, A, J. V. (1988) Ecological infrastructure: a conceptual framework for designing habitat
networks. In Schrieiber. K. -F. (ed), Connectivity in Landscape Ecology, Proceedings of the 2nd
International Seminar of the International Association for Landscape Eclogy. Ferdinand Schoningh.
Paderborn, 63-66.
Ahern, J. (1995) Greenways as a planning strategy. Landscape and Urban Planning, 33 131-155.
Yu, K.J. (1996) Security patterns and surface model in landscape planning. Landscape and Urban
Planning, 36 (5): 1-17.
Jongman, R.H.G. (1995) Nature conservation planning in Europe: developing ecological networks.
Landscape and urban planning, 32:169-183.
Jongman, R.H.G., & Mart Külvik, Ib (1995) Kristiansen, European ecological networks and
greenways, Landscape and Urban Planning 68 :305–319.
Benedict, M., & McMahon, E. (2002) Green Infrastructure: Smart Conservation for the 21st Century.
The Conservation Fund. Washington, DC: Sprawl Watch Clearinghouse. Retrieved June 13, 2003.
Flink, C.A. (1997) Greenway: Serving Infrastructure Needs in the 21st century. Paper Greenways
Incorporated, North Carolina.
Fromm, O. (2000) Ecological Structure and Functions of Biodiversity as Elements of Its Total
Economic Value, Environmental and Resource Economics, 16: 303-328.
Yu, K.J., Li, D.H. & Chao, L.M. (2001) Ten landscape Strategies to Build Urban Ecological
Infrastructure, Planner, 6: 9-17.
Yu, K.J. & Li, D.H. (2002) On Negative-planning Concept and Ecological Infrastructure, Proceedings
of Green City in Hangzhou, Beijing: China Fine Art. School Press.

403
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Impacts of landscape fragmentation on breeding habitats of Oriental White Stork

in the Sangjiang Plain, China

Liu HongYu1, Li Zhaofu2

1
College of Geography, Nanjing Normal University.122, Ninghai Road, Nanjing,
10097,China.
e-mail: [email protected]
2
College of Resources and Environmental Sciences, Nanjing Agricultural University.
Weigang Road, Nanjing 210095, China.

Introduction

Since the fragmentation of natural habitats is one of the most serious problems for many
species, it is highly interesting to study the impacts of landscape fragmentation on habitats of
endangered species (Farina, 1998). The Oriental White Stork (Ciconia boyciana), as a
representative wetland waterfowl in the Sanjiang Plain, China, has sharply declined in
numbers over the last decades apparently due to forest harvesting and fragmentation of
natural wetland landscapes (Li Xiaomim, 2002; Liu Hongyu, 2005). Conservation measures
have led to the establishment of a wetland reserve network to assure the maintenance of the
Oriental White Stork at a regional scale. However, little is known about the effects of
landscape structure and fragmentation of landscape on the habitats and the distribution of
Oriental White Stork.

Results

Distribution of breeding habitat of Oriental White Stork

Six breeding habitat variables were tested by field surveys conducted in spring 2004 and
2005 before they were used to establish a HSI model for evaluating the breeding habitats of
Oriental White Stork and their dynamic distribution over the past decades in the study area.
Distributions of the suitable breeding habitats were identified by integrating the HSI model
into GIS environments. Results show that the breeding habitats were originally distributed in
almost all of the area of the Sanjiang Plain in 1954, but habitats in the west-southern part
have been lost after the first large-scale reclamation when large area of wet meadows and
marshes were lost. The main suitable distribution area was located in the Northeast part from
1966 to 1983. But, the remaining habitats were only distributed in the northern part along
Heilongjiang River and eastern area along the Wusuli River from 1993 to 2005. Our results
also showed that the landscapes were in matrix states (wetland total area > 50% and
connectivity > 50) in periods of 1954~1976. This result suggested that the historical periods
of wetland changes could be divided into the matrix state (1954~1976) and fragmented state
(1983~2005) in a landscape scale.

Patch characteristics and dynamics

The sharply declined trend in areas for suitable breeding habitats of storks is clearly
shown in Table1. The total habitat area decreased 81.9% by 2005 with 73.6% of wetlands
being lost in the study area. The main reason for the change may be the significant variation
of patch characteristics and their sizes in landscapes. Our results show that the maximum
patch size of wetlands accounted for 86.3%~70.9% of the total wetland area in 1954~1983
and more than 15% of patches of wetlands were occupied. But only 9%~1.5% of wetland
patches were occupied after the wetland was fragmented in the period between 1983~2005.
This phenomenon indicates that larger patches have more positive effects on habitat quality
and patch characteristics have greater importance in maintaining higher quality habitats for
Oriental White Storks in the study area. Table 1 shows that larger patches have higher

404
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

diversity than small patches and suggests that habitat heterogeneity within a larger patch of
wetland could therefore provide more suitable and safety reproduction area for Oriental
White Storks.

Effects of connectivity of wetlands on habitats

Landscape connectivity had consistent positive effects on breeding habitats (Table.1).

Habitat connectivity is more than 50 in state of landscape matrix of wetlands between
periods of 1954~1976 which showing more than half of wetlands could be occupied for
breeding habitats by storks. But habitat connectivity had seriously declined when the
landscape matrix changed from wetland into farmland after 1983. Furthermore only 1.4% of
wetland patches would be occupied in 2005. This suggests that isolation by distance
between patches was also a significant factor affecting quality of breeding habitats for
oriental white storks, but the effect of distance was accentuated by fragmentation of the
landscape. Isolated patches may be absolute barriers for storks because the maximum
distance from a nesting patch to a feeding area is usually less than 2km or 3km in the study
area.

Table 1. Characteristics of habitat patches in historical periods in the Sanjiang Plain,China

Habitat/ Area (%) Mean patch size Diversity of Patch Patch
Wetland (km2) Max. patch (H) number connectivity

1954 40.8/71.1 178.6/67.7 3.63/3.63 104/478 69/88

1966 26.5/55.0 65.5/36.4 3.12/3.12 184/688 62/71
1976 23.8/50.7 70.9/22.7 2.98/2.98 153/1016 50/56
1983 19.4/37.4 61.2/11.5 1.56/1.66 144/1480 27/19
1993 14.8/22.3 27.0/4.4 0.95/1.48 249/2321 23/24
2005 7.4/18.8 21.4/0.76 0.86/1.36 157/11272 19/15

Discussion

Data on patch size and distance between patches is generally easy to obtain, and there
is no doubt that these variables are often important to explain the distribution of a species.
However, for studies of distribution and persistence of little-known species in patchy habitats,
it is prudent to evaluate both patch characteristics and landscape context of habitat patches
whenever possible. Efforts should be made to evaluate relevant patch characteristics and
factors that determine connectivity for such species.

Acknowledgements
This work was carried out with financial assistance provided by National Natural Science
Foundation of China (40471003).

References
lmo Farina.(1998) Principles and methods in landscape ecology. Chapman & Hall Ltd., pp.58~68.
Hanski I. & O. Ovaskainen. (2000) The metapopulation capacity of a fragmented landscape. Nature,
404:755-758.
Li Xiaomim. (2002) Wild animal resources and their changes. Liu Xingtu & Ma Xuehui. The Natural
Environmental changes and their conservation in the Sanjiang Plain. Science Press, pp232-234.
Liu Hongyu; Lu Xianguo; Zhang Shikui et.al.(2005) Fragmentation process of wetland landscape in
watersheds of the Sanjiang Plain. Chinese Journal of Applied Ecology, 16(2): 289~295.

405
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

A graph-based methodology for integrating habitat connectivity in landscape

conservation planning: application to the capercaillie in Catalonia (NE Spain)

L. Pascual-Hortal and S. Saura

Forest Landscapes and Biodiversity Research Group. Department of Agroforestry

Engineering. Higher Technical School of Agrarian Engineering. University of Lleida. Av.
Rovira Roure, 191. 25198 Lleida. Spain.
e-mail: [email protected]

Introduction

Large-scale habitat loss and fragmentation are critically hindering the dispersal and
survival of many species. The integration of landscape connectivity criteria in conservation
planning applications is therefore a need for preserving the biodiversity and the ecological
functions of natural ecosystems. Nevertheless, there is a lack of operative methodologies to
effectively tackle this problem.

Connectivity analysis through habitat availability indices and graph structures

We propose a practical and quantitative methodology for landscape connectivity analyses

(figure 1), based on graph structures and algorithms, GIS operations and new habitat
availability indices (integral index of connectivity, probability of connectivity) that overcomes
several computational and decisional limitations of other existing approaches. The new
developed indices present improved properties in the sense of being both sensitive to
different types of negative changes that can affect the landscape mosaic and effective
detecting which of those changes are more critical for its conservation from a spatial analysis
point of view (Pascual-Hortal and Saura, 2006a, 2006c). The habitat availability concept
considers a habitat patch itself as a space where connectivity exists, integrating habitat
abundance and connectivity between patches in a single measure; we suggest that the
connectivity problem should be considered within the wider concept of habitat availability in
order to be useful for conservation planning (Pascual-Hortal and Saura, 2006b). The
application of graph structures and algorithms is a powerful and effective way of overcoming
computational limitations that appear when dealing with large data sets and performing
complex analysis regarding connectivity (Urban and Keitt, 2001).

We developed a new version of the Conefor Sensinode software (version 2.2) as an

analysis support tool for identifying and ranking by importance the habitat patches that most
contribute to the maintenance of landscape connectivity. This software has been developed
by modifying, reprogramming and including new indices and features in the Sensinode 1.0
software (LandGraphs package) developed by Dean L. Urban (Duke University). Although
being developed under a complex theoretical and mathematical basis, Conefor Sensinode
2.2 is easy to use and interpret by users (not knowledge on graph-theory required). A free
copy of this software can be obtained by contacting the authors.

Conservation planning application to the capercaillie habitat in Catalonia

For demonstrative purposes, we apply this methodology to the analysis of the

endangered capercaillie (Tetrao urogallus) habitat in Catalonia (NE Spain). The habitat
distribution data was obtained from the Catalan Breeding Bird Atlas (Estrada et al., 2004),
which provides the probability of occurrence of this bird for each 1x1 km UTM square
covering Catalonia. By applying the new-developed index probability of connectivity (PC), we
determine the connected subregions existing within the landscape and the key habitat areas
for the connectivity of the habitat mosaic for this species. Furthermore, on the basis of these

406
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

results, we identify the public forests holding larger amount of high-priority sites for
capercaillie connectivity and evaluate the effectiveness of different protected areas networks
(e.g. Natura 2000) in this sense. We conclude providing specific recommendations for the
management of capercaillie habitat in these key locations and highlighting the potential and
practical interest of this methodology for successfully integrating connectivity in landscape
conservation planning applications.

SPECIES

Habitat requirements Dispersal

and distribution characteristics
GIS

Map of habitat patches Graph algorithms

(Spatial dataset)

GIS CONEFOR SENSINODE 2.2

software
Edge-to-edge distances
among patches Habitat availability index: PC
(Euclidean or minimum-cost)

Patches prioritization Overall

(relative patches la n d s c a p e
importances) connectivity

Map of prioritary areas for conservation GIS

(critical for connectivity maintenance)

Figure 1. Diagram of the methodology for the analysis of landscape connectivity and the
Conefor Sensinode 2.2 software in which it has been implemented.

References
Estrada, J.; Pedrocchi, V.; Brotons, L & Herrando, S. (2004) (Eds). Atles dels ocells nidificants de
Catalunya 1999-2002. Lynx Edicions, Barcelona, Spain.
Pascual-Hortal, L. & Saura, S. (2006a) Comparison and development of new graph-based landscape
connectivity indices: towards the prioritization of habitat patches and corridors for conservation.
Landscape Ecology 21: 959-967.
Pascual-Hortal, L. & Saura, S. (2006b) Integrating landscape connectivity in broad-scale forest
planning through a new graph-based habitat availability methodology: application to capercaillie
(Tetrao urogallus) in Catalonia (NE Spain). European Journal of Forest Research DOI
10.1007/s10342-006-0165-z.
Pascual-Hortal, L. & Saura, S. (2006c) Integrating landscape connectivity in broad-scale forest
planning: a methodology based on graph structures and habitat availability indices. R. Lafortezza
& G. Sanesi (Eds). Patterns and processes in forest landscapes: consequences of human
management. Proceedings of the IUFRO Landscape Ecology Conference. Academia Italiana di
Scienzi Forestali.Locorotondo, Bari, Italy, pp. 111-116.
Urban, D. & Keitt, T.H. (2001) Landscape connectivity: a graph-theoretic perspective. Ecology 82:
1205-1218.

407
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

Refining the fragmentation paradigm: the over-riding influence of land-use history

on present day ecosystems in fragmented agricultural landscapes

PG Spooner, ID Lunt

School of Environmental Sciences, Charles Sturt University, PO Box 789 Albury, NSW 2640,
Australia.
e-mail: [email protected]

Throughout the world, the development of agriculture has resulted in the rapid
transformation of continuous ecosystems to landscapes dominated by crops and pastures,
within which small remnants of the original vegetation are retained (e.g. Hobbs & Saunders,
1994; Yates & Hobbs, 1997a; de Blois et al., 2001, 2002). The analogy between remnant
patches in agricultural landscapes and islands in seascapes has enabled island
biogeography theory (MacArthur & Wilson, 1967) to be extended to agricultural landscapes
worldwide (Burgess & Sharpe, 1981; Gilpin & Soulé, 1986; Hanski & Simberloff, 1997).
However, a recent review of fragmentation experiments (Debinski & Holt, 2000) revealed that
most fragmentation studies exhibited species richness patterns contrary to predictions based
upon island biogeography theory. Levenson (1981), for example, showed that species-area
relationships were incapable of predicting plant species abundances. The contradictory
nature of many studies has led to ongoing questions about the utility of the fragmentation
paradigm (Simberloff & Abele, 1982; Gilpin & Soulé, 1986; Usher, 1987; Robinson et al.,
1992; Hanski & Simberloff, 1997; Davies et al., 2001; Watson, 2002; Hobbs & Yates, 2003;
Lindborg & Eriksson, 2004).
Landscape ecology considers spatial and temporal interactions across the landscape and
the influences of spatial patterns on biotic and abiotic processes (Levin, 1992; Wiens, 1997;
Baker, 1999; Turner et al., 2001). The discipline emerged from attempts to develop an all-
encompassing fragmentation paradigm (Turner, 1989; Naveh & Lieberman, 1990; Wiens,
1997; Naveh, 1998). Despite general acceptance by plant and animal ecologists, landscape
ecology has often been criticised for: (1) concentrating mainly on spatial patterns and
ignoring temporal changes (Hobbs, 1999; Turner et al., 2001), and (2) tending to exclude
people from the landscape (Hammett, 1992; Kirkpatrick, 1999). Recent studies have placed
greater emphasis on changes in functional aspects of remnant ecosystems in fragmented
landscapes (e.g. Saunders et al., 1991; Hobbs et al., 1993; Kearns et al., 1998), and have
highlighted that some landscapes are better described as variegated rather than fragmented,
since the landscape matrix is not necessarily hostile to all species (McIntyre & Barrett, 1992;
McIntyre & Hobbs, 1999). However many studies are still conceptually based on traditional
patch area–isolation concepts (Wilcove et al., 1986; Laurance & Bierregaard, 1997).
The apparent contradictory nature of many fragmentation studies may simply reflect the
short time frame of observations, which often cannot encompass the many indirect
feedbacks that occur in anthropogenic disturbed landscapes (Debinski & Holt, 2000).
Agricultural landscapes are characterised by dramatic changes on varying timescales. Some
short term changes are predictable in periodicity (such as annual cultivation of crops) and
even type (i.e. change in crops), but due to hysteresis (lag effects), longer term changes in
fragments are poorly understood (MacDonald & Smith, 1990). For example, many tree
species may live for centuries, and since adult plants are often less sensitive to changes in
environmental conditions than juvenile stages, current patterns may simply represent a slow
process of gradual extinction (Adamson & Fox, 1982; Hobbs, 1987; Bennett, 1993; Foster,
2000; Saunders et al., 2003). An alternate hypothesis for interpreting the contradictory nature
of many fragmentation studies is the over-riding imprint of spatially variable, historical
anthropogenic disturbances, which selectively advantage some taxa and disadvantage
others (Wilcove et al., 1986). Unfortunately our ability to detect the effects of short and long
term processes may be clouded by what Dovers (2000) describes as our ‘amnesia’, or
ignorance of past human events and their residual effect on landscapes.

408
Theme 3. Ecological Networks, fragmentation and connectivity
3.8 Open Session 13: Habitat Fragmentation and Mitigation

‘Historical ecology’ has been cited as a new paradigm in which ecologists view
ecosystems as historically and spatially influenced non-equilibrium systems that are complex
and open to human inputs (Hammett, 1992; Balée, 1998; Swetnam et al., 1999; Egan &
Howell, 2001). This framework recognises that: (1) historical events play a major role in
ecology, as changes in abiotic conditions or species composition that happened in the past
can have large and often irreversible effects on the structure and dynamics of present day
ecosystems (de Blois et al., 2001, 2002), and (2) remnant ecosystems may exist in various
non-equilibrium states due to a complex history of anthropogenic disturbance regimes
(Naveh, 1998; Pyne, 1998; Motzkin et al., 1999; Lugo & Gucinski, 2000; de Blois et al.,
2002). Historical ecology shares a similar approach to the ‘environmental history’ discipline,
since both are interdisciplinary, however the latter tends to focus on the development of a
narrative driven ‘story’ (Bowman, 2001; Griffiths, 2002), whilst historical ecology is more akin
to traditional biogeography in its attempts to quantify human relationships with the landscape.
Building on disturbance ecology, the historical ecology framework recognises that
landscape elements may have evolved with human inputs to such an extent that
abandonment of human interference may lead to impoverishment of structural and biological
diversity (Solon, 1995; Naveh, 1998; Kirkpatrick, 1999; Ross et al., 2002; Spooner et al.,
2004a, b). This approach highlights that not all human activity leads to degradation, and that
humans are an integral component of landscape dynamics (Gragson, 1998; Naveh, 2000;
Peterson, 2000). Agricultural landscapes contain remnant ecosystems that have evolved
from different historical management regimes, economies and political policies (Ihse, 1995).
By detailing the history of human activity and resulting changes in vegetation structure,
composition and pattern, we can examine biotic responses to novel anthropogenic
disturbance processes and devise appropriate conservation strategies (Reed, 1990; Foster
et al., 1998; Naveh, 1998; Swetham et al., 1999; Bowman, 2001; Miller & Hobbs, 2002; Ross
et al., 2002; Foster et al., 2003).
The aim of this paper is to enhance current attempts to understand patterns of
biodiversity in fragmented agricultural landscapes, by using an historical ecology perspective
to highlight the over-riding influence of land-use history in creating past, current and future
patterns of biodiversity across a range of spatial scales. We present a series of conceptual
postulates to enhance our understanding of biodiversity patterns in ways that cannot be met
by existing frameworks. We then illustrate each of these postulates with examples, largely
drawn from fragmented woodlands in south-eastern Australia. We summarise the key
implications of these postulates, and conclude by discussing the potential advantages of
using this approach to guide the development of future studies of biodiversity patterns in
other agricultural landscapes.

409
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

3.9 Posters

Seed rain in rainforest fragments: relating seed and fragment features

F.M. Jesus1, P.R. Pivello1, S.T. Meirelles1, J.P.Metzger1

1
University of São Paulo, Institute of Biosciences, Department of Ecology, Rua do Matão,
Travessa 14, Cidade Universitária Armando de Salles Oliveira, São Paulo, SP, Brazil
05508-900
e-mail: [email protected]

Introduction

Seed dispersal is one of the most important processes for forest dynamics (van der Pijl
1982). The arrival of diaspores in forest fragments through seed rain is a key-process to
direct gene flow, recruitment rates, species turnover, and therefore, the community
composition and structure (Venable & Brown 1993). Besides this, the current composition of
plant communities allows for predictions on the community course, as its development or
recovery. However, in fragmented landscapes the spatial arrangement of remaining forest
patches may affect the seed flux according to habitat connectivity and patch size. Therefore,
our objectives were to investigate the influence of patch size and habitat connectivity on seed
rain composition, and to compare tree species composition and seed rain within patches.

Methods

We analyzed nine Atlantic rainforest fragments surrounded by an agricultural matrix (São

Paulo, Brazil) being three small and isolated, three small and connected to three large
fragments. Seeds were collected during 12 months through 108 nylon seed traps (0.5 m2)
arranged as triplets in the centre of each forest patch. Seeds were classified according to
dispersal mode (autochoric, anemochoric, zoochoric), species life form (tree, shrub, vine,
epiphyte), and functional group (early successional, late successional). Tree species in the
fragments were estimated inside 10 m2 plots around seed traps. The independence of seed
rain composition upon fragment features was tested by G-test (Yates correction, Zar 1996);
species similarity between the arboreal community and seed rain was verified with Jaccard
indices (Magurran 1988).

Results

We registered 20,142 seeds belonging to 115 species. The small-isolated patches

accounted for more than 50% of total seeds, the highest amount of anemochoric and early-
successional seeds. They also showed a high amount of epiphyte seeds, but a low amount
of liane seeds (Table 1). Large patches accounted for 17.6% of total seeds and the highest
amount of arboreal, zoochoric, autochoric, and late-successional seeds (Table 1). The small-
connected patches accounted for 32% of total seeds, and the highest amount of shrub seeds
(Table 1), being all other parameters intermediate between the values of large and small
patches. The small-isolated patches showed the lowest diversity, while the large ones
showed the highest equability (Table 2). Tree species similarity between seed rain and the
corresponding fragment was very low, gradually increasing from the small-connected to large
and to small-isolated (Fig.1).

Discussion

The results suggest that fragment size and connectivity affect seed rain composition. The
small-isolated patches seem to receive more wind-dependent seeds of initial successional

410
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

trees, mostly coming from nearby. Animals use more intensely the large patches, dispersing
later-successional seeds. A high influence of external factors such as wind in the small
patches may explain the very low similarity between seed rain and the arboreal community.
Small fragments, however, do contribute with the seed flow in the landscape.

Table 1. Values of G (G-test with Yates correction) and p (probability at 0.95) resulting from
comparison between the studied patches (SI= small-isolated patch; SC= small-connected
patch; L= large patch)

SI x SC SI x L SC x L
anemochoric G(0,05;gl=1)=596,64;p<0,00 G(0,05;gl=1)=860,87;p<0,00 G(0,05;gl=1)=68,48;p<0,00
authocoric G(0,05;gl=1)=53,55; p<0,00 G(0,05;gl=1)=4,79; p<0,02 G(0,05;gl=1)=33,29; p<0,00
zoochoric G(0,05;gl=1)=558,76; p<0,00 G(0,05;gl=1)=797,9; p<0,00 G(0,05;gl=1)=61,89; p<0,00
epiphyte G(0,05;gl=1)=212,17;p<0,00 G(0,05;gl=1)=87,82;p<0,00 G(0,05;gl=1)=1,35; p<0,24
shrub G(0,05;gl=1)=682,17;p<0,00 G(0,05;gl=1)=317,59; p<0,00 G(0,05;gl=1)=15,17;p<0,00
arboreal G(0,05;gl=1)=976,71; p<0,00 G(0,05;gl=1)=459,73; p<0,00 G(0,05;gl=1)=6,39; p<0,00
vine G(0,05;gl=1)=780,33; p<0,00 G(0,05;gl=1)=15,17; p<0,00 G(0,05;gl=1)=0,04; p<0,83
early-successional G(0,05;gl=1)=4873,73;p<0,00 G(0,05;gl=1)=73,04;p<0,00 G(0,05;gl=1)=1973,72; p<0,00
late-successional G(0,05;gl=1)=6,11; p<0,01 G(0,05;gl=1)=31,86; p<0,00 G(0,05;gl=1)=11,09; p<0,00

Table 2. Tree species richness and diversity indices in nine rainforest patches (São Paulo
State, Brazil) according to seed rain composition (n= number of seeds; S= species richness;
H’= Shannon´s diversity index; J= Pielou’s eqüability index; SI= small-isolated patch; SC=
small-connected patch; L= large patch)
Patch identification
CM TE DI LU MA AL TA PE PD
(SI) (SI) (SI) (SC) (SC) (SC) (L) (L) (L)
n 1209 3456 5499 2059 1671 2702 1432 667 1447
S 33 47 32 53 35 39 27 39 38
H’ 0,803 0,768 0,335 0,963 0,694 0,652 0,575 0,946 0,868
J 0,529 0,459 0,223 0,559 0,450 0,410 0,402 0,595 0,550

UPGMA T(I)
T(I)
CM(I)
CM(I)
L(C)
L(C)
A(C)
A(C)
Ps(L)
Ps(G)
P(L)
P(G)
D(I)
D(I)
Tk(L)
Tk(G)
M(C)
M(C)
0,28 0,4 0,52 0,64 0,76 0,88 1
Jaccard's Coefficient
Coeficiente de Jaccard

Figure 1. Dendrogram resulting from a MVAR (minimum variance criterion) cluster analysis
(Jaccard’s coefficient), and seed rain data of the analyzed patches.

References
Magurran, A. E. 1988. Ecological diversity and its measurement. Princeton University Press,
Princeton, New Jersey.
Van der Pijl, A. 1982. Principles of dispersal in higher plants. 2nd edition. Springer-Verlag, Berlin.
Venable, D.L. & Brown, J.S. 1993. The population-dynamic functions of seed dispersal. Vegetatio,
107/108: 31-55.
Zar, J. H. 1996. Biostatistical analysis. 3. ed. Nova Jersey, Prentice Hall.

411
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Agro-environmental structures in the countryside. Their role in local bird

movements within the Sile River Natural Park

D. Tocchetto, M. Borin

University of Padova, Department of Environmental Agronomy and Crop Production,

Agripolis, 35020 Legnaro (PD), Italy.
e-mail: [email protected]

Introduction

Within the framework of the National Research Programme “Ecological networks and
agriculture” begun in 2004, a pilot area was established in the Sile River Natural Park, in
Veneto Region, N-E Italy. Due to its geographical position, the Park is an important area for
biodiversity conservation. The habitats host a wide diversity of species of both flora and
fauna, many of which are rare and protected.

Methods

The research focused on assessing the presence and multiple roles of ecological network
structures in a farming area inside the Park. Three adjacent zones were identified, each
differing in shape, hydrology, history, farming activities, agro-environmental structures and
other aspects. At a local scale each landscape element, i.e. fields, hedgerows, ditches, field
roads, woods, etc., is part of and has a function in an ecological network system (as core
areas, corridors, stepping stones, buffer zones). To evaluate the functions related to animal
life, bird movement surveys were conducted in 2006. It is well known that landscape
structures have direct influences on bird populations. Small woods, isolated trees, hedges,
farm fields, etc., allow and encourage the movement of birds.
Four observation sites were chosen in the study area, each differing in land use, farm
management and neighbouring hedgerow structure and length. The bird surveys consisted of
30 minutes of direct watching in the observation site. During the survey period the following
flight movements were noted: i) high above the site, ii) from hedgerow to field, iii) between
two or more hedgerows, iv) from plant to plant (within the hedgerow), v) leaving the site, vi)
flying into the field, vii) flying into the hedgerow. Four surveys were conducted from June to
September 2006.

Results and conclusions

The observations show high mobility where there are many hedgerows and the land use
is varied. More than two thousand movements were counted during the surveys.
All the movement categories considered occurred in sites 1, 2 and 3, with an average of
2.8, 2.0 and 1.9 counts per 30 minutes, respectively. Due to the agro-environmental features
many movements were noted from the hedgerows to fields, between hedges, and flights
from outside into the hedges or fields. Site 4, located in a large maize-growing area, totalled
0.9 counts per 30 minutes. There are no elements in this site (such us wheat grain, berries,
etc.) to attract birds and more than 84% of the movements were “flying high above the site”.
The results show that the agro-environmental elements had a strong influence on the
birds daily activities and movement within the study area and in the nearby zones. These
movements depend both on the season and the specific features of the countryside.

412
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Applying the bidirectional cost-distance theory to the connectivity analysis of

forest bird species richness in a fragmented landscape in the Andean mountains

J. Aubad1. C. Martinez1. H. Gulinck2. M. A. Rodriguez1.

1
Departamento de Ecología, Universidad de Alcalá, 28871 Alcalá de Henares Madrid,
Spain; e-mail: [email protected]
2
Department of Land Management and Economics. Katholieke Universiteit Leuven. Belgium.

Forest fragmentation is one of the major ecological problems worldwide and specially in
areas with extraordinary species richness and endemicity. Among others, the connectivity
concept was introduced as a tool in the search for solutions, not only theoretical but also
practical, for fragmentation impacts and biological richness loss.

The study landscape is located in the Central Andean Mountains of Colombia and is
currently under evaluation by the State environmental office for conversion into a regional
conservation area.

The selection of the variables was in accordance with a previous model based on
Generalized Linear Models (GLZ). To apply this model we registered bird species of the
natural forest patches. We also recorded all landcover types that surround forest patches to
compare bird species richness similarity. This method allows the quantification of each
landcover polygon of the landscape in the resistance map.

We used two maps for the cost distance analyses; a forest sources map and the resistance
map for the input model. The output image of the cost distance analysis is a cost map with
values that surround the specific sources, where the value of each cell represents the
distance to the source and the displacement effort. We compared the scenarios of 2001 and
2021, a hypothetical landcover map that includes changes if current conservation concerns
persists. Using the bidirectional cost distance (BCD) algorithm we generated multiple values
and topological structures as a support basis for the design of ecological habitat-corridors in
this landscape.

The bidirectional cost distance model (BCDM) assumes the simple addition of the cost
surface calculated from a point A to the cost surface from another point B. Therefore BCDM
is the sum of a cost-map from one source to another, plus a cost-map from the other to the
first source, where the result is a continuous raster map with a main route created by the low
cost values (BCD Route)

Conclusions

BCD models allow the quantification of changes in connectivity surfaces. These models
estimate the effects of the current conservations strategies on future bird species richness of
Andean mountain forests. BCD models show consistent results that facilitate decision
making at this landscape level and the design of natural forest conservation areas.

413
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Direct and indirect effects of habitat heterogeneity on flower-visiting insects and

associated species interactions

T. Diekötter1,2, K.J. Haynes1, D. Mazeffa1, T.O. Crist1

1
Department of Zoology, Miami University, Oxford, Ohio 45056, USA. e-mail:
[email protected]
2
Justus-Liebig-University, IFZ - Department of Animal Ecology, Heinrich-Buff-Ring 26-32, D-
35392 Giessen, Germany

The direct effects of landscape structure on individual species or functional groups may be
transferred indirectly through webs of interacting species. To date, however, most studies on
the effects of habitat size and spatial arrangement have focused primarily on responses of
population size or diversity. In a factorial field experiment, we explored the direct and indirect
effects of habitat area, habitat fragmentation, and matrix composition on a community of
flower-visiting insects in a red clover (Trifolium pratense) agroecosystem. Field surveys
revealed matrix-dependent changes in the visitation rates of pollinators in clover patches.
The number of pollinator visits per clover inflorescence was greater in clover patches
embedded within a bare-ground matrix than in patches within a matrix composed of orchard
grass (Dactylis glomerata). This effect of the matrix on visitation rates was attributed to larger
numbers of pollinators and visits per individual pollinator in patches surrounded by bare
ground. Structural equation modeling indicated that these changes in pollinator visitation
propagated across a tri-trophic system of clover seeds, seed predators, and their wasp
parasitoids: higher visitation rates very likely increased seed set, which in turn seemed to
have resulted in higher abundances of seed predators and their parasitoids. Our results
strongly suggest that interactions among species can be influenced not only by internal patch
characteristics but also by landscape context. The very likely pollinator-mediated matrix
effect on seed-seed predator-parasitoid interactions observed in this study emphasizes the
importance of considering direct and indirect interactions in studies on community structure
and ecosystem functioning. Our findings also suggest that habitat spatial structure and matrix
effects may have important influences on ecosystem services such as pollination.

414
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Habitat Networks in England

R.D.J. Catchpole

Natural England, Science and Evidence Team, Bull Ring House, Northgate, Wakefield, West
Yorkshire, WF1 3BJ.
e-mail: [email protected]

Introduction
Over the last 65 years there has been an extensive fragmentation and loss of semi-
natural habitat beyond protected areas in the UK. The documentation of these changes, in a
series of studies, has been described by one author as a “grim litany of outrage and
complaint” (Adams, 2003 p.5). What has been largely ignored, however, is how the pattern
of land use intensification has affected the functional isolation of this remaining habitat. While
it has been possible to define patch locations, no evaluation of patch context has been
undertaken, in spite of the recognised importance of this factor on site integrity (Riffell et al.
2003, Ricketts, 2001, Goodwin & Fahrig, 2002). In response to this need, areas where the
functional connectivity of terrestrial habitats might still remain were defined so that
conservation action could be targeted and wider networks of habitat maintained.

Methods
Functional connectivity was evaluated by ‘least-cost’ methods which used expert
judgements to define relative movement cost estimates for a ‘generic focal species’
associated with four broad habitat types. The habitats consisted of deciduous woodlands,
heathlands, grasslands and mires, fens and bogs. They were derived from modified national
habitat inventories that were first published in 2004 (English Nature). The matrix was derived
from remotely sensed (Landsat Thematic Mapper) land cover data (Land Cover Map 2000,
ITE 2000).

Results
For each broad habitat, three nested GIS layers were produced across the whole of
England to indicate potential movement envelopes at each dispersal interval. The degree of
overlap between different network types, number of patches within networks and their extent
varied widely. In more intensively managed landscapes, single patches with small movement
envelopes were typical. At a strategic level, it was possible to identify more extensive areas
of landscape where habitat networks might be maintained.

Conclusions
While further work clearly needs to be undertaken to test this generalisation, the approach
offers a pragmatic solution to the question of where land use managers might seek to
maintain and enhance functional connectivity. It is envisaged that any action that seeks to
increase functional connectivity will be the exception rather than the rule.

References
Adams, W.M. (2003). Future Nature: a Vision for Conservation. Earthscan, London.
Riffell, S.K., Keas, B.E., Burton, T.M. (2003). Birds in North American Great Lakes coastal wet
communities: is landscape context important? Landscape Ecology 18: 95-111.
Ricketts, T.H. (2001). The matrix matters: effective isolation in fragmented landscapes. American
Naturalist 158: 87-99.
Goodwin, B.J. & Fahrig, L. (2002). Effect of landscape structure on the movement behaviour of a
specialised goldenrod beetle, Trirhabda borealis. Canadian Journal of Zoology 80: 24-35.
English Nature (2004). National UKBAP priority habitat inventories. English Nature, Peterborough.
ITE (2000). Land Cover 2000, Institute of Terrestrial Ecology, United Kingdom.

415
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Prioritizing forest restoration based on late-seral habitat connectivity

W.H. Richards

Seattle Public Utilities - 19901 Cedar Falls Rd SE, North Bend, Washington, United States.
e-mail: [email protected]

Introduction

The cost of implementing forest restoration projects, in terms of both time and money, in
all areas that would likely benefit usually outstrips the resources available for such projects.
When the goal of restoration is based on habitat, the question arises of how to prioritize
restoration based on the connectivity of that habitat (Richards et al., 2002). This study
addresses prioritizing late-seral forest restoration on a watershed scale in Washington State,
USA.

Steps of Analysis

Step 1.
Target forest characteristics that would likely benefit ecologically from restoration.
Step 2.
Spatially locate forest stands with targeted forest characteristics.
Step 3.
Create “base landscape” by simulating forest growth over the planning period.
Step 4.
Create “alternative landscape(s)” based on potential effects of restoration.
Step 5.
Simulate dispersal of late-seral forest dependent wildlife species in both landscapes. The
PATCH model (Schumaker, 1998) simulates wildlife dispersal based on:
1) the distribution of habitats on a landscape (see Step 4),
2) the affinity of wildlife species for habitats,
3) home range size (smaller home ranges mean bigger potential populations),
4) mortality during dispersal (increased mortality with increased distance), and
5) dispersal turning probabilities (increased probability when in proximity to habitat).
Step 6.
Compare spatially explicit dispersal activity (see Step 5) between landscape alternatives
to identify areas that most benefit habitat connectivity.

References

Richards, W.H., D.O. Wallin, and N.H. Schumaker. 2002. An analysis of late-seral forest
connectivity in western Oregon, U.S.A. Conservation Biology 16(5):1409-1421.

Schumaker, N.H. 1998. A user’s guide to the PATCH model. EPA/600/R-98/135.

Environmental Research Laboratory, U.S. Environemental Protection Agency, Corvallis,
OR. Available from: http://www.epa.gov/wed/pages/models/patch/patchmain.htm.

416
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Developing ecological networks – a functional approach

M.C. Fessey, N. Bailey

Spatial Ecology and Land Use, Oxford Brookes University, Oxford, UK.
E-mail: [email protected]

Introduction

This study applies and builds upon recently developed techniques for identifying optimum
areas for habitat creation so as to reduce the fragmentation of important habitat in the
landscape. The chosen study area is the Milton Keynes and South Midlands (MKSM) sub-
region in the East Midlands of England. The MKSM was identified by the UK Government
Sustainable Communities Plan in 2003 as one of 4 major growth areas in the South East of
England. The planned house building in the MKSM could lead to funding being made
available for habitat creation that serves the multifunctional purposes of green infrastructure.

Methodology

A GIS is developed to model the functional connectivity of existing patches of woodland,

wetland and grassland habitat. Functional connectivity is determined not just by the distance
between habitat patches, but also by the permeability to species movement of the
intervening landscape matrix (this can be thought of as a resistance surface). A range of
Generic Focal Species (GFS) are developed and used as a surrogate for species specific
behavioural information that this type of model would normally demand. A major challenge is
finding large scale data-sets that can reflect, relatively accurately, the real life heterogeneity
in the resistance surface. This study makes use of Land Cover Map 2000 as a base for the
development of the resistance surface as well as using qualitative descriptions from a
national landscape characterisation study.

At this point the model is useful in that it can identify where newly created habitat patches
could benefit ecologically by being functionally connected to another (currently isolated)
habitat patch or existing ecological network. This study is novel in that it then develops the
model further to identify where habitat creation might link existing isolated habitat patches, or
existing functionally connected networks, so as to create large scale ecological networks.

417
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

A multiscale framework for analysis of species and population distribution and

persistence in heterogeneous landscapes

M. L. Lorini1234, V.G. Persson123 , I. Garay13 and J. Xavier-da-Silva23

1
Laboratório de Gestão da Biodiversidade, Instituto de Biologia, Universidade Federal do
Rio de Janeiro (UFRJ), CCS, Bloco A, Sala 100, Ilha do Fundão, Rio de Janeiro (RJ), Brasil,
e-mail: [email protected]
2
Laboratório de Geoprocessamento, IGEO, UFRJ. 3 Programa de Pós-Graduação em
Geografia, IGEO, UFRJ. 4 CAPES/ PETROBRAS.

Developing and maintaining sustainable landscapes became a major challenge in the XXI
century. However, to face this challenge we must deal with interactive complex systems (abiotics,
biotics, cultural and socioeconomics) that reflect spatial and temporal heterogeneities in multiple
scales (Wu et al., 2006). From a theoretical and conceptual point of view, the task of facing
landscape complexity may be benefited by approaches as Hierarchical Patch Dynamics (Wu,
1999) or Cognitive Landscape and Eco-field (Farina and Belgrano, 2006). Implemented in a GIS.
In this conception, we developed a methodological multiscale framework to analyze pattern and
process of distribution and persistence of species, populations and habitats in heterogeneous
landscapes. Our study-case represents one of the 25 most endangered primates in the world, the
Black-faced Lion Tamarin (Leontopithecus caissara Lorini and Persson, 1990), which inhabits the
Brazilian Atlantic Forest, one of the 25 global biodiversity hotspots. The integrative and spatial
basis of the framework design combine multiple tools from RS and GIS; Machine-learning;
Ecological Niche Modelling; and Population Viability Analysis. A hierarchical multiple-scale
modeling and scaling strategy followed the “scaling ladder” approach (Wu, 1999; Burnett and
Blaschke, 2003) in order to assess the distribution and persistence of L. caissara from the global to
local scales. A spatially nested hierarchical system with distinctive scaling domains or levels of
organization was constructed using top-down (partitioning) and bottom-up (aggregation) schemes,
resulting in a three level architecture (tab. 1).

Table 1. Hierarchical multiscale architecture.

hierarchical level organization level spatial extension spatial grain
+1 distribution continental/biome/ecoregion 500m -1km
0 metapopulation regional landscape 30 - 90m
-1 Population local landscape 1 - 30m

This architecture is the framework for making observations and developing models at focal
levels, and then for extrapolating information across the domains of scale hierarchically. We
suggest that scaling ladder approaches may contribute to address questions like “how the dynamic
of abiotic, biotic and anthropic factors relate to landscapes’ spatial and temporal heterogeneity”;
“how this affects the dynamics of biodiversity distribution at local, regional and global scales” or
“how the dynamics of natural and human disturbances reflect on the evolution of biodiversity
geographic limits and landscapes’ transformation” and “which are the implications for processes as
species fragmentation, differentiation, extinction and invasion”.

References
Burnett, C. & Blaschke, T. (2003). A multi-scale segmentation/object relationship modelling methodology for
landsacpe analysis. Ecological Modelling 168: 233-249.
Farina, A. & Belgrano, A. (2006). The eco-field hypothesis: toward a cognitive landscape. Landscape
Ecology 21(1): 5-17
Wu, J. (1999). Hierarchy and scalling: extrapolating information along a scaling ladder. Can. J. R.. Sensing 25
(4): 367-380.
Wu, J.; Jones, B.; Li, H. & Loucks, O.L. (Eds). (2006). Scaling and uncertainty analysis in Ecology.
Springer, Dordrecht, The Netherlands.

418
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Multiple-scale analysis of species filtering processes and their determinants.

L. Blank, Y. Carmel

Division of Environmental, Water and Agricultural Engineering, Faculty of Civil and

Environmental Engineering, Technion – Israel Institute of Technology, Haifa 32000, Israel.
e-mail: [email protected]

Understanding the factors regulating the diversity of ecological communities is one of the
greatest challenges of community ecology. The processes generating the observed patterns
of species diversity can be conceptualized as 'filters' that determine the derivation of local-
scale species assemblages from the species pools of larger scales (Hillebrand, 2002,
Lawton, 1999, Zobel, 1997). These filters, represented by climatic conditions, disturbance
regime and biotic interactions, result in specific local assemblages that are non-random
subsets of the regional species pool (Schmid et al., 2002). Furthermore, these filters operate
along a range of scales resulting in different species composition at different spatial scales.
This 'filtering' concept has been used as a conceptual framework of ecological communities
(Keddy, 1992), yet empirical applications of this concept in natural communities have usually
been limited to only two scales: a 'local community' and a 'regional species pool'. The
distinction between these two scales is usually arbitrary. Moreover, the identity and relative
importance of the 'filters' responsible for the observed relationships were often left unknown.
In order to advance our understanding of the regional-local relationship of biodiversity we
developed a nested, hierarchical sampling design which enables us to analyze patterns of
biodiversity at spatial scales ranging from a few meters to thousands of square meters
(Noda, 2004). The sampled biodiversity that represent producers, herbivores, detritivores
and carnivores are analyzed with respect to three potential filters: soil, topography, and
woody vegetation spatial pattern.
Our preliminary results indicate that there is no apparent impact of woody vegetation
spatial pattern on woody species in the local-scales (1 m2-100 m2). However, at meso-scales
(103 m2-106 m2) woody vegetation exert high impact on woody species.
We believe that such a spatially hierarchical, multi-factor analysis of species diversity and
filtering will contribute significantly to understanding of the role of scale in determining the
organization of ecological communities.

References
EN.REFLIST

419
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Influence of inter-habitat gap size on the dispersal pattern of Xiphorhynchus

fuscus (Aves, Passeriformes, Dendrocolaptidae) in the Brazilian fragmented
Atlantic forest.

D. Boscolo, C.Candia-Gallardo, M. Awade and J.P. Metzger.

Departamento de Ecologia, Instituto de Biociências, Universidade de São Paulo - USP, Rua

do Matão, trav. 14, nº 321, Cid. Universitaria, São Paulo, Brazil, zip: 05508-900
e-mail: [email protected]

Introduction

In fragmented landscapes, habitat patches are usually isolated from each other by a
matrix environment unsuitable for several species (Wiens 1995). This may lead to severe
isolation and reduced habitat connectivity, hindering the movement of individuals between
patches and incurring in ineffective dispersion, high extinction risk and low species
persistence time (Lindenmayer et al. 1999). These processes are generally linked to the skill
of dispersing individuals to overcome the matrix and reach nearby patches (Brooker &
Brooker 2001). However, this ability is not known for several species living in fragmented
habitats. Our objective was to collect information on how inter-habitat gaps may impede the
flux of Xiphorhynchus fuscus, a bird species found in fragmented Brazilian Atlantic forest.

Methods

Twelve individual birds were captured inside six forest fragments using mist nets. After
tagging them with micro radio transmitters, the birds were translocated to varying distances
(50 to 260 m) across the landscape divided into two paired treatments: translocation inside
the capture patch and to a nearby fragment isolated by open fields. Birds were followed by
telemetry for three consecutive days or until they returned to their origin patch (displacement
across the matrix) or to the vicinity of the capture point (displacement within the forest).

Results and Discussion

The mean return time of birds displaced across the matrix was significantly higher than of
those inside the forest (F=11.02, p<0.01). Birds were able to cross at once habitat gaps of up
to 150 m. Larger distances across the matrix incurred in longer return time (p=0.04) being
overcome only with the use of isolated trees. This species can disperse through fragmented
habitat, but this ability is limited by increasing inter-patch distances. The implementation of
small steeping stones in the open matrix may improve landscape connectivity for this
species, enabling the individuals to reach more isolated patches.

References
Brooker, M. & Brooker, L. (2001) Breeding biology, reproductive success and survival of blue-
breasted fairy-wrens in fragmented habitat in the western Australian wheatbelt. Wildlife Research
28: 205-214.
Lindenmayer, D.B; McCarthy, M.A. & Pope, M.L. (1999) Arboreal marsupial incidence in eucalypt
patches in south-eastern Australia: a test of Hanski’s incidence function metapopulation for pacth
occupancy. Oikos 84: 99-109.
Wiens, J.A. (1995) Habitat fragmentation: island vs landscape perspectives on bird conservation. Ibis
137: 97-104.

420
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

The role of connectivity on species-area relationship: evidences from birds in the

Atlantic Rainforest

E. Hasui, M. C. Ribeiro, A. C. Martensen, R. G. Pimentel, J. P. Metzger

Landscape Ecology and Conservation Laboratory – LEPAC, Bioscience Institute, University

of São Paulo, Rua do Matão - travessa 14 no 321 CEP 05508-900,
São Paulo, Brazil.
e-mail: [email protected]

The Species-Area Relationship (SAR), S = c Az, has stimulated the interest of

naturalists since the early 1900’s and remained central in our understanding of biodiversity
patterns. Despite the generalities of the SAR predictions under different landscape
configurations, these predictions remain vague. A review of empirical studies had suggested
huge variations and frequent underestimations of richness in the remaining patches. The
species loss function, based on SAR, uses solely the proportion of habitat lost to estimate
the proportion of species expected to become extinct without considering any configuration
aspect where the habitat was removed. Here, we hypothesized that the slope (expressed by
the “z” value) of SAR can be modulated by connectivity following different possibilities: (A)
Higher influence of connectivity for intermediate size patches; (B) From Lomolino (2000)
hypotheses of sigmoidal species-area relationships, it was predicted that “small patches
effect” may correspond to a range o patch size where the influence of connectivity is
especially higher; (C) Linear models may be pointed out as the best models, showing a
constant rise in the species-area relationship and also connectivity influence along the entire
gradient of patch size; (D) Effects of connectivity depending of the size of the fragment. We
used data from previous studies of bird species richness obtained with similar survey
methods (point counts) in 73 fragments (median 39.4 ha; range 3.42 – 339.34 ha) localized
in the Southeast Atlantic Rainforest of Brazil. The connectivity of each fragment was
calculated using proximity indices as well as indices based on graph theory. The relationship
between species richness estimation (fixed by number of individual sampled among
fragments) and fragment sizes were evaluated with regression models applying the best fit
using linear, logarithmic, power, piece-wise and exponential models. Results pointed out the
regression model using multiple linear regression as the best model (r2 adj. = 0.24, p =
0.00003), and there was no evidence of asymptote in the small patches. Preliminary
connectivity analysis, considering distances as 60 m best explain influence on species-area
relationship. There was evidence to support the effect of connectivity depends of the size of
the fragment.

References
Lomolino, M. V. (2000). Ecology's most general, yet protean pattern: the species-area relationship.
Journal of Biogeography 27 (1): 17-26.
Lomolino, M. V. (2001). The species-area relationship: new challenges for an old pattern. Progress in
Physical Geography, 25 (1): 1-21.

421
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Species richness and frequency of large and medium mammals related to

landscape parameters in an agro-forested region (São Paulo State, Brazil)

G. Ciocheti, L.R. Tambosi, M.C. Lyra-Jorge, M.C. Ribeiro, V.R. Pivello

LEPaC–Laboratório de Ecologia da Paisagem e Conservação, Departamento de Ecologia,

Universidade de São Paulo, Rua do Matão, trav. 14, Cidade Universitária, CEP: 05508-
900, São Paulo, SP-Brasil.
e-mail: [email protected]

Introduction
Habitat loss has been changing the structure of landscapes, isolating populations, and
leading to reduced resources for fauna. To implement actions towards the protection of
native fauna species it is fundamental to understand how such structural changes in the
landscape affect their behavior. Having as a prime goal the conservation of large and
medium mammals in a fragmented landscape, our objective in this study was to relate the
activity patterns of such mammals with the present landscape structure.

Material and Methods

In a 50,000 ha area (47°45’W-21°38’S; São Paulo State, Brazil) composed of remaining
forest and savanna patches surrounded by sugarcane and eucalyptus plantations, we
selected seven patches – being two of woodland savanna, two of typical savanna, two of
semideciduous forest, and one of eucalyptus monoculture – to install 20 camera traps (the
number of camera traps was proportional to patch area) which were monthly checked,
during 12 months. Species richness and frequency of large and medium mammals were
then recorded.
We generated a land use map (scale 1:15,000) and calculated some landscape structural
indices for the study region (patch size, patch shape, edge contrast; landscape connectivity
and proximity) using the Fragstats program.

Results
Table 1. Spearman correlation coefficients between the mammal species frequency and
richness and the landscape metrics tested. (PasP = perimeter of pasturelands adjacent to
the focal patch; PasA = total area of pasture adjacent to the focal patch; EucA = total area
of eucalyptus plantation adjacent to the focal patch; Area = total area of the focal patch;
PARA = relationship between the focal patch perimeter and area; Prox [proximity index] =
sum of patch area divided by the nearest edge-to-edge distance squared between the focal
patch and all patches of the same type whose edges are within a specified distance of the
focal patch; ENN = euclidian distance between the focal patch and other patches of the
same type.)
Chrysocyon Puma Leopardus Didelphis Mazama Myrmecophaga Species
brachyurus concolor pardalis albiventris guazoubira tridactyla Pecari tajacu Richness
EucA NS NS NS NS NS NS NS 0.87623
PasP 0.7641 NS NS NS NS NS NS NS
PasA 0.809 NS NS NS NS NS NS NS
Area 0.8469 NS NS NS NS NS NS 0.770934
PARA -0.8289 NS NS NS NS NS NS -0.899423
Prox NS NS 0.775 NS 0.810 NS NS NS
ENN NS NS NS NS NS NS 0.8568 NS

Discussion
Although this study did not count on replicates and the results cannot be generalized, it
indicates that structural changes in the landscape may affect the behavior of large/medium
mammals.

422
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

The importance of patch size, connectivity and annual variation to understanding

the effects of fragmentation on leaf litter frogs in the Atlantic Rainforest.

M. Dixo, J.P. Metzger

Laboratório de Ecologia da Paisagem e Conservação da Natureza. Depto. de Ecologia, IB-

USP.
e-mail: [email protected]

Introduction
Connectivity and fragment size are major parameters of the landscape structure to
maintain species in fragmented landscapes (Fahrig, 2003). Few studies have focused on
amphibians and reptiles (McGarigal and Cushman, 2002). In the present study, we
investigated the effects of fragment size and corridor presence over two years on total
abundance, alpha diversity, beta diversity, evenness, and abundance of frog species in the
Atlantic Rainforest, one of the most diverse and threatened biomes in the world.

Methods
Leaf litter frog community of the Atlantic Plateau of São Paulo (Brazil) was sampled using
pitfall traps in 20 forest fragments and six sites within a continuous area, the Morro Grande
Reserve. Fragment size varied between 2 to 276 ha and conditions of connectivity varied
depending on the presence or absence of corridors.

Results and discussion

The community of frogs responded negatively to forest fragmentation. Fragments had
fewer species than the continuous forest and they were not as evenly distributed.
Furthermore, small isolated fragments were more instable and hyper-dynamic than larger
fragments and control areas. Although the fragmented landscape preserved a large number
of species, forested landscapes help conserving the anuran community more intact, richer
and stable. Contrary to expectations, the presence of corridors was not associated with
greater richness and/or abundance of frogs in the connected fragments, but beta diversity
was greater in isolated fragments, indicating that the presence of corridors makes the
connected fragments more homogenous. This result reinforces the thesis that small
vertebrates are able to utilize even narrow corridors, (Lima and Gascon, 1999).

Conclusion
This study highlights the importance of creating extensive public reserves, like Morro
Grande Reserve. Furthermore, corridors connecting fragments may buffer temporal
variations, providing greater stability to frog community. Additionally, our study demonstrated
the importance of considering temporal variations to understand the response of frog
community to fragmentation.

Acknowledgements
Support for this work was provide by FAPESP - (99/05123-4; 01/07916-3), Fundação O
Boticário de Proteção à Natureza (0564-20022) and CNPq (690144/01-6).

References
Fahrig, L. (2003) Effects of habitat fragmentation on biodiversity. Annual Reviews in Ecology,
Evolution and Systematics 34: 487-515.
Lima, M.G, Gascon, C. (1999) The conservation value of linear forest remnant in central Amazonia.
Biological Conservation 91: 241-247.
McGarigal, K., Cushman, S.A. (2002) Comparative evaluation of experimental approaches to the
study of habitat fragmentation effects. Ecological Applications 12: 335-345.

423
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Predicting badger (Meles meles) sett suitability across England and Wales

T.R. Etherington

Central Science Laboratory, Sand Hutton, York, YO41 1LZ, UK.

e-mail: [email protected]

Introduction

The occurrence of bovine tuberculosis in cattle herds across England and Wales has been
spreading at an exponential rate in recent years, and is now a serious economic problem.
The badger (Meles meles) is suspected to be a major vector for spreading this disease
amongst cattle herds. In order to conduct disease modelling to try and understand the
problem, an estimate of badger distribution and abundance is required. Knowledge of sett
suitability across the landscape would be an important component of any badger population
estimate.

Modelling Approach

A large number of setts surveyed across southwest England were used as the basis of the
sett suitability model. The landscape variables related to the sett data are known to influence
sett location (Neal and Cheeseman, 1996) and were created at two scales. The first scale
referred to conditions at the sett itself and included elevation, slope, cover and soil
conditions. The second scale described conditions around the sett and included the amount
of improved grassland. All data were referenced to 100×100m grid cells in line with the
precision of the sett location data.
The Mahalanobis D2 statistic (Jenness, 2003; Rotenberry et al., 2006) was used to
develop the sett suitability model as it avoids the need for absence data. This is
advantageous as sett locations are not spatially independent of one another, therefore
absence of a sett does not necessarily mean the area is unsuitable. The Mahalanobis D2
statistic also allows for extrapolation, which was required as the sett suitability model
developed was ultimately applied across England and Wales.

Conclusion

The approach used shows great promise for predicting the suitability of a landscape for
sett locations, and that it is possible to achieve this at both a fine grain and over a large
extent. However, the success of the extrapolation of the model outside of the southwest of
England where all the sett data was collected remains untested.

References
Jenness, J. (2003) Mahalanobis distances extension for ArcView 3.x, Jenness Enterprises. Available
at http://www.jennessent.com/arcview/mahalanobis.htm
Neal, E. & Cheeseman, C. (1996) Badgers. Poyser, London.
Rotenberry, J.T.; Preston, K.L. & Knick, S.T. (2006) GIS-Based Niche Modeling for Mapping
Species’ Habitat. Ecology 87: 1458-1464.

424
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Scale does matter!

N.C. Brouwers, A. Newton

Bournemouth University - School of Conservation Sciences, Talbot Campus, Fern Barrow,

Poole, Dorset, BH12 5BB,
e-mail: [email protected]

Introduction

Current woodland conservation policy stresses the importance to preserve, expand and
re-connect habitat fragments on a landscape scale. However, Dolman and Fuller (2003)
suggest that more studies are necessary on woodland specialists species to provide a firmer
basis for current management strategies. To address this knowledge gap, research was
undertaken on a specialist woodland invertebrate, the wood cricket (Nemobius sylvestris) on
the Isle of Wight, UK. In 2005, a landscape scale survey and in 2006 a more detailed study
within 3 separate woodland fragments was undertaken. The results of both studies were
used to assess the relevance of a landscape scale approach for conservation effort of this
species.

Method

In 2005, all mature broadleaf dominated woodland complexes, larger than 5 hectares, on
the northern part of the island were surveyed. Wood cricket presence or absence was
recorded together with several patch variables. In 2006, within the 3 selected woodlands, 1x1
meter grids were developed recording wood cricket presence and a series of habitat
variables. For each, several distance measures were computed using ArcGIS (version 9.1).

Results

The distribution of the species on a landscape scale and within woodlands showed a
similar patchy pattern. For the landscape scale, a significant positive relation was found for
the probability of wood cricket being present and woodland (patch) area, and a negative
relation with distance to the nearest neighbouring inhabited woodland. Within woodlands a
significant positive relation was found for the probability of wood cricket being present and
leaf litter volume, and negative relations with vegetation cover, vegetation height, South
oriented canopy closure and nearest neighbour distance to an inhabited permanent
woodland edge. On both scales nearest neighbour distance revealed the highest explanatory
power for wood cricket presence.

Discussion

Results reveal the importance of different explanatory factors (related to wood cricket
presence) when looking at different scales. Therefore, for wood cricket (and possibly others
woodland specialist species), conservation effort within woodlands seems equally relevant
besides the current landscape scale initiatives.

References
Dolman, P.M. & Fuller, R.J. (2003) The processes of species colonisation in wooded landscapes: a
review of principles. J. Humphrey, A. Newton, J. Latham, H. Gray, K. Kirby, E. Poulsom & C. Quine
(Eds). The Restoration of Wooded Landscapes. Forestry Commission, Edinburgh, 25-36.

425
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Distribution of large terrestrial mammals in an Atlantic Forest fragmented

landscape: Does mesopredators benefit from deforestation?

K.D. Espartosa1 and R. Pardini2

1
Departamento de Ecologia and 2Departamento de Zoologia, Instituto de Biociências,
Universidade de São Paulo, Rua do Matão, travessa 14, 101, CEP 05508-900, São Paulo,
SP, Brazil.
e-mail: [email protected]

Introduction and Methods

While large mammals require large areas, present low population densities, are frequently
hunted and thus commonly suffer local extinctions, some generalist medium-sized mammals
are common in anthropogenic landscapes and may lead to cascade effects, such as
increasing bird nest predation. However, factors that influence mesopredator distribution are
poorly known. Using a 10-day sampling with 3 camera traps in 24 forest patches in a 10,000-
o o
ha Atlantic Forest landscape (23 50'S; 47 20'W), we investigated if the distribution of
mesopredators is influenced by the amount of forest at a local scale (800 m circumference)
using logistic and linear regressions. We expect that mesopredators are favored by
deforestation and human activities, being present in sites with low amount of forest.

Results and Discussion

Four mesopredators were the most common species registered, representing 91% of the 330
total records. The common opossum, Didelphis aurita, was the most widespread specie
present in 17 sites, followed by the nine-banded armadillo, Dasypus novemcinctus,
registered in 13, the domestic dog, Canis familiaris, in 12, and the crab-eating fox,
Cerdocyon thous, in 7. The variation from 10 to 70% of forest surrounding sites did not
influence the opossum, armadillo and fox occurrence or abundance. However, both domestic
dog occurrence (-2LL= 28.388; P= 0.027) and abundance (R2= 0.172; P=.0.024) was
negatively affected by the amount of forest (Figure 1). The presence of this exotic species
inside remnants may lead to negative consequences, not only through the transmission of
diseases, but also through predation (Butler et al., 2004). Our results corroborate that
mesopredators benefit from altered landscapes and that deforestation and the consequent
increase in human activities favor the arrival and proliferation of at least one exotic
mesopredator, which probably increase predation rates and alters the structure of small
vertebrate assemblages in small isolated remnants.

Figure 1. Variation in the abundance of mesopredators as a function of the amount of

surrounding forest around 24 Atlantic forest sites in Brazil.

References
Butler, J.R.A.; du Toit, J.T. & Bingham, J. (2004) Free-ranging domestic dogs (Canis familiaris) as
predators and prey in rural Zimbabwe: threats of competition and disease to large wild carnivores.
Biological Conservation 115: 369–378.

426
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Effects of landscape composition and physical characteristics of the land on the

biological control of aphids
1,2
N. Roullé, 1 E. Lucas, 2 G. Domon, 2 J. Ruiz
1
Biological Sciences Department, University of Quebec in Montreal, Montreal, Quebec, H3C
3P8, Canada.
e-mail: [email protected]
2
School of Landscape Architecture and Chair in Landscape and Environmental Design,
University of Montreal, Montreal, Quebec, Canada.
Introduction
Studies on the natural control of pests by natural enemies are generally done at small scales,
such as a single plant, a plot, or more rarely a whole farm. Effects at larger scales remain
poorly studied. The objective of this study was to examine the effect of the spatial structure on
biological control of aphids by their natural enemies. The specific objectives were (1) to
determine if the landscape composition and the physical characteristics of the land derived
from an ecological classification are predictive of biological control of aphids by their natural
enemies and (2) to identify landscape composition variables that had a significant effect on
biological control.

Materials and Methods

Information provided by the ecological classification (Cadre Écologique de Référence) allowed
for the selection of 17 circular areas of 1 km in diameter with different landscape structure (type
and structure of land use), within the same watershed (Assomption stream, Québec, Canada). In
the corn field at the centre of each of these areas, we monitored aphid populations from June 7
to September 28, 2005. Land use was mapped for each of the 17 circular areas. We also
recorded information about the agricultural practices.
We tested relationships between insect data and landscape data using variation partitioning
redundancy analysis. In a first step, in each of the three groups of variables (landscape
composition, physical characteristics of the land, agricultural practice), we selected the
variables which contributed significantly in explaining variation in aphid abundance. In a
second step, we partitioned variation among the three groups of variables. In a third step, we
partitioned variation among the selected landscape variables.

Results
Variable selection:
Landscape composition variables that have a significant effect on the abundance of aphids
are: forested area, river area, riparian vegetation area, and area of strawberry fields.
Moreover, there is significantly less aphids in «sandy terrace».
None of the agricultural practices significantly affected aphid abundance.

Variation partitioning: landscape composition – physical characteristic - agricultural practice

Taken all together the selected variables explain 77% of the total variation in aphid abundance (see
Fig. on poster). Landscape composition explains 77% of the variation (single contribution: 52%)
and the physical characteristics of land unit explains 25% of the variation (single contribution nil).

Variation partitioning: forest - river - riparian vegetation

The forested area explains 32% of the variation (single contribution nil), the river area
explains 41% of the variation (single contribution: 33%) and the area of riparian vegetation
explains 44% of the variation (single contribution: 19%) (see Fig. on poster).

427
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Consequences of habitat availability and land-use change for the population

genetic structure of a grassland bush-cricket

S.I.J. Holzhauer, V. Wolters

Department of Animal Ecology, Justus Liebig University – Heinrich-Buff-Ring 26-32 (IFZ),

D-35392 Giessen, Germany. e-mail: [email protected]

The impact of temporal changes in habitat availability and land use on the present genetic
diversity of the bush-cricket Metrioptera roeseli (Saltatoria) was investigated in an extensively
used agricultural landscape (Lahn-Dill-Bergland, Germany). By integrating spatial and
temporal dimensions, this study thus contrasts to conventional approaches that usually
record landscape changes at discrete points in time.
The following hypotheses were addressed: (i) the genetic diversity of M. roeseli
populations in matured habitat patches is higher compared to younger patches, (ii) the
genetic differentiation among younger populations of M. roeseli is higher than among older
populations (iii) the constant availability of suitable habitat in the surrounding increases the
genetic diversity of local populations of M. roeseli, and (iv) the frequency of changes in land-
use in the surrounding landscape has a negative influence on the allelic richness of local
populations. We applied microsatellite markers for the analysis of the population genetic
structure of bush-crickets. Genetic diversity was estimated in terms of heterozygosity and
allelic richness.
Molecular data suggest population turnover and effects of age structure, though the
population genetic structure was more affected by geographical structure. Observed genetic
diversity was reduced, with populations from older grassland showing higher diversity and
less differentiation compared to those from younger grassland. Furthermore, genetic
heterozygosity was positively influenced by grassland age as well as by land-use change
towards higher amounts of grassland on the one hand, while on the other hand the influence
of grassland age within the study sites could be both negative and positive, presumably
depending on differing modes of dispersal. Moreover, allelic richness was negatively
correlated with land-use change.
These results prove the strong impact of habitat persistence and turn-over on genetic
diversity. Moreover, our findings point to the urgent need for approaches that integrate
spatiotemporal changes to determine the cumulative effect of subsequent changes in
landscape structure on population genetic structure.

428
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Prediction of future habitat distribution for song birds under climate change in
Switzerland

B. Schröder 1, J. von dem Bussche 1, R. Revermann 1, H. Schmid 2, R. Spaar 2 , N.

Zbinden2
1
Institute of Geoecology, University Potsdam, Germany.
e-mail: [email protected]
2
Swiss Ornithological Institute, Sempach, Switzerland.

Content

Statistical species distribution models are a useful tool to determine the distinctive habitat
factors governing the spatial and temporal distribution of species and to predict species
distributions under changing environmental conditions. The presented study comprehends
two parts: (i) large-scale species distribution models for several bird species with different
habitat requirements based on the Swiss breeding bird atlas (Schmid et al. 1998), and (ii) an
implementation of global change scenarios for climatic predictors in the Swiss Alps. We
simulated three scenarios to predict the future habitat of these species in 2030, 2050 and
2070, based on changes in July temperatures and annual precipitation rates according to
Frei et al. (2006).
We show that on the grounds of climate change there is an upward shift of the potential
habitat of ptarmigan (Lagopus muta helvetica) to higher mountain regions with lower
temperatures. But since its main food source is not shifting upwards as quickly as
temperature is rising, the discrepancy between temperature increase and vegetation
development may result in a great loss of adequate habitat (Revermann et al., submitted).
Our models result in a contrary trend of important habitat parameters for blackbird and
ring ouzel. Distribution of blackbirds is negatively correlated to precipitation rates and non-
vegetated areas but positively to temperature, closed forests and settlement proximity.
Sparse highland forests, and subalpine-alpine grasslands in turn reflect the realized niche of
the ring ouzel. This species shows also a positive trend concerning precipitation but a
negative correlation to temperature. Our scenarios indicate a decline of suitable habitat for
the ring ouzel. Blackbird distribution however, shows a slight range expansion to higher
altitudes (von dem Bussche et al., submitted).

References
Frei, C.; Schöll, R.; Fukutome, S.; Schmidli, J. & Vidale, P. L. (2006) Future change of precipitation
extremes in Europe: an intercomparison of scenarios from regional climate models. Journal of
Geophysical Research 111: D06105.
Revermann, R.; Schmid, H.; Zbinden, N.; Marti, C. & Schröder, B. (submitted) Suitable habitat for
ptarmigan (Lagopus muta helvetica) in the Swiss Alps and its response to rapid climate change in
the 21st century – a multiscale approach
Schmid, H.; Luder, R.; Naef-Daenzer, B.; Graf, R. & Zbinden, N. (1998) Schweizer Brutvogelatlas.
Verbreitung der Brutvögel in der Schweiz und im Fürstentum Liechtenstein 1993-1996.
Schweizerische Vogelwarte, Sempach.
von dem Bussche, J.; Spaar, R.; Schmid, H. & Schröder, B. (submitted) Modelling the recent and
potential future spatial distribution of ring ouzel and blackbird in Switzerland

429
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Dispersal traits and fragmentation of the habitats

M.O. Vandewalle1,3, L. Jönsson1, T. Reitalu2, M.T. Sykes1, K. Hall1, H. C. Prentice2,

S. Lavorel3 and E. Garnier4
1
Department of Physical Geography and Ecosystems Analysis, Lund University, Sölvegatan
12, 223 62 Lund, Sweden. e-mail: [email protected]
2
Department of Plant Ecology Sölvegatan 37, Ecology Building, Getingevägen 60
223 62 Lund, Sweden

Introduction

To predict plant biodiversity in a changing landscape, information on whether plants can

persist and regenerate in their existing habitats and/or can colonize new habitats is needed.
Both abilities depend on their biological traits, i.e. vegetative expansion and multiplication,
reproduction, seed bank longevity and dispersability. Dispersability or seed dispersal traits
correspond to the seed number per ramet, seed weight, seed shape, seed longevity,
releasing height and dispersal mode.
The process of isolation might have an unpredictable effect on dispersal capacity of the
populations. Recently it was shown that isolated populations of wind dispersed species can
lose large part of their dispersal capacity within few generations (Cody and Overton, 1996).
Species with long-range seed dispersal (LDD) are indeed expected to respond faster to
environmental change (Helm et al., 2006). A possible explanation for this observation is
natural selection against dispersal related traits in isolated populations. This can be
envisaged in the following way: in a patch isolated by a resistant or uninhabitable area, the
selection of producing seeds might be towards low dispersal capacity, as all the propagules
with high dispersal capacity will be wasted into the area surroundings. Therefore, it might be
expected that the older isolated population would have reduced dispersal capacity. The
assumption that fewer but also larger seeds are favoured in long-term isolated fragment is
supported by the common opinion that larger seeds have higher recruitment success and
that larger seeds may out compete smaller seeds if they end up together at the same site
Jakobsson and Eriksson, 2000). We are also expecting that specialist species will be more
affected by habitat loss and isolation of the habitat than generalist species (Adriens et al.,
2006).

Methods

During my fieldwork, I have been measuring dispersal traits of 32 species in 10 patches of

old semi-natural grasslands (more than 275 years old). Each of these patches has been
assigned with both an index of connectivity and an index of heterogeneity. Together with the
vegetation analysis and a theoretical database, our main goal will be to study the variation in
dispersal traits in relation to the fragmentation of the habitat and to identify groups of plant
responses to fragmentation and long-term isolation.

References
Adriaens, D; Honnay, O. & Hermy, M. (2006) No evidence of a plant extinction debt in highly
fragmented calcareous grasslands in Belgium. Biological Conservation 133: 212-224.
Helm, A; Hanski, I. & Partel, M. (2006) Slow response of plant species richness to habitat
loss and fragmentation. Ecology Letters 9: 72-77.
Cody, M. L. and Overton, J. M. (1996) Short-term evolution of reduced dispersal in island plant
populations. Journal of Ecology 84: 53-61.

430
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

NELI: A landscape index of effective habitat availability

R. Fraser, D. Pouliot, I. Olthof

Canada Centre for Remote Sensing, Earth Sciences Sector, Natural Resources Canada, 588
Booth St, Ottawa, ON, Canada K1A 0Y7
e-mail: [email protected]

Metapopulation theory has become a useful framework in conservation biology for

modeling the persistence of populations in fragmented landscapes. Its basis is that the
dynamics of interacting subpopulations are driven by local extinctions and recolonizations of
vacant patches (Opdam, 1991). Vos et al. (2001) proposed using two landscape indices to
represent how a metapopulation responds to landscape change: average patch carrying
capacity (i.e. habitat area) and average patch connectivity (i.e. dispersal distance for
recolonization). These indices are ecologically scaled, as they are calculated based on the
scale at which a given species perceives and interacts with the landscape.

We propose combining these indices into a simple, single measure of effective habitat
amount based on functional home range area. Functional home range is defined as the
potential dispersal envelope for each landscape position measured over a cost surface that
represents resistance to dispersal over different land covers. This non-Euclidean Landscape
Habitat Index (NELI) is then calculated from the area of suitable habitat can be accessed
within this envelope. NELI is therefore designed to provide a single, integrated measure of
both habitat amount and connectivity that is spatially explicit and species-specific.

We demonstrate how the index is computed for a single cell and over an entire landscape
for a hypothetical species profile (Eastern Wolf) within La Mauricie National Park and its
greater ecosystem, in Canada. Multi-temporal (1985-2005) habitat maps and dispersal cost
surfaces were derived using satellite-based land cover products and GIS vector layers.
Functional habitat amounts for five dates were then determined by calculating, for each
landscape cell, the total amount of habitat that can be accessed by dispersing over the cost
surface in all directions up to the home range limit of the species. The index, derived using
an IDL Virtual Machine executable, has modest input requirements for both spatial data and
information about a species’ habitat preference and dispersal ability. NELI could be used as
a coarse filter to monitor changes in habitat quality for focal species over large areas.

References
Opdam, P. (1991) Metapopulation theory and habitat fragmentation: a review of holarctic breeding
bird studies. Landscape Ecology 5: 93-106.
Vos, C.C; Verboom, J.; Opdam, P.F.M. & Ter Braak, C.J.F. (2001) Toward ecologically scaled
landscape indices. The American Naturalist 183: 24-41.

431
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Long distance dispersers and displacement kernels: effect on resource

utilization.

T. Lindström, N. Håkansson, L. Westerberg, U. Wennergren

Dep. of IFM, Theoretical Biology – Linköping University, SE-581 31 Linköping, Sweden.

e-mail: [email protected]

In ecology we have recognized that dispersal and especially long distance dispersers
are key features when species exploit resources in a landscape (Cain et al. 1998). Still one
has not explicitly studied the different features of long distance dispersal itself and their effect
on the ability of exploiting resources. In this study we have divided long distance dispersal
into two different features and by modeling and simulations analyzed populations
exploitations of resource in different kinds of lattice landscapes.
Different movement patterns results in different amount of long distance dispersers,
mean displacement etc., and can be characterized by displacement kernels. Studies of
animal movement often assume a Gaussian displacement since this is the result of random
walk and correlated random walk. There are however many reasons for deviations from
Gaussian displacement. Properties of the kernel can be categorized by variance which is an
over all measure of long or short distance displacement. A second property is the kurtosis
which determines whether the kernel is a Gaussian distribution or plateau shaped with a thin
tail, or more peaked with a fatter tail. Kurtosis is also a measurement of deviation from
Gaussian displacement. In our model we use a kernel equation that actually can separate
variance and kurtosis. Hence, it allows us to test the effect of the two features of the kernel
separately.
In nature, recourses are usually not distributed randomly. Instead some spatial
autocorrelation is expected. It has been shown that both the amount and arrangement of
preferred habitat largely influence the populations’ ability to exploit the resources in a
landscape (Westerberg et al. 2005). Individuals spend more time within their preferred
habitat when these are aggregated. To generate neutral landscapes with different amount
and aggregation of preferred habitats we use spectral density. We digitalize the landscape to
contain only two types of habitat, primary and secondary, in different ratios.
The displacement process was described with a matrix model where movement
behavior is assumed to differ between the habitats. This is reflected in a higher variance of
the kernel in the non preferred habitat. Exploitation of resources is analyzed via the dominant
eigenvector, which is the stable state distribution. We conclude that kurtosis of the
displacement kernel is not important for resource exploitation. Hence is the fatness of the tail,
which is one of the measures of long distance dispersers, not important for resource
exploitation. This has consequences for empirical studies of resources utilization. Variance of
displacement is easier to estimate than kurtosis.
Also we find that the landscape parameters (aggregation and amount of habitat) are
more important than the movement parameters. This means that much of the resource
utilization can be predicted simply from landscape parameters.

References:
Cain, M. L., H. Damman, et al. (1998). "Seed dispersal and the Holocene migration of woodland
herbs." Ecological Monographs 68(3): 325-347.
Westerberg, L., O. Ostman, et al. (2005). "Movement effects on equilibrium distributions of habitat
generalists in heterogeneous landscapes " Ecological Modelling 188(2-4): 432-447.

432
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Synchronization and noise colour: a threat to endangered species?

F. Lögdberg, U. Wennergren

Dep. of IFM, Biology – Linköping University, SE-581 31 Linköping, Sweden.

e-mail: [email protected]

Introduction
It is known that variation in growth rate will increase the extinction risk. Synchronized
variation will also influence the risk of local and global extinctions (Heino et al. 1997,
Palmqvist & Lundberg 1998, Engen et al. 2002). We test how the combined effect of these
two features effect the extinction rates in a landscape. Most models deal with an independent
stochastic variation, so called white noise. Here we also use positively and negatively
autocorrelated variation (red and blue noise respectively) by using 1/f-noise.

Synchrony and variation in space and time

Model and Result

A global growth rate for each time step constitutes the mean of a normal distribution, from
which all local growth rates are picked at random. The variance of these distributions sets the
level of synchrony among the local populations. A high variance gives a low level of
synchrony, and vice versa. We start with a spatial null model where all local populations are
equally connected. Then we introduce landscapes with more or less aggregation, by using
Fourier transform. We investigate the extinction risk when combining different levels of
synchrony and stochastic variation patterns (noise colours). The local dynamic is described
by either a density dependent or a density independent discrete function.
Our results show that synchrony, noise colour and impact of density dependence all have
a great influence on the extinction risk. A white noise model will underestimate the extinction
risk if the true variation is a reddish noise. It is important to be aware of synchrony among
local populations and synchronizing factors. A high degree of synchrony, in a reddish
environment, will increase the extinction risk even more. Together with high impact of
carrying capacity this will really be a threat to endangered species, especially when the
landscape configuration and composition affect the dispersal success.

Application
This general model will then be applied on the ecosystem of old oaks in pastures. Old,
hollow oak is a species-rich environment. Due to changed agricultural management the
amount of old oaks has decreased and many oak-dependent species, for example
Osmoderma eremita, are threatened today (Ranius, 2002). By analysing the landscape
characteristics as above, we identify areas with oak stands where the oak-dependent
species can or can not survive in a long-term perspective.

References
Engen, S; Lande R. & Sæther, B.E. (2002) The spatial scale of population fluctuations and quasi-
extinction risk. The American Naturalist 160: 439-451.
Heino, M; Kaitala, V; Ranta E. & Lindström, J. (1997) Synchronous dynamics and rates of extinction
in spatially structured populations. Proceedings of the Royal Society of London - Series B:
Biological Sciences 264: 481-486.
Palmqvist, E. & Lundberg, P. (1998) Population extinctions in correlated environments. Oikos 83:
359-367.
Ranius, T. (2002) Influence of stand size and quality of tree hollows on saproxylic beetles in Sweden.
Biological Conservation 103: 85-91.

433
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Habitat quality assessment and evaluation based on remote sensing – a case

study for red kite (Milvus milvus)

A. Buschmann, C. Kleinn

Institute of Forest Management, Georg-August-University Göttingen – Büsgenweg 5, 37077

Göttingen, Germany.
e-mail: [email protected]

Introduction

In large-area nature conservation programmes, the focus of monitoring activities is on the

status of populations of endangered species. The assessment and evaluation of habitat
quality represents one central element in that context and has to be carried out on different
spatial scales depending on the species examined.
On the landscape level and beyond remotely sensed data is an indispensable source of
information to perform the monitoring task. Habitat resources can be recognised on multiple
scales and multi-temporal data sets allow for the assessment of changes in habitat condition.
Besides the identification of essential habitat elements high resolution remote sensing further
on allows for the characterisation of their spatial configuration. This aspect of landscape
structure in many cases plays a decisive role when assessing a landscape´s habitat potential
for selected species.
Consideration of landscape structure in the assessment and evaluation of habitat quality
necessitates consistent and comparable measures of landscape configuration. Remote
sensing derived measures of landscape structure (landscape metrics) allow for an objective
and quantifiable assessment of the spatial composition and configuration of habitat elements.
This makes it possible to utilise the metrics besides other spatially explicit environmental
variables as input parameters in habitat suitability models.

Methods and results

The poster introduces an approach to integrate spatially explicit structure information in

habitat suitability evaluation. A case study for the predator bird species red kite (Milvus
milvus) in an EU bird protection area is presented in which an empirical habitat model is used
to derive decisive factors of breeding habitat choice of the species. The model considers
environmental variables related to landscape composition and configuration as well as land
use/land cover, topography, hydrology etc. By comparing the values of the variables at
species´ presence locations (nest sites) with the mean values for the entire landscape, as
well as the respective ranges of values, the ecological niche of the species is determined
against the background of the examined area.
It is pointed out how nature conservation could make use of the procedure in monitoring
habitat quality on landscape level. As a result, the integration of landscape metrics and its
inclusion (among other environmental variables) into an empirical habitat model has the
potential to make habitat quality assessment on landscape level more focused, more
operational and possibly also more meaningful.

434
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Preliminary study of ecological connectivity in the region Veneto (Italy)

G. Caravello, C. Gallo

Università degli Studi di Padova, Dipartimento di Medicina Ambientale e Sanità Pubblica –

Sede di Igiene, via: Loredan 18 – 35131 Padova.
e-mail: gianumberto.caravello@ unipd.it

Introduction

Conservation biologists generally agree that species viability and diversity are enhanced by
well-connected habitats. European projects especially Natura 2000 (92/43/EEC) intends to
preserve natural areas considered important for their flora and fauna characteristics.
However, these reserves are usually undersized compared with some animals needs and
under the negative effects of habitat fragmentation and patch isolation thus unable to
maintain viable populations over the long-term. Therefore, especially in intensively-managed
cultural landscape, biodiversity conservation strategies require a land use planning and a
following land management process.
Veneto region, located in the north-east of Italy, has a surface of 18.399 Km2 and circa 4.5
million of residents (2003).It mainly shows a residential, industrial and agricultural landscape,
even though 101 SIC and 70 ZPS cover the 20% of regional territory.
In order to get a primary understanding of regional connectivity we considered landscape
structural attributes at 3 different scales with an inter-jurisdictional perspective. At regional
scale (1:100.000), by using ArcMap (GIS), we divided the Corine Land Cover map (version
2000) of the region and its surrounding area in 9 squares and to get more detailed
information about potential entrance connections, we create 16 squares on the regional
border. On these squares, we selected those land-use categories able to support
connectivity enlarging our choice from exclusively Corine Land Cover category 3 to cat. 2.4.3
and cat. 2.2 creating numerous files. Successively, by using Idrisi (GIS), we analyzed these
files with Percolation method obtaining land cover values for the selected categories and
indirectly the probability dispersal success values for a hypothetical organism. At inter-
regional scale (1:1 million), we got a broaden vision collecting datasets from all the regions
surrounding Veneto. Using ArcMap we got the possibility to overlay spatial data regarding:
natural reserves, major transportation systems, major rivers system and main cities location.
At provincial scale (1:25000) we applied, with the same procedure and parameters, the
Percolation method on 2 adjacent provinces located on the plain regional border.

Results and conclusions

The numerical results, referred to the model threshold (pc= 0, 5927), allowed us to locate
areas with good or bad connectivity and at provincial scale showed that percolation values
get higher but not enough to ensure ecological connectivity. The visual results allowed us to
locate: semi-natural areas and potential ecological corridors, out of protection, but
fundamental for landscape connectivity; main challenging spots where potential ecological
connection crosses roads and cities.
Conclusions clearly point out the different situation between plain and mountains regional
areas, therefore different conservation strategies must be taken. In particular, at this scale,
Veneto presents a good connectivity only with other regions sharing its mountain border, in
the plain instead, semi-natural areas are located almost exclusively along the rivers, broken
up by urbanization centers.

435
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Habitat correlates and distribution pattern of the Tiger (Panthera tigris) in the
Terai Arc Landscape: a case for landscape ecology in surrogate conservation in
India

K. Ramesh, A.J.T. Johnsingh, Q. Qureshi, A. David, G.S. Rawat, S.P. Goyal

Wildlife Institute of India, Post Box # 18, Dehradun – 248001, Uttarakhand, INDIA
e-mail: [email protected]

The forests along the Himalayan foothills and the Gangetic flood plains in India and
Nepal constitute the Terai Arc Landscape (TAL). It is among the globally important
ecoregions and supports the highest densities of tiger populations in the world. Because the
tiger is a wide ranging species, occupies diverse biomes and its persistence is linked to
contiguous forests containing high prey biomass, efforts to conserve tigers at landscape
scale clearly underline the ‘focal species’ and require a surrogate conservation strategy. We
assessed the habitat condition and distribution status of the tiger in TAL during July 2002 –
March 2004, involving 1001.2km survey efforts, 1530 concentric vegetation plots, high
resolution (23.5m) satellite data and spatial analysis. It was found that tiger usage of the
habitats was highly variable across the landscape (12% ± 18.1 SD), with higher
concentrations in a few Protected Areas. Anthropogenic factors have reduced the once
contiguous habitats to nine fragmented units in the Indian side. However, these units
together with the habitat contiguity in Nepal comprise five larger units, with no or poor
connectivity between the subpopulations. Significantly, each of these units has at least one
key forest patch where tigers have higher probability of persistence and potentially act as a
source for adjoining patches. The large prey species, Chital (Axis axis), Sambar (Cervus
unicolor), Nilgai (Boselaphus tragocamelus) and Wild pig (Sus scrofa) occurred
interchangeably in considerable proportions (50% or above) in most parts of the landscape.
Logistic regression analysis (74% classification accuracy, n = 246) revealed that the species
with wider spatial distributions had significant positive influence on tiger occurrence, while the
domestic dog was negatively correlated, largely because the dogs closely accompany
human disturbances such as livestock herding, wood collection and poaching.
Global conservation agencies, on the basis of expert knowledge and macro analysis,
have categorized the extant tiger habitats into varying priority landscapes for rationalizing
human and fiscal resources. We argue that if such prioritization is to elicit intended success,
further details on patch characteristics, source-sink populations and prey biomass densities
need to be considered. Protected Area strategy has so far succeeded to a credible extent in
safeguarding tiger populations. Nevertheless, it is unlikely to yield success in the long-term
unless the effects of surrounding matrix including habitat connectivity, and securing large
contiguous habitats are brought under sincere focus. These requirements of the tiger
implicitly benefit several other species, reaffirming the relevance of a tiger centered
conservation philosophy despite its shortcomings. With adequate care and addressing local
people’s concern, the tiger could continue to be a powerful totem for conservation of
biodiversity in this landscape, and in the rest of its range.

436
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Setting corridor network priorities for Bogotá’s main ecological infrastructure

F. Remolina

Bogotá, District Planning Department, Carrera 30 Nº 24-90 Piso 13, Bogotá, Colombia
e-mail: [email protected]

Introduction

Bogota’s Main Ecological Infrastructure (BMEI) is designed to guide the land-use planning
decision process to ensure the establishment of existing natural areas and their ecological
services. Due to BMEI’s recent inclusion in Bogota’s land-use development plan, BMEI’s
conceptual framework and implementation are currently in a preliminary stage. An important
limitation of this green infrastructure is its corridor system, which neither connects all the
protected areas nor identifies which corridors are critical for conservation planning purposes.
A central idea of this project is to introduce a corridor system that connects Bogota’s
protected areas and classifies them according to ecological priorities.

Methods

The methods used in this project are adapted from the Geoplan Center, University of
Florida (2002). Landscape-level information is used as criteria in determining corridor
priorities and is classified by the level of ecological importance or potential land use
conversion. Data sets for ecological importance include: characteristics of critical habitats for
native species such as density, size, shape, and relative proximity to protected areas
connected by corridors; ecological community representation; presence of either fragile or
rare communities linked by the corridor; and the surrounding land-cover. Potential land-use
conversion data includes landscape features that encourage transformation from wildland or
rural land use to urban land use; pressure drivers include land ownership type, number of
primary and secondary crossings, and areas with potential for urban growth.

Ecological importance and potential land-use conversion data are analyzed using
Geographical Information Systems. Each data set is combined and weighted by means of
matrices, resulting in two sets of variables that are combined in a final matrix to obtain a
spatial outcome of corridor classification according to a high, medium or low priority.
Corridors with a high priority are described according to their significant ecological value and
land-use pressure.

References
Geoplan Center, University of Florida, (2002). Identification of critical linkages within the Florida
ecological greenways network. Report for the Florida Department of Environmental Protection,
Office of Greenways & Trails. Florida, U.S.A.

437
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Estimating species richness of cerrado (Brazilian savanna) carnivores by using

landscape structural metrics in predictive GARP models

L.R. Tambosi 1, M.C. Lyra-Jorge 1, M.C. Ribeiro 1, G. Ciocheti 1, V.R. Pivello 1

1
LEPaC–Laboratório de Ecologia da Paisagem e Conservação, Departamento de Ecologia,
Universidade de São Paulo, Rua do Matão, trav. 14, Cidade Universitária, CEP: 05508-090,
São Paulo, SP-Brazil.
e-mail: [email protected]

Introduction
Data on the occurrence of most native species are scarce and incomplete in most
regions, and therefore inadequate for conservation purposes. Predictive models may help to
fill up such information gaps. Recent studies show the use of the program DesktopGarp in
conservation biology to predict species distribution. A limitation of DesktopGarp is that the
models it generates do not consider effects of spatial dependence which may exist in the
landscape. Based on that, the objectives of this study were: i) to verify the contribution of
landscape metrics in defining models to predict the potential distribution of medium and
large mammals, and ii) to evaluate the power of species potential distribution maps to
estimate the regional richness of mammal species.

Material and Methods

The study area is in the NE of São Paulo State, Brazil, covering about 50,000 h. It
contains remnant forest and savanna patches surrounded by sugarcane and eucalyptus
plantations.
To estimate species occurrence and richness of medium and large carnivorous
mammals, data were collected through records in track plots and by camera traps during 18
months. Two groups of potential distribution models were generated for each carnivorous
species, one considering the environmental variables only (ENV), and the other including a
combination of landscape environmental and structural variables (ELS). The independent
variables were correlated with the extrinsic omission error by Spearman correlation
coefficients, and only those poorly correlated with the error were accepted to produce the
final models. For each species and type of model (ENV and ELS), 20 maps were generated,
representing the presence or absence of each species. The presence-and-absence maps
were then added up to produce two final maps of potential carnivore species richness, one
for the ENV models and one for the ELS models. Cluster analyses (Jaccard coefficient,
UPGMA) were employed to compare the sampled species according to the independent
variables used to generate both the ENV and ELS models.

Results
The values of species richness obtained through field sampling were very similar to
those predicted by the ENV models and by the ELS models. The comparison between
species richness maps generated by ENV and by ELS models showed an average
Spearman correlation coefficient of 0.94 (CI = 0.95). The cluster analysis reflected some
responses of the sampled species to the landscape structural variables. However, when
considering ENV independent variables, a single class was produced since the
environmental variables used for all species were the same.

Discussion
The high similarity obtained when comparing estimates of species richness through ENV
and ELS models with the value obtained directly from field samplings shows that the use of
landscape metrics is valid to generate DesktopGarp models.
Financial support: CAPES, FAPESP, CNPq and Neotropical Grasslands Conservancy.

438
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Landscape fragmentation as an indicator of coastal landscape quality: an

application along the Apulian coast (southern Italy)

F. Minunno1, V. Leronni1, C. Tarantino2, M. Mininni3, P Mairota1

3
ICAR Department Politecnico of Bari, Italy 1 Dipartimento di Scienze delle Produzioni
Vegetali Università di Bari, Italy 2 CNR. ISSIA Bari, Italy .
e-mail: [email protected]

This work has been carried out within the framework of the IMCA (Integrated Monitoring
of Coastal Areas) Research Project, among the activities aimed at drawing coastal
landscape quality maps through the use of indicators derived from satellite RS images. The
present contribution focuses on fragmentation as this phenomenon, as well as the loss of
heterogeneity, initiated by urban settlement processes of dislocation and diffusion,
represents the main cause of the landscape ecological efficiency decrease, of the area
decay and of the beginning of diseconomy in its management (Forman, 1995).
In order to quantify fragmentation, at a given spatial scale (defined in terms of both grain and
extent), a set of LPI was computed at the landscape level on a sample plot population,
extracted via an unaligned random sampling procedure from the whole southernmost part of
the Apulian peninsula (Southern Italy) and for which intepretation of recent aerial
photographs had already been performed within the framework of the IMCA research project
(Miacola et al. 2006). The same protocol was applied to categorical maps of the same area,
derived, both by past aerial photo-interpretation and by segmentation, from medium
resolution satellite images of two time steps.
Preliminary results are encouraging in many respects. The distribution analysis performed
on the indexes computed on the different data sets show, for this particular landscape at the
given scale, a significant trend towards a normal distribution model, thus contributing to the
ongoing debate (Remmel and Csillag 2003) on the uncertainties about the possibility of
statistically comparing indexes computed at different times and places, derived from a lack of
knowledge about their distribution. Principal Component Analysis performend on the indexes
obtained from the different data sets, yielded the ordination of sample plots along a
fragmentation gradient, that might be used to construct fragmentation intensity maps at the
subregional scale, as well as to interpreting the change processes and obtain intelligent
maps based upon the integration of field (aerial-photo interpetation) and RS data, thus
achieving the twofold purpouse of performing a phenomenological study aimed both at
modelling coastal landscape transformations and identifying new survey categories that may
have the temporal dimension as the main parameter. As far as the relations between the
indexes computed on the different data sets are concerned, they allow for the assessment of
the potential for using unsupervised categorical maps for the description and monitoring of
landscapes fragmentation, as well as for testing hypotheses concerning fragmentation
scaling relations in both space and time (Wu, 2004; Jelinski and Wu 1996).

References
Forman R.T.T. 1995 Land Mosaics. The ecology of Landscapes and Regions. Cambridge University
Press. Cambridge.
Jelinski D.E. and Wu J. 1996. The modifiable areal unit problem and implications for landscape
ecology. Landscape Ecology 11: 129–140.
McGarigal K. Marks B.J. 1995. FRAGSTATS: spatial pattern analysis program for quantifying
landscape structure. USDA Forest Service, Pacific Northwest Research Station. 122 p.
Remmel T.K. and Csillag F. 2003 When are two landscape pattern indices significantly different? J.
Geograph. Syst. 5:331-351
Wu J. 2004 Effects of changing scale on landscape pattern analysis: scaling relations. Landscape
Ecology 19: 125–138, 2004.

439
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Process-based connectivity metrics for conservation corridors

B. Rayfield1, A. Fall2, M.-J. Fortin1

1
Department of Ecology and Evolutionary Biology, University of Toronto – 25 Willcocks
Street, Toronto, ON, M5S 2L4, Canada.
e-mail: [email protected]
2
Gowlland Technologies Ltd. – 220 Old Mossy Road, Victoria, B.C., V9E 2A3, Canada.

Introduction

An ongoing challenge for conservation planning is to identify conservation corridors that

connect fragments of wildlife habitats. To this end, least-cost path analysis has been
combined with graph-theoretical techniques to relate wildlife movement to the spatial pattern
of the landscape (Chetkiewicz et al., 2006; O’Brien et al., 2006). This spatial graph approach
represents landscape connectivity according to species-specific dispersal abilities by
connecting pairs of habitat patches (graph nodes) with a single least-cost route (graph edge)
if the effective distance between them is less than a species’ dispersal ability. An important
modification to this approach has been to identify multiple least-cost routes between pairs of
habitat patches that collectively delineate spatial zones (areas) accessible for probable
movement within the intervening landscape (Theobald, 2006). In this research we present
novel, process-based metrics of connectivity on these spatial graphs that will improve our
ability to identify, design, and evaluate corridors for their conservation value. We partitioned
landscape connectivity into three essential components that can each be quantified
separately: (1) dispersal likelihood, (2) route redundancy, and (3) route vulnerability.
Dispersal likelihood is generally thought to decrease exponentially as the effective distance
between patches increases and is often also a function of the size, shape, and quality of the
connected patches (Urban and Keitt, 2001). Route redundancy is important in stochastic
environments when the presence of alternate routes between patches can maintain dispersal
if one or more routes become blocked. Route vulnerability can be quantified in terms of the
presence of bottlenecks along potential movement routes that are key areas to focus
conservation efforts. The components of graph connectivity can be computed locally
between patches that are connected by direct least-cost routes or globally between patches
that are connected by indirect routes along connected nodes. We generated artificial
landscapes to test the predictability of local and global corridor metrics given controlled
changes in composition and spatial configurations of habitat and other landscape features.
Graph-theoretic techniques have the potential to vastly improve our ability to design and
manage corridors (Chetkiewicz et al., 2006) but only if the measures we use to quantify the
components of connectivity of these spatial graphs are well-grounded in our current
ecological understanding of species-specific movements through heterogeneous landscapes.

References
Chetkiewicz, C.-L.B; Cassidy St. Clair, C. & Boyce, M. (2006) Corridors for conservation:
integrating pattern and process. Annual Review of Ecology, Evolution and Systematics 37: 317-
342.
O’Brien, D.; Manseau, M.; Fall, A.; Fortin, M.-J. (2006) Testing the importance of spatial
configuration o f winter habitat for woodland caribou: An application of graph theory. Biological
Conservation 130: 70-83.
Theobald, D.M. (2006) Exploring the functional connectivity of landscapes using landscape networks.
K.R. Crooks & M. Sanjayan (Eds.) Connectivity conservation, Cambridge University Press,
Cambridge, pp. 416-443.
Urban, D.L. & Keitt, T..H. (2001) Landscape connectivity: a graph-theoretic perspective. Ecology 82:
1205-1218.

440
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Seed flow from a dry calcareous grassland community into the adjacent
landscape

J.C. Bolli1, H.H. Wagner1, H. Bugmann2, Ch. Scheidegger1, P.J. Edwards2

1
Swiss Federal Research Institute WSL, Zürcherstrasse 111, 8903 Birmensdorf,
Switzerland, e-mail: [email protected]
2
Swiss Federal Institute of Technology (ETH) Zürich, Universitätsstrasse 16, 8092 Zürich,
Switzerland

Dry calcareous grasslands, which are one of the most species-rich habitats in Central
Europe, are threatened by habitat loss and fragmentation leading to isolated communities.
Conservation or restoration of dry grassland communities requires a seed flow of target
species between communities. Actual connectivity between communities, however, does not
only depend on the distance between communities but also on the amount of dispersed
seeds and the species composition, which is likely to change with dispersal distance. We
studied the actual seed flow by wind out of an isolated dry grassland community into the
adjacent landscape in the Schaffhauser Randen, Switzerland in summer 2006. Along 10
transects leading in two directions through the dry grassland patch up to 40m into the
surrounding landscape, a total of 230 low (0.2m) or tall (0.7m) funnel traps were installed to
study the spatial variation of local and middle-distance seed rain. The results suggest that
most species disperse the majority of their seeds locally within 1 m of the source plant. With
increasing distance from the source patch, the composition of the seed rain changed
markedly, and soon did not reflect the species composition typical for the dry grassland patch
anymore. The amount of dispersed seeds and mean dispersal distance is species-specific
and depends mainly on plant height, seed productivity and seed morphology. Tall grasses
followed by tall forbs have the highest dispersal ability as far as distance and quantity is
concerned. We conclude that seed dispersal from a dry grassland patch into the surrounding
landscape beyond 20m by wind is marginal and does not contain the species composition
typical for dry grassland. Unless there are other important vectors, a recolonisation after local
extinction might therefore be prevented for many species in many dry grassland
communities.

441
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

The influence of temperature and irradiation on the behaviour of butterflies

H. Malinowska1, A. Cormont2
1
Wageningen UR, Environmental System Analysis, Droevendaalsesteeg 4, 6708 PB
Wageningen, The Netherlands,
e-mail: [email protected],
2
Wageningen UR, Landscape Centre, Land Use Planning / ALTERRA Green World
Research, Droevendaalsesteg 3, 6708 PB Wageningen, The Netherlands.

Global climate change is a well-acknowledged problem. IPCC (2001) projects

higher climate variability and higher frequencies of extreme weather conditions, such as
heavy rains, El Niño events or droughts. One of the interesting questions to ask is how this
climate change will influence living organisms. Many species in adaptation to climatic
conditions change their niche or their distribution ranges (Opdam et Wascher, 2004). A study
of non-migratory European butterflies (Parmesan et al., 1999) revealed that during last few
decades the ranges of most of them shifted northwards by 35-240km. As genetic adaptations
to climate change might be too slow, the spatial configuration of suitable habitat will be of
crucial importance to many species (Opdam et Wascher, 2004). Range shifts in response to
climate change can be restricted by low habitat coherence. It is likely that the distribution of
suitable habitats for some species will also change, being outside of their dispersal
possibilities (Walther et al., 2002). Climate change together with anthropogenic land use
change can further increase habitat loss and fragmentation. In such a reality population
survival depends heavily on successful dispersal. Perhaps the dispersal behaviour itself can
be altered by climate change as well.
The behaviour of ectothermic animals is to a large extend influenced by
temperatures, therefore can be altered as a result of global climate change. Studies of
dispersal, which is a particular kind of behaviour, are of great importance for population
viability, especially in fragmented landscape where migration pressures are large.
Consequently, these studies can contribute to the conservation and better management of
the species.
In our study, the behaviour of four butterfly species Coenonympha pamphilus,
Plebejus argus, Melitaea athalia and Maniola jurtina living in the open habitats of the National
Park 'De Hoge Veluwe' was investigated in relation with temperature and irradiation.
According to a research hypothesis butterflies are most active in a species-specific thermal
optimum. Field time-budget data were analysed by means of Survival Analysis (Cox's
regression). In most cases significant models indicate that the rise in temperature and/or
irradiation has a positive effect on butterflies’ activity. Active bouts (defined as flying or
feeding) as well as flying bouts are longer in higher temperatures. However, it is possible that
the range of temperatures recorded was not wide enough to detect optimum trend. More
research, focusing on butterflies’ behaviour in extremely hot temperatures is thus needed.

References
IPCC (2001) Climate Change 2001: Synthesis Report. Summary for Policymakers.
Opdam P & Wascher D (2004) Climate change meets habitat fragmentation: linking landscape and
biogeographical scale levels in research and conservation. Biological Conservation 117, 285-297.
Parmesan C & Yohe G (2003) A globally coherent fingerprint of climate change impacts across
natural systems. Nature 421, 37-42.
Walther G, Post E, Convey P, Menzel A, Parmesan C, Beebee TJC, Fromentin J, Hoegh-
Guldberg O, Bairlein F (2002) Ecological responses to recent climate change. Nature 416, 389-
395.

442
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Plant species co-occurrence patterns on different spatial scales in dry, semi-

natural grasslands

T. Reitalu1, H. C. Prentice1, M. T. Sykes2, M. Lönn3, L. Jönsson2, K. Hall2

1
Plant Ecology and Systematics, Department of Ecology, Lund University, Sölvegatan 37,
SE-223 62 Lund, Sweden.
e-mail: [email protected]
2
Department of Physical Geography and Ecosystem Analysis, Lund University, Sölvegatan
12, SE-223 62 Lund, Sweden,
3
School of Life Sciences, Södertörn University College, SE-141 89, Huddinge, Sweden

Introduction
Landscape-scale factors (habitat area, connectivity, habitat history) are widely recognized as
determinants of community diversity (e.g. Bruun, 2000). But few studies have considered the
influence of landscape factors on species co-occurrence patterns (Badano et al., 2005). We
studied the extent to which patterns of plant species co-occurrence differ from random
expectations and explored different landscape and environmental factors.

Material and methods

The study was carried out in a 5 × 5 km “landscape” on the Baltic island of Öland and used
three, nested, spatial scales: 50 × 50 cm plots (N = 516), grassland patches (mean area of
ca 90 × 90 m, N = 109) and the whole landscape (N = 6). We used a null model approach,
with the C-score as the test statistic characterizing species co-occurrence (Stone and
Roberts, 1990). The results from the null model analysis were combined with ordination
analyses of the species, as well as linear models including information on environment,
landscape structure and history.

Results and conclusions

The C-score was significantly higher than expected by the null model in 21 % of the plots, 19
% of the grassland patches and in four, out of six, landscape-scale analyses, indicating that
pairs of species were co-occurring less than expected by random. On the plot scale, the
species-segregation was in terms of competitive interactions. The degree of segregation was
significantly associated with the plots' positions within the grassland patches and with the
shrub cover of the grasslands – both variables can be assumed to influence the intensity of
the competitive interactions. On the grassland patch scale, we interpreted the species-
segregation in terms of environmental heterogeneity within the patches. The degree of
segregation was significantly associated with the area of the patches and with vegetation
height – both variables are likely to be related to environmental heterogeneity. Species-
segregation on the landscape scale was interpreted in terms of environmental heterogeneity
among grassland patches and was significantly associated with land-use history. Previously-
arable grasslands showed a significantly higher degree of segregation than previously-
forested grasslands. Our results demonstrate the importance of incorporating replicated data
on multiple spatial scales and including information on environment, landscape structure and
history when investigating the patterns of species distributions.

References
Badano, E.I.; Regidor, H.A.; Núñez, H.A.; Acosta, R. & Gianoli, E. (2005) Species richness and
structure of ant communities in a dynamic archipelago: effects of island area and age. Journal of
Biogeography 32: 221-227.
Bruun, H.H. (2000) Deficit in community species richness as explained by area and isolation of sites.
Diversity and Distributions 6: 129-135.
Stone, L. & Roberts, A. (1990) The checkerboard score and species distributions. Oecologia 85: 74-
79.

443
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Analysis on Landscape Fragmentation in the Coastal Area of Pearl River Estuary

Based on RS and GIS

Z. F. Wu 1, J. Cheng1, C.P. Chang 2 and H.F Wan1

1
Guangdong Institute of Ecology, Environment and Soil Sciences, Tianyuan Road 808,
Guangzhou 510650, China.
e-mail: [email protected]
2
Chung-Chou Institute of Technology,–6,Line2,Sec.3,Shan-Chiao Rd, Yuanlin
Changhwa 51003, Taiwan, China.

The speed of urbanization has been most prominent in the Pearl River Delta in South
China during the past two decades. The rapid urbanization process caused an
unprecedented scale and rate of land use change in the area (Seto, 2005; Li,2004). The
Coastal Area of Pearl River Estuary (CAPRE) is the core and sensitive ecotone of Pearl
River Delta. As urban built-up areas sprawl, transform, and envelop the surrounding
landscape, one of the important changes in land-use were agricultural land loss and
landscape fragmentation. Combining GIS with landscape metrics, we attempted to quantify
urban expansion and landscape Fragmentation in CAPRE based on the TM remote sensing
data. The intent of this analysis is to understand the Spatial-temporal patterns regarding how
urbanization and physical characteristics affect landscape fragmentation over time.
Our study region CAPRE (including 4 cities: Zhongshan, Panyu, Dongguan, Shenzhen) is
located in the core of Pearl River Delta. Land cover data for CAPRE were captured using the
landscape thematic mapper (TM) sensor on 1988,1995,1998,2002. Images were classified
into 7 types land cover: urban, developing area, farmland, forest, orchard, grassland, barren ,
and water (wetland). FRAGSTATS (McGarigal et al., 2002) was used to calculate landscape
fragmentation index.
With the economy and population increasing, Urban and developing area underwent the
great growth during the entire study period (Table 1). The maximum urban land increase
occurred in the period 1988-1995.
Results from the landscape metrics analysis indicated that patch density (PD), edge
density (ED), patch number fragmentation index (PNFI) and diversity index increased. At the
same time, mean patch area (MPA), mean patch nearest neighbour distance (MPNND)
decreased. Landscape fragmentation degree became higher in CAPRE from 1988 to 2002.
Landscape fragmentation mainly took place in the end of the1980's and the beginning of the
1990's. The tendency landscape fragmentation became slowly after 1995.
There is obvious correlation between landscape fragmentation and the degree of
urbanization, society-economy development in CAPRE. Landform structure and drainage
density are intrinsic natural factors influenced landscape fragmentation degree and process.

Acknowledgements

This research was supported by National Natural Science Foundation of China (40571164)
and Guangdong Province Natural Science Foundation (04201163, 06105518).

References
Seto K.C. and M. Fragkias. (2005) Quantifying Spatiotemporal Patterns of Urban Land-use Change
in Four Cities of China with Time Series Landscape Metrics. Landscape Ecology 20:871-888.
Xia Li and A. G. Yeh. (2004) Analyzing Spatial Restructuring of Land use Patterns in a Fast Growing
Region Using Remote Sensing and GIS. Landscape and Urban Planning 69:335-354.
McGarigal, K., Cushman, S.A., Neel, M.C., Ene, E. (2002) FRAGSTATS: spatial pattern analysis
program for categorical maps, www.umass.edu/landeco/research/fragstats/fragstats.html

444
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Formation of necessary ecological conditions for sustainable socio-economic

development in Armenia

E. Shahbazyan

Maastricht School of Management, Endepolsdomein 150, 6229 EP, Maastricht, The

Netherlands.
e-mail: [email protected]

Since the beginning of human activity nature has had to undergo many changes,
strengthening from day to day. Today we stay before imbalance in the whole nature, which
undoubtedly influence on people life. Since the collapse of the Soviet Union a decade ago,
Armenia has been catapulted into a severe economic crisis. Nevertheless, Armenia is one of
the richest natural areas in the world, notable for its extraordinary biodiversity and wealth of
endemic species. Armenia also contains many rare and relict species of animals and plants.
The diversity of ecosystems in Armenia includes deserts, semi-deserts, mountain meadows
and steppes, forests, wetlands and alpine lands. The Armenian government is in the process
of developing policies regarding protected areas, biodiversity conservation, and sustainable
use of natural resources. However it still has to improve its natural legislative system, to
asses the real situation in ecosystems of Armenia and to monitor and manage them. The
ecological situation of Armenia demands urgent measures and long-term strategies for
biodiversity management. It is necessary to coordinate conservation strategies across
national borders.
Biodiversity of Armenia has maintained a complicated way of conservation measures. In
1958 a system was established of specially protected areas to protect ecosystems, habitats
and rare, endemic and threatened species. These areas currently cover around 311,000ha,
or 10% of the total area of the country.
For maintaining of balance in nature I suggest the creation of an Ecological Network which
will support biodiversity conservation and the achievement of the sustainable development.
The formation of Econet between reserves national parks and sanctuaries is a new approach
for managing the landscape for the wildlife of Armenia. One of the main issues of this
proposed work will be including people either from commercial, governmental and/or
nongovernmental sectors, which will, undoubtedly, support nature conservation and will
increase human life level in Armenia. This will promote sustainable, economic development
either in urban, or in rural areas and will also increase community capacity development.
If we are to go on enjoying wildlife and wildlife habitats in Armenia, new ways need to be
found to manage the landscape more sustainable. Ecological networks offer a possible
solution to this challenge. And the principal aims of the project should be:
• promote sustainable development,
• halt and reverse the disappearance, fragmentation and isolation of wildlife habitats,
• integrate ecological networks in land use planning and management,
• share the approach and findings with other countries.
This project will work to improve the Armenian landscapes both for people and for wildlife.
Constructing the network includes involving people. The network will also complement
existing initiatives and agencies working in wildlife conservation. The establishment of
ecological network for Armenia should be based on the support and cooperation of local
people. The realization of an ecological network will help local authorities and land managers
to integrate environmental considerations in land use planning and management.

445
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Increased Human Management Negatively Affects Beetle (Coleoptera) Species

Richness in Swiss Cities

T. Sattler1, 2, P. Duelli3, M.K. Obrist3, F. Bontadina2, R. Arlettaz2, M. Moretti1

1
Swiss Federal Research Institute WSL, Ecosystem Boundaries Research Unit, Via
Belsoggiorno 22, 6500 Bellinzona, Switzerland.
e-mail: [email protected]
2
Zoological Institute, Division of Conservation Biology, University of Bern,
Baltzerstrasse 6, 3012 Bern, Switzerland.
3
Swiss Federal Research Institute WSL, Biodiversity and Conservation Biology Research
Unit, Zürcherstrasse 111, 8903 Birmensdorf, Switzerland.

Introduction
Urban landscapes represent a heterogeneous environment that is constantly changing
and expanding due to developing human needs. The biodiversity of such mosaic habitats is
influenced by many environmental and man-made characteristics as well as urban-specific
processes (Shochat et al., 2006). In a strong ongoing debate, scientists argue which vari-
ables are essential for urban species diversity and on which scale (e.g. Clergeau et al.,
2006). We contribute to this discussion by analysing the relationship between invertebrate
diversity and urban environmental factors in three Swiss cities (Lugano, Lucerne, Zurich). In
particular, we tested the influence of (i) sealed area, (ii) human management by lawn- and
meadow-mowing (short time scale) and (iii) age of settlement on beetle diversity.

Methods
Recently, Duelli & Obrist (2005) developed a method called ‘Rapid Biodiversity
Assessment’ (RBA) which we applied at 106 study sites. Invertebrates were collected with
standard pitfall and flight traps during seven weeks in June/ July, 2006, which is the most
favourable time to sample invertebrates due to the maximum number of active species. We
separated invertebrates into 14 taxa groups (order and family-level) and subsequently
identified beetle specimens to morphospecies level within 14 different families. The various
orders and families correspond to different trophic guilds (i.e. herbivores, carnivores,
pollinators). The number of morphospecies correlates highly with the total species number.
Thus, this so-called RBA-index is an indicator for local α – diversity.

Results & Implications

Analyses of the beetle morphospecies numbers revealed that human management exerts
the greatest influence on morphospecies richness. Increasing numbers of lawn cuts nega-
tively affect the beetle diversity. Age of settlement and proportion of sealed area show only
minor influence on beetle morphospecies richness. Analyses of the trophic guilds showed
that human management negatively affects herbivore and pollinator, but particularly
carnivore numbers within beetles. In order to maintain and enhance beetle species richness
in urban areas, we recommend reducing the number of lawn and meadow cuts.

References
Clergeau, P., Jokimaki, J. & Snep, R. (2006) Using hierarchical levels for urban ecology. Trends in
Ecology & Evolution 21: 660-61.
Duelli, P. & Obrist, M.K. (2005) Eine preiswerte Methode zur Abschätzung der lokalen Arthropoden-
fauna: "Rapid biodiversity assessment" (RBA). Schriftenreihe der Forschungsanstalt Reckenholz
56: 132-38.
Shochat, E., Warren, P.S., Faeth, S.H., McIntyre, N.E. & Hope, D. (2006) From patterns to emerging
processes in mechanistic urban ecology. Trends in Ecology & Evolution 21: 186-91.

446
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Experimental deliberate releases of genetically modified plants near NATURA

2000 sites – using buffer zones to simplify the risk assessment

A. Benzler, U. Euler

Federal Agency for Nature Conservation, Konstantinstrasse 110, 53179 Bonn, Germany.
e-mail: [email protected]

Introduction

Recent progress in genetic engineering and increasing efforts in the development of

transgenetic plants evidence the need of extended numbers and extended area of
experimental sites for the deliberate release of GMOs. Placing experimental sites in the
wider agricultural countryside may lead to risks for the agrobiodiversity. In particular nearby
or within the experimental area located protected areas like Natura 2000 sites may cause
concern. According to the European law, “any plan or project […] likely to have a significant
effect thereon [expl: the Natura 2000 sites], either individually or in combination with other
plans or projects, shall be subject to appropriate assessment of its implications for the site in
view of the site's conservation objectives” (Art. 6 Habitats Directive). In theses cases an
extended environmental risk assessment according to the Habitats Directive may be
required. This may lead to delay within the regulation process. Therefore the German
competent authorities discuss buffer zones around Natura 2000 sites as a measure to
facilitate approval of deliberate release.

Approach

In a preliminary approach GIS based buffer zones of 1.000 m were created around Natura
2000 sites. To understand the consequences of this modus operandi for the choice of
experimental sites, we overlayed the geographical data with the arable land data of Corine
Landcover to identify the share of arable land which is situated within the buffer zones.
The calculation was undertaken separately for every federal state in Germany.

Results

We present the data separated in Sites of Common Interest (SCI) according to the
Habitats Directive and the Special Protected Areas (SPA) according to the Birds Directive.
Surprisingly, the calculation results had an extremely wide range. The arable land inside the
buffer zones (including the arable land inside of Natura 2000 sites) differed between 18%
(North Rhine-Westphalia) and 53% (Saarland) for SCI and between 3% (Thuringia) and 33%
(Saarland) for SPA, based each on the total area of arable land of the federal state.

Conclusions and consequences

The results indicate the importance of regional distinctions in land use, extent and
dispersion of protected areas in the wider agricultural landscape. They may have relevance
for other plans or projects, which may have adverse effects on Natura 2000 habitats.
The suitability of this “blanket distance” approach vs. a case-specific approach in respect
of the various potential vectors of different GMOs is discussed.

447
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Genetic factors in metapopulation survival - introduction to a PhD-project

M.M.P. Cobben1, P. Arens1, J. Verboom2, M.J.M. Smulders1

1
Plant Research International, Wageningen UR Droevendaalsesteeg 1, Wageningen, the
Netherlands.
e-mail: [email protected]
2
ALTERRA, Wagenbingen UR. Droevendaalsesteeg 3, 6708 PB Wageningen, the
Netherlands

Introduction
In view of the expected climate change, species may need to either shift elevation or latitude
ranges, or genetically adapt to their changing environment. Habitat fragmentation might have
partially impeded both options. Furthermore, both responses to climate change can have
future genetic consequences in prevailing populations. We will use a spatially explicit
metapopulation model to simulate genetic processes in fragmented populations under
climate change for the Dutch National Ecological Network (EHS).

Story Content
Genetic factors, particularly affecting small populations, contribute to extinction risk
(Frankham 2005). To prevent loss of genetic diversity, Ne should be 500-1,000 (Franklin and
Frankham 1998). To maintain long-term genetic security, Lynch and Lande (1998) advise an
Ne of 1,000-5,000. The ratio of Ne to N averages 0.10-0.11 (Frankham 1995).1
Subpopulations in a metapopulation structure suffer maximum inbreeding due to sequences
of extinction and recolonisation2 (Frankham et al. 2002), further reducing the Ne to N ratio
(Nunney 1999). So 10,000 individuals would be a fairly secure lower threshold for a
sustained genetically healthy population. However, on average only about 1,000 individuals
remain when an animal species makes the endangered list, and only about 120 individuals of
plants (Culotta 1995). Threshold for receiving the vulnerable status of the IUCN Red List, is
10,000 mature individuals worldwide for a species living in fragmented populations.3 In
contrast, 35 species with the status least concern are considered threatened by intrinsic
factors indicating unhealthy genetic background (IUCN 2006). We plead for more focus on
species that are not on the Red List, to prevent them from getting there in the future.
picture content
1. statement 1 from the above visualised in 1 figure.
2. statement 2 from the above visualised.
3. statement 3 from the above visualised in 1 figure.
4. schematic overview of all IUCN red list categories with their criteria

References
Culotta, E. (1995). Endangered species - minimum population grows larger. Science 270(5233): 31-
32.
Frankham, R. (1995). Effective population-size adult-population size ratios in wildlife - a review.
Genetical Research 66(2): 95-107.
Frankham, R. (2005). Genetics and extinction. Biological Conservation 126(2): 131-140.
Frankham, R.; Ballou, J. D. & Briscoe, D.A. (2002). Introduction to Conservation Genetics.
Cambridge University Press, Cambridge.
Franklin, I. R. & Frankham, R. (1998). How large must populations be to retain evolutionary
potential? Animal Conservation 1: 69-73.
IUCN (2006). 2006 IUCN Red List of Threatened Species. <www.iucnredlist.org>. Downloaded on 23
February 2007.
Lynch, M. & Lande, R. (1998). The critical effective size for a genetically secure population. Animal
Conservation 1: 69-73.
Nunney, L. (1999). The effective size of a hierarchically structured population. Evolution 53(1): 1-10.

448
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Spatial analysis of landscape patterns and their relevance for large mammal
conservation in the dry-deciduous forests of Central India

A. Paliwal, V.B. Mathur

Wildlife Institute of India, Post Box # 18, Dehradun – 248001, Uttarakhand, India
e-mail: [email protected]

Introduction and methodology

Ecological processes and landscape structures are interlinked (Milne,1992). Spatial
structure and configuration of the landscape govern the distribution and abundance of the
organisms. The importance of spatial dynamics is reflected in the availability of numerous
landscape metrices (Ritters et al., 1995). It is useful if the species-environment relationship is
explained in spatially explicit form, specifically for conservation and management of natural
resource management. In this study, we analyze the landscape pattern in the context of large
mammal conservation in Tadoba-Andhari Tiger Reserve (TATR) of Central India.
Unsupervised classification of Indian Remote Sensing P6 data (5.8m spatial resolution) was
done in ERDAS IMAGINE 8.7 supported by hierarchical clustering of 520 sampling
vegetation plots. Accuracy assessment was done based 700 random points. The classified
image was then subjected to analysis using Fragstats software (McGarigal and Marks, 1994)
to determine the structure and composition of landscape. Information distribution and relative
abundance of large mammals were obtained by 300 line transects of 2 km each, and also the
daily monitoring data of the Forest Department. Encounter rates and density of these
mammals were related to land cover type, and patch properties.

Results, discussion and conclusion

Twelve land cover classes, including six major vegetation types were mapped. The
resultant map had 84.8% accuracy. The landscape matrix was of mixed forests thickly
interspersed with bamboo (Dendrocalamus strictus). This also offered intact habitat with
largest patch size, while riparian habitat though under represented (3%) appeared to provide
viable connectivity for large mammal assemblages. The entire landscape consists of highly
heterogeneous elements, represented by 9000 forest patches with mean patch size of 0.6ha
(± 62.5 SD). Among the habitat mosaic, mixed trees with bamboo is more fragmented with
17.8 patches per hectare, while teak (Tectona grandis) dominant mixed forest is the most
intact habitat with lowest patch density of 0.6 patches per hectare. The variation in the patch
properties is significantly linked to status of distribution and abundance of wild ungulates in
the study area. It can be concluded that the northern parts of study area which have high
level of heterogeneity and interspersion, supports substantial wild ungulate populations.
Given that northern portion of TATR contains large number land cover types along with wide
range of patch sizes, it depicts high diversity in composition and configuration, which in turn
offers favourable habitat for the large mammals. Interestingly, though habitat fragmentation is
an issue at higher thresholds, but at the small scales, it reflects habitat diversity, and that
current pattern of management effort, with focus in northern portion is possibly justified.

References
McGarigal, K. & Marks, B., (1994) Fragstats- Spatial Pattern Analysis Programme for
Quantifying Landscape Structure. Forest Science Department, Oregon State University,
Cornvallis
Milne, B. (1992) Spatial aggregation and neutral models in fractal landscapes. Amer. Natur.
139(1): 32-57.
Ritters, K.H.; O’Neill, R.V.; Hunsaker, C.T.; Wickham, J.D.; Yankee, D.H.; Timmins, S.P.;
Jones, K.B.; and Jackson, B.L. (1995) A fractal analysis of landscape pattern and structure metrics.
Landscape ecology.10: 23-39.

449
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Indicative map of the Pan-European Ecological Network in Western Europe

R.H.G. Jongman, I.M. Bouwma, A. van Doorn

Alterra, Wageningen UR, PO Box 47 Wageningen, The Netherlands

e-mail: [email protected]

The objective of the Pan-European Ecological Network is to develop a vision for a coherent
network of internationally and nationally protected areas and other suitable habitat areas for
long term favourable conservation of Europe’s key ecosystems, habitats and species. As the
European strategy to reach the goals of the Convention on Biological Diversity the
establishment of the Pan-European Ecological Network (PEEN) has been one of the priority
issues for nature conservation in Europe since 1995 as formulated in the Pan-European
Biological and Landscape Diversity Strategy (PEBLDS).
The project resulted in an indicative map of PEEN which identifies the core nature areas of
European importance, existing corridors between these areas, and where new corridors
could and should be established to meet the connectivity requirements of key species.
The map illustrates the relevance of national and regional biodiversity within a European
context; it communicates the concept of nature as a coherent entity, rather than an
agglomerate of individual sites and species. The map also draws attention to the changes in
land use and infrastructure development that can have an impact on biodiversity, even when
core nature areas are not directly affected. As such, the indicative map of the Pan-European
Ecological Network in Western Europe is a powerful communication and education
instrument.
The indicative map of the Pan-European Ecological Network for Western Europe shows
areas that are vital for biodiversity in this part of Europe. It indicates possibilities to reinforce
the long term existence and possible return of internationally important species following the
strategy of a coherent and robust network. It summarises insights and data in a way that is
readily understandable, useful and inspiring for policy makers responsible for nature
protection and land use planning.
The map is strictly indicative, i.e. it only gives a tentative indication of the possible or likely
location of core areas and corridors of Pan-European importance on a scale of 1: 3,000,000.
Therefore the map cannot and should not be used to draw conclusions concerning the actual
location and boundaries of core areas and corridors of the Pan-European Ecological Network
or the location of (inter)nationally designated or acknowledged sites. The map does not
suggest that the identified areas should be designated under international or national
protection instruments, nor does it wish to comment on or influence the way in which national
governments apply their sovereign rights to designate areas for nature conservation
purposes.
The map is no blueprint for decisions and implementation; it indicates important areas where
further investigations, arguments for concrete decisions should lead to more concrete and
balanced plans taking into account interests of different stakeholders. The indicative map is
based upon many data, insights, assumptions and targets which are explained in the
following chapters. The map can be used together with other maps presenting underlying
and more detailed data on habitat types or designated areas with an international status.
One of the main conclusions of the project is that due to the high degree of fragmentation
there is a huge task in Western Europe for reconstructing coherence in nature. For this area
the challenge is to reconstruct coherence in biodiversity areas of importance.

450
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

The Development of a Site Inventory for Triturus cristatus in Cheshire

J. Hollinshead, A.P. Hull

Ponds Research Unit, Liverpool John Moores University, Liverpool, UK.

e-mail: [email protected]

Triturus cristatus, or Great Crested Newt, is the most stringently protected amphibian species
in the UK. Declining throughout its range, T. cristatus is locally common in North West
England. The species is vulnerable to pond loss due to changes in agricultural practice and
succession, since it is reliant on vegetated mid succession ponds and the availability of
woodland and un- or semi improved grassland terrestrial habitat, for foraging and refugia.
The protection afforded the species demands licensing of activities which will disturb or
destroy individuals or the species’ aquatic habitat and rigorous mitigation measures where
habitat is lost. Often, the presence of the species is only discovered at a development site
once work has commenced, resulting in unnecessary ecological damage and, unexpected
delay and expense for developers. The Site Inventory is a compilation of records from a wide
variety of sources and has been developed for use by researchers, ecological consultants,
planners and developers. It maps known T. cristatus breeding sites, providing details
including capture figures and population estimates, details of survey methodology, site
characteristics and levels of protection afforded the site. Additionally, details of ponds located
within home range and short distance dispersal distance of the breeding sites, and details of
surveys showing negative results for T. cristatus presence are held in the database. The
inventory is kept in database and GIS formats, and served via the web through the National
Biodiversity Network Gateway. The MAGIC Land based data portal may be used to
disseminate GIS and other data formats not compatible as yet with the NBN Gateway
systems. The data compiled has provided the basis for the GIS modeling of T. cristatus
distribution and habitat correlates presented.

451
Theme 3. Ecological Networks, fragmentation and connectivity
3.9 Posters

Landscape effects on anuran pond occupancy in an agricultural countryside:

barrier-based buffers predict distributions better than circular buffers

F. Zanini1,2, A. Klingemann3, R. Schlaepfer4 and B.R. Schmidt5,6

1
Geographical Information Systems Laboratory, Swiss Federal Institute of Technology
Lausanne, CH-1015 Lausanne, Switzerland.
e-mail: [email protected]
2
Swiss Federal Research Institute for Forest, Snow and Landscape (WSL), CP 96,
Lausanne, Switzerland.
3
Geobotanical Institute, Swiss Federal Institute of Technology Zurich, CH-8008 Zurich,
Switzerland
4
Institute for Environmental Sciences and Technologies, Swiss Federal Institute of
Technology Lausanne (EPFL), 1015 Lausanne, Switzerland.
5
Zoologisches Institut, Universität Zürich, Winterthurerstrasse 190, CH-8057 Zürich,
Switzerland
6
KARCH, Passage Maximilien-de-Meuron 6, CH-2000 Neuchâtel, Switzerland

Species movement and consequently occupancy of suitable habitat patches are dependent
on landscape permeability. Some land use types, such as canalized rivers and roads, may
be complete barriers to animal movement. We therefore expect that such barriers affect
species distributions. In analyses of landscape effects on animal patch occupancy it is the
common practice to use landscape variables extracted from circular buffers around patches.
The main assumption of this methodological approach is that species are affected by a
particular landscape element equally in every direction from a given pond. This assumption is
likely not to hold if animal movement is restricted by barriers. Barriers or inhospitable land
use types may reduce movement patterns and reshape the ideal circular surface into a non-
circular buffer. In this study, we developed a method to determine the effect of landscape
variables on amphibian distribution by considering physical barriers on their movement into
the landscape surrounding breeding ponds. We studied two amphibian species, the common
toad (Bufo bufo) and the common frog (Rana temporaria), in a highly fragmented landscape
in Switzerland. We extracted landscape variables (up to 3 km from ponds) within (i) “circulars
buffers” (CB) and (ii) “barriers-based buffers” (BBB). BBB were produced by reducing the
boundaries of CB according to major impassable barriers in the study area (highways and
canalized rivers). Our results show that the BBB approach almost doubles the explained
deviance of multiple regression models and clearly fits the distribution data better than the
CB approach. This suggests that the proposed BBB approach is ecologically more useful
than the traditional CB analyses of species-habitat relationships. Moreover, the BBB method
can be used to identify putative barriers. Our study shows the necessity to consider
ecological barriers in species distribution models in order to improve their explicative power
and avoid incorrect conclusions.

452
Theme 4: Ecohydrology, water and rivers

Theme 4 Ecohydrology, water and Rivers

453
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

454
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

4.1 Symposium 8: Landscapes and rivers

Landscape characteristics and aquatic habitats

E.A. Steel1, D. Jensen1, B.E. Feist1, M.B. Sheer1, I.A. Lange2, A. Odle2, R.
Brannom2, M. Danielsdottir2
1
NW Fisheries Science Center, NOAA Fisheries, 2725 Montlake Blvd East, Seattle, USA.
e-mail: [email protected]
2
University of Washington, Seattle, WA 98119, USA

Introduction
Viewing river systems within a landscape context is a new and rapidly developing
approach to river ecology (Allan et al. 2004, Weins 2002, Richards et al 2006, Hughes et al.
2006). Although the linkages among landscapes and associated physicochemical and
biological characteristics of rivers have long been recognized, the development of conceptual
frameworks and tools for measuring and synthesizing such linkages is relatively recent. We
present a series of spatial and statistical analyses to test the hypothesis that landscape
conditions have a controlling influence on aquatic habitats by examining four response
variables: (1) stream width and the density of pool habitats; (2) water quality in lowland
streams; (3) the variability of water flow and temperature patterns; and (4) salmon population
status. These analyses identify several key mechanisms by which large-scale land-use
patterns may impact aquatic resources.

Predicting aquatic habitat features over vast extents

Many aquatic taxa rely on pool habitats, including threatened and endangered salmonids.
Because the formation of streams, rivers, and pool habitats is driven by landscape-scale
features such as underlying geology and climate as well as by landscape-scale processes
such as sediment and water routing, we hypothesized that strong correlative relationships
would exist between large-scale landscape features and both stream width and percent pool
habitats. We generated a stream network for the Willamette basin, OR, USA, from a digital
elevation model and attributed each stream reach with field-based measurements of percent
pools and stream width, where available, and existing remotely-sensed data such as climate,
geology, and land-use. We were able to develop strong correlative models for both percent
pools and stream width that allow large-scale predictions of both habitat quantity and
distribution.

Watershed versus riparian predictors of water quality

We present new data on water quality in floodplain streams in King and Skagit Counties,
WA, USA. We compare the predictive ability of land-use (traditional agriculture, organic
agriculture, golf course, natural, high-impervious
1500

commercial) in the riparian area to land-use in the

watershed draining to each stream.
Sqrt Wavelet Coef. Variance (Scale 1)
500 1000

Figure 1: Increase in flow variability, as measured

using the variance of the wavelet coefficients at scale 1
= 1 day, as a function of the percent of the watershed
0

-8 -6 -4 -2 0 2 draining to the gage that is urbanized.

Log(Percent Urban Area)

Impact of landscape alterations on water flow and temperature variability

455
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Alterations of flow and temperature regimes are another potentially common, yet poorly
understood, impact of landscape-scale anthropogenic change (Steel and Lange In Press).
We used wavelet analysis to quantify fluctuations in both water temperature and flow at a
series of USGS gages in the Willamette basin and correlated these patterns with descriptors
of anthropogenic change within the watersheds draining to those USGS gages; early results
indicate that development within a watershed may impact water flow patterns (Figure 1).

Predicting population viability using simple landscape features

Many complex habitat models aim to predict the historical, current, or potential
performance of endangered salmonid populations. Using only simple landscape-scale data
on percent of habitat lost to anthropogenic barriers and the number of barriers in lowland
habitats, we were able to provide reasonable estimates of population viability (r2 = 0.80) for
spring chinook salmon and to classify watersheds as to the type of impact from barriers: a
large number of downstream barriers blocking a small amount of habitat, a large area of
upstream habitat being blocked, or a mixed, intermediate impact (Figure 2, Table1).

Linear regression model to used to predict square root

of the spring Chinook population performance score

Parameter Value Standard t- P-

error statistic value
Intercept 0.1352 0.2703 0.5002 0.634
Fraction
of
historical 1.6823 0.4119 4.0840 0.006
habitat 5
Number of
barriers - 0.0013 - 0.017
under 500 0.0042 3.2300 9
m

Figure 2 and Table 1: Watershed classification; shading indicates type of barrier impact:
downstream (light gray), moderate (dark gray), large area (black). Watersheds are
labeled with the modeled population performance score. Cross-hatch pattern indicates
natural exclusion zones and diagonal line shading indicates blocked areas. Table
describes parameters of the linear regression model that best predicted the square root
of the spring Chinook population performance score. (Figure from Sheer and Steel 2006).

References
Allan, J.D. 2004. Landscapes and riverscapes: the influence of land use on stream ecosystems.
Annual Review of Ecology, Evolution, and Systematics 35:257-284.
Hughes, R.M., L. Wang, & P.W. Seelbach, editors. 2006. Landscape influences on stream habitat
and biological assemblages. American Fisheries Society, Symposium 48.
Richards C., Johnson L.B., & Host G.E. 1996. Landscape-scale influences on stream habitats and
biota. Canadian Journal of Fisheries and Aquatic Sciences 53(Supp.1):295-311.
Sheer, M.B. & E.A. Steel. 2006. Lost watersheds: barriers, aquatic habitat connectivity, and species
persistence in the Willamette and Lower Columbia basins. Transactions of the American Fisheries
Society 135:1654-1669.
Steel, E.A. & I.A. Lange. In press. Alteration of water temperature regimes at multiple scales: Effects
of multi-purpose dams in the Willamette River basin. River Research and Applications.
Wiens, J.A. 2002. Riverine landscapes: Taking landscape ecology into the water. Freshwater Biology
47:501-515.

456
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Altered river landscapes in the alpine region: patterns of land use as a crucial
factor for the status of the aquatic environment

M. Poppe, S. Muhar, A. Melcher, C.Trautwein

Institute of Hydrobiology & Aquatic Ecosystem Management; BOKU – University of Natural

Resources and Applied Life Sciences, Max Emanuel Straße 17, A-1180 Vienna, Austria
e-mail: [email protected]

Introduction

The pattern of river landscapes is characterised by the composition and structure of

natural as well as human induced elements and patches. Under natural conditions, the
patterns and the spatial heterogeneity of these patterns are determined by physiographic
characteristics and by particularly driven hydro-morphological dynamics. However, since
hundreds of years these patterns and processes have been altered by human uses. In
Europe and North America, up to 90% of flood plains are already “cultivated” and therefore
functionally extinct (Tockner and Stanford, 2002).
The objective of this paper is to illustrate the land use- and habitat composition of near
natural as well as anthropogenically altered riverine landscapes at three different spatial
scales as a determinant attribute of the status of the aquatic environment.

Methods

For analysis at the catchment scale remote sensing data (SINUS - Spatial Indicators for
landUSe sustainability - Wrbka, 2003) are used. With a resolution of 30 m, the land cover of
the whole Austrian catchment area, divided into four biogeographic regions (Muhar et al.,
2004) is characterised by three land use intensity classes: (1) low use intensity, (2) moderate
use intensity, (3) high use intensity. The status quo of all large Austrian rivers (n=53, with a
catchment area >500 km2 and a total length of 5625 km) within these four regions is
described by three habitat quality classes (Muhar et al., 2000): (1) high habitat status, (2)
good habitat status, (3) moderately to heavily impacted.
Concerning the flood plain scale, the adjacent area (defined by a 50 m wide riparian buffer
within the flood plain) along the large rivers of Austria is categorised by four land use
classes: (1) forest, (2) grassland, (3) cropland and (4) urban land. The status of the large
river systems is assessed by large scale surveying (minimum stretch length is 1 km) using
historic data to characterise reference conditions (Poppe et al., 2004). As an integrative
parameter, the river type index is calculated, which reflects the degree of impacts on the
river: (1) only local stabilisation measures, (2) continuous stabilisation measures, however
the river course still reflects the morphological river type, (3) moderately altered
morphological river type (4) altered morphological river type, (5) heavily
channelised/straightened river course or impoundment.
The investigations at the reach scale comprise 13 rivers (with a total length of 1155 km) in
Lower Austria, one of the nine federal states of Austria. Land use pattern is again
characterised by the land use classes within the 50 m riparian buffer as explained above.
The rivers are investigated by detailed field mapping (minimum stretch length is 100 m) using
five parameters to describe the physical environment: (1) current river course, (2) river bed
morphology (3) instream structures (4) riparian structures, (5) land use structures. The total
ecomorphological evaluation index is calculated as the arithmetical mean of all five
parameters within a range of 1 (high quality) to 5 (bad quality - NRL, 2001).

Results

Catchment scale

457
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

At this nationwide scale the differences between the four biogeographical regions in
Austria are evident. Within the Alpine region and the Bohemian Massif the share of “low use
intensity” is between 50 and 70%. Corresponding, in these two mountainous regions there
are still 25-30% of the river stretches with “high and good habitat status”. In the lower
situated Alpine Foothills and the lowlands of the Tertiary Basin and Hill Country 42-28% of
the regions are “intensively used” by agriculture and settlement/infrastructure. In both regions
more than 90% of the large rivers are “moderately or heavily impacted”.

Flood plain scale

An hierarchical cluster analysis (Ward-Method) results in five land use clusters: (1) >80%
forest, (2) approx. 50% forest, 50% grassland, (3) >50% grassland, (4) up to 40% urban
land, (5) >40% cropland. These land use clusters reflect the different degree of impacts on
river morphology, expressed by the river type index (Fig. 2). Clusters with low or moderate
land use intensity types (Cluster 1-3) are characterised by river sections showing minor
impacts on river morphology (river type index 1-3).

Reach scale

The same land use clusters described above are now linked to the ecomorphological
evaluation index of the 13 investigated rivers in Lower Austria. Also at that small spatial scale
the intensity of land use within the 50 m riparian zone corresponds with the evaluation index
(see Fig. 3). Land use clusters dominated by forest and/or grassland (Cluster 1-3) are
connected to river stretches with an evaluation index <3 (moderate quality).

Conclusions
At a nationwide spatial scale, comprising the entire catchment areas of Austrian large
rivers, land cover types and their proportion within four biogeographic regions document the
different degree of human uses within these regions and indicate the pressure on the
physical environment of the rivers depending on the regions characteristics. Both at the flood
plain and the reach scale the intensity of land alteration in the flood plain is reflected in the
habitat status of the river. These results correspond to Wang et al. (2006), who describe
“forested rivers” with the least degraded physical habitat conditions, and “agricultural rivers”
lacking buffers as the most impacted.

References
Muhar, S.; Schwarz M.; Schmutz, S. & Jungwirth, M. (2000) Identification of rivers with high and
good habitat quality: methodological approach and applications in Austria. Hydrobiologia 422/423:
343-358.
Muhar, S.; Poppe M.; Egger, G.; Schmutz, S. & Melcher, A. (2004) Flusslandschaften Österreichs.
Forschungsprogramm Kulturlandschaft. bm:bwk 16, Wien.
Niederösterreichische Landesregierung (NLR; 2001) NÖMORPH. Strukturkartierung ausgewählter
Fließgewässer in Niederösterreich. Endbericht Teil I: Methodik. Freiland Umweltconsulting. Wien.
Poppe M.; Muhar, S.; Egger, G. & Schmutz, S. (2004) Status quo der österreichischen
Flusslandschaften. Österreichische Wasser- und Abfallwirtschaft. 55 (7-8): 122-129.
Tockner, C. & Stanford, J.A. (2002) Riverine flood plains: present state and future trends.
Environmental Conversation 29 (3): 308-330.
Wang, L.; Seelbach, P.W. & Hughes, R.M. (2006) Introduction to landscape influences on stream
habitats and biological assemblages. American Fisheries Society Symposium 48: 1-23.
Wrbka, T. (2003) Landschaftsökologische Strukturmerkmale als Indikatoren der Nachhaltigkeit.
Endbericht zum Forschungsprojekt SINUS (Spatial INdicators for Land USe Sustainability).
Forschungsprogramm Kulturlandschaft - bm:bwk 13, Wien.

458
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Management of nutrient fluxes in large river basins – the River Danube as an

example

M. Zessner

Institute for Water Quality, Vienna University of Technology, Karlsplatz 13,Vienna, Austria.
e-mail: [email protected]

Introduction

During the period between the seventies and the early nineties the North-Western and
Western Black Sea coastal area has suffered from chronic harmful algal blooms, permanent
hypoxic situations, as well as mass mortalities of benthic and pelagic organisms including
fish. An excessive input of nutrients (nitrogen (N) and phosphorus (P)) was the main reason
for this development. The river Danube can be identified as the major source for nutrients in
this part of the Black Sea ecosystem. Especially eutrophication problems close to the mouth
of the Danube Delta as well as on the coast south of the Delta, are a direct result of the
influence of the Danube. However, moving north along the Ukrainian coast the influence from
the rivers Dniestr and Dniepr increases.

Good news – for the moment

The situation in the Western Black Sea has improved considerably since the early 90s
due to: (i) reduced eutrophication (reduced phytoplankton biomass, frequency of blooms and
extension of high chlorophyll area), (ii) a considerable increase in water transparency, (iii)
improvement of river bed oxygen regime (Figure 1), (iv) regeneration of phytoplankton
species (Diatoms) diversity, (v) regeneration of phytobenthos and (vi) regeneration of
macrozoobenthos with an increase in species number. The zooplankton community in the N-
W and W Black Sea is still controlled by the gelatinous macrozooplankton (Mnemiopsis,
Aurelia, Pleurobrachia), with corresponding consequences on the recovery of the pelagic fish
stocks (daNUbs, 2005).

Figure 1. Mytilus galloprovincialis in front of the Danube Delta as an indicator, that anoxic
conditions have disappeared (Horstmann et al., 2002)

The limiting factor for phytoplankton growth in the eutrophic areas of the N-W-Black Sea
is P (since 1997). In the off shore waters N mainly limits primary productivity. The
improvement of the shelf ecosystem is a result of decreasing nutrient discharges (especially
phosphorus) to this part of the Black Sea (Figure 2). Current low discharges of N and P to
the Black Sea by the Danube river are the result of (i) improved nutrient removal from waste
water in Germany, Austria and the Czech Republic, (ii) reduced phosphate discharges from
detergents and (iii) the consequence of the economic crisis in central and eastern European
countries which lead to: (iiia) closure of large animal farms (agricultural point sources), (iiib) a

459
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

dramatic decrease of the application of mineral fertilizers and (iiic) closure of nutrient
discharging industries (e.g. fertilizer industry). (daNUbs, 2005)

Figure 2. Changes of nitrogen and phosphorus emissions into the river system of the
Danube from 1955 to 2000, (Behrendt et al., 2005)

New challenges

For sustainable development of the western Black Sea ecosystem, the nutrient
discharge from the Danube River should be further lowered or at least maintained at the
present level. It has been shown that the economic development in the Danube Basin may
reverse the improvement of the quality of the north-western and western Black Sea
ecosystem, if nutrients are not managed properly. In order to avoid deterioration of the
current situation, national governments should declare the total area in the Danube Basin as
sensitive area to facilitate the financial support of investments for waste water treatment with
nutrient removal from international donor funds. Furthermore, a consequent implementation
of measures to limit nutrient emissions from agriculture is necessary. These measures
should be based on the best available agricultural practices for reduction of nutrient losses
from agricultural areas and a limitation of the intensity of agricultural production. (daNUbs,
2005)

References
Behrendt, H; van Gils, J.; Schreiber, H. & Zessner, M. (2005) Point and diffuse nutrient emissions
and loads in the transboundary Danube River Basin. Large Rivers Volume: 16, No. 1-2 Arch.
Hydrobiol. Suppl. 158/1-2, p. 221-247.
daNUbs (2005) Nutrient Management in the Danube Basin and its Impact on the Black Sea, Institute
for Water Quality and Waste Management, Vienna University of Technology, final report,
supported under contract EVK1-CT-2000-00051 by the Energy, Environment and Sustainable
Development (EESD) Programme of the 5th EU Framework Programme,
http://danubs.tuwien.ac.at/, 78 pages
Horstmann, U., Davidov, A., Cociasu, A. & Velikova, V. (2003) Der Einfluss verringerter
Nährstofffrachten der Donau auf das Schwarze Meer, Österreichische Wasser- und
Abfallwirtschaft, Heft 11-12, 55 Jahrgang, 205-211.

460
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Impacts of climate and land use change scenarios on water quality of stream

C.P. Tung1, Y.J. Chen 2, Y.P. Lin1, N.M. Hong3

1
Department of Bioenvironmental Systems Engineering, National Taiwan University, No. 1
Sec. 4 Roosevelt Rd., Taipei 106, Taiwan.
e-mail: [email protected]
2
Environmental Restoration and Disaster Reduction Research Center, National Chung-
Hsing University, No. 250 Kuo Kuang Rd., Taichung, 402, Taiwan.
3
Department of Environmental Resources Management, The Overseas Chinese Institute of
Technology, No. 100 Chiao Kwang Rd., Taichung 407, Taiwan

Introduction

Climate and landuse changes could have a potentially adverse effect on water quality
and then ecosystems. This study focuses on the watershed scale. The possible impacts of
climate and landuse changes on water quality are assessed in this study. Besides,
assimilative capacity of a stream and allowable pollutant discharges of sub-watersheds may
also be changed and require re-assessment. At last, this study provides several important
suggestions in water quality management to decision makers to mitigate the effects of
climate and land-use changes.

Methodology

Two simulation models and one multi-objective optimization model are utilized to
assess the impacts of climate change on water quality and assimilative capacity. A
hydrologic model, Generalized Watershed Loading Functions (GWLF), is applied to simulate
the stream flow under different climate and landuse scenarios. Meanwhile, a water quality
model, Enhanced Stream Water Quality Model (QUAL2E), is used to estimate the
concentrations of Biochemical Oxygen Demand (BOD) after the hydrological model provide
simulated design flows. The QUAL2E is then applied to determine a response matrix for an
optimization model. The optimization model for estimating the assimilative capacity has two
objective functions. One is maximum allowable discharge pollutant, while the other one is
achieving equity between different sub-watersheds. The Gini coefficient is an index to assess
socio-economic equitable in economics and is used to assess the equality of allowable
maximum discharge pollutant in this study. Constrains here are that pollutant discharge must
meet the water quality standard and conservation of mass function.
The climate change scenarios are derived from transient experiments. The two
transient experiment scenarios are A2 (Medium-High Emissions) and B2 (Medium-low
emissions) of Special Report on Emissions Scenarios (SRES), respectively. Four GCMs
(General Circulation Models) were used to construct climate change scenarios: the HadCM3
(Hadley Centre for Climate Prediction and Research), the CGCM2 (Canadian Center for
Climate Modelling and Analysis), the CCSR (Center for Climate Resrearch Studies), and the
GFDL30 (Geophysical Fluid Dynamics Laboratory). In this study, the changes of climate at a
location are assumed to be the same as those of a nearest grid of GCMs. The changes are
applied to modify historic weather statistics for generating weather data.

Analysis and results

The Tochen River watershed is selected as a case study. The Tochen River is a
principal water resource for the Hsinchu water supply district, which is located in the Northern
Taiwan. The Hsinchu Science Park, a high tech industrial zone, is located in this watershed.
This park has a primary role in Taiwan’s spectacular economic growth. The Tochen River
watershed is divided into 10 sub-watersheds to analyze the impacts of climate change on
BOD concentration and assimilative capacity. Except of the results based on the CGCM

461
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

scenarios, the BOD concentration is raised conspicuously in other three GCM models. The
results also show that BOD concentration deterioration in downstream is more significant
than in upstream. Even the BOD concentration has exceeded the water quality standard in
some of the downstream reaches. The results indicate that total assimilative capacity is
reduced in B2 scenarios, but is increased in the A2 run of CGCM2 and HADCM3 model due
to increase of streamflow. Analyzing the results, it is concluded that impacts caused by
changing flow are more significant than changing water temperature.
A hypothetical case which a 100ha developing area in upstream NaHo sub-watershed is
shifted to downstream CanShia sub-watershed is applied to assess the variation of the
allowable discharge pollutant with considering equity. According to the Figure 1, it is found
that if the equity coefficient is increased, the allowable discharge pollutant will be decreased.
Furthermore, after the landuse change, the allowable discharge pollutant is decreased.
Hence, the developing area shifted from upstream to downstream does not improve the
allowable discharge pollutant.

Before land used change

Allowable discharge pollutant(100kg/day

After land used change

12.0
11.5
11.0
10.5
10.0
9.5
9.0
8.5
8.0
7.5
0 0.01 0.02 0.03 0.04 0.05 0.06
Equitable coefficient

Figure 1. The allowable discharge pollutant variation curve of before and after landuse
change

Conclusion

A simulation-optimization approach is applied to assess the impacts of climate and

landuse changes. The results show that impact of climate change on BOD assimilative
capacity at Tochen river watershed may requires decision makers to redistribute assimilative
capacity or reduce the pollutant to avoid that pollutant discharge exceed water quality
standard in the future. Based on the simple case study of impact of landuse change on
assimilative capacity, when a developing area is changed to downstream, where stream flow
is low and there are many water withdrawals, the assimilative capacity does not increase.
Climate and landuse changes significantly influence water assimilation capacity and then
feedback to limit land developing. Landuse and climate changes are a long-term processes,
and thus it is necessary to establish a long-term early warning system to provide information
to modify landuse plans and/or enhance adaptive capacity of environmental management.
Besides, there are still many uncertainties which may influence the conclusions of this study
and certainly require further studies.

462
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

How do landscape scales influence riverine fish in Europe?

S. Schmutz, C. Trautwein

Institute of Hydrobiology & Aquatic Ecosystem Management; BOKU – University of Natural

Resources and Applied Life Sciences, Max Emanuel Straße 17, A-1180 Vienna, Austria;
e-mail: [email protected]

Introduction

Spatial patterns of fish assemblages are poorly understood at large scales. Landscape
classifications (LC), e.g. ecoregions, are supposed to reflect fish assemblage patterns at
large scales, however, the significance of different LC commonly used in Europe never has
been tested for the ability to explain spatial structures of fish assemblages. Here, we analyse
the relationship between fish diversity and LC associated with different spatial scales and
compare the results with grid-based classifications reflecting species-area relationships.

Methods and data

We used 6 ecological and hydrological LC. (1) The European Biogeographic Regions
(EBR) of the Habitats Directive are based on a map of natural vegetation (EEA, 2005). (2)
Illies’s Ecoregions (IER), representing the only aquatic LC, is used for delineating river
typologies for the Water Framework Directive (WFD). (3) The Digital Map of European
Ecological Regions (DMEER) is based on climatic, topographic and geobotanical data (EEA,
2000). (4) Catchments of European Sea Outlets (CESO) were derived from the project
“Catchment Characterisation and Modelling – CCM” (Vogt et al., 2003). (5) European River
Catchments (ERC) are based on an aggregation of CESO and represent main European
ocean watersheds (EEA, 2006). (6) The Main River Groups (MRG) have been developed in
the FAME project based on fish assemblage similarity among catchments (FAME consortium
2005, Pont et al. 2006).
The 6 ecological and hydrological LC were compared with 7 grid-based LC computed as
squares with progressing grid side lengths of 31.25, 62.5, 125, 250, 500, 1000 and 2000 km.
Fish data have been provided by the FIDES database collated in the FAME project (FAME
consortium 2005). Only fishing occasions with no or minimal human pressures (1 608
reference sites) were selected for analyses as defined in the FAME project (Pont et al. 2006).
All data were available as GIS layers. The study area was defined by overlaying the 6 LC
and extracting the common cutting area. The area and the total number of fish species was
calculated for each LC unit. Finally, the median area and mean number of fish species (+-CI)
was calculated for each LC classification.

Results and Discussion

In total 80 fish species are recorded in FIDES for selected reference sites considering the
entire study area. At the largest spatial level ERC LC splits the study area into 5 large
catchment units (median area: 641 371 km2) with a mean number of fish species of 47. EBR
LC consist of 6 regions (median area: 436 596 km2) with a mean of 43 species. IER LC with
12 regions (median area: 203 402 km2) hosts 30 species at the average. DMEER LC
(median area: 103 426 km2) splits Europe in 16 regions with a mean of 26 species. There are
11 MRG LC units (median area: 96 603 km2) with 28 species at the average. The smallest
units (N=141) are derived from CESO LC (median area: 1 273 km2) with a mean number of
species of 10 (Figure 1).
The species-area curve derived from the grid-based LC reveals the form of an upper part
of a power function. Number of grid units increases from the largest scale (2000 km grid side
length) with 4 units to 454 units at the smallest scale (31.25 km grid side length) and the
mean number of species decreases from 46 to 9 species, respectively.

463
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

LC located above the curve indicate a higher number of fish species than suggested by the
area of the LC. Only EBR and MRG are above the curve and host about 5 species more than
expected, however, due to the high variation within LC and low number of samples (i.e. LC
units) the difference is statistically not significant (Figure 1). On the other hand, IER, used for
river classifications in the WFD, do not contribute to species diversity beyond the information
given by the species-area relationship. Therefore, we conclude that for analysing spatial
patterns of fish diversity or for developing European ecological river typologies EBR, MRG or
similar LC should be used instead of the IER.

Figure 1 Mean number of fish species (± CI) in different landscape classifications of Europe (solid
circles) in relation to median area of landscape classifications and compared with grid-
based classifications (squares and hand-fitted curve). EBR: European Biogeographic
Regions, IER: Illies’s Ecoregions,DMEER: Digital Map of European Ecological Regions,
MRG: Main River Groups, ERC: European River Catchments, CESO: Catchments of
European Sea Outlets.

Acknowledgements
We would like to thank all institutions delivering data for the FAME project (Contract n°:
EVK1 -CT-2001-00094).

References
European Environmental Agency (EEA) (2005) Biogeographical regions, Europe. Retrieved on
20.11.2006 from: http://dataservice.eea.europa.eu/dataservice/metadetails.asp?id=839.
European Environmental Agency (EEA) (2000) Digital Map of European Ecological Regions -
DMEER. Retrieved on 14.12.2006 from:
http://dataservice.eea.europa.eu/dataservice/metadetails.asp?id=192
FAME Consortium (2004) Manual for the application of the European Fish Index - EFI system. A fish-
based method to assess the ecological status of European rivers in support of the Water
Framework Directive. Version 1.1, January 2005. http://fame.boku.ac.at. 3.3.2005.
Pont, D; Hugueny, B. Beier, U; Goffaux, D; Melcher, A; Noble, R; Rogers, C. & Roset, N. (2006)
Assessing river biotic condition at a continental scale: a European approach using functional
metrics and fish assemblages. Journal of Applied Ecology 43:70-80.
Vogt, J.V; Colombo, R; Paracchini, M.L; Soille, P & De Jager, A (2003) CCM River and Catchment
Database for Europe, Version 1.0. EC-JRC. EUR 20756 EN.

464
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Landscape characteristics and Pacific salmon distribution: consistent patterns

across watersheds

B.E. Feist, E.A. Steel, G.R. Pess, D. Jenson

Northwest Fisheries Science Center, 2725 Montlake Blvd E., Seattle, WA 98112, USA.
e-mail: [email protected]

Introduction

Habitat loss and alteration is the leading cause of species’ declines worldwide, therefore,
habitat restoration and protection is an important conservation strategy. The current dogma
is that there is a strong connection between habitat and threatened or endangered species;
however, conservationists are often hard-pressed to explicitly link abundance or population
health with habitat quality or quantity. Given that habitat relationships with species are often
characterized at a spatial scale that does not account for the functional relationships between
habitat and populations, it’s not surprising that the habitat-population disconnect persists.
There have been a series of analyses completed in various subbasins of the Pacific
Northwest of the United States, where coarse-grained GIS representations of habitat have
been linked to Endangered Species Act (ESA) listed anadromous Pacific salmon
(Oncorhynchus spp.) population performance (Pess et al. 2002; Feist et al. 2003; Steel et al.
2004). Lacking in these studies is synthesis of across-subbasin patterns in habitat
significance as well as the importance of observation scale. In this paper, we attempt to
synthesize and build on the work of the aforementioned authors by using their methodology
to analyze habitat and population connections in three additional subbasins in the Columbia
River basin. We posed three questions in this project: at what extent is each potential
predictor variable best correlated with salmon performance; at what extent does one find the
best model fit; and how does a combined extent model compare with a single extent model?

Methods

In order to explore the influence of spatial scale on the apparent relationship between
habitat and animal populations, we examined the relationship between GIS-based habitat
data and various anadromous Pacific salmon (Oncorhynchus spp.) populations in three
subbasins of the Columbia River basin, USA. We characterized habitat and ran our models
at three different spatial extents: local or stream reach ( X = 11.9, SE = 2.4 km2),
intermediate ( X = 230.2, SE = 33.8 km2), and watershed or landscape ( X = 644.5, SE =
221.7 km2). Our population data were derived from annual redd (spawning nests built by
females) count surveys (StreamNet 2002), collected for many decades by various state
agencies. Our “habitat” GIS data included land use and land cover, geology, terrain,
precipitation, temperature, dams, diversions, mining hazards, and grazing allotments.

We used mixed models to explore the relationships between redd density and candidate
explanatory variables. We then used the Bayesian Information Criterion (BIC) to select the
most appropriate covariance structure, as well as to compare models (Littell et al. 1996). All
statistical analyses were run using SAS/STAT (V 8.2).

Four random-coefficients models were evaluated. The slope and intercept were initially
modeled as random effects. The slope and intercept for each year was allowed to vary
randomly from an average slope and intercept. These are also referred to as hierarchical
linear models (Byrk & Raudenbush 1992). The appropriateness of these models was
assessed by noting whether the variance of the slope and/or intercept was significantly
different from zero, which was assessed with Wald tests (Casela & Berger 1990), as well as
by comparing the BICs of the various models. The intercept was also modeled as a random

465
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

function of both year and reach. Lastly, the intercept was allowed to vary randomly by year
only.

Models were fit to the covariates from each of the three spatial extents independently, and
the best models from each were compared at the end of the model-selection process. The
variable selection process primarily identified those models with undesirable properties and
those models were eliminated. Among the criteria and methods used to evaluate models
were high pairwise correlation, large Cook’s distance (Cook 1977), multicolinearity and a
cross-validation procedure.

Results

Overall, redd densities were most often correlated with land use variables, while land
cover, structure and geology were also correlated, but not as frequently. Climate variables
were not included in any of the final models. Certain categories of variables were better
correlated with salmon redd density at a particular extent. However, our intermediate-extent
models usually provided poor predictive power compared with our local- and landscape-
extent models, and r2 values for these models were higher as well. In addition, the suite of
habitat characteristics most often associated with spawner density changed as a function of
our observation extent.

Discussion

It appears as though certain types of habitat are correlated with Pacific salmon redd
density in the three subbasins of the Columbia River basin we studied. It is also clear that the
amount of variation explained in redd density over time is contingent upon the size of our
analysis window. Based on these results, we conclude that our perception of which habitat
attributes are important is clearly a function of our extent of observation, and that restoration
efforts should focus on different habitat characteristics depending on the “scale” of the
restoration effort. This is a significant result in general for conservation efforts, and an
important practical result for salmon recovery planning in particular.

References
Byrk, A.S. & Raudenbush, S.W. (1992) Hierarchical Linear Models. Sage Publications, Newberry
Park, CA, USA.
Casella, G. & Berger, R.L. (1990) Statistical Inference. Duxbury Press, Belmont, CA, USA.
Cook, R.D. (1977) Detection of influential observation in linear regression. Technometrics 19: 15-18.
Feist, B.E.; Steel, E.A.; Pess, G.R. & Bilby, R.E. (2003) The influence of scale on salmon habitat
restoration priorities. Animal Conservation 6(3): 271-282.
Littell, R.C.; Milliken, G.A.; Stroup, W.W. & Wolfinger, R.D. (1996) SAS® System for Mixed Models.
SAS Institute Inc., Cary, NC. USA.
Pess, G.R.; Montgomery, D.R.; Bilby, R.E.; Steel, E.A.; Feist, B.E. & Greenberg, H.M. (2002)
Correlation of landscape characteristics and land use on coho salmon (Oncorhynchus kisutch)
abundance, Snohomish River, Washington State, USA. Canadian Journal of Fisheries and
Aquatic Sciences. 59: 613-623.
Steel, E.A.; Feist, B.E.; Jensen, D.; Pess, G.R.; Sheer, M.; Brauner, J. & Bilby, R.E. (2004)
Landscape models to understand steelhead (Oncorhynchus mykiss) distribution and help prioritize
barrier removals in the Willamette basin, Oregon, USA. Canadian Journal of Fisheries and Aquatic
Sciences 61: 999-1011.
StreamNet Spatial Data (2002) Current anadromous chinook salmon and steelhead distribution at a
1:100,000 scale. Pacific States Marine Fisheries Commission. URL: <http://www.streamnet.org/>.

466
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Limiting factors controlling the spatial distribution of redband trout

(Oncorhynchus mykiss gairdneri) and their implications across a basin

L.F. Madriñán ¹-², S. White ¹, H.W. Li ² and G. Giannico ¹

¹ Department of Fisheries and Wildlife, 104 Nash Hall, Oregon State University, Corvallis,
OR, 97331, USA . ² Oregon Cooperative Fishery Research Unit, 104 Nash Hall,
Oregon State University, Corvallis, OR, 97331, USA.
e-mail: [email protected]

Introduction

Temperature relationships between aquatic organisms and their surrounding environment

need to be understood to insure protection of threatened stocks of fish. Aquatic organisms
respond to thermal heterogeneity and require specific ranges of temperature to survive and
reproduce. Several studies in stream ecology (Fausch et al., 2002) have recognized that
streams are strongly influenced by the surrounding landscape in which they flow. Elevated
stream temperatures are the result of ecological and physical processes interacting at the
same time, and are directly related with land-use practices, with negative consequences for
aquatic ecosystems.
Monitoring longitudinal summer stream temperatures may be useful to asses the carrying
capacity of threatened and endangered salmonids. However, to understand expressions of
habitat heterogeneity in stream systems it is necessary increase research efforts to
understand how multi-scale patterns of stream temperature and habitat quality affects fish
community distribution (Torgersen et al. 1999).

This study attempts to classify habitat quality for Oncorhynchus mykiss gairdneri (Redband
trout) in the south fork of the John Day River, Oregon, USA using spatially continuous data
on fish assemblage structure and habitat data.

Methods

To standardize the extent of our sampling design we developed a method to classify

valley river segments (reaches) using a 10 m digital elevation model in which we divided the
SFJD and the two tributaries using three metrics (elevation, slope of the stream channel and
aspect). After this initial classification we overlapped the FLIR (Forward looking infrared)
imagery and subdivided some of the reaches if the difference in temperature was greater
than 3°C per each section
Once the reaches were defined we identified potential limiting factors for Redband trout
using an information-theoretic approach based on Akaike’s information criteria (AIC), were
the relative importance of the predictor variables of each model was ranked using data from
different spatial scales.

Existing land use - land cover maps, LiDAR (light detection and ranging) geomorphic
information, longitudinal temperature profiles from FLIR images and extensive in situ habitat
and fish community surveys were used in these models

Once all the fish information was collected, we used a cluster and outlier analysis (Anselin
Local Moran's I) to identify possible spatial autocorrelation among sets of clusters of
distribution points with values similar in magnitude and clusters of distribution points with
very heterogeneous values. To identify possible high density areas for fish distribution we
used the “Hot Spot Analysis” with rendering to calculate the Getis–Ord Gi statistic by
rendering features quantitatively

467
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Results

Our results confirmed the presence of different strata in the watershed. Differences in
landscape and land use attributes showed that continuous sampling is most useful that site-
specific sampling in order to detect limiting factors and cumulative effects. We found that
Redband trout distribution in the principal tributaries and in reaches within tributaries have
different limiting factors.
The cluster and outlier analysis shows that there was not any clustering in fish distributions
that was not detectable at the reach scale. It was also found three principal “hotspots” for
Redband trout distribution that are principally associated with cooler water temperatures,
higher sinuosity and stream gradient.
Conclusion
Reach analysis (1 to 8 km) is the most appropriate scale to understand Redband trout
distribution patterns in the South Fork Jon Day River, Or, USA. We found that among
reaches Redband trout distribution is delimited first by physiological tolerance to temperature
and secondly by competition with warm water fishes.
Redband trout was principally correlated with relative higher gradients and shallow reaches
in the warmer sections of the SFJD, in contrast in the coolest tributary (Black canyon creek)
RBT was associated with smaller gradients and deeper pools.
This study also found that the cross over point between warm water – cool water fish
assemblages was 22 Cº.
This approach will provide managers with hypotheses for restoration experiments in
specific areas (e.g., wood additions, beaver dams, culvert replacements). This is consistent
with the concept of Adaptive Management.

References
Fausch, K. D.; Torgensen, C. E.; Baxter, C.V.; Li, H.W. (2002). Landscapes to riverscapes:
Bridging the Gap between Research and Conservation of Stream Fishes. BioScience. Vol. 52
No. 6. pp 1 – 16.
Torgersen, C.E; Price, D.M; Li, H.W; and McIntosh, B. A. (1999). Multiscale Thermal Refugia and
stream Habitat associations of Chinook salmon in Northeastern Oregon. Ecological Application. 9
(1) pp. 301 – 319.

468
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Relating fish assemblages to environmental patterns at three multistate scales

R.M. Hughes1, A.T. Herlihy 1, and J.C. Sifneos 2

1
Department of Fisheries & Wildlife, Oregon State University, 200 SW 35th St., Corvallis, OR,
97333, USA.
e-mail: [email protected]
2
Department of Geosciences, Oregon State University, 200 SW 35th St., Corvallis, OR,
97333, USA..

Introduction
Wang et al. (2006a) identified several key challenges to studying and managing
landscape-river systems. These included understanding how factors measured at various
spatial-scales influence aquatic biota. Biological assemblage responses to natural gradients
and anthropogenic disturbances have been studied at multiple spatial scales (Roth et al.
1996) because large scale patterns are believed to influence local conditions (Frissell et al.
1986). However because of study objectives, assemblage responses to environmental
conditions are often largely focused at the catchment (Steedman 1988) or local (Southwood
1977) scale. Such studies hinder critical evaluation of the relative importance of proximal and
distal environmental factors to aquatic assemblages. In addition, the types, intensities, and
locations of major environmental gradients alter the relative importance of environmental
factors (Wang et al. 2006b), as do the assemblages studied (Allen et al. 1999).
Aquatic ecosystem classification has long been used for predicting fish assemblages in
streams and rivers. These have included fish zones (Hawkes 1975), physiographic regions
(Pflieger 1971) ecoregions (Hughes et al. 1994), and river basins (Hocutt and Wiley 1986).
However, large quantitative fish assemblage data sets linked with quantitative physical and
chemical habitat data and landscape data are becoming increasingly available (Pont et al.
2006).
Our objectives were to determine (1) fish clusters at three spatial scales in the western
USA, (2) how predictor variables for those clusters changed with spatial scale, and (3) the
amount of variability in the fish clusters that could be explained by those predictor variables.

Methods
Using cluster and indicator species analysis, we analyzed a 780-site database obtained
from USEPA’s EMAP western survey to determine 12-15 fish clusters from those sites at
three spatial scales (Pacific Northwest mountains, all mountains, all 12 conterminous states).
We next determined the predictor variables for those clusters through use of discriminant
function analysis and classification tree analysis. We applied Bray-Curtis mean similarity
analysis to assess fish assemblage similarity within a number of geographic and data
classification schemes.

Results & Discussion

Longitude, dams and temperature were the best predictors for all sites; longitude, dams
and catchment area were the top predictors for mountain sites; and latitude, turbidity, and
canopy density ranked highest for Pacific Northwest mountains. However, landscape and
site variables alone only accounted for half of the mean within-cluster similarity demonstrated
by the fish clusters. We conclude that, when available, using large fish assemblage, physical
and chemical habitat, and landscape data sets to predict fish assemblage patterns is
preferable to using preexisting landscape classifications, or only one set of site-scale
quantitative data.

469
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

References
Allen, A. P., Whittier, T. R., Kaufmann, P. R., Larsen, D. P., O'Connor, R. J., Hughes, R. M.,
Stemberger, R. S., Dixit, S. S., Brinkhurst, R. O., Herlihy, A. T., and Paulsen, S. G. (1999)
Concordance of taxonomic composition patterns across multiple lake assemblages: effects of
scale, body size, and land use. Canadian Journal of Fisheries and Aquatic Sciences 56: 2029-
2040.
Frissell, C. A., Liss, W. J., Warren, C. E., and Hurley, M. D. (1986) A hierarchical framework for
stream habitat classification: viewing streams in a watershed context. Environmental Management
10:199-214.
Hawkes, H.A. (1975) River zonation and classification. B. A. Whitton (Ed.). River Ecology. University
of California Press, Berkeley, pp. 312-374.
Hocutt, C. H., and Wiley, E. O. (1986) The Zoogeography of North American Freshwater Fishes.
Wiley, New York, New York.
Hughes, R. M., Heiskary, S. A., Matthews, W. J., and Yoder, C. O. (1994) Use of ecoregions in
biological monitoring. S. L. Loeb and A. Spacie (Eds.). Biological Monitoring of Aquatic Systems.
Lewis, Boca Raton, Florida. pp.125-151.
Pfleiger, W.L. (1971) A distributional study of Missouri fishes. University of Kansas Publication
Museum Natural History 20:225-570.
Pont, D., Hugueny, B., Beier, U., Goffaux, D., Melcher, A., Noble, R., Rogers, C., Roset, N., and
Schmutz, S. (2006) Assessing river biotic condition at a continental scale: a European approach
using functional metrics and fish assemblages. Journal of Applied Ecology 43:70-80.
Roth, N. E., Allan, J. D., and Erickson, D. L. (1996) Landscape influence on stream biotic integrity
assessed at multiple spatial scales. Landscape Ecology 11:141-156.
Southwood, T. R. E. (1977) Habitat, the templet for ecological strategies? Journal of Animal Ecology
46:337-365.
Steedman, R. J. (1988) Modification and assessment of an index of biotic integrity to quantify stream
quality in southern Ontario. Canadian Journal of Fisheries and Aquatic Sciences 45:492-501.
Wang, L., Seelbach, P. W., and Hughes, R. M. (2006a) Introduction to influences of landscape on
stream habitat and biological assemblages. R.M. Hughes, L. Wang, and P.W. Seelbach (Eds.).
Landscape Influences on Stream Habitat and Biological Assemblages. American Fisheries
Society, Bethesda, Maryland. pp. 1-23.
Wang, L., Seelbach, P. W., and Lyons, J.(2006b) Effects of levels of human disturbance on the
influence of catchment, riparian, and reach-scale factors on fish assemblages. R.M. Hughes, L.
Wang, and P.W. Seelbach (Eds.). Landscape Influences on Stream Habitat and Biological
Assemblages. American Fisheries Society, Bethesda, Maryland. pp. 199-219.

470
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Impact of barriers on aquatic species composition in Japan

M. Fukushima1, S. Kameyama1, M. Kaneko2, K. Nakao3

1
National Institute for Environmental Studies, Tsukuba, Ibaraki 305-8506, Japan.
e-mail: [email protected]
2
Faculty of Environment Systems, Rakuno Gakuen University, Ebetsu, Hokkaido Japan.
3
Hokkaido Aquaculture Promotion Corporation, Sapporo, Hokkaido, Japan

Introduction

Instream barriers such as dams have affected freshwater fish populations around the
globe (Bunn and Arthington, 2002). The effects of barriers are particularly detrimental to
diadromous fish species because they need to perform marine-freshwater migration during
their life cycles (Joy and Death, 2001). Previous studies have therefore focused on the
effects of dams on migratory species. In this study, however, we assessed the effects of
dams on both migratory and nonmigratory freshwater fish taxa in Hokkaido, Japan. Our aim
was to identify fish taxa that are either negatively or positively impacted by dams and to
predict the spatial extent and magnitude of the impacts.

Methods

Fish and dam data

We prepared a fish database that contains information about the presence or absence of
46 dominant fish taxa based on 13,701 surveys conducted during 1953–2003 (Fukushima,
2005; Fukushima and Kameyama, 2006). Based on the presence/absence data, we obtained
species richness data by counting the number of taxa captured in each survey. We combined
167 large dams (>15 m in height) and 1040 low-head dams that existed in Hokkaido as of
2000. These dams were mapped in a geographical information system (GIS), and sub-basins
fragmented by the dams were delineated across Hokkaido (Kameyama et al., 2004). All the
survey sites were then superimposed on the map of habitat fragmentation in the GIS to
determine whether each survey was conducted above dam(s) or not at the time sampling
was conducted, producing a covariate representing the presence or absence of the impact of
dams. Another covariate representing the length of time after being isolated from the sea due
to dams was prepared by subtracting a year of dam construction from a year of fish sampling
for every survey site with the impact of dams.

Statistical analysis

Both fish species richness and occurrence of each of the 46 taxa were modelled with a
series of generalized additive models (GAMs) using dam variables described above together
with other environmental variables as candidate predictors. The models were cross-
validated, and the discrimination performance was assessed by producing the receiver
operating characteristic curves.

Results and discussion

Fish species richness was plotted against elevation above sea level after dividing the data
into 3 groups with different survey periods (Fig. 1). The data were further divided into those
with no impact of dams (top 3 panels of Fig. 1) and those with the impact of dams (bottom 3
panels). The species richness exponentially decreased with increasing elevations regardless
of survey periods and damming impact. However, the richness appears to be slightly higher
during 1990s and after than the previous periods. This may be due to increased fish

471
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

sampling efficiency with the advent of electrofishing in Japan in early 1990s. Most
importantly, the fish species richness with the impact of dams was much lower at lower
elevations than species richness without the damming impact regardless of survey periods.
The GAMs of the fish taxa revealed that the occurrence probability of 19 taxa had been
either positively or negatively influenced by dams. All but one of the 10 taxa on which
damming had a negative influence were migratory taxa, indicating a strong barrier effect on
their migration. A positive damming impact was noted for five fish taxa, comprising three
strictly freshwater, one landlocked, and one anadromous (but commonly propagated) taxa.
The influence of dams on the fishes either occurred immediately after dam construction or
was gradual over time.

Figure 1. Freshwater fish species richness plotted against elevation for three periods
with and without dam(s) below the survey sites.

References
Bunn, S.E. & Arthington, A.H. (2002) Basic principles and ecological consequences of altered flow
regimes for aquatic biodiversity. Environmental Management 30: 492-507.
Joy M.K. & Death R.G. (2001) Control of freshwater fish and crayfish community structure in
Taranaki, New Zealand: dams, diadromy or habitat structure? Freshwater Biology 46: 417-429.
Kameyama, S; Fukushima, M.; Shimazaki, H.; Takada, M. & Kaneko, M. (2004) The watershed
fragmentation by dams and its impacts on freshwater fishes. N. Sappington (Ed). ESRI MAP
Book vol. 19. ESRI Press, New York, p. 89.
Fukushima, M. (2005) The dam-related decline of freshwater fish diversity—analyses of the data
collected from Hokkaido during the last half century. Japanese Journal of Ecology 55: 349-357 (in
Japanese).
Fukushima, M. & Kameyama, S. (2006) The effects of damming on masu salmon and the Sakhalin
taimen and the assessment of their conservation areas based on predictive habitat models.
Ecology and Civil Engineering 8: 233-244 (in Japanese).

472
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

The influence of population dynamics and landscape condition on pacific salmon

(Oncorhynchus spp.) re-colonization

G.R. Pess1,2, T. Quinn2, R. Hilborn2, K. Kloehn1

1
United States Department of Commerce (DOC), National Oceanic and Atmospheric,
Administration (NOAA), National Marine Fisheries Service (NMFS), Northwest Fisheries,
Science Center (NWFSC), 2725 Montlake Blvd East, Seattle, WA, U.S.A. 98112,
e-mail: [email protected]
2
University of Washington, School of Aquatic and Fishery Sciences (SAFS), Box 355020,
Seattle, WA, U.S.A. 98195-5020

Introduction

The colonization or re-colonization of aquatic habitats by salmonid fishes (Salmo,

Oncorhynchus, Salvelinus) at a watershed scale is related to the compatibility between
specific life history variation and landscape conditions that determine aquatic habitat
characteristics. Salmonids are well-suited for rapid colonization of newly created or opened
habitats because of two converse behaviours – homing and straying. Homing is defined as
the return of mature salmon to the general location of where their parents bred to spawn
(natal site), whereas straying is the return of mature salmon to a non-natal site to spawn. The
majority of salmon home to their natal sites, however straying has allowed salmonids to
colonize new habitats over their evolutionary history. Salmon can thus disperse, colonize
habitats, and establish self-sustaining populations. Natal sites are not static because habitats
change with anthropogenic alterations and other disturbances that force dispersal and
colonization of new habitats. What causes the strays to succeed and become colonists in
some cases? When do colonists become self-sustaining populations?

Background and Objectives

To answer these two questions we investigate how the establishment of pink salmon
(Oncorhynchus gorbuscha) populations in the Fraser River, British Columbia, Canada in
newly reopened habitats is related to specific life history variation and landscape conditions.
Pink salmon in the Fraser River were cut off from most of the watershed between 1913 and
the 1940s due to a rockslide at Rkm 209 that altered flow conditions and made adult fish
passage impossible (Roos 1996). Local spawning populations above the slide area
disappeared. Fish passage facilities developed in the 1940s allowed adult pinks to migrate
past the flow barrier and re-colonize the Upper Fraser. Re-colonization allowed pink salmon
to establish large self-sustaining populations above the landslide in one decade.
We hypothesize self-sustaining populations of colonists can be established when the
population growth rate of the colonizing population is greater than one. This occurs when
specific population and landscape factors are met. This includes an increasing population
size, a distance from source population that is comparable to existing migrations, a lack of
barriers, and the appropriate habitat type, quantity, and quality in the newly colonized
habitats.

Methods

We developed a general population model with multiple parameters to fit observed

spawning population growth for each watershed above the historic barrier. The models
include the colonizing population growth rate, harvest rate, immigration to the colonizing
population from the source population, a distance-dependent dispersal effect, an annual
relative population size effect, and a habitat area and habitat quality effect. We use

473
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

maximum likelihood techniques to estimate each of the parameters, and likelihood ratios to
compare each of the models to determine the one which best fit the observed data.

Results and Discussion

We found that self-sustaining spawning populations of pink salmon can be established

within 10 to 30 years of habitat being reopened to access. However, a self-sustaining
population in the uppermost watershed was not established during the dataset time period,
and instead became a sink population for two of the nearest spawning populations. Factors
that enabled pink salmon to successfully re-colonize included a physiological ability to swim
and efficiently migrate to newly opened habitats (MacNutt et al. 2006), an expanding
population size irrespective of the newly opened habitats, and life history variation that was
compatible with the newly opened habitat types and quantity. Differences in colonization
rates between the self-sustaining and non self-sustaining populations were related to
distance from source population, population growth rate, habitat area, and annual relative
population size. Environmental variables such as stream discharge also effected annual
variation redistributing upstream populations below natural flow barriers during low stream
discharge years. The results suggest that the combination of distance from source
population, habitat suitability, natural barriers, and population dynamics helped determine the
spatial and temporal patterns of Fraser River pink salmon re-colonization.

References
Roos, J. F. (1996) Restoring Fraser River Salmon. Pacific Salmon Commission, Vancouver, British
Columbia, Canada.
MacNutt, M.J.; Hinch, S.G.; Lee; C.G. Phibbs; J.R., Lotto; A.G. Healey, M.C. and Farrell, A.P.
(2006) Temperature effects on swimming performance, energetics, and aerobic capacities of
mature adult pink salmon (Oncorhynchus gorbuscha) compared with those of sockeye salmon
(Oncorhynchus nerka). Canadian Journal of Zoology. 84: 87-91.

474
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Multiscale analysis of the relationship among land use cover and streams water
quality in the Venice lagoon watershed

L. Favero1, A. Zuin 2, E. Mattiuzzo 2, G. Zanetto1, P. Ghetti3, D. Franco3

1
IDEAS – San Giobbe 873, 30121 Venice (Italy).
e-mail: [email protected]
2
Planland – Cannaregio 3841, 30121 Venice (Italy).
3
Università Cà Foscari – Dorsoduro 3246, 30123 Venice (Italy).

Introduction

The Venice lagoon is a fragile habitat threatened by water pollution. A regional plan has
been created to reduce nutrients inputs to Venice lagoon to 3000 t yr-1 for nitrogen and 300 t
yr-1 for phosphorus. The analyses of the relationships between the watershed Land Use (LU)
and the surface water quality could provide important information for management and
planning purposes. The LU patterns can be represented by simple percent values or by
metrics that consider the landscape spatial patterns. Together with LU, natural watershed
characteristics such as soil texture and permeability can influence water quality. The aim of
this study was (1) to analyse the relationship between streams nutrients concentrations and
LU patterns; (2) to verify the role of soil properties in this relationship; (3) to analyse the role
of landscape pattern; (4) to highlight a scale effect, (5) valuing the hedgerow network effect.

Methods

The Venice lagoon watershed covers about 2,124 Km2 and is 70% cultivated, mainly for
cereal crops. It is an alluvial, almost entirely flat plain, with clay loam sediments and is
divided into 8 monitored sub-basins (ca. 90% of the total watershed). Three years (2002-
2004) of ammonium and nitrate loads at the sub-basins outlets, digitised land use/land cover
maps, digitised soil characteristics maps, streams and basins boundaries were obtained from
Veneto Regional Agency for the Environment Protection.
LU types were aggregate into 7 classes: urban, agriculture, industrial, tree farming and
orchard, natural zones, vineyards, zootechnics. Among soil characteristics, soil texture
classes and soil hydrologic groups were used. Landscape metrics were selected from
literature studies; in particular, we used the Shannon-Wiener index for heterogeneity
(Franco, 2000) and the Effective Mesh Size index for fragmentation (Jaeger, 2000).
Landscape proximity analysis was conducted by buffering concentric zones around the
streams within each sub-basin. Zone widths (0-50 and 0-100 m from streams) were chosen
based on the size of the sub-basins and the literature analysis (Sliva and Williams, 2001).
The autocorrelation among variables was studied by means of correlation and Principal
Components Analysis (PCA). The selected independent set of variables was then regressed
against the water quality variables by means of multivariate models. The significance of the
obtained models was evaluated first for their ecological relevance and then using the Akaike
Information Criterion, corrected for small sample size (AICc, e.g.: Hamer et al., 2006).

Results

The correlation analysis revealed a strong correlation among soil texture and hydrologic
groups, which characterized also the first extracted component of PCA and MDS (data not
shown). The soil characteristics were thus represented by the PCA first extracted component
in the regression analysis (F1), in order to reduce the redundancy among predictors.
The models obtained by the regression analyses, selected for their ecological relevance
and their AICc are listed in Table 1. The two competing models for ammonium indicated
that NH4 loads were dependent on industrial LU at the whole watershed scale and on urban
LU at the 50m buffer scale, and on soil characteristics at both scales. Nitrate loads were

475
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

dependent on agriculture at the three scales, and on heterogeneity at the whole watershed
scale. The regressions performed using the most significant variables at the three spatial
scales were not significant (p>0.05, data not shown).

Discussion and conclusions

Ammonium loads were dependent mostly on industrial and urban LU: this may reflect a
higher concentration of sewage and waste disposal associated with urban and industrial
areas (Jones et al., 2001). The relationship with soil characteristics indicated that finer
textured, low permeable sub-basins have a higher NH4 content. In facts, clay minerals and
clay humics have a larger potential for adsorption of nutrients such as ammonia, and the low
permeability enhances overland flow that may easily carry particulates with adsorbed
nutrients into rivers (Sliva and Williams, 2001). The relationship among nitrate loads and
agriculture is frequently reported in literature (e.g. Sliva and Williams, 2001) and represents
the contribution of fertilizers to the non-point source pollution. The negative relation with
heterogeneity at the watershed scale can be explained by the impact of ecotone density
(generally matching with ditches in this ancient reclaimed land) on nitrate dynamic.
Several researchers have addressed the issue of whether LU near rivers is a better
predictor of water quality than LU over the whole watershed (e.g. Sliva and Williams, 2001),
obtaining contrasting results. Our results show that in the Venice watershed there were not
significant differences among spatial scales. This could be due to the highly impacted
structure of this landscape, where agriculture was 60-75%, urban LU 9-28% and natural
habitats covered less than 8%, even in the 50m buffer zone.

Table 1. Competing models. AG: agriculture; URB: urban; IND: industrial; F1: PCA factor
(soil texture and permeability); Shannon: heterogeneity index; AICc: corrected
Akaike Information Criterion. In brackets the sign of the relationship.
Nutrients Variables R2 AICc AICc Scale
Ammonium IND(+), F1(+) 0.83 -60.53 0.00 Watershed
URB(+), F1(+) 0.81 -58.81 1.72 50m

Nitrate AG(+) 0.69 -26.85 0.61 Watershed

Shannon(-) 0.67 -25.70 1.76 Watershed
AG(+) 0.71 -27.47 0.00 100m
AG(+) 0.67 -25.69 1.78 50m

References
Franco, D. (2000) Paesaggio, reti ecologiche e agroforestazione. Il Verde Editoriale, Milano, p. 24.
Hamer, T.L; Flather, C.H. & Noon, B.R. (2006) Factors associated with grassland bird species
richness: the relative roles of grassland area, landscape structure, and prey. Landscape Ecology
21:569-583.
Jaeger, J.A.G. (2000) Landscape division, splitting index, and effective mesh size: new measures of
landscape fragmentation. Landscape Ecology 15:115-130.
Jones K.B; Neale A.C; Nash M.S; Van Remortel R.D; Wickham J.D; Riitters K.H. & O’Neill R.V.
(2001) Predicting nutrient and sediment loadings to stream from landscape metrics: a multiple
watershed study from the united states mid atlantic region. Landscape Ecology 16:301-312.
Sliva L. & Williams D.D. (2001) Buffer zones versus whole catchment approaches to studying land
use impact on river water quality. Water Resouces 35: 3462–3472.

476
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Aquatic ecosystems in the core Cerrado: environmental impacts and

conservation

F.L. Tejerina Garro

Centro de Biologia Aquática, Universidade Católica de Goiás, Campus II, Av. Bela Vista s/n
Jardim Olímpico, CEP 74605010, Goiânia, Goiás, Brazil, Phone/Fax 62 32271722,
e-mail: [email protected]

Introduction

The Cerrado is the second greatest Brazilian biome covering originally 2·106 km2, that is,
fourteen Brazilian States (Ratter et al., 1997). Actually, the Cerrado core region (Goiás and
Tocantins State) is represented by 45.1% of the originally vegetation cover, but only 22.7%
remains in the Goiás State (Galinkin, 2003). It is classified as a hotspot (Myers et al., 2000).
Recent efforts have been conducted to qualify and quantify regional biodiversity, but these
have been limited to the terrestrial environment. This study aims to characterize the aquatic
ecosystems in Goiás based on the character of the hydrographic basins, their conservation,
biodiversity and the associated environmental impacts.

Hydrographic basins and human occupation

Four drainage basins are present in Goiás: the Araguaia River (86,109 km²), the
Tocantins River (102,120.6 km²), the Paraná River (149,488 km²) and the São Francisco
River (3,400 km²). The human occupation of these basins has increased since 1900 in the
south-north direction. Actually, the demographic density varies from low (e, g., São Miguel do
Araguaia 3.7 inh/km²) to high numbers (e. g., Brasília 352.16 inh/km²) (IBGE, 2002).

Conservation of the aquatic environment

The objectives of conservation in Goiás were determined between the 1960’s and the
1980’s, but do not include specified parts of the aquatic regional ecosystems. This seems to
be related to a) the predominant viewpoint that the aquatic environment is a component of
the terrestrial landscape; b) the fact that people are more familiar with terrestrial than aquatic
environments. However, based on actual knowledge about aquatic landscape ecology,
because the present Brazilian conservation policy is influenced by a terrestrial conservation
viewpoint, it is not very effective for the aquatic environment. On the other hand, a new
approach based on the hydrographic drainage basin (MMA, 2002) seems to be more efficient
to regional aquatic conservation process, despite the biological, political and social-economic
problems that it represents (Agostinho and Gomes, 2002).

Aquatic biodiversity

Data about aquatic biodiversity in Goiás is scarce. Among the nineteen aquatic taxonomic
groups considered for Brazilian continental water ecosystems, twelve (63.2%) do not have
available data in Goiás drainage basins. Systematized studies are irregular and mainly
focused on the Paraná River. This situation reflects the regional lag between the slow
generation of scientific knowledge and the rapid environment transformation because of
anthropogenic activities. Indeed, the aquatic environments in Goiás include headwaters,
where the probability of finding endemic species is considerable (Sá et al., 2003).

Environmental impacts
The deforestation, agriculture, the implantation and expansion of urban and industrial
areas, the construction of water reservoirs with different outcomes, and the introduction of

477
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

exotic species are the main anthropogenic activities that affect the aquatic environment in
South America (Pringle et al., 2000). This is also observed in the Goiás State.

Conclusion
There is a lag between the knowledge about the aquatic diversity and the rate of
modification of the Cerrado biome. Urge systematized studies about the aquatic flora and
fauna considering conserved and impacted environments.
The actual governmental policy do not has effective contribution to the conservation of the
aquatic environment. It is necessary a broad approach considering not only the terrestrial
environment but also the aquatic one.

References
Agostinho, A.A. & Gomes, L.C. (2002). Biodiversity and fisheries management in the Paraná River
Basin: successes and failures. In: World Fisheries Trust (org.). The Blue Millennium Project
Workshop: Managing fisheries for biodiversity. Victoria, BC., Canada. pp. 30.
Galinkin, M. (2003). Geo-Goiás 2002. Agência Ambiental de Goiás, Fundação Centro Brasileiro de
Referência e Apoio Cultural (CEBRAC), PNUMA, Secretaria do Meio Ambiente e Recursos
Hídricos de Goiás (SEMARH), Brasília, pp. 272.
IBGE Instituto Brasileiro de Geografia e Estatística (2002). Atlas Nacional do Brasil. 4ta. Edição.
Available in: <http://mapas.ibge.gov.br>.
MMA. Ministério do Meio Ambiente (2002). Recursos Hídricos: conjunto de normas legais. 2.ed.
Brasília: Secretaria de Recursos Hídricos, pp. 141.
Myers, N.; Mittermeier, R.A.; Mittermeier, C.G.; Fonseca, G.A.B. & Kent, J. (2000). Biodiversity
hotspots for conservation priorities. Nature 403(24):853-858.
Pringle, C.M.; Scatena, F.N.; Paaby-Hansen, P. & Núñez-Ferrera, M. (2000). River conservation in
Latin America and the Caribbean. In: P. J. Boon, B. R. Davies, G. E. Petts (eds.). Global
perspectives on river conservation: science, policy and practices. John Wiley & Sons. Chichester,
pp. 41-78.
Ratter, J.A.; Ribeiro, J.F. & Bridgewater, S. (1997). The Brazilian Cerrado vegetation and threats to
its biodiversity. Annals of Botany 80:223-230.
Sá, M.F.P. de; Fenerich-Verani, N. & Fragoso, E.N. (2003). Peixes do cerrado em perigo. Ciência
Hoje 34(200): 68-71.

478
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Rivers and roads controlled by, and central players in, urban regions

R.T.T. Forman

Harvard University, Graduate School of Design, Cambridge, MA 02138, USA e-mail:

[email protected]

Introduction

Two central questions are explored: (1) the effect of an urban region on rivers; and (2)
the role of rivers and roads in the region.

Analysis and Discussion

Analyzing 38 urban regions of large-to-small cities on all continents (plus an intensive

ecological planning study of the Barcelona Region; Forman 2004) found 40% of the cities by
major rivers, and another 10% at the intersection of major rivers (Forman 2007). Cropland is
the predominant human land use around rivers and major streams in almost all of the urban
regions, suggesting extensive sediment and agricultural chemicals in river water. Forty
percent of the regions have considerable (10-40%) built area around rivers/streams, implying
much urban-pollutant runoff and other human impacts. Most regions have <33% natural
vegetation cover near rivers/streams, while only a fifth of the regions has >80% to support
good water quality, aquatic ecosystems and fish populations; another fifth has 40-70%
natural cover.

The typical urban region worldwide offers many unusual and powerful attributes that affect
the river, including: widespread built areas and hard surface on slopes; diverse heavily used
infrastructure along the river; intense straightening and squeezing of river and tributaries;
elimination of most wetlands; a huge human population with manifold daily effects; low
percent of natural vegetation cover remaining; very little stream-corridor vegetation
remaining; intense conversion of groundwater and river-water to water vapor; a huge human
wastewater source; commonly a single pipe system for draining stormwater and wastewater;
an enormous density of big and little roads; and a large and rapidly growing mass of moving
vehicles.

Little roads cross and run along floodplains for maintenance and repair of the city’s major
infrastructure conduits. Grids of little roads in the city and on the slopes help accelerate
stormwater and pollutant flows, raising peak flows and flooding (Forman et al. 2003). In
general, little roads cross streams, while highways parallel rivers. Roads and vehicles
provide linear streaks of pollutants that reach rivers. Fill on the lower side of roads provides
much sediment to streams and rivers. In addition, highways cut the regional connectivity of
movement for terrestrial wildlife, highlighting the importance of wildlife underpasses and
overpasses.

All this results in normal low-water flows, periodic floods with major damage to human
structures, heavily polluted downslope flows, blockage of fish migration, cleaning of the city
by river carrying away wastes, and severely degraded water quality, aquatic ecosystems and
fish populations. Contrasting patterns normally appear upriver and downriver of the major
city.

479
Theme 4: Ecohydrology, water and rivers
4.1 Symposium 8: Landscapes and rivers

Conclusion

In brief, urban region characteristics, including roads, almost overwhelm river

ecosystems, or render them outliers to be mainly modeled by urban-region parameters. At
the same time these rivers and roads are the major conduits and barriers in the urban region,
effectively determining where species and people live and move. Landscape ecologists
seem to have more to offer for understanding and solutions for society than any other field.

References
Forman, R.T.T., Sperling, D., Bissonette, J. A., et al. (2003) Road Ecology: Science and Solutions,
Island Press, Washington, D.C.
Forman, R.T.T. (2004) Mosaico territorial para la region metropolitana de Barcelona, Editorial
Gustavo Gili, Barcelona.
Forman, R.T.T. (2007) Urban Regions: Ecology and Planning Beyond the City, Cambridge University
Press, Cambridge (in press).

480
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context:

implications for catchment water management

Approaches to Link Microbial Structures and Processes at the Microscale to

Ecosystem Functioning

A. Ogram1, R. Corstanje2, M. Christman3

1
Soil and Water Science Dept., University of Florida, Gainesville, FL 32611-0290 USA.
e-mail: [email protected]
2
BAB Division, Rothamsted Research, Harpenden, AL5 2JQ, UK
3
Statistics Dept., University of Florida, Gainesville, FL, 32611-0339, USA

The functions of soil microbial communities constitute essential links in biogeochemical

cycles, such that knowledge of the factors that control the structures and activities of these
communities is essential for a mechanistic understanding of biogeochemical cycling. Thanks
to the development of increasingly sophisticated tools based in molecular biology, the ability
to characterize the structure of soil microbial communities has increased dramatically over
the last ten years. Environmental factors that determine the structures and activities of these
extraordinarily complex communities are poorly understood, however, as are the quantitative
relationships between individual community structures and activities, and the pathways and
rates of the major biogeochemical cycles.
Of particular interest are quantitative relationships between microbial communities and
the processes they control at landscape scales. Many DNA-based procedures utilize sample
sizes of 0.25 g soil. Samples of this size provide great opportunities for spatial resolution of
microbial communities on small scales, but present challenges when attempting to
extrapolate to the landscape scale. More research, including development and application of
advanced methodological and statistical approaches, is required to successfully establish
quantitative links between DNA-based data concerning community structure and
environmental parameters across large scales.
A limitation to scaling from 0.25 g samples to landscape scales has been the high cost
and labor involved in most standard molecular ecological methods, which preclude their use
in analysis of large numbers of samples typically required for landscape analysis. These
limitations are being overcome, however, thanks in large part to advances made in
genomics. Rapid fingerprinting approaches such as terminal restriction fragment length
polymorphism (T-RFLP) analysis are in common use, and microarray analyses are being
developed for ecological applications. Perhaps of greatest potential for analysis of microbial
community structure is high-throughput sequence analysis of clone libraries from disparate
sources. DNA-based approaches such as these have little meaning, however, if they are not
combined with biogeochemical parameters that control, and are controlled by, their activities.
Most studies that attempt to link DNA-based analyses with environmental parameters
have relied on relatively simple analytical approaches such as Principal Components
Analysis (PCA). PCA can be useful when analyzing aspects of community structure that are
linked to a dominant characteristic of the landscape, such as in the cases of sulfate reducing
prokaryotes (SRP) and methanogens along nutrient gradients in the Florida Everglades
(Castro et al., 2005). In this study, the distribution of genes representative of SRP and
methanogens in monthly samples were characterized via T-RFLP analysis. Distributions of
genotypes characteristic of SRP and methanogens depended on position along the gradient,
and knowledge of the general physiological attributes of cultured representatives of those
genotypes allowed development of hypotheses to explain the distribution. In general, genes
characteristic of SRP capable of utilizing a broad range of electron donors (“generalists”)
dominated soils from high nutrient regions, while genes characteristic of SRP limited to a few
electron donors (“specialists”) dominated the low nutrient soils. Possible reasons for
selection of specific methanogens may be related to differing H2 concentrations along the

481
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

nutrient gradient. This general approach has been used in many applications, such as to
define relationships between microbial community structure and plant community (Blume et
al, 2005), and between methanotrophs land use and (Ogram et al., 2006).
PCA and related approaches can be valuable but are limited in scope; they provide
information on correlations between observed species distributions an environmental factors,
but are of less value in identifying complex relationships that control community structure and
activity. A drawback of some of the ordination techniques is the ambiguity and general
descriptive nature of the factor, specifically that PCA factor scores have a large degree of
associated ambiguity, too large to be used as composite environmental indexes (Yu et
al.1998). As a result, alternative multivariate approaches have been used in many soil
microbial ecology studies. One such study was to investigate the controls on microbial
community structure on global scales (Fierer and Jackson, 2006). The analytical approaches
for community structure were similar to those described above (T-RFLP analysis), but the
links between T-RFLP derived diversity indices and environmental parameters were defined
using non-metric multidimensional scaling and partial Mantel tests. Noteworthy in their
findings was that microbial community diversity is determined primarily by soil
biogeochemical characteristics related to pH.
Biogeochemical processes and multiple constituent measures have complex inter-
relationships. The relationship between those processes and the diversity and composition of
soil bacterial communities is equally complex. The common thread in many studies linking
communities with processes is that there are some underlying factors or latent variables that
link the observed variables describing the biogeochemistry or community composition.
Relationships between community composition and biogeochemistry are established by
determining the degree one explains the variation in the other through either regression or
another measure of association or dependence. This process can be formalized using
structural equation modeling (SEM) (Shipley, 2000). SEM is based on the premise that there
is a directed graph (or web) describing the relationships among the variables of interest (a
path model) but that the underlying latent causal variables are described by other variables
observed with measurement error. For example, microbial community composition is not
directly measured but is represented by T-RFLP fingerprinting. In a statistical analysis of an
SEM, the observed variables ‘load’ onto the latent factors, which are the constructs
representing some meaningful process or characteristic. The relationships between the latent
factors then are described by path coefficients. To illustrate an analysis using SEM, we
constructed a structural equation model for the soil biogeochemistry and microbial
community using the same dataset as used by Castro et al. (2005).

References
Castro, H., S. Newman, K.R. Reddy, and A.V. Ogram (2005) Distribution and stability of sulfate
reducing prokaryotic and hydrogenotrophic methanogenic assemblages in nutrient-impacted
regions of the Florida Everglades. Appl. Environ. Microbiol. 71: 2695-2704.
Fierer, N. and R.B. Jackson (2006) The diversity and biogeography of soil bacterial communities.
PNAS USA 103:626-631.
Shipley, B. (2000) Cause and Correlation in Biology: A User’s Guide to Path Analysis, Structural
Equations and Causal Inference. Cambridge University Press: Cambridge, UK.
Yu, C.C., Quinn, J.T., Dufournaud, C.M., Harrington, J. J., Rogers, P. P., and Lohani, B. N. (1998)
Effective dimensionality of environmental indicators: a principal component analysis with bootstrap
confidence intervals. J. of Environ. Manage. 53:101–119.

482
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

Progress and challenges in demonstrating riparian buffer effects on nutrient

discharges from whole catchments

D.E. Weller1, M.E. Baker2, T.E. Jordan1

1
Smithsonian Environmental Research Center, 647 Contees Wharf Rd., Edgewater, MD,
21037-0028, USA. e-mail: [email protected].
2
Utah State University, Dept. of Watershed Sciences, 5210 Old Main Hill, Logan, Utah
84322-5210, USA.

Introduction

Riparian forests and wetlands are important nutrient-retaining landscape elements. Many
field studies report strong nutrient retention along sampling transects from source areas
through riparian buffers to streams. Efforts to demonstrate the impacts of buffers at the
whole catchment scale often use statistical models which relate observed nutrient discharges
from many catchments to variables describing catchment land use and riparian buffers.
These studies have not consistently demonstrated that riparian buffers are important to
predicting nutrient discharges from whole catchments (Johnson et al. 1997, Weller et al.
1998, Jones et al. 2001). We have developed new approaches to testing for the effects of
buffers in GIS-based statistical models of whole catchment nutrient discharges.

Functionally-based riparian buffer metrics

Efforts to quantify the effects of riparian buffers in whole catchments have been
confounded by a commonly used analysis which estimates buffer potential as the percentage
of forest or wetland within a fixed distance of streams. This method of estimating buffer
potential is not based on a clear conceptual model and does not match the scale at which
field studies have documented buffer nutrient uptake. To better quantify the distribution of
riparian buffers in a catchment, we developed new functionally-based metrics that account
for the length of buffer traversed by every hydrologic flow path connecting a nutrient source
area to a stream (Baker et al. 2006). This approach correctly “scales up” from transects (flow
paths) considered in field studies to whole catchments. We applied the new approach to 503
catchments in four different physiographic provinces of the Chesapeake Bay Drainage. For
comparison, we also analyzed land cover within a fixed distance of streams. There were
distinct patterns of relationship between catchment and near-stream land cover in each
physiographic province and strong correlations with catchment land cover confounded fixed-
distance metrics. In contrast, our flow-path metrics were more independent of catchment
land cover than fixed-distance measures and provided greater detail, interpretability, and
flexibility than the fixed-distance approach (Baker et al. 2006)

Effects of stream map resolution

Geographic studies of buffer effects in whole catchments have also used stream maps of
different resolutions without explicitly considering the possible effects of map resolution on
estimates of source-stream connectivity and overall buffer effectiveness. For 503
catchments of the Chesapeake Bay drainage, we applied our functionally-based riparian
metrics (Baker et al. 2006) to explore how differences in stream map resolution affect
inferences about the effective placement of buffers within a catchment (Baker et al. in press).
Increasing stream map resolution reduced the mean distance from source areas to streams,
reduced mean buffer width, and increased the frequency of buffer gaps. Overall, increasing
stream map resolution led to reduced estimates of nutrient retention potential in riparian
buffers. In some catchments, switching from a coarse resolution to a fine resolution stream
map completely changed the perception of a stream network from well buffered to very

483
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

poorly buffered. Previous broad-scale analyses of riparian buffers based on coarse-

resolution stream maps may have overestimated landscape-level buffer prevalence and
effectiveness.

Relating buffer prevalence and distribution to catchment nutrient discharges

We developed statistical models that explicitly test the statistical significance of a riparian
buffer effect as an addition to non-spatial models relating nutrient discharges to land use
proportions. In the Chesapeake Bay drainage, information on riparian buffer distribution
explained a small but statistically significant amount of the variation in nitrate discharges
among rural Coastal Plain catchments, but not among rural catchments in the other
physiographic provinces of the Bay’s drainage. This result is consistent with our previous
finding that Coastal Plain catchments retain more riparian forest along flow paths connecting
croplands to streams than do catchments in the other physiographic provinces (Baker et al.
2006, in press).

Conclusions and challenges

Our new functional riparian metrics and our analysis of the effects of stream map
resolution address problems that have confounded efforts to demonstrate and quantify the
effects of riparian buffers on catchment nutrient discharges. Despite these advances, our
statistical models relating buffer patterns to stream nutrient measurements still found only
weak evidence of buffer effects on catchment nitrate discharges and only in one of four
physiographic provinces of the Chesapeake Bay drainage. It may be possible to document
stronger buffer effects after improving analyses by using land cover data of finer spatial
resolution and greater categorical accuracy, by incorporating methods to identify and map
buffers that are more likely to be biogeochemically active, and by considering watersheds
with broader ranges of riparian configurations. In the meantime, our new methods already
offer improvements in accounting for buffers in statistical nutrient models, in quantifying
spatial patterns for process-based modeling, and in targeting management actions such as
buffer restoration (Baker et al. 2006, in press).

References
Baker, M.E.; Weller, D.E & Jordan, T.E. (in press) Effects of stream map resolution on measures of
riparian buffer distribution and nutrient retention potential. Landscape Ecology.
Baker, M.E.; Weller, D.E & Jordan, T.E. (2006) Improved methods for quantifying potential nutrient
interception by riparian buffers. Landscape Ecology 21:1327-1345.
Johnson, L.B.; Richards, C.; Host, C.E.; & Arthur, J.W. (1997) Landscape influences on water
chemistry in Midwestern stream ecosystems. Freshwater Biology 37:193-208.
Jones, K.B.; Neale, A.C.; Nash, M.S.; Van Remortel, R.D.; Wickham, J.D.; Riiters, K.H. &. O'Neill,
R.V. (2001) Predicting nutrient and sediment loadings to streams from landscape metrics: A
multiple watershed study from the United States Mid-Atlantic region. Landscape Ecology 16:301-
312.
Weller, D.E.; Jordan, T.E. & Correll D.L (1998) Heuristic models for material discharge from
landscapes with riparian buffers. Ecological Applications 8:1156-1169.

484
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

Linkages between surface and subsurface hydrology and ecological functioning

of mangrove ecosystems in Ft. Pierce, Florida

D.F. Whigham1,3, C.E. Stringer2, M.C. Rains2, J.T.A. Verhoeven3,P. Baas3, P.J.M.
van der Ven3
1
Smithsonian Environmental Research Center, P.O. Box 28, Edgewater, MD 21037, USA.
email:[email protected]
2
Department of Geology, University of South Florida, Tampa, USA
3
Landscape Ecology, Institute of Environmental Biology, Utrecht University, P.O. Box 80084,
3508 TB Utrecht, The Netherlands

Introduction

Hydrological processes have a dominant impact on the structure and function of wetland
ecosystems, including coastal wetland that in the tropics are dominated by mangroves.
Globally, mangroves have been impacted by a variety of activities that result in significant
modifications of the original hydrologic conditions. Mangrove systems comprise ~2,500 km2
of south Florida and many are located in the Indian River Lagoon, a series of connected
estuaries that extends ~250 km along the east coast of Florida. Most of the mangrove
systems in the Indian River Lagoon have been ditched and impounded for mosquito control
but are still hydrologically linked by surface water and groundwater pathways to the Indian
River Lagoon. Our study site is a mosquito impoundment located on a carbonate barrier
island near Ft. Pierce, Florida. A consequence of mangrove alterations has been the
development of a complex matrix of vegetation within the impoundment that ranges from
high salinity salt pannes in which no mangrove trees grow to areas adjacent to open water
that support tall mangroves.

This paper focusses on the relationships between the structure and functioning of mangroves
and variations in hydrologic conditions. A goal of the research is to understand the
relationships between hydrologic conditions at a range of scales from small-scale processes
associated with nitrogen, the nutrient that limits the growth of mangroves at this site (Feller et
al. 2003), cycling to the relationship between surface and subsurface hydrology and the
distribution, structure and function of different variants of vegetation. Dominant mangroves
at the study site are Rhizophora mangle (Red mangrove), Avicennia germinans (Black
mangrove), and Laguncularia racemosa (White mangrove).

Hydrologic setting

The common conceptual model of carbonate barrier island hydrogeology is that a freshwater
lens rests on top of saline groundwater and that salinity variations are largely due to spatial
and temporal variations in precipitation and subsequent freshwater runoff and recharge.
Results of our preliminary hydrologic studies suggest that this conceptual model is incorrect.
Salinities vary spatially, with surface water and groundwater salinities ranging from ~10 ppt in
the upland groundwater, to ~30 ppt in the regularly-flushed mangrove surface water and
groundwater, to ~75 ppt in the irregularly-flushed mangrove surface water and groundwater.
These high salinities extend to depths of many meters. However, salinities do not
significantly differ temporally, and relative cation and anion concentrations do not vary
spatially or temporally. These results indicate that salinity variations are largely due to
variable mixing between precipitation and lagoon water and the subsequent
evapoconcentration in the irregularly-flushed areas rather than to variations in precipitation
and subsequent freshwater runoff and recharge.

Hydrology and vegetation patterns

485
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

The spatial variations in salinities correlate with spatial variations in species composition and
abundance, with maritime hammock, Red mangrove, dense Black mangrove, sparse Black
mangrove, and salt pannes being arranged on a gradient of increasing salinities. Vegetation
structure varies from one habitat type the other.

Hydrology and nitrogen cycling

Spatial variations in nutrient concentration in soils and plants and rates of nitrogen
mineralization, denitrification, and nitrification correlate with spatial variations in species
composition and abundance and vary from one habitat to another. Based on preliminary
DNA analyses of ammonia-oxidizers in the β-proteobacteria there are also clear differences
across the gradient in the composition of the microbial community responsible for nitrification
(Figure 1), especially in the habitats (sparse Black mangrove, salt pan) with high salinities.

Figure 1: Dendrogram illustrating the similarity in species composition at t=0 and the fertilized
sub-plots at t=2 in five different mangrove habitats. The results of t=0 and t=2 are shown in
one dendrogram to indicate a possible fertilization caused clustering. Between brackets the
measurement time at which the samples were collected is indicated. R: R. mangle; SP: salt
panne; DA: dense A. germinans; SA: sparse A. germinans; F: forest. The circles indicate
branches in the Dendrogram indicting a cluster with a similarity of above 60 percent.

References
Feller, I.C., D.F. Whigham, K.L. McKee and C.E. Lovelock. 2003. Nitrogen limitation of growth and
nutrient dynamics in a disturbed mangrove forest, Indian River Lagoon, Florida. Oecologia 134:
405-414.

486
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

From the micro- to the macroscale: effects of microbial Fe(III) and sulfate
reduction on element fluxes in acidic peatlands

K. Küsel

Limnology Research Group, Institute of Ecology, Friedrich Schiller University Jena,

07745 Jena, Germany.
e-mail: [email protected]

Introduction

Since wetlands are considered a major unknown regarding the influence of global climate
change on element dynamics, knowledge of processes and conditions controlling the sink
and source function of redox processes is crucial. Over more than a century wetlands
received sulfate, nitrate and acid impact either from direct atmospheric deposition or from
the desorption and release of matter from upland sites (Wieder, 1985). Athough peatlands
are known to be significant sources of the green house gas methane, minerotrophic acidic
fens can receive nitrate, sulfate and Fe(II) from groundwater flow: Thus, nitrate, Fe(III) and
sulfate can be available as alternative electron acceptor for the oxidation of organic matter
especially under alternating redox conditions in minerotrophic fens.

Redox processes and microbial populations in fens

Fens located in northern Bavaria receive sulfate and Fe(II) from groundwater flow. Thus,
considerable amounts of Fe(III) (up to 18 g kg-1 soil) and small amounts of sulfate (up to 300
µM) are available (Paul et al., 2006). Biogeochemical parameters measured in the fen soil
solution during 5 years documented that the upper 5 to 10 cm can be exposed to drying and
oxygenation. Periodic oxygenation reached a maximum depth of 25 cm in the water
saturated fen and was concomitant with relative high concentrations of nitrate and sulfate.
The fen soil was permanently anoxic below 30 cm depth with average concentrations of
sulfate below 40 µM and maximum concentrations of methane. Anoxic incubation
experiments with peat samples from the oxidized surface layer demonstrated high Fe(III)-
reducing but negligible sulfate-reducing or methanogenic activities. Reduction of Fe(III) can
be accompanied by a release a phosphorous to the soil water leading to an export to
adjacent surface waters.
Numbers of Fe(III)-reducing bacteria cultured at pH 5.5 approximated 105 to 106 cells g (fresh
weight soil)-1 and did not decrease with soil depth. Detection of sulfate reducers with
Terminal-Restriction Fragment Length Polymorphism (T-RFLP) analysis of amplified
dissimilatory (bi)sulfite reductase genes (dsrAB) separated three subgroups along the depth
profile (Schmalenberger et al., 2007). Cloning of dsrAB PCR products yielded a total of 84
distinct dsrAB restriction patterns (genotypes). Most of the genotypes were unique.
Differences in the sulfate-reducing community structures suggested that sulfate reducers
present in the upper fen soil have to tolerate O2 and even drying, whereas sulfate reducers
present in deep anoxic zones may act as synthrophic fermentors in cooperation with H2-
utilizing methanogens. Under short term conditions, fixation of sulfur as inorganic reduced
sulfur seems to play a major role in peatlands. However, this pool appears to be rapidly
oxidized, and sulfate is exported to adjacent ecosystems during heavy rainfall. The majority
of reduced sulfur, which is stored as organic sulfur, seems to be more stable.

Conclusions

Thus, despite ongoing Fe(III)- and sulfate-reducing activities, these peatlands can not be
considered as effective sinks for sulfur. Drying or oxygenation of the top soil favors Fe(III)- or
sulfate-reducing activities due to a renewal of the electron acceptor pools. Enhanced drying

487
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

events caused by climate change might accelerate not only carbon mineralization processes
but also further inhibit methanogenesis. An increase of extreme meterological conditions can
turn peatlands temporarily to a source for carbon which might effetc the global carbon carbon
budget. In addition, rewetting of drained Fe(III)-rich peatlands can enhance eutrophication of
adjacent surface waters.

References
Paul, S., Küsel, K., & Alewell, C. (2006). Reduction processes in forest wetlands: tracking down
heterogeneity of source/sink functions with a combination of methods. Soil Biology &
Biochemistry, 38: 1028-1039.
Schmalenberger, A., Drake, H. L. & Küsel, K. (2007). High unique diversity of sulfate-reducing
bacteria in a depth gradient in an acidic fen, Environmental Microbiology (in press)
Wieder, R.K. (1985). Peat and water chemistry at Big Run Bog, a peatland in the Appalachian
mountains of West Virginia, USA. Biogeochemistry 1, 277-302.

488
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

Spatio dynamic modelling of the hydrological processes to investigate the

functionality of nutrient-retaining landscape features

P. Merot, P. Durand

INRA, Agrocampus Rennes, UMR1069, SAS. F-35000 Rennes, France.

e-mail: [email protected]

It is largely recognised that the functionality of Nutrient-Retaining landscape Features

(NRLFs) depends both on the local hydrological processes and on the hydrological setting of
the catchment (Haycock et al. 1997). Due to the complex nature of a catchment,
understanding the NRLF impact at the catchment scale, and accounting for these structures
in the management of water resources call for the use of spatio-dynamic hydrological
modelling.
The purpose of this paper is threefold: First, a generic scheme of the buffer functions is
given to understand the multiple ways NRLF can act as a buffer (Viaud et al. 2004); then, the
major part of the paper proposes a review of the NRLF hydrological modelling at the
catchment scale; lastly we present some critical issues that modelling is facing.
Two types of modelling are distinguished, status driven and structure driven; the status
driven models do not consider separately the buffering landscape structures, but are based
on a continuous representation of the spatial distribution of the soil and subsoil water content
in the catchment. MODFLOW (Mc Donald and Harbaugh 1988) or TOPMODEL (Beven and
Kirkby 1979) are such models that allow to include the impact of some natural buffer zones
such as riparian wetlands without any a priori description or conceptualisation on buffer
structures. By contrast, in structure driven modelling, the natural or the man-made landscape
structures are considered as distinct entities. Depending of the objective, these models are
used either as heuristic models (Weller et al. 1998), generally applied to crudely described
catchments, testing quite specific assumptions on some buffering structure (Beaujouan et al.
2001), or as scenario based modelling, taking into account complex interactions (Figure 1,
Viaud et al. 2005). The machine learning approach for evaluating the impact of land use and
management practices is an original approach based on a high-level scenario language.
30 250
No hedge
relative decrease of annual discharge (%)

thinly distributed network

25 27 m/ha
thinly distributed network downhill 200
60 m/ha
Annual N output (t)

dense network
20 dense network downhill 100 m/ha
150
200 m/ha
15
100
10

50
5

0 0
400 600 800 1000 1200 1400 93-94 94-95 95-96 96-97 97-98 98-99 99-00 00-01 01-02
annual precipitation (mm)

Figure 1 : Simulations of the effect of different hedge network density and distribution on
annual discharge (left) and annual N flux (right) at the outlet of a Brittany rural catchment.

The current development of virtual landscapes for modelling is the more recent attempt to
capture the complexity of the landscape, and specifically nutrient-retaining landscape
features. They allow controlled virtual experiments based on environmental, economic or
social scenarios. These different models and ways to apply them are explored (Ruiz et al.
2004, Weiler and McDonnell 2004).
Whatever the model, modellers of NRLFs hydrology are facing difficult issues. The first one
is the question of the topology of the water pathways and their connectivity with the NRLFs.
The second one is the question of the spatial scale: to what extent does the NRLF affect the
catchment as a whole? Does it depend on the size –or order- of the stream? The last point -
and from a hydrological perspective it may be the most important-, is the question of what we

489
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

can call the hydrological time (figure 2). Is the model able to take into account the non-
linearity and the thresholds of the NRLF functioning? Is the model able to integrate a
topology and connectivity that is variable in time? Is the model of NRLF able to simulate the
different temporal phases of the hydrological system, during which the NRLF functioning is
really different, namely the dry period, the flow resume period, the wet period, and the crisis
periods (periods of rainfall-runoff event)?

120 1
retention
Nitrogen load (kg/ha/year)

100 0.8

Nitrate fate factor

80
0.6
60
0.4
release
40
0.2
20 storage
0 0
1960 1990 2020 2050 2080 2110 1960 1990 2020 2050 2080 2110
year year

Figure 2 Simulated time series of nitrogen input (left) and nitrogen fate factor
(output/input, right) for a high rainfall, high nitrogen input catchment. (Basset-Mens et al.
2006)
Finally, a large set of models is available to simulate the functionality of NRLFs. They vary
widely in complexity and have been applied for a diversity of objectives, from a heuristic to a
management perspective. They confirm that the hydrological processes at the catchment
scale are key factors for the functioning of retaining nutrient landscape features. However,
there is still a strong need for experimental studies and observations in the real world to
improve and validate these models.

References
Basset-Mens, C., L. Anibar, P. Durand, and H. M. G. van der Werf. (2006) Spatialised fate factors
for nitrate in catchments: Modelling approach and implication for LCA results. Science of the Total
Environment 367:367-382.
Beaujouan, V., P. Durand, and L. Ruiz. (2001) Modelling the effect of the spatial distribution of
agricultural practices on nitrogen fluxes in rural catchments. Ecological Modelling 137:93–105.
Beven, K., and M. J. Kirkby. (1979) A physically based variable contributing area model of basin
hydrology. Hydrological Sciences Bulletin 24:43-69.
Haycock, N. E., T. P. Burt, K. W. T. Goulding, and G. Pinay. (1997) Buffer zones: their processes
and potential in water protection. Quest Environmental, Hardfordshire UK.
Mc Donald, M. G., and A. W. Harbaugh. (1988) A Modular Three-Dimensional Finite-Difference
Ground-Water Flow Model. U.S. Geological Survey.
Ruiz, L., P. Aurousseau, B. J., V. Beaujouan, P. Cellier, P. Curmi, P. Durand, C. Gascuel-Odoux,
P. Leterme, J. L. Peyraud, C. Thenail, and C. Walter. (2004) Conception de bassins versants
virtuels : vers un outil pour l'étude de l'influence de l'organisation spatiale de l'activité agricole et
du milieu physique sur les flux d'azote dans les bassins versants. pp 337-354 in P. Monestiez, S.
Lardon, and B. Seguin, editors. Organisation spatiale des activités agricoles et processus
environnementaux. INRA Editions.
Viaud, V., P. Durand, P. Merot, E. Sauboua, and Z. Saadi. (2005) Modeling the impact of the spatial
structure of a hedge network on the hydrology of a small catchment in a temperate climate.
Agricultural Water Management 74:135-163.
Viaud, V., P. Merot, and J. Baudry. (2004) Hydrochemical buffer assessment in agricultural
landscapes: from local to catchment scale. Environmental Management 34:559-573.
Weiler, M., and J. McDonnell. (2004) Virtual experiments: a new approach for improving process
conceptualization in hillslope hydrology. Journal of Hydrology 285:3-18.
Weller, D. E., T. E. Jordan, and D. L. Correll. (1998) Heuristic models for material discharge from
landscapes with riparian buffers. Ecological Applications 8:1156-1169.

490
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

The Operational Landscape Unit concept as a guidance for landscape restoration

and water quality enhancement

J.T.A. Verhoeven, M.B. Soons

Landscape Ecology, Institute of Environmental Biology, Utrecht University, P.O. Box 80084,
3508 TB Utrecht, The Netherlands.
e-mail: [email protected]

Introduction

The scope of nature conservation has widened to include spatial considerations in nature
conservation strategies, aimed first predominantly at preservation of large areas and later
also at preservation of networks of habitat areas, the ‘ecological networks’. The need for this
approach to nature conservation in densely populated regions arises from the increasing
degree of landscape fragmentation. This fragmentation has been shown to have dramatic
consequences for biodiversity and for environmental quality. In this paper, we want to
emphasize the importance of lateral connections and their pivotal role in maintaining
biodiversity and ecological functioning at a much higher level than would be possible for
unlinked ecosystems within the same landscape. We propose the Operational Landscape
Unit concept, which combines biotic and hydrological connectivity to analyze the best
options for restoration of wetland ecosystem functioning and plant biodiversity in fragmented
landscapes.

Consequences of landscape fragmentation

A first consequence of landscape fragmentation is the transformation of large natural

landscape patches connected by wide, continuous corridors into a large number of small
patches, many of which are not connected to each other (Soons et al. 2005). This
development results in a loss of area because of habitat destruction, e.g. for housing or
agriculture, and a decrease in connectivity because of destruction of corridors or the
construction of barriers such as roads. As predicted by the ecological theory on island
biogeography, the fragmentation of the landscape into small patches leads to high extinction
and a high degree of isolation, resulting in low rates of colonization and a loss of species
richness from patches and the landscape as a whole (Hanski 2005).
A second consequence of fragmentation is the drastic change in the hydrological
functioning of catchments. Measures to drain agricultural land have led to lower groundwater
tables, loss of groundwater discharge in (semi-)natural landscape patches, straightening of
lower-order streams, dampening of stream water level fluctuations, loss of floodplain habitat,
loss of meandering as a natural stream habitat process and a deterioration of stream water
quality (Brierley & Stankoviansky 2002). These changes have led to drastic changes in
ecosystem functioning, often reducing (the diversity of) habitat site conditions all across the
landscape, even in protected reserves with appropriate internal management. It is evident
that landscape-scale modifications, including hydrological measures and land amelioration
such as levelling of microtopography, have resulted in additional losses of biodiversity, which
have probably been just as severe as those caused by fragmentation per se.

Operational Landscape Units

In fragmented landscapes, the lack of colonization or movement of organisms or particles

between sites may reversed through the restoration of connectivity by corridors or by water
flow. Two types of spatial processes (biological and hydrologic), both mostly lateral
landscape flows parallel to the earth surface are important. In population biology, the
dispersal of organisms is important and dispersal processes operates at a variety scales,

491
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

depending on life history strategies (i.e, dispersal mechanisms) of individual species. At the
landscape scale, the larger the number of target species, the larger the variety of dispersal
mechanisms and dispersal capabilities, and the larger the variety of relevant spatial scales.
In wetlands, hydrological processes shape the diversity of habitats and linkages between
wetlands typically operate at a larger scale than the scale of a single site as materials (i.e.,
sediments, nutrients) move in ground and surface water. Groundwater flows have a lateral
component and flows in the landscape from groundwater recharge areas at higher elevations
towards groundwater discharge areas at lower elevations may be important. In many stream
catchments, hydrology has been drastically altered. The main goal was often to quickly drain
water from precipitation from the major part of the catchment and bypass the normal routes
through groundwater recharge and (sub)surface runoff.
We define Operational Landscape Units as ‘combinations of landscape patches with their
hydrogeological and biotic connections’. An example of an OLU in a stream valley is (from
high to low elevation) a combination of flat or sloping patches with groundwater recharge, the
patches with groundwater discharge to which they are connected through local groundwater
flow or overland run-off, the floodplain or stream bank patches and the stream itself (Fig.1).
OLU’s in stream catchments are characterized by connections through surface water flow in
the stream or through regional groundwater flow. An important notion is that the connections
that characterize OLU’s are often disrupted in fragmented landscapes: groundwater recharge
areas are being drained so that water short-circuits to the stream and groundwater discharge
is strongly diminished; streams have been straightened to maximize hydraulic flow rates;
floodplains have strongly decreased because of flood control schemes by building dikes or
water level control structures. This has also a direct bearing on water-dispersed organisms,
and, through effects on fauna dispersal, on animal-dispersed organisms. The OLU concept
can be a valuable tool to analyze the scales on which key landscape connections for species
richness operate, and to define the best strategy for restoration and conservation of
biological diversity in stream catchments. The OLU concept has been tested in the Dinkel
catchment in East Twente, The Netherlands, in the process of developing nature restoration
targets, water management policies and land use changes at a regional scale.

Figure 1. The OLU concept. Lateral connections in a

stream corridor: hydrological flowpaths and dispersal
of organisms and diaspores among (semi-)natural
landscape patches.

References
Brierley, G., & Stankoviansky, M. (2002) Geomorphic responses to land use change: Lessons from
different landscape settings. Earth Surface Processes and Landforms 27:339-341.
Hanski, I. (2005) Landscape fragmentation, biodiversity loss and the societal response - The longterm
consequences of our use of natural resources may be surprising and unpleasant. Embo Reports
6:388-392.
Soons, M. B., Messelink, J. H., Jongejans, E. & Heil, G. W. (2005) Habitat fragmentation reduces
grassland connectivity for both short-distance and long-distance wind-dispersed forbs. Journal of
Ecology 93:1214-1225.

492
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

Scaling up denitrification in heterogeneous landscapes

G. Pinay1, T.P. Burt2

1
CEFE – CNRS, 1919 route de Mende, F-34293 Montpellier cedex – France;
e-mail: [email protected]
2
Department of geography, University of Durham, Sciences Laboratories, South Road,
Durham, DH1 3LE, UK

Biogeochemical hot spots are areas, or patches, that exhibit disproportionately high
reaction rates relative to the surrounding area, or matrix (McClain et al. 2003). Hot spots can
be delineated at spatial scales ranging from molecular to global, while hot moments can be
delineated at time scales ranging from instants to millennia. As with all other forms of
heterogeneity, the identification of hot spots and moments depends on the system of interest
or the chosen scale for a study. As the extent under consideration increases, new, ‘hotter’
hot spots may be encountered from the surrounding area. With increases in grain, hot spots
might also disappear as they fall below the resolution of the study. Here, we describe how
hot spots change with scale for the process of denitrification
Denitrification is the conversion of NO3- to gaseous N (N2O or N2). It is performed by
particular groups of ubiquitous heterotrophic bacteria that have the ability to use NO3- as an
electron acceptor during anaerobic respiration. The factors controlling denitrification rates are
C and NO3- supply and anoxia. Appropriate conditions for the formation of denitrification hot
spots are found at oxic-anoxic interfaces crossed by a continual water flow; oxic conditions
are needed for NO3- production by nitrification, denitrification requires anoxic conditions, and
water serves as the transport medium. The underlying physiological basis for denitrification
remains the same irrespective of the scale of analysis. However, since direct measurement
of denitrification is impossible at larger scales, different metrics can be used as indices of the
denitrification occurring. In the following paragraphs, we review the different metrics used to
assess denitrification hot spots across spatial scales ranging from soil profiles to large
basins.
At the scale of a soil profile (1-10 m), denitrification hot spots exist around patches of labile
organic matter, e.g., plant detritus, manure, at the anaerobic center of large soil aggregates
or in earthworm casts. Reactants are transported into these hot spots by percolating soil
water or ground water. Note that in the unsaturated zone, hydrological flowpaths will be
intermittent, with strong seasonal variations. Thus, denitrification hot spots within unsaturated
soil profiles will be active during hot moments.
At the catena scale (10-100 m), the distribution of anoxic zones is controlled by differences in
soil texture and natural drainage that affect the duration and timing of soil saturation and the
accumulation of organic matter. These factors also indirectly influence carbon and nitrate
availability through their influence on plant community type and microbial activity.
At the scale of the upland to stream toposequence (circa 100-1000 m) the interface between
the upland and riparian zones is typically a hot spot for denitrification. Denitrification is
triggered by allochthonous NO3- input from uplands along ground water flow paths. As
mentioned earlier, in most cases the hot spots for denitrification are a only few metres wide
at the upland margin of these features, though they can occur at the riverbed-wetland
interface or within the wetland or riparian zone, depending on the location of groundwater
flow paths and seasonal variations. At the same scale, hot spots of denitrification have been
identified within rivers in association with hyporheic zones, regions of saturated sediments
and subsurface water beneath and lateral to streams that are hydrologically connected to
surface and ground waters. Instream denitrification is most prevalent at downwelling sites,
i.e. locations of surface water infiltration into hyporheic zones, where anoxic, organic-carbon-
rich subsurface zones receive downwelling NO3- from surface water.

493
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

At the scale of 10 to 100 km, the width of the riparian zone is no longer resolvable and must
be replaced by the length of contact between upland and wetland, where NO3- from upland
sources is delivered to anoxic sites.
At very large scales (100 km and above), the amount of N lost due to denitrification is related
to the percentage of land covered by wetlands.

References
McClain, M.E., Boyer, E.W., Dent, C.L., Gergel, S.E., Grimm, N.B., Groffman, P.M., Hart, S.C.,
Harvey, J.W., Johnston, C.A., Mayorga, E., McDowell, W.H., Pinay, G. 2003. Biogeochemical
hot spots and hot moments at the interface of terrestrial and aquatic ecosystems. Ecosystems, 6:
301-312.

494
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

Hotspots for denitrification and nitrous oxide emission in riparian zones

M.M. Hefting, R.N. van den Heuvel

Landscape Ecology, Institute of Environmental Biology, Science Faculty, Utrecht University,

e-mail: [email protected]

The hot spot approach

Riparian wetlands bordering lower order streams and river floodplains have a high
potential to remove nitrogen from through flowing water. Denitrification is the dominant
process in the effective removal of nitrate by conversion of nitrate to nitrous oxide and
nitrogen gas. However, the denitrification process is extremely variable in space and time
which complicates the extrapolation of process rates to larger surfaces and longer time
scales. A hot spot and hot moment approach focusing on patches and periods where the
process is operating at disproportionately high rates might be of major importance in
quantifying process rates in many ecosystems (McClain et al.,2003).

Water as a master variable

Generally, denitrification hot spots occur where hydrologic flow paths carry nitrate into
a substrate containing easily degradable carbon. This indicates the importance of water flow
in determining the process rates involved with denitrification. The presence of water also
causes anoxia and affects the diffusion of denitrification end products. Many studies on
driving factors for denitrification showed with multivariate techniques that water-related
factors such as water level, water filled pore space or soil moisture content are the most
important predictors of denitrification rate, while nitrate concentrations are second in
importance (Machefert and Dise, 2004; Pinay et al.,2006; Hefting et al.,2003). Hence, water
availability is the master variable determining the height of the potential process rates of
denitrification and nitrous oxide emission. Nitrate and carbon availability, temperature and pH
determine the actual rates.
A lot of data on denitrification and nitrous oxide emission from riparian zones are
available nowadays. It is, however, still challenging to extrapolate these data to the
catchment scale at which water resources management operates. In this presentation we
focus on the role of water in controlling denitrification and nitrous oxide emissions at multiple
scales and investigate whether hydrological models can be used to scale up process rates to
the catchment scale.

Water related parameters at different scales

At the landscape scale riparian zones are hotspots of nitrogen transformation and removal.
The nitrate removal capacity of these zones is strongly related to catchment characteristics
which influence hydrology such as differences in lithology and topography (Vidon and Hill,
2004, Burt et al.,2002).
An analysis of a large data set on nitrogen transformation rates in riparian zones and
floodplains measured across continental gradients of climatic, hydrological and land use
conditions showed that hydrology, rather than climate and land use, tended to be the most
important control of the nitrogen transformation processes. Flat riparian zones with their
water table close to the soil surface were more effective at removing nitrogen compounds
than drier sites where water tables were below 30 cm (Burt et al.,2002, Sabater et al.,2003,
Hefting et al.,2004). In another study performed in a range of European river floodplains the
most important control variable of denitrification appeared to be soil moisture, which was
determined by flooding and precipitation events (Pinay et al.,2006)

495
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

Within riparian zones hydrological flow paths are important carriers of nitrate into the
biologically active riparian system. The actual pathway of water flow through riparian
substrates is often complex and creates spatial differences in residence time and material
encountered. Hence the riparian zone contains a mosaic of suitable and unsuitable flow
paths for nitrate removal with distinct spatial and temporal patterns of denitrification and
nitrous oxide emissions (Hefting et al.,2006).
Even within active patches of denitrification and nitrous oxide emission within flow paths,
there is substantial spatial and temporal variability. Nunan et al.,(2003) identified water and
nutrient movement through soil columns as a possible structuring agent for short scale (<1
mm) spatial patterns. At this scale the percentage of water filled pores is an important
variable showing an exponential relationship with denitrification rates above a threshold of
60-80%. This variable proved to be robust and comparable across a range of soil texture
classes and ecosystems (Machefert and Dise, 2004).

References
Burt, T.P., Pinay, G.,Matheson, F.E., Haycock, N.E., Butturini, A., Clement, J.C.,
Danielescu, S., Dowrick, D.J., Hefting, M.M., Hillbricht-Ilkowska, A. & Maitre, V. (2002). Water
table fluctuations in a riparian zone: comparative results from a pan-European experiment. Journal
of Hydrology 265: 129-148.
Hefting M.M., Bobbink R. & Caluwe, H. (2003). Nitrous oxide emission and denitrification in
chronically nitrate-loaded riparian buffer zones Journal of Environmental Quality 32:1194-1203.
Hefting M.M., Bobbink R. & Janssens M.P. (2006). Spatial Variation in Denitrification and N2O
emission in relation to nitrate removal Efficiency in a N-Stressed Riparian Buffer Zone.
Ecosystems, 9: 550-563.
Hefting, M.M., Clement, J.C., Dowrick, D., Cosandrey, A.C., Bernal,S.,Cimpian, C., Tatur, A.,
Burt,T.P. & Pinay, G. (2004). Water table controls on soil nitrogen cycling in riparian wetlands
alomg a European climatic gradient. Biogeochemistry 67:113-134.
Machefert, S.E. & Dise, N.B. (2004). Hydrological controls on denitrification in riparian ecosystems.
Hydrology and earth System Sciences, 8: 686-694.
McClain, M.E., Boyer, E.W., Dent, C.L., Gergel, S.E., Grimm, N.B., Groffman, P.M., Hart, S.C.,
Harvey, J.W., Johnston, C.A., Mayorga, E., McDowell, W.H. & Pinay, G. (2003).
Biogeochemical hot spots and hot moments at the interface of terrestrial and aquatic ecosystems.
Ecosystems, 6: 301-312.
Nunan, N., Wu,K., Young, I.M., Crawford, J.W., Ritz, K., Spatial distribution of bacterial communities
and their relationships with the micro-architecture of soil. FEMS microbiology ecology 44:203-215.
Pinay,G., Gumiero, B., Tabacchi, E., Gimnez, O., Tabacchi-Planty, A.M., Hefting, M.M., Burt, T.,
Black, V.A., Nilsson, C., Iordache, V., Bureau, F., Vought, L., Petts, G.E. & Decamps, H.
(2007) Patterns of denitrification rates in European alluvial soils under various hydrological
regimes. Freshwater Biology, 5: 252-266.
Sabater, S., Butturini, A., Clement, J.C., Burt, T., Dowrick, D., Hefting, M.M., Maitre, V., Pinay, G.,
Postolache, C., Rzepecki, M. & Sabater, F. 2003 Nitrogen removal by riparian buffers along a
European climatic gradient: patterns and factors of variation. Ecosystems 6:20-30.
Vidon, P. & Hill, A.R. (2004). Denitrification and patterns of electron donors and acceptors in eight
riparian zones with contrasting hydrogeology. Biogeochemistry 71:259-283.

496
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

Learning from the past: the impact of human-induced changes to landscape

(dis)connectivity on biophysical fluxes

G. Brierley

School of Geography, Geology and Environmental Science, University of Auckland, NZ.

email: [email protected]

In a sense, swamps and wetlands such as discontinuous watercourses, trapped tributary fills,
and backswamps at floodplain margins, are topographic expressions of landscape
disconnectivity. These low relief landforms accumulate and store materials (sediments,
nutrients, contaminants, etc), acting as environmental filters that influence water transfer and
water quality. In many instances, swamps and wetlands are found in cut-and-fill landscapes
that are characterised by long periods of sediment accumulation interspersed by short
intervals of incision, channel expansion and ‘release’ of materials. Among other names,
these landscapes are referred to as arroyos (American south-west), ndongas (southern
Africa), oued and wadis (North Africa/Middle east), chains of ponds, dells and swampy
meadows (Australia). Shifts in climate, extreme storms, or exceedance of intrinsic (slope-
induced) thresholds may interrupt the ‘natural’ tendency of these landscapes to accumulate
materials over the long term. However, human efforts to ‘make dry lands wetter and wet
lands drier’ have promoted much more consistent and permanent connectivity, disrupting the
storage and filtering roles of these landscape elements (Ward, 1989).

Conceptually, swamps and wetlands can be considered as barriers that impede the
movement of materials through landscapes (Fryirs et al., in press (a)). Analyses of slope-
valley floor, tributary-trunk stream, and channel-floodplain connectivity based on DEMs can
be used to map these forms of landscape connectivity (Fryirs et al., in press (b)). However,
other sets of resources must be used to place contemporary conditions in context of past
environmental relationships. Such insights present a critical basis with which to assess the
trajectory of landscape change, and associated implications for environmental health.

In many parts of the colonial world the impacts of human disturbance have been so profound
over such a short period of time that the vestiges of former landscape behaviour can be
unravelled from historical documents and field evidence. Extensive research of this ilk has
been used to unravel the operation of biophysical fluxes in the period since European
settlement of Bega Catchment in southeastern Australia (i.e. since 1860; see Brierley et al.,
1999; Brierley & Fryirs, 2005; Fryirs & Brierley, 2001, 2005).

Since European settlement, the channel network in Bega Catchment has become
longitudinally connected, as discontinuous watercourses that characterised base of
escarpment and mid-catchment locations prior to European settlement have been
transformed into continuous ephemeral channels. An associated increase in channel
capacity along the lowland plain ensures that higher magnitude flows are now retained within
the channel. This increase in longitudinal connectivity has increased the rate and efficiency
of downstream transfer of flow, sediment, organic matter and other nutrients, decreasing
their retention within the system. Base flow contributions from the middle and upper
catchment have been reduced. Deposition of extensive sand sheets in downstream reaches
has induced a greater proportion of subsurface flow, affecting the viability of backswamps.

Changes to channel form have also altered lateral exchanges of water, sediment, organic
matter and nutrients between the channel and its floodplain. Deeply incised channels that cut
through former swamps at the base of the escarpment have severed lateral linkages,
inhibiting inundation of perched valley floors. In contrast, deposition of coarse sands atop the
formerly fine-grained floodplain of the lowland plain has partially infilled backswamps,
inhibiting the capacity of this area to act as a nutrient sink.

497
Theme 4: Ecohydrology, water and rivers
4.2 Symposium 9: Biogeochemical Hotspots in a Landscape Context: implications for catchment water
management

The development of a continuous channel network, associated changes to channel-

floodplain relationships, and the highly degraded riparian vegetation cover, have altered
vertical water exchange and depth of the water table along river courses throughout Bega
catchment. Incision of swamp deposits at the base of the escarpment has lowered the water
table. In contrast, aggradation in lowland areas has raised the water table. These changes,
along with the homogeneity of bed material size (contrasting with the fine-grained swamps,
bedrock pools and sand bars of the pre-disturbance channel), have modified the hydrologic
regime, hydraulic diversity and hyporheic zone functionality, impacting profoundly on the
ecological condition of these rivers (Chessman et al., 2006).

Geomorphic changes have compromised the potential for rivers in Bega catchment to regain
structural and functional attributes that characterized their pre-European settlement
condition. These insights have been used to predict realistic target conditions for river
rehabilitation, using a conservation-first catchment-scale prioritisation framework to protect
remnant swamps and develop rehabilitation strategies that extend from areas that are in
good condition (Fryirs and Brierley, 2005).

Three primary implications of this work are noted:

1. It is contended that we seldom appreciate the extent to which discontinuous
watercourses in lower order basins have been modified by human disturbance, in a wide
range of environmental settings.
2. Changes to landscape configuration provide a biophysical template with which to
appraise geoecological and water quality considerations. As swamps and wetlands are
key components of landscape (dis)connectivity, changes to their distribution over time
exerts a critical influence upon hydrologic regimes and water quality.
3. Fine-resolution mapping using DEMs presents a remarkable opportunity to apply
conceptual frameworks to model catchment-scale changes to landscape connectivity,
and associated biophysical responses, over time (e.g. Fryirs et al., in press (b)).

References
Brierley, G.J., Cohen, T, Fryirs, K. & Brooks, A. (1999) Post-European changes to the fluvial
geomorphology of Bega catchment, Australia: implications for river ecology. Freshwater Biology
41, 1-10.
Brierley, G.J. & Fryirs, K.A. (2005) Geomorphology and River Management: Applications of the
River Styles framework. Blackwell Science, Oxford, UK. 398pp.
Chessman, B.C., Fryirs, K.A. & Brierley, G.J. (2006) Linking geomorphic character, behaviour and
condition to fluvial biodiversity: implications for river rehabilitation. Aquatic Conservation: Marine
and Freshwater Ecosystems 16, 267-288
Fryirs, K. & Brierley, G.J. (2001) Variability in sediment delivery and storage along river courses in
Bega catchment, NSW, Australia: Implications for geomorphic river recovery. Geomorphology 38,
237-265.
Fryirs, K. & Brierley, G.J. (2005) Practical application of the River Styles® framework as a tool for
catchment-wide river management: A case study from Bega catchment, New South Wales,
Australia. Retrieved on 28.11.06, from www.riverstyles.com.
Fryirs, K., Brierley, G.J., Preston, N. and Kasai, M. (In press (a)) Buffers, barriers and blankets:
The (Dis)continuity of catchment-scale sedimentary cascades. Catena.
Fryirs, K. Brierley, G.J., Preston, N. & Spencer, J. (In press (b)) Catchment-scale (dis)connectivity
in sediment flux in the Upper Hunter catchment, New South Wales, Australia. Geomorphology.
Ward, J.V. (1989) The four-dimensional nature of lotic ecosystems. Journal of the North American
Benthological Society 8, 2-8.

498
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

4.3 Open Session 17: Ecohydrology, rivers and wetlands

Quantifying lost opportunities in conservation: an evaluation of current and

proposed scenarios to protect the Pantanal wetland in Brazil.

R. Lourival1,2, H. McCallum2, G.Grigg 2 , C.Arcangelo,3, R. Machado 3,

H.Possingham 2
1
The Capes Foundation – Ministry of Education – Brasília - Brazil
2
University of Queensland - Goddard building – St. Lucia Campus.
Brisbane – QLD – 4072 – Australia –
[email protected]
3
Conservation International – Brasília - Brazil

Introduction

The Pantanal is the world’s largest contiguous freshwater wetland. It is shared by Brazil,
Bolivia and Paraguay. Designated by the Brazilian constitution as a National Heritage and
considered by the RAMSAR convention for wetland conservation a key site for wetland
protection, it combines healthy populations of endangered species with the greatest wildlife
densities of the Neotropics. Whilst the majority of threats to the Pantanal are generated
externally (e.g. siltation, pollution and water diversion), there are many internally generated
threats, such as expansion of cultivated pasture, selective logging and damage from invasive
species (Mittermeier 1990), all connected in some way to global issues (Harris, Tomas et al.
2005).
There is an array of conservation scenarios proposed to protect the Pantanal in all three
countries. We evaluate in this article the efforts on the Brazilian side, exemplified by: (1) the
existing protected areas, (2) the core areas proposed by the Cerrado & Pantanal-Priority
Setting Workshop CP-PSW, (3) the Pantanal Biosphere Reserve and (4) the core areas
added to corridors, recommended by the CP-PSW, (BRASIL 1999). Our primary aim is to
evaluate the complementarity and compare the efficiency of these conservation scenarios in
achieving the protection of its physiognomies.
We use 17 classes of vegetation as appropriate surrogates for regional biodiversity (Da
Silva, Mauro et al. 2000). We tested each surrogate’s representation for each conservation
scenario at three target levels; 10%, 20% and 50% achievement (Pressey and Taffs 2001)
using irreplaceability scores generated by the conservation planning software MARXAN (Ball
and Possingham 2000).

Results and Discussion

We found that large fractions of the area of each conservation scenario considered are
not needed to meet targets, while important planning units are missing from them, and we
use this metric to determine and to compare lost opportunities under each scenario (Figure
1).
We conclude that none of the scenarios were simultaneously comprehensive and efficient in
representing the vegetation classes. Although scenario (4) was able to represent all
vegetation classes at 10% and 20% targets; it comprised 50% of all planning units, making it
the least efficient of all scenarios.
In summary, neither the existing Pantanal reserve system nor any the proposed
conservation scenarios guarantees efficient and comprehensive protection of the vegetation
classes and, therefore, none is effective for safeguarding regional biodiversity.

499
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

Scenario 1 Scenario 2
(Existing reserves)
(Core CP-PSW)
Opportunity
replac. excluded missed 25%
41% 28%

replac. included 7%
5%

28%
irrep. excluded 21%
23%

23%
irrep. included
3%
negot. excluded 4%
negot. included 4%
25%
3%

Scenario 3 Scenario 4
(Biosphere Reserve) (Core+corridors CP-PSW)
25%
34%

14% 22%
13%
33%
30%

16% 11%
18%
15%

9% 14%
9%

Figure 1 – Opportunities lost by each conservation scenario in the Pantanal, represented by

the percentage of irreplaceable planning units excluded by scenario and replaceable PU
included in each scenario, according to their selection frequency in 10000 runs of MARXAN-
CLUZ.

References
Ball, I. R. and H. P. Possingham (2000). Marxan v1.8.6 Marine Reserve Design; using Spatially
Explicit Annealing.
BRASIL (1999). Ações Prioritárias para a Conservação da Biodiversidade do Cerrado e Pantanal.
Brasília, PROBIO: 27p.
Da Silva, M. P., R. Mauro, et al. (2000). "Distribuição e quantificação de classes de vegetação do
Pantanal através de levantamento aéreo." Revista Brasileira de Botânica 23(2): 143-152.
Harris, M. B., W. Tomas, et al. (2005). "Safeguarding the Pantanal wetlands: Threats and
conservation initiatives." Conservation Biology 19(3): 714-720.
Mittermeier, R. A. C., I.G.; Pádua, M.T.J. and Blanck J. (1990). "Conservation in the Pantanal of
Brazil." Oryx 2(24): 103-112.
Pressey, R. L. and K. H. Taffs (2001). "Scheduling conservation action in production landscapes:
priority areas in western New South Wales defined by irreplaceability and vulnerability to
vegetation loss." Biological Conservation 100(3): 355-376.

500
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

The relationship between floodplain sedimentation and the succession of rare

riverine grasslands in the Netherlands; implications for conservation strategies

H.P. Wolfert, G.J. Maas, B. Makaske, P.W.F.M. Hommel, B.S.J. Nijhof

Alterra, Wageningen University and Research Centre, P.O. Box 47, 6700 AA Wageningen,
the Netherlands,
e-mail: [email protected]

Introduction

Medicagini-Avenetum occurs on relatively high sandy ridges on embanked floodplains in

the Netherlands. The rare riverine grassland type Medicagini-Avenetum, which is reported to
have occurred more extensively in the past, is generally appreciated because of its abundant
colourful flowers in spring. Nowadays, remaining patches of Medicagini-Avenetum seem to
degrade slowly, despite local conservation measures. This study addresses the prerequisites
for development of Medicagini-Avenetum, and an appropriate conservation strategy.

Methods

In our research (Maas et al., 2003) we investigated the geomorphological processes

leading to suitable physiotopes for Medicagini-Avenetum. We applied various methods for
measuring floodplain sedimentation rates in the field (Maas and Makaske, 2003) and linked
these data to soil analyses and vegetation data (Nijhof and Hommel, 2003). Sedimentation
on embanked floodplains was measured in three study areas along the rivers Waal and
IJssel, in order to investigate its influence on vegetation succession. The amount of
sedimentation during the flood season of 2001 was measured using sediment traps (Fig. 1).
Grain size analysis was carried out on each of the trapped sediment samples. Average
sedimentation rates for the period 1960-2001 were deduced from radiometric properties
(137Cs activity) of undisturbed sediment cores. In total 62 vegetation relevées, following the
simplified Braun-Blanquet methodology, were made. These were clustered by Twinspan and
further ordinated by hand.

Conceptual model

A conceptual model, following from our field research, illustrates the link between fluvial
morphodynamics and vegetation succession. Overbank deposition results in natural levees,
a physiotope which is favoured by Medicagini-Avenetum (Fig. 1) because of its sandy,
nutrient-poor and calcareous soil conditions (Wolfert et al., 2002). Optimal conditions occur
when the levees have grown to a height corresponding with an inundation frequency of once
every 4 to 5 years. However, with increasing height deposition of calcareous sand will
decrease, thus favouring enrichment by nutrients during occasional floodings, and a decline
of the vegetation type. Optimum floodplain conditions for Medicagini-Avenetum thus are
shown to be the result of a long landform and substrate evolution, and are only a temporary
situation. In natural river systems the inevitable degradation of Medicagini-Avenetum at one
place is compensated for by the development of new suitable physiotopes elsewhere in the
river system.

Conservation strategies

With restriction of natural river dynamics, due to groynes and revetments, this cyclic
process has come to an end. Any conservation strategy for Medicagini-Avenetum should aim
at (partial) restoration of natural river morphodynamics, or at least should incorporate a

501
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

floodplain management mimicking basic geomorphological processes. Our work enabled to

identify the locations along the river where conservation of existing Medicagini-Avenetum
vegetation should deserve priority on the short term, but also raised awareness that we have
to prepare sites for future development. Maps are being made to support nature conservation
strategies.
SSE NNW
MA Ar flood level 27 March 2001
7 Ar
elevation (m + OD)

MA Ar
6 Ar
LP LP Ar
5 Al
MA = Medicagini-Avenetum pubescentis LP
Waal R. Ar = Arrhenatheretum elatioris
4
Al = Alopecurion pratensis
3 a LP = Lolio-Potentillion anserinae

sediment thickness (mm)

35,4 kg/m2
clay, silt, sand (%) sedimentation (kg/m2)

10 = 29 mm 8
8
6
6
4
4
2 2
b
0 0
100
80 sand
60
40
silt
20
c clay
0

Figure 1. (a) Surface topography in a studied transect in the Rijswaaard study area on the
Waal River, the Netherlands. Black triangles indicate locations of sediment traps. Grassland
types at locations are also indicated. (b) Sedimentation in the transect during the flood
season of 2001 measured with sediment traps. (c) Proportions of sand, silt and clay of the
samples from the sediment traps.

References
Maas, G.J., & Makaske, B. (2003). Sedimentation on embanked floodplains of the rivers Waal and
IJssel. R.S.E.W. Leuven, A.G. van Os, & P.H. Nienhuis (Eds.). Proceedings NCR-days 2002;
current themes in Dutch river research. NCR, Delft, pp. 122-124.
Maas, G.J., Makaske, B., Hommel, P.W.F.M., Nijhof, B.S.J. & Wolfert, H.P. (2003). Verstoring en
successie; rivierdynamiek en stroomdalvegetaties in de uiterwaarden van de Rijntakken. Alterra-
rapport 759, Alterra, Wageningen.
Nijhof, B.S.J., & Hommel, P.W.F.M. (2003). The relation between the Medicagini-Avenetum
vegetation and sedimentation. R.S.E.W. Leuven, A.G. van Os, & P.H. Nienhuis (Eds.).
Proceedings NCR-days 2002; current themes in Dutch river research. NCR, Delft, pp. 130-134.
Wolfert, H.P., Hommel, P.W.F.M., Prins, A.H. & Stam, M.H. (2002). The formation of natural levees
as a disturbance process significant to the conservation of riverine pastures. Landscape Ecology
17, Supplement 1, pp. 47-57.

502
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

The spatial organisation of savanna landscapes within drainage networks

C.J. Cullum

School of Animal, Plant & Environmental Sciences, University of the Witwatersrand,

Private Bag 3, Wits 2050, Johannesburg, South Africa.
e-mail: [email protected]

In water-controlled ecosystems, spatial and temporal patterns of soil moisture not only
constrain the distribution of vegetation, but are also implicated in system productivity and the
provision of habitat (Scholes and Walker, 1993; Rodríguez-Iturbe and Porporato, 2004).
Landscape patches defined in terms of their soil moisture regime are therefore likely to
reflect fundamental ecosystem units, providing a powerful basis for a land classification that
meets diverse management needs.
However, processes controlling the spatial and temporal distribution of water, soil and
vegetation are interdependent, such that patch character and behaviour cannot be
described, explained or foresighted by considering any of these three factors in isolation
(Rodríguez-Iturbe and Porporato, 2004). Furthermore, interactions between water, soil and
vegetation are scale-dependent, notoriously variable in time and space and are modified by a
wide range of other factors (e.g. humans, fire, herbivory, biological engineering) that also
vary across multiple spatial and temporal scales (O'Connell, et al., 2000).
This paper focuses on network geometry as a control on the spatial distribution of
Vegetation/ Soil/ Hydrology (VSH) patches in the semi-arid savannas of Kruger National
Park, South Africa. Network geometry describes the intensity and direction of flow paths for
water, sediment and associated nutrients. Thus, in water-controlled ecosystems, drainage
network configuration can be viewed as an organising template for the ecohydrological and
biogeochemical processes that determine vegetation patterns and dynamics (Caylor, et al.,
2005).
In an idealised river basin, consistent climate and geology produce a consistent fluvial
erosion regime that generates fractal patterns of landscape dissection. Geometric
relationships between area drained, stream length, stream density, bifurcation ratios, channel
gradient and hillslope length elegantly describe basin structure at multiple scales using a
limited number of parameters (Hack, 1957; Horton, 1945). For example, just two exponential
components, d and h, effectively describe basin morphology and shape (Equations 1 and 2)
for an idealised network (Dodds and Rothman, 2000).
L α Ah where L = mean length of main stream channel
A = mean area of watershed (from any point along a channel) Equation 1
L α Ll d where Ll = mean longitudinal diameter of a watershed Equation 2

Allometric relationships will be calculated for the Sabie river network within Kruger National
Park and the spatial variance in exponential components examined at various resolutions.
Since the Sabie traverses areas with different geologies and with a West-East rainfall
gradient, it is expected that the exponential components should vary significantly within the
network. However, patches with relatively similar exponential components should indicate
areas with similar climate and geology, reflected in distinct vegetation, soil and hydrological
characteristics. Variability within each group of VSH patches defined in this way may be
accounted for by local conditions and the history of disturbance.
The ecological implications of contrasting network geometry are explored at catchment
and reach scales. At the catchment scale, for example, drainage density is much lower on
the relatively permeable basalts in the East than on the less permeable granites in the West
of the catchment. On the basalts, water falling on hillslopes tends to infiltrate vertically, with
relatively little (or very slow) lateral surface movement of water and sediment. By contrast,
higher levels of runoff in the granites generate a higher stream density and more lateral
surface movement of water and sediment. Differences in the lateral fluxes of water and
sediment into the river channels result in comparatively sandy/gravelly beds in rivers fed from

503
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

the granites, compared with frequent bedrock or cobbles in streams flowing from and over
the basalts. The lateral movement of water and sediments on the granites also favours the
formation of soil catenas (Milne, 1935), where fine clays and water-soluble nutrients are
progressively leached out and transported downhill over long geological time scales. Distinct
bands of soil, parallel to the contours, support different vegetation and in turn favour
0.different fauna. In the basalts, there is far less downslope variation in soil, with large areas
of relatively homogenous vegetation. The relatively low stream density in the basalts also
means that a smaller percentage area is occupied by channels, tributary junctions and fertile
riverine vegetation, with larger distances between rivers.
At the reach scale, the area drained, the length and gradient of hillslopes (which are all
associated with network position) are important controls both on the distribution of soil
moisture (and hence on vegetation) and on the discharge of water and sediment. Thus, in
an idealised catchment, the flow/sediment regime varies with network position to produce the
longitudinal patterns of channel morphology that have long been observed in river systems
(Schumm, 1977; Petts and Amoros, 1996).
Such examples illustrate the wide-ranging implications of network geometry and hence the
potential descriptive and explanatory power of differences in the fractal dimensions of
landscapes. Within-catchment variations in geology and climate that lead to breaks in the
theoretical river ‘continuum’ (Vannote, et al., 1980; Minshall, et al., 1983) should be reflected
in differences in the exponential components that succinctly describe river network structure.
Further research is needed to establish the scales at which such an approach can effectively
delineate distinct VSH patches that can inform landscape classifications for conservation
management in savannas.

References
Caylor, K. K., Manfreda, S. & Rodriguez-Iturbe, I. (2005) On the coupled geomorphological and
ecohydrological organization of river basins Advances in Water Resources 28: 69-86.
Dodds, P. S. & Rothman, D. H. (2000) Scaling, universality and geomorphology Annual Review of
Earth and Planetary Sciences 28: 571-610.
Hack, J. T. (1957) Studies of longitudinal stream profiles in Virginia and Maryland US Geological
Survey Professional Paper 45-97.
Horton, R. E. (1945) Erosional development of streams and their drainage basins; hydrophysical
approach to quantitative morphology Bulletin of the Geological Society of America 56:
Milne, G. (1935) Some suggested units for classification and mapping, particularly for East African
soils Soil Research 4: 183-198.
Minshall, G. W., Petersen, R. C., Cummins, K. W., Bott, T. L., Sedell, J. R., Cushing, C. E. &
Vannote, R. L. (1983) Interbiome Comparison of Stream Ecosystem Dynamics Ecological
Monographs 53: 1-25.
O'Connell, D. A., Ryan, P. J., McKenzie, N. J. & Ringrose-Voase, A. J. (2000) Quantitative site and
soil descriptors to improve the utility of forest soil surveys Forest Ecology and Management 138:
107-122.
Petts, G. E. & Amoros, C. (1996) Fluvial hydrosystems. Chapman & Hall, London.
Rodríguez-Iturbe, I. & Porporato, A. (2004) Ecohydrology of water-controlled ecosystems: soil
moisture and plant dynamics. Cambridge University Press, Cambridge.
Scholes, R. J. & Walker, B. H. (1993) An African savanna: synthesis of the Nylsvley study.
Cambridge University Press, Cambridge.
Schumm, S. A. (1977) The Fluvial System. Wiley, New York.
Vannote, R. L., Minshall, G. W., Cummins, K. W., Sedell, J. R. & Cushing, C. E. (1980) River
Continuum Concept Canadian Journal of Fisheries and Aquatic Sciences 37: 130-137.

504
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

Influences of changing scale of modelling land use on simulating hydrological

components in watershed land use planning assessments

Y.P. Lin 1, P.J. Wu1, N.M. Hong2

1
Department of Bioenvironmental Systems Engineering, National Taiwan University, No.1
Sec. 4 Roosevelt Rd. Taipei 106 Taiwan.
e-mail: [email protected]
2
Department of Environmental Resources Management, The Overseas Chinese Institute of
Technology, No. 100, Chiao Kwang Rd., Taichung 407, Taiwan

Introduction

Land-use changes can influence hydrological processes including infiltration, groundwater

recharge, base flow and runoff. The dynamic of these hydrological processes even influence
most of the abiotic and biotic processes in a watershed. Integrated land-use and hydrological
simulation models are useful to evaluating watershed land-use planning strategies. However,
scale can influence the measurement and quantitative description of land-use patterns.
Therefore, the scale effects of simulating land-use on modelling hydrological components are
important to impact analyses of land-use changes on hydrology for assessing watershed
planning strategies.

Method

Study area and data

The Wu-Tu watershed is located in east of Keelung River Basin in the northern Taiwan.
This watershed was urbanized by extensions of Taipei and Keelung Cities. Following land-
use modelling in Lin et. al (2006), the driving factors were demography, social, infrastructure,
geomorphology and soil-related variables, including altitude, slope, distance to river, ,
distance to major roads, distance to built-up area, distance to urban planning areas, soil
erosion coefficient, soil drainage and population density for land-use modelling. Data for
land-use and driving factors were generated and digitized by the Soil and Water
Conservation Bureau using 1:5000 aerial photographs taken in 1999. Grain size in maps of
land-use and driving factors were aggregated from 50×50 m to 75×75 m, 100×100 m,
125×125 m and 150×150 m resolutions.

Land use change model

In the CLUE-S model, the non-spatial module calculates the area of change for all land-
use types at the aggregate level, while the spatial module translates demands into land-use
changes at various locations within the study region (Verburg et al., 2002). Empirical analysis
results are used within the model when simulating the competition between land-use types
specified by the user to restrict conversion of each land use that can take place based on
actual land-use patterns (Verburg et al., 2002). Probability maps for all land-use types are
calculated using logistic regression models. The relationships between land uses and driving
factors are then obtained using stepwise logistic regression as follows:
⎛ P ⎞ m
Log⎜⎜ i ⎟⎟ = β 0 + ∑ β j X ij
⎝ 1 − Pi ⎠ i =1

where i is the i-th grid cell, Pi is the probability of a land-use type occurring in a grid cell, j is
the j-th driving factors, X ij is the driving factor, m is the number of driving factors, β 0 is the

505
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

intercept of the regression model and β j is the coefficient for each driving factor in the
model.

Landscape Metrics

To identify the effects of changing the resolution of spatial patterns for simulated land-use,
landscape metrics—Number of Patches (NP), Mean Patch Size (MPS), Total Edge (TE),
Mean Shape Index (MSI), Area Weight Mean Shape Index (AWMSI), Mean Patch Fractal
Dimension (MPFD), Area Weight Mean Fractal Dimension (AWMPFD), Mean Nearest
Neighbor (MNN) and Interspersion and Juxtaposition Index (IJI)—at landscape and class
levels are calculated using the Patch Analyst in the GIS software ArcView (McGarigal and
Marks, 1995).

Hydrological Model
In this study, the Generalized Watershed Loading Functions (GWLF) model was used to
simulate hydrological components. This model is a combined distributed/lumped parameter
watershed model for simulating runoff, sediment, and nutrient loadings in watersheds. Daily
water balances are calculated for unsaturated and saturated sub-surface zones (Haith and
Shoemaker, 1987).

Results and discussion

The results of landscape metrics indicated that compositions of simulated land-use

changed when grain size changed. The results of hydrological modelling with various land-
use planning strategies indicated that the hydrological components (streamflow, surface
runoff and groundwater discharge) were impacted by land-use planning strategies in all
cases of grain size, especially streamflow was. Nevertheless, the purposed integrated
approach systematically provides useful information in effective scales to assess watershed
land-use planning strategies.
The information provided by integrating models using an appropriate scale are more
effective than that when using inappropriate scales for land-use planers. Moreover, the scale
effects and selections for simulating land-use on modelling hydrological components are also
important to impact analyses of land-use change on hydrology when assessing watershed
strategies. This study integrates an empirical land-use change model, landscape metrics and
a hydrological model to simulate and assess future land-use, land-use patterns and
hydrological processes under various land-use planning strategies in various scales (grain
sizes).

References
Haith, D.A. & Shoemaker, I.L. (1987) Generalized watershed loading functions for stream flow
nutrients. Water Resources Bull. 107: 121-137.
Lin, Y.P.; Hong, N.M.; Wu, P.J.; Wu, C.F. & Verburg, P.H. (2006) Impacts of land use change
scenarios on hydrology and land use patterns in the Wu-Tu watershed in Northern Taiwan.
Landscape and Urban Planning. In Press.
McGarigal, K. & Marks, B.J. (1995) FRAGSTATS: spatial pattern analysis program for quantifying
landscape structure. USDA For. Serv. Gen. Tec. Rep. PNW-351.
Verburg, P.H.; Soepboer, W.; Veldkamp, A.; Limpiada, R. & Espaldon, V. (2002) Modeling the
spatial dynamics of regional land use: The CLUE-s model. Environ. Manage. 30: 391-405.

506
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

Simulated annealing with landscape metrics to optimally simulate landscapes in

watershed landscape planning and management

Y.P. Lin, M.H. Chen

Department of Bioenvironmental Systems Engineering, National Taiwan University, No.1

Sec. 4 Roosevelt Rd. Taipei 106 Taiwan. e-mail: [email protected]

Introduction

One of the major land-use changes that have an impact on the hydrological cycle and
regulation are changes in the spatial patterns caused by deforestation and forest
fragmentation by human development within a watershed. Optimizing landscape or land use
aims to obtain landscape or land use patterns which well sustain landscape functions and
processes. Landscape models simulate alternative landscape dynamics which provide
information to evaluate and test hypotheses or scenarios of landscape or land use
management. Landscape or land use composition, configuration, and connectivity are
primary descriptors of the landscape pattern which can be quantified using spatial landscape
indices or metrics to characterize and quantify landscape composition and configuration. This
study developed an optimal landscape simulation model by optimizing (maximizing or
minimizing) the landscape metrics using simulated annealing for alternative land use
management strategies in Wu-Tu watershed in Northern Taiwan at the forest class level.

Method

The Wu-Tu watershed is an urbanizing watershed in the Keelung River Basin, located
between the Taipei metropolitan area and Keelung Harbor in northern Taiwan. The Wu-Tu
watershed has an area of approximately 204.41 km2, with a mean elevation and mean slope
of 242.00 m and 0.005°, respectively. The population in this watershed has increased owing
to the expansion of the Taipei metropolitan area, Keeling city and the need for labor in
Keelung Harbor during 1987-1997. In this study, an optimal technique, Simulated Annealing
(SA), was used to solve a single-objective Spatial Pattern Optimization Problem. Optimal
objectives were to identify the landscape patterns with the largest mean forest patches by
maximizing a landscape metric Mean Patch Size (MPS), the most compact forest patterns by
minimizing a landscape metric Mean Shape Index (MSI), and the most connected forest
patches by minimizing a landscape metric Mean Nearest Neighbor (MNN) under various
deforested scenarios that are removed 5%, 10%, 15%, 20%, 30%, 35%, 40%, 45% and 50%
forest to be replaced by built-up area, agricultural land and grassland within the study
watershed.

Results and Discussions

Mean patch sizes of forest patches ranged between 400 and 500 ha by optimal
removal of 5 to 20% forest in the study area. Patch number increased significantly when 50%
of forest was removed in the case of maximizing mean forest patch size. However, optimal
mean patch size of forest decreased when the removed forest percentage was greater than
20% of the forest cover of the study watershed. Moreover, patch shapes tended to complex
by optimal removing 5-35 % forest under the optimal mean patch size objective. In the
objective of minimizing mean patch shapes of forest patches mean patch sizes decreased by
removing 5 to 50% of forest in the study area. Optimal mean patch shapes tended to regular
when the removed forest percentage increased in the study watershed in the case of
minimizing mean shape indices of forest patches. Moreover, patch number increased by an
optimal removal of 5-50 % of forest cover under the optimal Mean Patch Shape objective. In
the case of minimizing mean nearest neighbor of forest patches Mean Patch Sizes
decreased significantly by removing 5 to 50% of forest in the study area. Moreover, patch

507
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

number increased significantly by optimal removal of 5-50 % of forest under the optimal
Mean Nearest Neighbor objective.

Conclusions

The optimal landscape simulation model was developed optimizing (maximizing or

minimizing) the landscape metrics using simulated annealing which are capable for
simulating alternative land use management strategies. Various optimal objectives such as
maximizing Mean Patch Size, minimizing Mean Patch Shape index and minimizing Nearest
Neighbor Distance obtained various forest compositions and configurations in the study area.
The optimal simulation approach can be integrated with environmental models, such as a
hydrological model, in future land use planning assessments.

References
McGarigal, K. & Marks, B.J. (1995) FRAGSTATS: spatial pattern analysis program for quantifying
landscape structure. USDA For. Serv. Gen. Tec. Rep. PNW-351.

508
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

A recent landscape evolution of the Shiyang river basin---patterns, changes and

causative factors

L. Lu, X. Li, E.S. Kang, T. Wang

Cold and Arid Regions Environmental and Engineering Research Institute, Chinese
Academy of Sciences, Lanzhou 730000, P. R. China.
e-mail: [email protected]

Introduction

Many investigations have shown that most arid inland basins of northwest China are
experiencing ecosystem degradation (Wang and Cheng, 1999; Kang et al., 2004). The
Shiyang river basin (SRB), which is located between 37°07’-39°28’N and 101°22’-104°05’E
and covers the area of 41600km2 with a population of 2.37 million, is the one under greatest
threat from desertification and the one with the greatest degree of poverty and environmental
stress (ADB, 2003). Landscape theory with remote sensing technology facilitates our
understanding of the changing environment over a broad range of spatial and temporal
scales (Lu et al., 2003).

Data and method

In this study, two landscape maps of SRB with 30m resolution in 1986 and 2000 were
compiled from several Landsat TM/ETM+ images through interactively manual interpretation.
The interpretation accuracy of the landscape classification for the 14 types (Table 1) was
more than 80% compared to intensive field surveys. The landscape changes of SRB
between 1986 and 2000 were then derived from the two maps via GIS platform. Some major
and important transitions were generalized from the variety of possible transitions (Figure 1).

Results and analysis

The recent landscape pattern of SRB displayed a significant process of expanding the
existing oases into the surrounding grasslands and desert (Figure.1). As shown in Table 1,
the most pronounced changes occurred in the increased area of farmland (over 38,000 ha
increased in the 15-year period). Most of them were converted from grasslands, woodlands,
gobi, sandy and salinized lands. The loss of a large area of grasslands and woodlands is
suggested to be irrational because the grasslands and woodlands are important buffers to
the oasis and provide water preserve in the upper reaches. In addition, much of the new
cultivation which was suspected to be illegal was at the margins of the oases or outside the
boundaries of protected woodlands. The landscape changes in these marginal areas
displayed complex shapes and were remarkably fragmentary and disturbed. In any case
because of its low productivity, this temporary cultivation is likely to be abandoned within a
short time as commodity prices continue to fall or groundwater becomes too costly to pump.
The results of abandoned farmlands existed in the whole basin are a legacy of irreparably
damaged soil on which nothing will grow and follows in sandy desertification and dust
storms.

Conclusion and discussion

The recent landscape pattern and change in SRB are strongly controlled by local human
activities in terms of overextended reclamations and poor water management. The natural
resources of land and water in the basin are being used beyond their sustainable limit.
Therefore, it is necessary to carry out a comprehensive authorized management strategy of
land use development and water resource allocation among different sectors and
stakeholders in the whole basin.

509
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

Table 1. Total areas of changes in different landscape types in SRB between1986-2000

Landscape type Total area in Change 2000- Total % Change

2000 (Ha) 1986 (Ha) (2000-1986) /2000
Farmland 679,653 38,986 5.74
Woodland 263,115 -4618 -1.76
Grassland 1,117,917 -19,403 -1.74
Water area 4989 -26 -0.52
Occasional flooding 9916 247 2.49
Glacier & perpetual snow 64 0 0.00
Residential area 38,609 2289 5.93
Sandy desert 922,809 -7536 -0.82
Gobi 446,555 -2799 -0.63
Salinized land 200,635 -5871 -2.93
Marsh 30,501 0 0.00
Bare soil 18,763 -1346 -7.17
Bare rock 261,010 26 0.01
cold desert 72,788 51 0.07

Figure 1. Landscape changes in SRB between1986-2000

References
ADB. (2003) The Project of ADB Technical Assistance to the People's Republic of China for
Optimizing Initiatives to Combat Desertification in Gansu Province. T.A.No.3663-PRC
Kang, E.S; Li, X.; Zhang, J.S. & Hu, X.L. (2004) Water resources relating to desertification in the
Hexi area of Gansu province, China. Journal of Glaciology and Geocryology. 26(6): 657-667.
Lu, L.; Li, X. & Cheng, G.D. (2003) Landscape Evolution in the Middle Heihe River Gasin of
Northwest China during the Last Decade. Journal of Arid Environments. 53(3): 395-408.
Wang, G.X. & Cheng, G.D. (1999) Water resource use and its eco-environmental problems in arid
zone of northwest China. Journal of Natural Resources. 14: 109-116.

510
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

Response of ecological process after water deliveries to dry watercourse of lower

reaches of the Tarim River Basin

Chen Y.N, W.H. Li, Y.P. Chen, C.C. Xu, X.M. Hao

Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences,40-3 South

Beijing Road, Urumqi 830011, Xinjiang, China.
e-mail: [email protected]

Introduction

The recent fast growth of population and economic activities in the Tarim River Basin has
dramatically changed the temporal and spatial distributions of water resources in the area,
and has made the existing problem of scarcity of water resources even more striking. Even
worse, more than 321 km river is dried up in the lower reaches of the Tarim river so that the
groundwater depth in the vicinity of the lower reaches has dropped from 1-3m to 5-12m. The
groundwater is also salinized and the ecosystems have been suffering serious damage.
Vegetation relying on the groundwater has seriously degenerated, the “Green Corridor”
between Kuluke desert and Taklimakan desert is shrinking continuously. Since 2000, the
local government has conducted ecological water deliveries four-times to the lower reaches
in order to lift the groundwater level for rehabilitating the damaged ecosystems, and
protecting the “Green Corridor”. The paper discusses the response effect in the groundwater
depth, NDVI, growth and physiology of natural vegetation in order to provide a scientific basis
for protecting the environment.

Response of ecological process to water delivery

Change of groundwater depth

According to the investigation of monitoring well in sections, the change of groundwater is
sensitive (Chen, 2004; Xu, 2003). The depth of groundwater was raised from 9.87m before
water delivery to 2.66m after the fourth stage of water delivery (Table 1). The increasing
height of the water table at each time of water delivery is 21.6% 66.4% 42.3% and 55.74%,
respectively. The scope of response in transverse section is from 250 m after the first stage
to 1050 m after the fourth stage of water delivery.

Response of growth and physiology of natural vegetation to water delivery (NDVI)

Spatial distribution change of vegetation to water delivery can be fully reflected by NDVI.
With water deliveries being carried out, the NDVI of all sections in the lower reaches have
increased, that is, vegetation growth improved. In a lengthways direction, vegetation of upper
excelled that of lower. In transverse direction, the NDVI decreased with the increasing
distance away from the river, indicating that the water delivery had a regressive effect in
distance. The NDVI was highly correlated with groundwater depth. With the groundwater
depth increasing, the NDVI decreased.

Vegetation cover and growth

According to the investigation from both transverse and lengthways directions, the
vegetation cover and growth were improved along the river course after water delivery.
Owing to the different depths of groundwater needed for vegetation to grow, the degree and
response to less water are different with diverse vegetation (Li, 2004). The sensitive area of
Phragmites communis is about 100-150m by measuring the characters of the leaves. The
sensitive area of Populus euphratica was about 200-250m after the first stage of water
delivery and extended to 800m after the fourth stage of water delivering by testing new
branch's length and the quantity of leaves. The average quantity of leaves increase by 57%,

511
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

22.2% and 9.2% respectively comparing with the end section from the upper reaches to the
lower reaches in lengthways.

Floral physiology
According to the index test of floral physiology, the physiological index of Chinese
Tamarix such as PRO, POD, SOD does not change very much as the depth of groundwater
changed from 2 to 5 meters, which indicates that this enables to Chinese Tamarix to survive.
The content of PRO of Phragmites communis will increase as the ground water level goes
down, while its POD will reduce as the ground water level brings down. The scope of
response is 150-200m in transverse section.
Populus euphratica is the only tree in the lower reaches of the river. The index of Populus
euphratica’s SOD takes on the change processes of water gradient in both lengthways and
transverse direction. The average SOD activity of Populus euphratica is from 60.66, 54.96,
53.77 to 51.15 activity unit as the depth of groundwater depth changed from 9.16, 8.41, 8.34
to 7.65 meters from upper to the lower reach area respectively, inosculating with the change
of water gradient. The influence range in transverse can amount to 250-300 metres, which is
accordant with the result by the testing on the new branch's length, and the quantity of leaves
mentioned above.
In summary, the necessary groundwater depth for Chinese Tamarix survival is from 5 to 6
metres, Populus euphratica is 4-5metres and Phragmites communis is about 3 metres.

Table 1. The fluctuation of groundwater depth in each time of water delivery in lower reaches
of Tarim River.

First Second Third Fourth

Water delivering
time time time time
Groundwater depth (m) 7.74 3.79 3.16 2.66
Uplift extent (%) 21.6 66.4 42.3 55.4
Duration of water delivering (days) 61 104 67 110
Amount of water delivering per day (×104m3) 162.02 211.54 294.03 266.37

References
Chen Y.N.; Zhang X.L.; Zhu X.M.; Li W.H.; Zhang Y.M.; Xu H.L.; Zhang H.F. & Chen Y.P. (2004)
Ecological effect of water delivery in dried-up watercourse in lower reaches of Tarim River,
Xinjiang. Sciences in China (D) 34(5): 475-482.
Xu H.L.; Song Y.D. & Chen Y.N. (2003) Dynamic change of groundwater after ecological water
transport at the lower reaches of Tarim River. China Environmental Science 23(3): 327-331.
Li W.H.; Chen Y.P.; Zhang H.F. & Hou P. (2004) Response of vegetation to water input at lower dry
Tarim River. Journal of Desert Research 24(3): 301-305.

512
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

Site selection in dynamic landscapes: a probabilistic model to protect hydroseres

in the Pantanal wetland

M.Drechsler 1, R.Lourival 2,3, H.P. Possingham 3

1
University of Leipzig - Germany, e-mail:
[email protected]
2
The Capes Foundation – Ministry of Education Brasilia - Brazil
3
The Ecology Centre - The University of Queensland – Brisbane 4072 - Australia,

The great majority of conservation planning algorithms ignores an important feature of

the real world: ecosystem dynamics. These algorithms have been a support tool for decision
makers in terrestrial, marine and freshwater ecosystems. They offer a spatially explicit
optimization mechanism capable of reducing most common bias from ad hoc reserve
selection, through a transparent, target oriented and defensible arguments for negotiation
(Pressey and Cowling 2001).
Incorporating ecosystem dynamics is currently regarded as one of the major challenges
in reserve design (Cabeza and Moilanen 2003). Broadly speaking, ecosystem dynamics are
driven by two processes: disturbance (Sousa 1984), through events like fire of flooding, and
succession. There have been many attempts to classify disturbance regimes based on the
extent, intensity and frequency of disturbance (Tutin 1941). Usually, for catastrophic events,
subsequent succession is regarded as a unidirectional process (Platt and Connell 2003),
where a disturbance event sets back the system to the earliest state, from which it evolves to
later states. A less frequently considered type of succession is what might be termed
“bidirectional succession”. An important type of system that exhibits this type of dynamics are
wetlands (Ward 1998; Junk 1999). In a wetland a flood is not an event that destroys
everything and sets the system “back to zero”. Instead, there is only movement back and
forth on a terrestrialization-paludification gradient (Klinger 1996). Hence in our site selection
framework we combine succession and disturbance into one process – bidirectional
succession.
We consider two different levels of state variables: an abiotic and a biotic one. At the
abiotic level we simply define a site by the time since the last disturbance event (“site age”).
From a biotic perspective a stage is defined by the presence of a particular seral community.
After integrating the biotic to the abiotic models we investigate the performance of various
site selection strategies for a given target where the target is the probability of having a
particular abiotic or biotic state protected. We describe the transition probabilities as a
Markov process and we use simulated annealing to optimize the selection of sites for
conservation.
Our result shows that the relationship between site age and the observed species
community is not deterministic (Figure-1), such that for a given site age several species
communities can be observed with some likelihood. This is plausible, since communities that
that survive in such fluctuating environments do not simply disappear; they expand and
contract their distribution responding to the disturbance history of the site (Junk 1999).
Another important issue is the spatial correlation between sites, which play an important role
in selection, since neighbouring sites are assumed be equally affected by disturbance. Our
results do not support such view, and sites showed independency regarding the communities
they sustain. This is also plausible because differences in micro-topography and flood
frequencies seem to counterbalance those deterministic forces.

Figure 1: Probability of observing a community (1-5) on a site if the last flood was t years
ago. Negative time intervals, t<0, mean floods in every of the previous |t|+1 years. The
probability of flooding is q=0.5.

513
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

References
Cabeza, M. and A. Moilanen (2003). "Site-selection algorithms and habitat loss." Conservation
Biology 17(5): 1402-1413.
Junk, W. J. (1999). "The flood pulse concept of large rivers: learning from the tropics." Archiv Fur
Hydrobiologie(3): 261-280.
Klinger, L. F. (1996). "The myth of the classic hydrosere model of bog succession." Arctic and Alpine
Research 28(1): 1-9.
Platt, W. J. and J. H. Connell (2003). "Natural disturbances and directional replacement of species."
Ecological Monographs 73(4): 507-522.
Pressey, R. L. and R. M. Cowling (2001). "Reserve selection algorithms and the real world."
Conservation Biology 15(1): 275-277.
Sousa, W. P. (1984). "The Role of Disturbance in Natural Communities." Annual Review of Ecology
and Systematics 15: 353-391.
Tutin, T. G. (1941). "The hydrosere and current concepts of the climax." Journal of Ecology 29: 268-
279.
Ward, J. V. (1998). "Riverine landscapes: Biodiversity patterns, disturbance regimes, and aquatic
conservation." Biological Conservation 83(3): 269-278.

514
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

Indicators for the linkage of forest, river, village and ocean ecosystems

Y. Natuhara, A. Imanishi, K. Imai

Graduate School of Global Environmental Studies, Kyoto UniversityYoshida-Honmachi,

Sakyo, Kyoto 606-8501 JAPAN.
e-mail: [email protected]

Introduction

Landscape ecology focuses on interacting ecosystems. Linkage of different ecosystems

produces unique ecosystem services that are not produced by a single ecosystem. For
example, gravel riverbeds provide spawning sites for salmon that grow up in the ocean. Then
the salmon migrate upstream from the ocean to eventually provide nutrients for forests
(Helfield & Naiman 2001). However, the linkages between ecosystems have been interrupted
by urban development, road construction, dam, land reform in paddy field, etc. Some
ecosystems deteriorate by the interruption of the linkage. We reviewed studies on the
ecosystem linkages in landscapes. We explore indicators of the linkages to find the priority
sites for restoration.

Function of the linkage and problems identified in Japan

Catchments
Many studies have reported that the increased agricultural land within catchments leads to
declines in water quality, habitat, and biological assemblages (e.g. Richards et al. 1993).
Fine sediment increases by decreasing flow resulted from dam construction and river
improvement, and by increasing sediment production resulted from deforestation. The
density of macroinvertebrates in agricultural catchments was only 10-20% of that in the
forested catchments (Nagasaka et al. 2000).

Riparian forest
Riparian forests have diverse functions. Approximately 30 m is required for maintaining
shading effects and the provision of organic litter and woody debris. But 100-200 m is
needed for the habitat of animals (Takahashi et al. 2003). Riparian forests decreased in the
middle and lower reaches in Japan.

Dam
Interruption of the continuity of sediment transport by dams causes sediment-starved and
prone to erode the channel bed and banks, producing channel incision, coarsening of bed
material, and loss of spawning gravels for salmon and trout (Kondolf 1997). Dams change
down stream ecosystems and the effect reach ocean ecosystem (Humborg et al. 1997).

Floodplain
Floodplains provide spawning and nursery sites for fish (Halyk & Balon 1983) and habitats
for aquatic plants (Tremolieres 2004). But floodplains have decreased because of
modification of the channel and dredging, flow control by dams and bank construction in
Japan. The results of experimental flooding suggest that diversity of floodplain morphology
should be preserved to maintain larval habitats (Sagawa et al. 2005). Paddy fields also have
functioned as floodplains for fishes and plants for 2000 years. Catfish and loach move into
paddy fields from rivers after irrigation, and spawn (Naruse & Oishi 1996). However,
implementation of land consolidation projects in the paddy fields interrupts this linkage
between rivers and paddy fields (Suzuki et al. 2001).

Shore erosion

515
Theme 4: Ecohydrology, water and rivers
4.3 Open Session 17: Ecohydrology, rivers and wetlands

Sandy shore area decreased by 5059 ha in the 70 years until 1978 and by 2395 ha from
1978 to 1992 due to shore erosion. The forest area in the basins has increased and the
sediment load in the river channel had decreased. Tsutsumi (2005) suggested that supply of
sand from river to shore increased mangan concentration of sediment that causes negative
impact on clam production.

References
Halyk, L. C. & Balon, E. K. (1983) Structure and ecological production of the fish taxocene of a small
floodplain system. Canadian Journal of Zoology 61: 2446-2464.
Helfield, J. M. & Naiman, R. J. (2001) Effects of salmon-derived nitrogen on riparian forest growth
and implication for stream productivity. Ecology 82: 2403-2409.
Humborg, C; Ittekkot, V; Cociasu, A. & Bodungen, B. V. (1997) Effect of Danube River dam on
Black Sea biogeochemistry and ecosystem structure. Nature 386: 385-388.
Kondolf, G. M. (1997) Hungry Water: Effects of Dams and Gravel Mining on River Channels.
Environmental Management 21: 533–551
Nagasaka, A; Nakajima, M; Yanai, S. & Nagasaka, Y. (2000) Influences of substrate composition on
stream habitat and macroinvertebrate communities: a comprehensive experiment in a forested
and an agricultural catchment. Ecology and Civil Engineering 3: 243-254
Naruse, M. & Oishi, T. (1996) Annual and daily activity rhythms of loaches in an irrigation creek and
ditches around paddy fields. Environmental Biology of Fishes 47: 93-99.
Richards, C; Host, G. & Arther, J. W. (1993) Identification of predominant environmental factors
structuring stream macroinvertebrate communities within a large agricultural catchment.
Freshwter Biology 29: 285-294.
Sagawa, S; Kayaba, Y; Arai, H. & Amano, K. (2005) Habitat selection by larval cyprinid fishes:
relationship between larval habitats and experimental flooding. Ecology and Civil Engineering 7:
129-138
Suzuki, M; Mizutani, M. & Goto, A. (2001) Trial manufactures and experiments of small-scale
fishways to ensure both upward and downward migration of freshwater fishes in the aquatic area
with paddy fields. Ecology and Civil Engineering 4: 163-178
Takahashi, K; Hayashi, S; Nakamura, F; Tsuji, T; Tsuchiya, S. & Imaizumi, H. (2003) A review on
buffer width required for ecological functions of riparian forests. Ecology and Civil Engineering 5:
139-168
Tsutsumi, H. (2005) Marked decline of clam harvesting fisheries and environmental changes on the
tidal flats facing the Ariake Sea in Kumamoto Prefecture. Ecology and Civil Engineering 8: 83-102
Tremolieres, M. (2004) Plant response strategies to stress and disturbance: the case of aquatic
plants. Journal of Biosciences 29: 461-470.

516
Theme 4: Ecohydrology, water and rivers
4.4 Posters

4.4 Posters

Landscape models predict distribution and assess protective status of freshwater

fishes

K. M. Mattson1, P. L. Angermeier2 , S. D. Klopfer3

1
Department of Fisheries and Wildlife Sciences, Virginia Polytechnic Institute and State
University, Blacksburg, VA 24061-0321, USA,
e-mail: [email protected].
2
United States Geological Survey, Virginia Cooperative Fish and Wildlife Research Unit,
Virginia Polytechnic Institute and State University, Blacksburg, VA 24061-0321, USA.
3
Conservation Management Institute, 1900 Kraft Drive, Blacksburg, Virginia 24061, USA.

Introduction
Freshwater biota are highly imperiled worldwide due to intensive human land and
water uses, which impact the flow regime, water quality, physical habitat, biotic interactions,
and energy dynamics of aquatic ecosystems. Protecting freshwater biodiversity requires
innovative approaches that enable conservationists to retain and restore valued biotic
components and use a landscape perspective to model ecological conditions. Such
approaches would inform decision-makers about the spatial relations among biodiversity
targets, human uses of land and water, and the biotic risks associated with those uses. We
employed a geographic information system framework to integrate maps of the type, extent,
and severity of anthropogenic stressors with maps of fish distributions, and to assess the
long-term sustainability of species within catchments. We applied our approach to 107
catchments in the upper Tennessee River basin (UTRB),

Methods
We combined projected occurrences of fish species, based on habitat descriptors,
with projections of threats to the major drivers of biological integrity, then assessed the extent
to which current land/water uses impair the long-term persistence of each species. We
derived habitat associations from relations between known occurrences and landscape
descriptors such as river drainage, physiography, stream size, and elevation, then used them
in logistic regression models to predict reach-specific occurrences for 118 species. Readily
available data on mining, impoundments, urbanization, agriculture, road density, and other
human activities were used to rank the extent and severity of threats within catchments
according to their expected impact on stream integrity (1). Each catchment was assigned a
protective status (highly vulnerable, somewhat vulnerable, not vulnerable) based on
catchment-level risk (low, moderate, high) and occurrences of conservation targets.

Results and Conclusions

Although most (76%) of the fish species we modeled are not imperiled overall, many,
including some ubiquitous species, are at risk in the UTRB. Several species are highly
vulnerable due to their limited distributions. Of the remaining species, 42% were under-
represented in low-risk catchments and considered somewhat vulnerable. The human
land/water uses posing the most risk to biota are manufacturing and effluent sites, bridges,
and mining activity. Our approach integrates natural and anthropogenic landscape features
with catchment-specific risks and ranks protective status of catchments based on the
juxtaposition of anthropogenic stressors with valued freshwater biota.

References
Mattson, K. M., and P. L. Angermeier. 2007. Integrating human impacts and ecological integrity into
a risk-based protocol for conservation planning. Environmental Management 39:125-138.

517
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Landscape Ecology in the Pantanal, integrating science and society

R.H.G. Jongman1, M. van Eupen1, B. Makaske1, C.R. Padovani 2

1
Alterra Wageningen UR, PO Box 47, Wageningen, The Netherlands, e-mail:
[email protected]
2
Embrapa CPAC, Caixa Postal 109, 79320-900 Corumbá, Mato Grosso do Sul Brazil

The Pantanal in Brazil measures about 250.000 km2 and consists of a number of large rivers
in a joint wetland area. The economy is based on cattle breeding, fishing and ecotourism.
Large areas are dominated by the river regime of the Paraguay and its tributaries. In the wet
season large areas of the savannah are flooded. At present erosion and silting up make the
one of its rivers, the Taquari into an unstable braided system. This is at the moment a major
problem causing a more or less permanent inundation of 5.000 to 8.000 km2 in the subregion
Paiaguas in stead of periodic inundation. The result is a decline of the fish populations and a
decline of the area for cattle breeding.
The important added value of the project is that knowledge has been set into context of the
river as an ecosystem and a socio-economic unit. Within that context the links between
science fields of hydrology, ecology and economics have been made. It has been concluded,
that biodiversity can be important for regional economy: Less aquatic biodiversity means less
fish, less fishing tourism, less ecotourism, less income and more isolation. The relationship of
the hydrological behaviour of a river system and its ecological functioning (the flood pulse)
can also be an important lesson to be learned for river management in Europe.
In a situation where politics is important, it is essential that all are involved and discuss
matters using political and scientific arguments and the right economic and hydrological
models to explain the situation. Proper knowledge appeared the only convincing argument
for taking decisions as it is important that the results of the project will be accepted both in
the region and by authorities that supervise the region.
We concluded that technical solutions are not always the best solutions. Building a dam have
been proposed by different stakeholders and the project was capable of showing the
consequences, both positive and negative especially the negative ecological consequences
and the long term problems that they cause. The detailed conclusions were:
− The important added value of the project is that knowledge has been set into context of
the river as an ecosystem and a socio-economic unit.
− Biodiversity can be important for regional economy: Less aquatic biodiversity means less
fish, less fishing tourism, less ecotourism, less income and more isolation. The
relationship of the hydrological behaviour of a river system and its ecological functioning
(the flood pulse) can also be an important lesson to be learned for river management in
Europe.
− It is essential that all are involved and discuss matters using political and scientific
arguments and the correct economic and hydrological models to explain the situation.
Knowledge appeared the only convincing argument for taking decisions.
− It is important that the results of the project will be accepted both in the region and by
authorities that supervise the region.
− Technical solutions are not always the best solutions: Technical solutions such as building
a dam have been proposed by different stakeholders and the project was capable of
showing the consequences, both positive and negative.
In general, understanding the hydrological dynamics and related ecology of rivers at the
basin scale and communicating this with the organisations and people involved is the basis
of economically and ecologically sustainable river management.

518
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Channel and vegetation changes in the Sharda River - conservation implications

for swamp deer in Kishanpur Wildlife Sanctuary, Uttar Pradesh, India

N. Midha1, P.K. Mathur2

1
WII-MoEF-NNRMS Project, Wildlife Institute of India, Post Box #18, Chandrabani, Dehra
Dun - 248 001, India.
e-mail: [email protected]
2
Department of Landscape Level Planning and Management, Wildlife Institute of India, Post
Box #18, Chandrabani, Dehra Dun - 248 001, India.

Introduction

Jhadi taal, an important swamp in Kishanpur Wildlife Sanctuary (KWLS) on the

floodplains of Sharda River is one of the few surviving strongholds of a key species of Terai
ecosystem i.e. Northern swamp deer (Cervus duvauceli duvauceli). Changes in the river
course during recent decades have brought the channel dangerously close to it. Jhadi taal
and eventually swamp deer future is considered precarious owing to potential flooding,
siltation or loss of habitat. Present study assessed channel changes in ca. 10 km stretch of
Sharda River and its likely implications on conservation of swamp deer.

Methodology

Changes in bank line position, length, area, sinuosity and land use/cover were
documented using toposheets (1948 and 1965) and satellite imageries of Landsat (MS -
1977, TM -1990 and 1999) and IRS ID (LISS III - 2001). Visual interpretation was carried out
to prepare river channel and land use/cover maps. Generated maps were overlaid.
Locational Probability Model was also developed (Wasklewicz et al., 2004).

Results

Analyses revealed channel shift by 3.1 km from 1948 to 2001 towards Jhadi taal and
distance in 2001 remained just 100 m. Channel length, area and sinuosity oscillated
considerably with a net increase in sinuosity by 15% and area by 96%. Loss of Sal (Shorea
robusta) forest and grassland habitats by 12% and 3% respectively was also registered. Net
gain of 5% was documented in agriculture area. Locational Probability Model predicted that
51% area of the studied channel as ‘unstable’, 45% ‘moderately stable’ whereas only 4%
area as ‘stable’.

Conclusions

Study revealed noticeable fluctuation in channel characteristics in short span of 53–

year of assessment period. The probable reasons for such behavior could be attributed to
rapid land use changes particularly deforestation on Nepalese side during 1960s-70s and
channelization of Sharda River in upper reaches. Enchanced dynamism of river has definitely
imperiled one of the strongholds of already endangered deer.

References
Wasklewicz, T.A.; Anderson, S. & Liu, Pin-Shou (2004) Geomorphic context of channel locational
probabilities along the Lower Mississippi River, USA. Geomorphology 63: 145-158.

519
Theme 4: Ecohydrology, water and rivers
4.4 Posters

From catastrophe to conservation: Protecting the pond complex of Tommelen,

Belgium

J. Hollinshead1, T. de Bie2, A.Hull1

1
Ponds Research Unit, Liverpool John Moores University, UK.
e-mail: [email protected]
2
University of Leuven, Belgium

At Tommellen, near Hasselt, Belgium an important wildlife site consists of over 117 ponds,
plantation woodland and grassland. The ponds are fairly uniform morphologically, being in
the main circular, and occurring in two basic sizes, the smaller around 5m in diameter, and
the larger around 8m in diameter, with depth 2 – 4m respectively. Age structure is completely
uniform, being created more or less simultaneously during an air raid, which took place on
April 8th. 1944. One hundred and ninety eight B-26 Marauder bombers, and 32 P-47
Thunderbolt fighter-bombers of the 9th USAAF attacked rail installations at Hasselt, Belgium,
in preparation for the Normandy landings, creating a substantial crater field, some of which
has survived as the pond complex. Despite the uniformity of morphology and age,
ecologically these ponds display extreme variation across the site and between neighbouring
ponds, in both floral and faunal communities, which can only partially be explained by
differences in size and hydroperiod. The site as a whole is notable for invertebrate species,
particularly Odonata, and amphibian species richness and abundance. The local community
appreciate the value of the site, which is now almost completely isolated by road and rail
corridors, industrial and housing development. Older members of the community, and
American veterans who participated, recognise the site as a link to their past and the
experiences of the 1940’s. The importance of the historic and social dimension of pond
conservation is emphasized by the role explanation of these aspects is playing in the
campaign for statutory protection for the site, which is being carried out by local
conservationists, researchers at LJMU and KU Leuven and supported by the European
Ponds Conservation Network.

520
Theme 4: Ecohydrology, water and rivers
4.4 Posters

A research agenda for the Pantanal and the Upper Paraguay River basin

R.H.G. Jongman1, J.M. Leeuwestein2, P. Girard 3

1
Alterra Wageningen UR, PO Box 47, 6700AA Wageningen, The Netherlands
e-mail: [email protected]
2
Ecobusiness, SHIN 08, Conj 11, C18, Brasilia, Brazil; 3 CPP, Av. Fernando da Costa s/n
Cuiaba, 78060-900, Brazil.

The Pantanal is the largest wetland of the world except for rain forests. It measures about
250.000 km2 and is one of the biodiversity hotspots in the world. The main objective of
INREP is to strengthen the cooperation between policy makers, stakeholders and scientific
institutions in Brazil, Bolivia and Paraguay and the European science community and NGOs
working in the region for sustainable water management at river-basin scale and efficiency in
water use in the Pantanal.
To achieve this goal, two preparatory workshops were held which counted on a participation
of about 80 people, as well as a Bi-regional Symposium in Campo Grande, where 76
persons – representing the three countries of the UPRB and the international community –
worked on an international research agenda for the watershed (see http://www.inrep.nl and
http://www.ec.europa.eu/research/water-initiative).
The approach that has been chosen was the participatory methodology analysing Strengths,
Weaknesses, Opportunities and Threats (SWOT), focusing on the potentials for both
sustainable management and development. The analysis was based on a country document
for each country. The result was an number of priorities for action This information was the
basis for the Bi-regional Symposium of Campo Grande. The documents where discussed
during the Symposium, which resulted in a final SWOT for the region as a whole and over 70
priority actions resulting in a in a Research Agenda for the region.
On the initiative of the Bolivian participants of the Symposium, an extra workshop was
realized in Santa Cruz, Bolivia, on the 7th of November of 2006, providing the opportunity to
include more stakeholders from this country to define the research activities. At this event, 22
persons participated.
The outcomes of the Symposium, the following major issues are identified for the UPRB:
1. Environmental change: diagnosis and remediation
1.1. The Paraguay Hidrovia
1.2. The climate functions and land use of the Pantanal
2. Assessment and mapping
3. Natural resources management
4. Governance, laws, institutions and policy
4.1 Social organizations in the Pantanal
4.2 Research organizations in the Pantanal
4.3 Governmental organisations and legislation in the Pantanal
5. Sustainable development: diagnosis and solutions;
6. Capacity building and education.
For these major topics, 32 main research activities have been identified. This Research
Agenda is the result of a participative effort involving the main stakeholders from the three
countries, European research institutions and NGOs working in the region. It is a planning
document for governmental, social and research institutions to elaborate future projects. In
being a joint action it establishes the key-elements of knowledge required for sustainable
development in this region. Research projects that have been integrated within the scope of
this Agenda do have a conscientious basis in the region and are supported by governmental
agencies, NGOs and civil society.

Acknowedgement: This project is carried out in the framework of EU FP6

521
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Landscape scale influences on the streams habitats and biota: the riverine
system of Natisone (Italy)

R. Zorza 1, M. Sigura 2, G. Oriolo 3, P. Bonfanti 2, G. Honsell 1

1
University of Udine, Department of Agro-Industrial Economics and Biology: Plant Biology
Section, via Cotonificio 108, 33100 Udine, Italy. e-mail: [email protected]
2
University of Udine. Department of Agricultural and Environmental Sciences, via delle
Scienze 208, 33100 Udine, Italy.
3
University of Trieste, Department of Biology, via Giorgieri 9/10, 34127 Trieste, Italy.

Introduction

Landscape ecology emphasizes the interaction between spatial pattern and ecological
processes and has conceptual and technical tools relevant to monitoring rivers. Quantifying
landscape structure is a prerequisite to the study of landscape function and change. A new
trend in the calculation of landscape metrics is to integrate the traditional calculation program
into the geographic information system (GIS) so that data models and spatial analysis tools
can be utilized effectively. Our objectives were to analyse the characteristics of landscape
structure and its influences on the stream habitats and biota of the riverine system of
Natisone. To understand the spatial complexity of the river we have used physical habitat
measure, a number of bioindicators following the Water Framework Directive (WFD
2000/60/CE), vegetation and landscape indicators.

Methods and Results

The Natisone is a torrential prealpine river in the province of Udine (Friuli Venezia-Giulia,
Italy) with a bed characterized by gravel and stones.
A vegetation map (1:5000) was prepared to analyze the habitat of each vegetation type
and to define six homogeneous vegetal landscape units along the river using 75 transects
perpendicular to the broken line of the stream (40 Km).These transects described the relative
presence of 34 mapped classes. We used landscape metrics calculated by FRAGSTATS on
the vegetation and land use map (1:25000) to evaluate the influence of human pressure and
to confirm the status of riverine system.
Upstream we found units characterized by high connectivity and low diversity, due to the
dominance of woods, large and species-rich patches types that reflect the high biodiversity of
this area. Downstream there are units with a lower naturality and a decrease in quality of
water as effect of a higher human pressure by intensive agriculture and urban development.
The downstream landscape pattern is an heterogeneous mosaic of fragmented patches
characterized by a small number of natural patches strictly connected with the bed river that
underline the decrease of naturality along it and that the gradient of the naturality is
confirmed in the integrated reading of the data. A good-acceptable water quality was
assessed in ten sampling stations by the Diatom Eutrophication/Pollution Index (EPI-D)
(Dell’Uomo, 2004). In this study we have noticed that Natisone has a Good Ecological State
in the sense of WFD with a lot of natural habitats. Attempts have been made to understand
disturbance factors in the landscape and to underline the impact of land uses on spatial
habitats modifications along Natisone river.

References
Dell’Uomo, A. (2004) L’indice diatomico di Eutrofizzazione/Polluzione (EPI-D) nel monitoraggio delle
acque correnti. Linee guida, APAT, Roma, 102 pp.

522
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Corridor effect of the spatial changes of landscape patterns in arid areas: a case
of the river corridor areas in the middle and lower reaches of Tarim River

H. Zhou, 2, D. Xiao, 2 & K. Zhou

1. Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Urumqi

830011, China (e-mail: [email protected]);
2. Shenyang Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016,
China

Introduction

The radiative effect and the characteristics of the river corridors on landscape and patch
levels were analysed in the middle and lower reaches of the Tarim River.
Study area
This river corridor areas (38°47′52″~41°38′30″N, 85°33′38″~88°40′30″E) are dominated by
the traditional green corridors along the river section from the Qara Reservoir in Yuli County
to the Taitema Lake in Ruoqiang County and is bounded by the piedmont and the large-area
dunes. The study area has a length of 413 km and width of 87 km with a total area of 34526
km2.
Study Method
The buffer zones perpendicular to the watercourse were divided in a 5-km interval away
from the watercourse along both riversides 10 km wider using the method of the buffer
zone analysis.The landscape pattern indexes of each buffer zone were calculated, and the
indexes of the main landscape patterns were analyzed so as to reveal.
Results
(1) The radiative width is generally a 30-km buffer zone in the study area, and can be up
to 50~70 km in some areas. (2) The landscapes of woodlands and wetlands are generally
distributed within the 10 km buffer zone along both riversides.(3) The closer the buffer zone
to the watercourse, the higher the patch density and the largest patch index are. (4)In the
ecological and environmental regeneration for the study area, the extent of ecological effect
of the river corridors must fully be considered.

References
Ward J.V., Maland F. & Tockner K., Landscape ecology integrates pattern and process in river
corridors, Issues in Landscape Ecology (ed. Jon A. Wiens & Michael Moss R.), International
Association for Landscape Ecology Fifth World Congress, Snowmass Village, Colorado, USA,
1999, 97-102.
Xiao D., Hu Y., Li X., et al., Study on Landscape Ecology in the Bohai Delta (in Chinese), Beijing:
Science Press, 2001.
Zhou Huarong Study on landscape ecology for the river corridor of the middle and lower reaches of
Tarim River, Xinjiang, China, PhD thesis, Shenyang Institute of Applied Ecology, Chinese
Academy of Sciences, Shenyang , 2004, 156.

523
Theme 4: Ecohydrology, water and rivers
4.4 Posters

A landscape ecological study of the Ganges-Brahmaputra-Meghna delta basin,

Bangladesh

T. Byomkesh1, N. Nakagoshi1, M.S. Rashid2

1
Graduate School for International Development and Cooperation, Hiroshima University,
Japan.
e-mail: [email protected]
2
Department of Geography and Environment, Jahangirnagar University, Dhaka, Bangladesh.

Bangladesh is a land of extremes and home to an estimated 140-plus million people in

144,000 sq. km. Continuous severely degradation of terrestrial, aquatic and floodplain
ecosystems, land covers change, deforestation, land degradation, water and air pollution,
salinity, affect of global climate change and frequent natural calamities (e.g. flood and
cyclone) are some of the environmental issues of Bangladesh. Estimates in 1990 revealed
that Bangladesh had less than 0.02 ha of forestland per person. The rapidly increasing
Bangladesh population has dramatically amplified the demands for natural resources and
has caused significant changes in the quantity and quality of natural resource. To solve
these problems well thought out strategies are needed. Presently Environmental
management practices are trending away from simple, local-scale assessments toward
complex, multiple-stress or regional assessments. Landscape ecology provides the theory
behind these assessments while GIS with Remote Sensing supply the tools to implement
them. As changes throughout landscapes occur, the overall structure of ecological
communities is affected. In order to understand landscapes, it is important to recognize the
patterns that are created along both spatial and temporal scales. On the basis of the above
background the objectives of this study is to (i) Finding land use pattern through time and
space in the delta (ii) Characterized the types, rates and temporal variability of land use and
land cover changes over time and (iii) Identify the driving forces and consequences of
changes. This study will make use of remotely sensed data and GIS technologies to evaluate
qualitatively and quantitatively outcome of Bangladesh land cover/use distribution. Through
the data analysis the land use pattern model of Bangladesh will be established. It is
understood that this study will be useful for future environmental planning and management
that conserves and sustains the ecosystem support needed for the livelihood of
Bangladesh’s poor dense population’s livelihoods.

References
Loveland T., (2001). Strategic plan for the U.S. climate Change, Science Programme. Chapter 6,
Land-Use/Land-Cover Change, U.S.A. pp 63-70.
National Adaptation Programe of Action (NAPA). (2005). Final report. Ministry of Environment and
Forest, Government of the People’s Republic of Bangladesh.
State of Environment, Bangladesh (2001). Report. UNEP. ISBN: 92-807-2017-1.
http://landcover.usgs.gov/pdf/anderson.pdf
http://www.epa.gov/esd/land-sci/pdf/overview-fs.pdf
http://www.epa.gov/esd/land-sci/region-assess.htm#project-summary

524
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Quick scan system analysis as a basis for ecological restoration measures

J.C.J de Hoog1, M.H. Jalink2

1
Water board De Dommel, Postbus 10001, 5280 DA Boxtel, The Netherlands,
email: [email protected]
2
Kiwa Water Research, Postbus 1072, 3430 BB Nieuwegein, The Netherlands.

Introduction
In the convention of Cork (2004) the water boards, provincial authorities and other
organizations in the province of Noord-Brabant (the Netherlands) have stated that they will
prioritize the ecological restoration of the most important and critical nature reserves, the so-
called nature pearls. The water boards have taken the responsibility to realize the required
ground water levels and surface water levels for restoration of wetland plant communities.

Before starting detailed research and ground water modeling in about 40 nature pearls,
Water board De Dommel wanted to know which restoration measures are likely to be
successful, and what characteristics of the hydro-ecological system should be incorporated
in the ground water models. Kiwa Water Research and Water board De Dommel developed
a method for a quick scan ecosystem analysis that requires only a few days per nature pearl.

Method
General knowledge of hydrological, hydrochemical, ecological processes and indicators,
and a framework of hydro-ecological system types in Noord-Brabant is used to identify the
local hydro-ecological system types of each nature pearl, the causes of non-optimal
functioning and promising restoration measures. The information used consists of soil type,
elevation, depth of drainage, hydrolic head, groundwater quality data and the presence of
certain plant communities or species. Additionally, a field excursion is necessary to acquire
specific information of the location such as soil structure. The findings are presented in
hydro-ecological cross-sections showing the most important gradients and underlying
processes. This information forms a basis for further, detailed modeling

Case study
The quick scan ecosystem analysis was tested in the area Bossche Broek-south (near ’s-
Hertogenbosch). In this area, results of an earlier hydrologic model indicate seepage from a
deep aquifer, so it was assumed this would support base rich hay meadows. However, the
groundwater quality data show the dominance of local groundwater systems, rich in sulfate
and potassium, in the upper four meters. No signs of groundwater from the deeper aquifer
were found. Additionally the thick layers of clayey peat and clay show the former influence of
flooding and the Carex acuta vegetation is typical of (formerly) flooded wet meadows. It was
concluded that the thick peat and clay layers and loam layers in the subsoil may prevent the
seepage from the deep aquifer. These layers should be incorporated in the model. Nature
goals should be adapted to this (formerly) flooded system instead of being dependent on
seepage from the deeper aquifer.

Perspectives
In the Bossche Brook south the quick scan appeared to be an effective tool to gain insight in
the hydro-ecological system and the possibilities for ecological restoration. The method is
now being carried out for 17 of the 40 nature pearls of Water Board De Dommel. The poster
depicts the main goals, the main principles and the method of the quick scan. With the case
study of Bossche Broek-south we will illustrate the method and focus on the implications for
further actions and research.

525
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Changing Landscape Mosaic of the Mid-Paraíba do Sul River Valley: Geo-

Hydroecological Responses to Eucalyptus Growth (1)

A.M. Sato, L.G.G. Vianna, R.C.G. Almeida, A.S. Avelar, A.L. Coelho Netto

GEOHECO/Geo-Hydroecology Laboratory, Geography Department, Institute of

Geosciences, Federal University of Rio de Janeiro, 21941-590, Brazil.
e-mail:[email protected]

Introduction
Since 2000 the landscape mosaic of Middle Paraíba do Sul river valley has been rapidly
modifying by the spreading of Eucalyptus patches in substitution to cattle grazing grasslands.
Evidences of landscape instability are given by the occurrence of deep gullies which are
eroded by the action of groundwater flows and landslide scars as previously shown by
Coelho Netto (1999). Our current research interests are driven to understand the role played
by Eucalyptus controlling the hillslope hydrology and erosion and explain how these changes
may affect the regional landscape system. This paper presents the initial studies with special
attention to the effects on the biota-soil-water interactions.

Study Area and Methods

Field work has been conducted in the Sesmarias river basin (149 km2), a tributary of
Paraíba do Sul river (SE Brazil), at the Eucalyptus Farm Monte Alegre (hybrids E. urophylla x
grandis, 3x2 m spacing, planted in April 2004). Average annual precipitation is 1700 mm and
rainfall concentrates from October to March. A land use map of the basin (rainforest,
grassland, Eucalyptus and urban zones) was generated via CYBERS images (August 2006)
and geo-referenced with LANDSAT. A representative hillslope was selected within a first
order basin to examine litter production, storage and water retention (WR) and to ascertain
the physical characteristics of soil and hydrological measurements (gross rainfall,
interception and overland flow production) at both the hillslope divide (8o) and side-slopes
(24o). Sampling sites followed the spatial pattern of tree planting and included samples under
tree canopies (UC) and between tree trunks (BT). Hydrologic records refer to the period from
October to December 2006, while the litter records refer from October 2006 to January 2007.

Results and Discussion

Grasslands predominate in 63.5% of the basin area; Eucalyptus concentrates in the hilly
lowlands and occupies 3.1% of total area; rainforest remnants (31.7%) occur only in the
mountainous domain; and a small urban zone (1.6%) is adjacent to Paraíba do Sul river.
Eucalyptus litter storage is spatially non-uniform: at BT (8.925 Mg.ha-1) is greater than UC
(7.421 Mg.ha-1). Average side-slope WR is greater (342%) than at hillslope divide (183%)
explained by greater presence of dry grass remnants on the side-slope. Deus (1991) found
that grass WR may attain 500% in the dry period. Throughfall UC is frequently higher than in
the open area due to convergent flows from branches and leaves: mean throughfall at UC
sites is 140% ranging from 567% to 55%; at BT sites the average value is 84% ranging from
176% to 51%. Overlandflow production is very low on steep side-slopes; it can be neglected
in the hillslope divides. So the Eucalyptus growth is highly favorable to infiltration.
(1)
Financial support: CNPq & FAPERJ

References
Coelho Netto, A.L. (1999) Catastrophic Landscape Evolution in a Humid Region (SE Brazil):
inheritances from tectonic, climatic and land use induced changes. Supplementi di Geografia
Fisica e Dinamica Quaternária III. 3: 21-48.
Deus, E. (1991) The Role of Ant (Atta spp.) Excavation in Pasture Area Slope Hydrology – Bananal
(SP). Master’s Thesis, PPGG/IGEO/UFRJ. 135p.

526
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Analysis of the quality of the riparian forest as a bioindicator in mediterranean

river basins monitoring

S. Sànchez Mateo, M. Boada Juncà

Environmental Science and Technologies Institute. Autonomous University of Barcelona

(UAB). Campus de la UAB. Edifici Cn. Torre 5 (parells). 08193 Bellaterra (Cerdanyola del
Vallès). Barcelona. Spain.
e-mail: [email protected]

The ecological study of the riparian communities has become in one of the main keys to
understand the dynamics of the fluvial systems. In this context, the European Water
Framework Directive (Directive 2000/60/EC) defines new sustainability approaches to protect
the hydric resources through river basin analysis, in terms of ecosystems and anthropic
activities repercussions, in order to assess the ecological state by monitoring indicators.

And this is the framework in which L’Observatori project is submitted; a regional pioneer
initiative started up in 1996 that aims to develop and monitoring sustainability indicators on a
river basin scale in Catalonia (Spain). L’Observatori proposes a holistic approach towards
both the global comprehension of the fluvial ecosystem functioning and the dynamic
transformation at which Tordera river basin is subjected. It aims a continued and integrated
evaluation of the ecological, hydrological and social status of the basin. One of the research
lines that L’Observatori includes, referring to the ecological aspect, is the analysis of the
quality of the riparian forest as a measure of sustainability, which defines the status of
riverine resources.

The role of the riparian forest in the fluvial ecosystem dynamics can be defined in a
multifunctional scope since its hydrological, ecosystemical and economical implications.
Taking into account these aspects, the main goal of this study is the proposal of a
methodology for the monitoring of the mediterranean riparian forests at the Tordera river
basin, based in the evaluation of its quality by applying the QBR index (Riparian Forest
Quality Index) in serial transects throughout the area of study.

527
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Anthropogenic interference on fish assemblages, Meia Ponte River, Goiás, Brazil

M. Pazete de Oliveira, F.L. Tejerina-Garro

Centro de Biologia Aquática, Universidade Católica de Goiás, Av. Engler s/n, Jardim
Mariliza, CEP 74650-010, Goiânia, Goiás, Brazil.
[email protected]

The Meia Ponte River basin (Upper Paraná River) occupies an area of 12180 km²,
supplying water for industrial and agricultural activities (Fialho and Tejerina-Garro, 2004), but
also receiving domestic seweage. This work aims to compare the fish assemblage
composition and structure of five river stretches (each one of 1000 m) Considering four
ecological descriptors (abundance, richness, Shannon-Wiener diversity, equitability) and
similarity.
Fish were collected each two months between March and November/2006 using four
set of nets (meshes 12, 15, 20, 25, 30, 35, 50 and 70 mm) following the protocol suggested
by Tejerina-Garro and Mérona (2001). Ecological descriptors and fish assemblages similarity
(cluster analysis using the Ward method followed by a Monte Carlo test, 1000 interactions)
were calculated using the software ADE-4 (Thiolouse et al., 2001).
They were collected 2454 individuals distributed in 71 species, including eight exotic
ones. Only richness display significant differences among stretches (p=0.006), but this
descriptor does not increase in the headwaters-rivers’ mouth sense. The comparison of the
similarity indicates that the three river stretches upstream the Rochedo dam are different
from downstream ones.
Despite the presence of important tributaries of the Amazon and Paraná basin in Goiás
State and aquatic landscape modifications, systematized information about its ichthyofauna
or the influences of anthropogenic activities on fish assemblages are poor. This kind of
information can contribute to the conservation planning of the local aquatic environment,
which represents goods and services for the society (Richter et al., 2003).

References
Fialho, A. & Tejerina-Garro, F. L. 2004. Peixes do Rio Meia Ponte, GO. Série didática 12. Editora da
Universidade Católica de Goiás, Goiânia.
Richter, B. D.; Mathews, R.; Harrison, D. L. & Wigington, R. (2003) Ecologically sustainable water
management: managing river flows for ecological integrity. Ecological Applications, 13(1): 206–
224.
Tejerina-Garro, F. L. & Mérona, B. (2001) Gill net sampling standardisation in large rivers of French
Guiana (South America). Bulletin Français de la Pêche et de La Pisciculture, 357/360: 227–240.
Thioulouse, J.; Chessel, D.; Doledec, S. Oliver, J.M.; Goreaud, F. & Pelessier, R. (2001).
Ecological data analysis: exploratory and Euclidian in Environmental Sciences. Version 2001
©CNRS 1995 – 2001.

528
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Interaction between land use and fish assemblage in the River João Leite basin,
Goiás, Brazil

A.M. Dias1, F.L. Tejerina-Garro 1

1
Aquatic Center Biology, University Catholic of Goiás, Campus II, Av. Engler s/n, Mariliza
Garden, CEP 74650-010, Goiânia, Goiás, Brazil.
e-mail: [email protected]

This study aims to compare the fish assemblages of creeks located into the Altamiro of
Moura Pacheco State Park and those draining regions where dominated cattle raise and
agriculture activities (Novaes-Pinto, 1993). All creeks are tributaries of the João Leite River,
high Paraná River basin (GALINKIN, 2003).. Fish was sampled each two month between
August/2004 and July/2005 using electric fishing in twelve creeks in a stretch 50m. Each
creeks was grouped by and region similarity, a group considered preserved, a group in the
average region, a group in the intermediate region and a group in the inferior region. The
evaluation of the communities was given through analysis of correspondence of program AD-
4 where it presented differences enters the communities of fish in the different regions
(conserved, average, intermediate and inferior). Mount Carlo’s test identified significant
difference, p<0,005. The ecological describers (Abundance, Shannon, uniformity and
Richness) had been different for each region, reflecting in differences between the
communities of fish of the sub-basin of the River João Leite. The analysis shown abundance
of 10.838 individuals, being bigger in the average region, 4509 individuals collected and
lesser region in the intermediate region, 1687 individuals. The average of the wealth was
bigger in the medium region 22,58 species and minor in the inferior, 14,33 species. The
index of Shannon presented a average diversity of 3,13 in the medium region, and minor in
the preserved region 2,8 bits. The uniformity showed a minimum average of 0,69 in
preserved region and 0.81 in the medium region. The results show significant differences
between conserved areas and not conserved however the richness, the abundance, diversity
and the uniformity it is not bigger in the areas conserved. Moreover, one observed that it has
an influence of the position of the creeks in the groups not conserved (direction headboard-
estuary), however the ecological describers do not follow the standard predicted for the
concept of the continuous river (Vannot, et al.; 1980), or either, these do not increase in the
direction headboard-estuary, what it suggests an influence of the anthropics activities in the
inferior intermediate region and inferior region.

References
Galinkin, M. (2003). GeoGoiás 2002. M. Galinkin (ed). Agência Ambiental do Estado de Goiás,
Fundação CEBRAC, PNUMA, SEMARH. Brasília, 272 p.
Novaes-Pinto, M. (1993) Caracterização geomorfológica do Distrito Federal. 285-320. In NOVAES-
PINTO, M. (Org). Cerrado: caracterização, ocupação e perspectivas. 2ª ed., Editora UnB, Brasília.
Vannote, R. L.; Minshall, G.W.; Cummins, K. W; Sedell, J. R.; Cushing, C. E. (1980) The river
continnun concept. Can J. Fish Aquatic. Sci., 37: 130-137.

529
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Composition and spatial distribution of the fish assemblage in the tributaries of

the Ribeirão João Leite sub-basin, Goiás, Brazil

A.P. Fialho1, L.C. Gomes2, F.L. Tejerina-Garro1.

1
Universidade Católica de Goiás, Centro de Biologia Aquática, Campus II Av. Engler s/n,
Jardim Mariliza, Cx. Postal 86 CEP 74605-010 Goiânia, GO, Brazil.
e-mail: [email protected]
2
Universidade Estadual de Maringá, DBI/Nupélia. Av. Colombo, 5790, Maringá, PR.

The present study was conducted in the Ribeirão João Leite sub-basin, Center-West
Region, Brazil. It aims to describe the composition and spatial distribution of fish
assemblages in nine tributaries distributed along the mentioned sub-basin.
Five samples were conducted between August/2004 and September/2005. Fish were
collected using electrofishing along a 50 m stretch. A Detrended Correspondence Analysis
(DCA) was applied on transformed abundance data (square root) as suggested by Gauch Jr
(1982). Spatial distribution of species was tested using a two-way ANOVA.
The 5277 individuals collected are distributed in 49 species, 14 families, and 6 Orders.
Two axes from DCA with eigenvalues above 0.20 (Matthews, 1998) were retained for
interpretation (eigenvalues 0.35 and 0.24, respectively). The bifactorial ANOVA (position)
was significant (p< 0.05) demonstrating difference in the composition and space distribution
of fish assemblages along the sub-basin (upper section - 23.15% of abundance, middle
section - 48.95% and low section - 27.80%)
Although the small space scale of the sub-basin sampled, the results demonstrate a great
heterogeneity in fish assemblage composition among sections. This result can contributes to
planning conservation of this sub-basin, which is used for water supply of the metropolitan
area of Goiânia city.

References
Gauch Jr., H. G. (1982). Multivariate analysis in community ecology. Cambridge University, USA.
Matthews, W. J., (1998). Patterns in freshwater fish ecology. International Thomson Editors, USA.

530
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Reconstruction of paleoenvironment to understand involved processes on

landscape development; an important tool for long term research and
management at the Valdivian urban wetlands (Chile).

H. Rubilar1,3; B. Jessel1; G. Rojas2; C. Ramírez3

1
Geoecology Institute - Landschaftsplanung. Universität Potsdam. PF 601553; Potsdam. Germany.
2
Geoscience Institute - Universidad Austral de Chile. Casilla 567 Valdivia. Chile.
3
Botany Institute - Universidad Austral de Chile. Casilla 567 Valdivia. Chile.
[email protected]

Abstract

To know baseline conditions is a high need for nature management. Using basic
geological studies, historical information, the testimonies of people and the knowledge of
current vegetation situation entered into GIS, it was possible to create a picture of the
Valdivian urban wetlands (Chile) development, and quantify the land use change. The results
should be checked with future research, nevertheless is a concrete contribution as starting
point. The processes involved in the landscape evolution were elucidated, giving guidelines
for the wetlands management (Figure: 1).

Figure 1: Natural and anthropogenic events that led to the current Valdivian urban wetland
distribution. (*Hualve: wet-forest, Blepharocalyo-Myrceugenietum exsuccae).

Comments
The loss of wetlands is a fact and it will continue being a usual situation in Valdivia.
Considering the important functions and values of these wetlands, we propose conservation
management and soon deeper studies of the area. According with the social features of the
region, an integrative approach is crucial, which takes into account environmental and social
factors. Special emphasis should be given to educational and cognitive aspects of the
wetlands.

531
Theme 4: Ecohydrology, water and rivers
4.4 Posters

The application of an artificial wetland on the treatment of river water

D.F. Juang1, P.C. Chen2, Y.C. Lee2

1
Department of Healthcare Administration, Mei-Ho Institute of Technology, 24F, 230, Ming-
Chuan Second Road, Kaohsiung 806, Taiwan, R.O.C. e-mail: x2060@
email.meiho.edu.tw
2
DHV Planetek Co., LTD, 4F, 505, Chung Shan Second Road, Kaohsiung 801, Taiwan,
R.O.C.

Introduction

In this study, we selected a new constructed free water surface (FWS) wetland located at
the Chia Yi County of Taiwan and observed its treatment efficiency by periodically analyzing
the water quality in the influent and effluent. The influent of this wetland was pumped from
the Ho-Bou-Yu Drainage which is next to the wetland, and the treated water was then
discharged to the downstream of Drainage by gravity. Table 1 showed the characteristics of
this FWS wetland.

Table 1. Characteristics of FWS wetland.

Wetland Zone Flow Rate between 3672 CMD and 5352 CMD
Items Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Total
Water Depth (m) 1.2 0.6 1.2 1.2 1.2
Area (m2) 1800 1000 6300 4000 2400 15500
Effective Volume (m3) 2160 600 7560 4800 2880 18000
Hydraulic Retention
0.4-0.59 0.11-0.16 1.41-2.06 0.90-1.31 0.54-0.78 3.36-4.90
Time (day)
Covering Ratio of
70% 90% 20% 10% <5%
Macrophytes

Treatment Efficiency of FWS Wetland

The results of this study showed that the FWS wetland could effectively treat the
pollutants in the river water, and the relationships between the effluent concentration and the
loading rate of pollutants were illustrated in Fig. 1-2.

effluent
effluent
5.0 Linear Regression 12.0
Linear Regression
4.5 y = 0.7745x - 1.1186
10.0
4.0 R2 = 0.8453
3.5 8.0
BOD5(mg/L)
NH 4+ -N(mg/L)

3.0
y = 1.5785x - 0.1192 6.0
2.5
2.0 R2 = 0.7539 4.0
1.5
1.0 2.0
0.5 0.0
0.0 0.0 2.5 5.0 7.5 10.0 12.5 15.0
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50
Fig. 1. Relationship between
NH4+ -N Loading efflu. BOD
Rate (g/m2/day) Fig. 2. Relationship between efflu. NH4+-N
BOD Loading Rate (g/m2/day)

conc.and BOD loading rate. conc. and NH4+-N loading rate.

References
Jing, S.R., Lin, Y.F., Lee, D.Y., and Wang, T.W. (2002). Performance of constructed wetlands
planted with various macrophytes and using high hydraulic loading. Journal of Environmental Quality,
31(2), 690-696.
USEPA (2000). Constructed Wetlands Treatment of Municipal Wastewaters, EPA Manual,
EPA/625/R-99/010, USA.

532
Theme 4: Ecohydrology, water and rivers
4.4 Posters

A human-modified ecological hotspot: Jabbul salt lake (Syria)

F. Turkelboom 1, Z. Masri 1, A. Saeed 2, B. Kasmo 3

1
ICARDA, P.O. Box 5466, Aleppo, Syria.
e-mail: [email protected]
2
Chamber of Industry, Aleppo, Syria.
3
GOLD, Ministry of Irrigation, Aleppo, Syria.

Jabbul salt lake (or sebkha) is a hydrologically-closed agro-ecosystem in NW Syria, which

includes a seasonal lake (28,000 ha), seasonal shorelines and surrounding agricultural
villages. This place is rich in agricultural production, indigenous culture and biodiversity.
However, over the last decades, many changes have taken place, which threaten the
sustainability of both livelihoods and the environment. The objectives of this study were to
assess the changes in the environment and the institutional setting, and to design a way
forward for sustainable management of this valuable agro-ecosystem.
Sebkhat al-Jabbul experienced a lot of changes as a result of human interventions during the
last 50 years. Historical records and maps indicate that Sebkhat al-Jabbul was a seasonal
salt lake, which was fed mainly by Al-Dahab river, and which dried out for 90% during the hot
summer season. Until the 1960’s, the sabkha was surrounded by rainfed agriculture and
rangeland. Beside salt extraction, there was no major human interference and the only
institutions involved in the area were local village councils. During the 1970’s, water
extraction of the streams and overpumping from the groundwater around the lake resulted in
a decline of the water level in the lake. A major turning point for the eco-system was the
extension of the Euphrates irrigation scheme to the north and northeaster parts of the lake
during the 1980’s and 1990’s. It resulted in a population explosion and expansion of
intensive high-input irrigated agriculture. The inflow of drainage water from the irrigation
schemes caused a more permanent presence of water in the lake, while the building of
dykes resulted in several quasi-independent lakes with different salinity content. The new
lake conditions attracted an extremely rich birdlife. Among them are the greater flamingo
and the global threatened white-headed duck (Serra et al., 2006). On the other hand,
growing rural towns, new factories and increased use of agricultural inputs resulted in a fast
increase of inflow of polluted sewage water. For many villages around the lake, salt
collection was and remains an important source of income, despite the pollution of the salt.
Other side effects of this changing situation are the increasing water table, secondary
salinisation of some agricultural land, decline of the halophyte vegetation along the shoreline,
illegal hunting and unregulated fishing.
The institutional landscape has also changed dramatically over the last 3 decades. The
major stakeholders now include 5 Ministries, 12 government institutions, 5 research centers,
one NGO, the private sector and farmer organisations. A major challenge is that government
agencies have overlapping and unclear mandates related to Sebkhat al-Jabbul. This
resulted in a lack of common-agreed vision and integrated master plan, in single-disciplinary
interventions, lack of coordination, legal vacuum and scattered knowledge of the lake. The
key to get out of this impasse is to start-up a multi-stakeholder process (MSP), which
provides leadership for envisioning and planning, while accommodating the diverse
perceptions and aspirations of all involved, and taking into account the environmental
limitations and opportunities of the ecosystem. This MSP approach is expected to lead to
conservation of the rich biodiversity, while ensuring local livelihoods and strengthening the
local environmental governance institutions.
Reference
Serra G., Murdoch D., Turkelboom F., Travert F., Mujawer Y. and Scott D., 2006. Sabkhat al-
Jabbul, a threatened Ramsar wetland, Syria. Sandgrouse, 28(2): 127-141.

533
Theme 4: Ecohydrology, water and rivers
4.4 Posters

Influence of groundwater system on landscape ecology: a case study in Beijing,

China

P. Yun1, N. Nakagoshi 1, G. Huili 2,3, X. Zaiping1

1
Graduate School for International Development and Cooperation, Hiroshima University,
Higashi-Hiroshima, 739-8529, Japan.
e-mail: [email protected]
2
Key Lab of 3-D Information Acquisition and Application, MOE, Beijing,100037,China.
3
Key Lab of Resource Environment and GIS of Beijing, Capital Normal University, Beijing,
100037, China.

Introduction

Since groundwater has been greatly exploited as the primary water resource in Beijing,
the influence of groundwater system on landscape ecology must be well understood to
achieve a well urban ecosystem. GIS(Geographic Information System), RS(Remote
Sensing)and some other space technologies now provide us with advanced tools to analyze
the influence of groundwater on landscape ecology.

Methods

This study chose the plain of Beijing as the study area, shallow groundwater table and
remote sensing images of this area in the time series from year 1990 to 2004 were collected
as the basic data source. Firstly, water table distribution was interpolated, and then its spatial
variation was analyzed with geostatistics methods. Then, NDVI (Normalized Difference
Vegetation Index) and some landscape metrics were calculated from processed TM images,
outputted to raster layers, and analyzed on heterogeneity with Kriging semi-variogram. After
this, overlay, gradient analysis, and visual comparison were implemented to find the spatial
and temporal relationship between groundwater and landscape ecology.

Results and discussion

Based on the spatial and temporal analysis, some relationship between groundwater and
landscape ecology was established by this study for the first time. One is about spatial
heterogeneity. Shallow groundwater table and landscape has the similar variation in
orientations. Another is about temporal dynamics. By comparing the value of variogram
formulation parameters, which are nugget, sill, and range of spatial dependence, it can be
concluded that spatial structure of landscape metrics and NDVI changes following the
groundwater table dynamics. This kind of spatial and temporal relationship is considered to
be the result of interaction between groundwater and vegetation.

References
Bernaldez F.G., Rey Benayas J.M., Levassor C., etc.(1989). Landscape ecology of uncultivated
lowlands in central Spain. Landscape Ecology 3:3-18.
Caruso B.S.(1989). Temporal and spatial patterns of extreme low flows and effects on stream
ecosystems in Otago, New Zealand. Journal of Hydrology 257: 115-133.

534
Theme 5: Monitoring and classification

Theme 5: Monitoring at the Landscape scale

535
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

5.1 Symposium 10: Monitoring at the landscape scale

Monitoring patterns of land cover change in cultural Irish landscapes using

landscape metrics and SPOT 5 imagery

S.J.P. McKenzie, A. Cooper, T. McCann

School of Environmental Sciences, University of Ulster, Coleraine, Northern Ireland.

e-mail: [email protected]

Introduction

Early approaches in environmental Remote Sensing (RS) concentrated on low-resolution

satellites that cannot satisfactorily detect fine-level variations inherent in cultural landscapes.
Research focussed upon spectral responses rather than considering the role of spatial
characteristics. Calibration between satellite image models and land cover was largely in the
form of sparse ground truth samples. Our research investigates landscape metrics, RS and
Geographic Information Systems (GIS), calibrated against field survey data from sample grid
squares of the Northern Ireland Countryside Survey (Cooper et al., 2003). This was a project
which mapped and estimated changes in land cover in Northern Ireland over a ten year
period between 1988 and 1998. Our current research considers how effectively landscape
metrics add value to interpreting land cover change.

Methods

A baseline and resurvey sample of 35 quarter kilometre grid squares in a lowland

agricultural landscape of County Down was selected for study from the Northern Ireland
Countryside Survey (NICS). Land cover field maps from each survey date (c.1988 and 1998)
were digitised then rasterized using ArcScan software. Landscape metrics were generated
using Fragstats software to assess the patterns of change in the shape of mapped land
cover. SPOT 5 satellite imagery, provided through the OASIS scheme, was used to create
pan-sharpened images of the sample grid squares. Imagery was available from July 2004
and October 2005 in 2.5m panchromatic and 10m multispectral modes. Each scale was
classified using a maximum likelihood classifier. Each image was divided into four main land
cover types; agricultural grassland and crops, semi-natural vegetation, woodland and hard
surfaces. Data was also available from the 2001 Landsat sensor. Aerial photography from
2004 was digitised to serve as a check on values generated from the satellite imagery.

Results

The landscape metrics applied included area, contagion, patch-per-unit area (PPU),
number of patches, total edge (Zeng Wu, 2005), percentage land area (PLAND) and
weighted mean patch fractal dimension. Contagion values suggested a highly aggregated
landscape. When considered alongside metrics such as PLAND and absolute area they
showed that agricultural grassland and crops predominated. Contagion values did not
change significantly despite changes in the spatial configuration of land cover parcels. The
PPU statistic suggested by Frohn (1998) was much more sensitive to changes in landscape
structure showing that as the number of patches increased in a landscape, the level of
fragmentation also increased. The Euclidean Nearest Neighbour (ENN) statistic was
consistently low for agricultural land cover while the value was much higher for semi-natural
and woodland land cover. Our research found that different types of land cover have unique
spatial signatures. Agricultural land cover is characterised by high contagion and PLAND
values while ENN and mean fractal dimension values are low. Semi-natural vegetation is
characterised by high contagion values but low PLAND values while ENN and mean fractal

537
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

dimension values are generally much higher. Patch-level statistics such as mean shape
index and fractal dimension were able to distinguish between agricultural and non-
agricultural land cover patches.
Spectral similarities between land cover parcels meant that it was difficult to maintain the
separability of field-mapped land cover recorded by the NICS. However, there was a high
level of correlation between SPOT imagery, aerial photography and NICS data.

Discussion

When applied to NICS field data, landscape metrics significantly aid the interpretation of
land cover change. Interesting trends are apparent when applying metrics such as ENN,
PPU and total edge. Investigation of metrics from satellite imagery suggests a high level of
correlation between woodland, agricultural land cover and hard surfaces in terms of area.
Other metrics such as contagion are less reliable as a number of misclassified pixels pepper
the landscape. The level of correspondence between NICS data and classified satellite
imagery was relatively high despite differences in typology and dates.

References
Cooper, A; McCann, T. & Meharg, M. J. (2003) Sampling Broad Habitat change to assess
biodiversity conservation action in Northern Ireland. Journal of Environmental Management 67:
283-290.
Frohn, R.C. (1998) Remote Sensing for Landscape Ecology. New metric indicators for monitoring,
modelling, and assessment of ecosystems. Lewis Publishers, Boca Raton, Florida.
Zeng, H. & Wu, X.B. (2005) Utilities of edge-based metrics for studying landscape fragmentation.
Computers, Environment and Urban Systems 29: 159-178.

538
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

The NILS programme, monitoring the Swedish landscapes for biodiversity

assessment

A. Allard,¹, P-A, Esseen¹, A. Glimskär², G. Ståhl¹, S. Sundquist¹, O. Inghe³

¹ Department of Forest Resource Management, SLU,SE-901 86 Umeå, Sweden

e-mail: [email protected]
² Department of Conservation Biology, SLU, SE-750 07 Uppsala, Sweden
³ Swedish Environmental Protection Agency, SE-106 48 Stockholm, Sweden

Introduction

The National Inventory of Landscapes in Sweden, (NILS) is a programme for monitoring

the Swedish landscapes and it has now run for some years. It was created to meet the
increasing demands for information on national resources and environmental conditions. An
example of such a demand is a set of 16 Environmental Quality Goals that the Swedish
Environmental Protection Agency has decided upon as guidelines, and information is needed
to assess if these objectives are being met. The awareness of environmental stresses has
increased globally and international collaboration is needed in order to mitigate effects are on
the rise. There are many and varied end-users of information and this places special
demands on the type of data captured, both in terms of scales and variables.

Design
The NILS programme has chosen to capture data without pre-classification and at several
scales, thus to be able to conform to, and constitute a platform for as many other
programmes as possible. One inventory is made in the field and a parallel inventory is made
using interpretation of colour infrared air photos, both using quantitative variables in a
context-dependent flow that captures processes, habitats, structures as well as species, see
Table 1. Data from the two inventories are then used for a two-phase estimation of conditions
and quantities. The design is derived from a grid of landscape squares, statistically sound,
which covers representative land and lake areas in the entire nation. Examples are given in
Figure 1.

Collaborations and projects for development

NILS is pursuing many types of projects, from internal development projects to
commissions and collaborations with research specialists and other environmental
monitoring programs. Examples are to gather information about small biotopes and their
management in rural landscapes for the follow-up made by the Swedish Board of Agriculture.
Another example is the construction of inventories of the cultural heritage of the country.
Collaboration is also being carried out to asses Fungi and the spread of disease from vector
animals in an European perspective.

Always looking forward

A variety of techniques are being examined in the NILS programme involving end-users,
mainly at the Remote Sensing Laboratory at SLU, for example to speed up the time used in
the inventory based on aerial photography. Examples are to use laser scanned data for
image matching to extract tree height and possibly get information of the floor of dense
forests and thereby be able to obtain information on hydrology. Another example is automatic
tree recognition, which seems promising for the northern half of Sweden, where there are
few tree species.

Contact
More information is found at: http://nils.slu.se/

Table 1. The variable content in NILS (DPSIR-compatible).

539
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Processes (Pressure) Structures (State)

Ground disturbance Vegetation structure
Hydrological changes Dead wood and canopy structure
Grazing and mowing Hydromorphological mire structures
Forestry Linear and point features
Climate changes and air pollution Soil properties

Habitats (State) Species (Impact)

Forest Vegetation-forming plants
Wetlands and shores Epiphytes
Grassland and heath Grasslands indicators (e.g. grazing impact)
Cliffs, rocks and other bare substrates Game (droppings, etc.)
Man-made habitats (parks, etc.)

Figure 1. Examples of
Swedish habitat types
monitored by the NILS
programme.

References
Esseen, P-A; Glimskär, A; Ståhl, G. & Sundquist, S., (2006) Fältinstruktion för nationell inventering
av landskapet i Sverige, NILS, (in Swedish), Department of Forest Resource Management,
Swedish University of Agriculture, Umeå
Allard, A; Nilsson, B; Pramborg, K; Ståhl, G. & Sundquist, S., (2003) Manual for aerial photo
interpretation in the National Inventory of Landscapes in Sweden, Department of Forest Resource
Management, Swedish University of Agriculture, Umeå
Allard, A; Marklund, L; Glimskär, A. & Högström, M., (2006) Utveckling av nationellt
uppföljningssystem för småbiotoper vid åkermark, (in Swedish), Report 158, Department of Forest
Resource Management, Swedish University of Agriculture, Umeå
Glimskär, A. (2005) Indikatorsystem för småbiotoper – metodutveckling för nationell övervakning av
biologisk mångfald. (in Swedish) Swedish Board of Agriculture, Report 2005:7, Jönköping,
Sweden
Glimskär, A; Allard, A. & Högström, M., (2005) Småbiotoper vid åkermark -indikatorer och
flygbildsbaserad uppföljning i NILS, (in Swedish), Report 134, Department of Forest Resource
Management, Swedish University of Agriculture, Umeå
The National Heritage Board, (2004) Karaktärsdrag och bebyggelsemönster - Modell för karakterisering och
uppföljning inom miljömålsarbetet, (in Swedish), Report 2004:8, Stockholm, Sweden.

540
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Small-scale extraction of non-timber forest products in developing countries – a

framework for ecological monitoring

A. Fröde

German Development Service (DED), Southern Alliance for Indigenous Resources

(SAFIRE), Research and Development Section, P.O. Box 3362, Belvedere, Harare,
Zimbabwe,
e-mail: [email protected]

Context

Small-scale commercialisation of Non-Timber Forest Products (NTFP) is one of the most

important livelihood strategies in many woodland areas in developing countries. It was often
claimed to be environmentally sustainable even without regulation (e. g. Wong et al., 2001).
However, significant long-term decreases in ecosystem resilience, biodiversity and functional
ability of the ecosystems were detected over time in numerous small-scale NTFP extraction
systems. In case-specific studies, these proved to be hard to be predicted due to the
complexity of woodland dynamics and human impacts (e. g. Sunderland et al., 2004). In
order to minimize destructive effects, ecological monitoring, i. e. the continuous collection,
analysis and interpretation of data related to the environment in order to understand,
determine or detect ecological trends in an area (Walsch, 2000), is of crucial importance in
extraction systems. This article draws upon findings in the implementation of ecological
monitoring in small-scale NTFP extraction projects led by SAFIRE (Southern Alliance for
Indigenous Resources), a Zimbabwe-based non-governmental organisation.

Objectives and methodology

Against the mentioned background, an ecological monitoring framework had to be

developed, which had to be easy and cost-efficient to implement and needed to provide the
necessary results for decision support and NRM steering. It had to ensure compliance of the
NTFP-extraction activities to ecological criteria, be based on the participation of the collectors
and response to the challenging political, social and economical context in Zimbabwe.
The framework development was based on a report analysis, semi-structured interviews
with stakeholders from the project context and experiences in three pilot implementations.

The monitoring framework

For all NTFP extraction activities, an expert-led ecological baseline survey is carried out
at the project start. On this base, NRM plans are developed with the local users in a
participatory way featuring on ecological risks and possibilities for mitigation (Sola, 2005).
For the ecological monitoring, tentative indicators are finalised by a participatory
methodology into a monitoring plan, which schedules for 6-monthly assessments of critical
ecosystem features by key informant interviews and expert surveys. These also include
Permanent Sampling. Evaluation of the monitoring results and determination of action
options are carried out in community meetings and in staff workshops.

Outputs and Outcomes

An example of indicators developed and ecological risks addressed are provided in

Table 1 for the mountainous woodland area of Nyanga. There, leaves from Fadogia
ancylantha (Makoni Tea Bush) and Myrothamnus flabellifolius (Resurrection Tea Bush) are
collected by small-scale entrepreneurs for processing and regional sale.

541
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Table 1. Extract from Nyanga Ecological Monitoring Plan (SAFIRE 2006)

Indicator Specifications Ecological challenges

addressed
Early/late fires per Number of fires per village, Uncontrolled fires
dry season woodland/grassland area burned
Erosion gullies Number of gullies wider than 1.5 m Erosion
and/or longer than 3 m
Alien plant Number of Eucalyptus sp, Acacia Alien plants
populations marnsii and Lantana camara
individuals in sampling plots
Number of Fadogia Average estimated distance, time Target species
plots walked development
Area under Area under tobacco cultivation per Wood off-take
Tobacco village
Trade in Uapaca Amount of Uapaca fruits traded by Uapaca population
kirkiana fruits collectors, distance development

Significant effects from changes in land-use practices were already detected in the pilot
implementation. Regulatory or corrective means are being developed accordingly. Amongst
others, the considerable increase of the number and extent of erosion gullies over the last
few years is being addressed by the local communities by reclamation and contour digging.
Facts on the sharp increase of wood off-take caused by extended small-scale tobacco
farming (which requires firewood for tobacco drying) calls for action by local authorities.

Conclusions

The process and the product are indicative of dealing with the challenges faced by
applied landscape research and ecological monitoring in NTFP-related initiatives in
developing countries: the lack of reliable baseline data, the need to define and use highly
aggregated core indicators, limited financial, technical and human resources and difficulties
in deducting evidence for long-term trends from locally and temporally limited natural
phenomena. However, the pilot implementation has underlined the sustainability and
effectiveness of the approach and calls for transfer into other projects and programmes.

References
SAFIRE (2006) Ecological Monitoring Plan for the Makoni Tea Extraction Project in Nyanga. SAFIRE,
Harare. Internal document.
Sola, P. (2005) The Community Resource Management Plan: A tool for integrating IKS into natural
resource management. Ethnobotany Research and Applications 3: 143-153.
Sunderland, T.C.H., Harrision S.T. 7 Ndoye, O. (2004) Commercialisation of non-timber forest
products in Africa: History, context and prospects. Sunderland, T. & Ndoye, O. (Eds). Forest
products, Livelihoods and Conservation: Case Studies of Non-Timber Forest Product Systems:
Volume 2 – Africa. CIFOR, Bogor, pp. 1-24.
Walsch, A. (2000) Participatory Environmental Monitoring: A Facilitators Manual. German Foundation
for International Development, Bonn.
Wong, J.L.G, Thornber, K. & Baker, N. (2001) Resource Assessment of non-wood forest products:
Experience and biometric principles. FAO, Rome.

542
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Monitoring landscape changes in the Netherlands

A.J.M. Koomen1, W. Nieuwenhuizen1, D.J. Brus2, L.J. Keunen1, G.J. Maas1, T.N.M.
van der Maat1, T.J. Weijschede1.
1
Wageningen University and Research Centre, Alterra, Landscape Centre, P.O. Box 47,
6700 AA, Wageningen, Netherlands.
E-mail: [email protected]
2
Wageningen University and Research Centre, Alterra, Soil Centre, P.O. Box 47, 6700 AA,
Wageningen, Netherlands.

Introduction

Monitoring landscape changes in the present-day landscape are generally approached

from the perspective of biodiversity related to climate change. Another important approach
for monitoring landscape is from a perspective of landuse, geomorphology and historical
geography. These characteristics of the landscape function as a base for biodiversity and are
therefore of equal importance. The geomorphological and historical geographical values are
recognized as important by the Dutch government as most of the recent spatial
developments have a deteriorating effect. Therefore a monitor system has been developed
primarily aiming to describe changes in relation to these landscape qualities.

Method

A new monitor (Steekproef Landschap) for the Netherlands has been developed aiming to
describe recent changes over the period 1996-2003 in landscape using probability sampling
in 72 study areas with an extent of 1 square kilometre, using the characteristic Dutch
landscape types as strata. In addition to this a map displaying ‘spatial pressure’ for future
claims was used to further stratify the study areas so that the majority would be situated in
areas with a higher ‘spatial pressure’. The year 1996 was used as a reference for
establishing changes. Within the study remote sensing (aerial photographs, digital elevation
data) in combination with detailed topographic maps and field investigation was used to
describe changes in land use, patterns of cultural history (such as old roads, waterways,
green lineair elements, land reclamation patterns) and geomorphology. Assessing data from
the field was expected to give valuable and additional information next to the use of digital
information and aerial photographs. All data has been statistically processed taking into
account the stratification used.

Results

Results on the development of the Dutch landscape have been obtained for the period
1996-2003. Due to ground works related to urbanisation, agricultural adaptations and nature
development (table 1) significant areas have lost their geomorphological identity. This is also
the case for the features that reflect the historical development of landscapes (figure 1).
Urbanisation obviously has a large influence on the land use that reflects a growing urban
environment including additional changes linked to the transformation from rural to semi-
urban in rather large areas surrounding the urban zones.
Table 1 shows that 1.5% of the rural landscape has been transformed into urban areas. The
numbers are lower for agriculture and nature development but the combined effect makes
clear that more than 2.4% of the rural landscape has been changed in a 7 year period. The
effects these spatial developments have on the the geomorphological and historical
geographical values are significant. The small scale relief elements and lineair green and
blue historical elements rapidly disappear. This means for instance that small scale relief
elements will have been completely erased within the next 40 years.

543
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Table 1. Major changes in land use for the Netherlands 1996-2003.

Type of land use Change in ha/100 ha Total in ha

(extrapolation)
Urbanisation 1.5 46.064
Agriculture 0.6 17.594
Nature development 0.3 8.957

Discussion

Method and results trigger a discussion that narrows down to two topics. The first is the
number of study areas in relation to the level at which conclusions can be drawn. The
number of 72 study areas allows significant conclusions to be drawn on the level of the
Netherlands but not or only partly for the different landscape types that have been used to
stratify. How many study areas does one need to incorporate to enable conclusions of the
level of landscape types? This is dependent on the specific indicator and the quantity of
change it shows.
The second topic for discussion is the use of field work. Is it really necessary (high costs) and
does it contribute to the final results? Analysis shows that without field work about 40% of the
small scale changes that influence geomorphological and historical geographical values will
be missed when complete relying on digital land use maps. This primarily concerns small
scale ground works (levelling) related to agricultural land use.

Conclusions

Four main conclusions can be drawn on basis of the result of this monitoring study:
1 – The rate of changes in land use in the Netherlands between 1996 and 2003 is high and
dominated by urbanisation, agriculture and nature development;
2 – The changes in land use lead to significant loss of geomorphological and historical
geographical values
3 – Fieldwork is essential for a good assessment as national digital data generally do not
signal local and small scale changes.
4 – To enable conclusions for landscape type the number of study areas should be increased
to allow significant conclusions.

References
D. J. Brus, W. Nieuwenhuizen and A. Koomen, 2006. Can we gain precision by sampling with
probabilities proportional to size in surveying recent landscape changes in the Netherlands?
Environmental Monitoring and Assessment (2006) 122: 153–169.
Koomen, A.J.M., W. Nieuwenhuizen, D.J. Brus, L.J. Keunen, G.J. Maas, T.N.M. van der Maat,
T.J. Weijschedé, 2004. Steekproef landschap; actuele veranderingen in het Nederlandse
landschap. Alterra report 1049, Wageningen, the Netherlands

544
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Why is strategic conservation monitoring so rare in Europe?

R.G.H. Bunce, R.H.G. Jongman

Alterra, PO Box 47, 6700 AA Wageningen, The Netherlands.

e-mail:[email protected]

Introduction

Landscape ecology has still not made a significant impact on policy makers and planners in
Europe. who are concerned with strategic conservation planning.. In contrast, many papers
in this congress provide worked examples of the successful implementation of landscape
ecological principles into general planning. The present paper examines some of the reasons
for this discrepancy.
The main reason is that many conservation agencies do not appreciate the need for
objective figures on change nor the need to sample landscape complexity. Conservation
managers are also not familiar with the methods of sampling, statistical procedures and
analysis developed by landscape ecologists in recent years. Conservation priorities have
been determined mainly by expert committees –for example the categories of the Habitats
Directive ( Council Directive 92/43/EEC ) involve such diverse criteria that they are difficult
to record consistently in the field.

European databases

The European databases that are available do not have the level of detail required to cover
the necessary information on biodiversity eg habitats and vegetation. They do however have
potential for converting in situ information into synoptic coverage .For example the CORINE
Land Cover Map has complete coverage of the EU but the basic unit of 25ha inevitably
misses any changes which involve smaller units as shown by Levin (2006).Other biodiversity
databases do not contain sufficient detail for monitoring eg CORINE biotopes covers only
the presence of categories and varies according to national definitions eg France has e few
large sites and Italy many small ones.
Another difficulty is that there are no common standards between countries because member
states have pursued there own agenda. However, recent work on habitats has shown that
general habitat categories Bunce et al (2005) can be used to incorporate extant data, (
Bloch-Petersen et al (2006) ) indicating that, if sufficient financial resources were made
available, then European figures could now be produced. It is now technically possible to link
in situ data with satellite imagery, so that if all the sources of information were coordinated
and modern high resolution satellites were used, then major advances could be made.
However, such methods are unlikely to have sufficient replicates to adequately estimate
changes in the small area of land under protection.

Natura 2000

The main European initiative specifically related to nature conservation is the Natura 2000
series of sites (http://ec.europa.eu.environment/nature/home.htm) and it is obligatory for
member states to undertake monitoring in these sites to ensure that Favorable Conservation
Status (FCS) is maintained. Whilst the requirements of the EU are laid out eg setting
priorities for further monitoring, these have been interpreted in different ways by the member
states. Local site managers are also directly involved with their individual conditions and
have no reason to be concerned about transfer of any data collected to a European
database. There is therefore not only the problem of compatibility of data but also any
European monitoring scheme will need to deal with dispersed sites covering under 10% of
the land and often in unique situations. Such sites should not be considered in isolation and
it is not only necessary to set them in context, but also to have controls against which to

545
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

measure the efficiency of the protection measures. This has been recognized in the
assessment of the efficiency of European agri-environmental schemes eg in the UK and
Austria, where farms inside and outside schemes were compared.
The ECOLAND Forum, a IALE working group on countryside and landscape monitoring in
Europe, has had two meetings to discuss monitoring in Natura 2000 sites because of their
importance. At the first meeting the general characteristics of the series were discussed.
The sites vary in size from hundreds of square kilometers to a few hectares and, whilst
some contain only semi-natural vegetation, others have a wide range of habitats including
urban and industrial developments. The reasons for designation vary from individual species
to extensive habitats and the information on their distribution and character is variable. There
are therefore both scientific and practical issues involved. In developing monitoring
protocols. The second meeting was held in the field to examine actual rather than
theoretical problems of monitoring a large site (about 250 square kilometers) in south-east
Spain. Initial field visits showed that unless local knowledge was available, it was not
possible to assign habitats to those described in The Habitats Directive and which are
needed for determining the FCS of the site. Many of those habitats had a very restricted
distribution in the site. Mapping of habitats was tested in stratified 1 kilometre squares and,
with local experts present, it was possible to identify and map widespread key habitats. It
was concluded that a range of stratifications would need to be applied for scarce habitats or
a separate procedure involving local knowledge could be used to identify very rare point
features and a separate monitoring protocol would be then be needed for objective
monitoring -eg random vegetation plots. It was also decided that an expert system
formalizing local experience was required enable the information from the Habitats Directive
to be converted to a repeatable system that could be transferred between sites and
countries. The first stage of such a system could use the General Habitat Categories of
Bunce et al (2005) as a framework, followed by site data, geographical location, species
information from the Habitats Directive and finally where necessary, expert knowledge.
Examples of such an expert system have been prepared.

Conclusion

Europe therefore has no coordinated programme for monitoring the wider countryside or
key conservation sites. However, landscape ecologists in several countries have now
developed sufficient experience to design appropriate protocols to fill this gap.

References
Bloch- Petersen M., Brandt J., & Olsen M. (2006). Integration of European habitat monitoring based
on plant life form composition as an indicator of environmental change and biodiversity.
Geog.Tids.106. p61-74
Bunce,R.G.H., Groom, G.B., Jongman, R.H.G. and Padoa Schioppa, E. (2005) Handbook for
Surveillance and Monitoring of European Habitats .Alterra-Rapport 1219.Wageningen.The
Netherlands
Levin,G., (2006). Farm size and landscape composition in relation to landscape change in Denmark.
Geog.Tids. 106 p 45-60.

546
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Indicators to Monitor changes and conservation in Natura 2000 sites: A focus on

drivers, pressures and states

S. Luque 1, J. Pino 2, C. Basnou 2, R. Swetnam 3, C.A Mücher4, G. Hazeu 4, L. Halada

5
, F. Gerard3
1
Cemagref, Groupement de Grenoble, UR AMM, 2, rue de la papeterie, BP 76, F-38402,
St-Martin-d'Hères, cedex, France,
e-mail: [email protected]
2
Center for Ecological Research and Forestry Applications, CREAF, Universitat Autonoma
de Barcelona, E-08193, Bellaterra, SPAIN, 3.
3
Centre for Ecology and Hydrology, CEH, Monks Wood, Abbots Ripton, PE28 2LS,
Huntingdon, Cambridgeshire, United Kingdom
Alterra, Centre for Geo-Information, PO Box 6780 AA, Wageningen, The Netherlands
5
Institute of Landscape Ecology of the Slovak Academy of Sciences, ILE SAS, Stefanikova
3, PO Box 254, 814 99 Bratislava, Republic Of Slovakia

Introduction

In 2004, the European community adopted a framework within which indicators could
be identified to facilitate the assessment of progress towards the 2010 target for biodiversity
and communication of this evaluation (EEA-ECNC 2004). A few indicators were considered
ready for immediate use. However it was recognized that new indicators were required to
enable the assessment of progress towards the agreed goals and targets, taking into
account indicators that were developed through other processes for regional (e.g. Pan-
Europe and European Union) and national use. To contribute to this framework, in this work
changes were examined over the last 50 years in selected transects over Europe.
The research was carried out within the framework of the BIOPRESS (EC-FP5
project to support GMES ‘Global Monitoring for Environment and Security’) project and
AlterNET (A Long-term Biodiversity, Ecosystem and Awareness Research Network – EU-
FPVI).
The paper focus on the discussion of the results obtained from analysing historic
Land Cover Changes (LCC) in relation to patterns and processes for selected Natura 2000
sites. We quantified pressures on biodiversity (urbanisation, arable intensification,
afforestation, deforestation, abandonment and drainage) associated with land cover changes
within the framework of DPSIR (DPSIR - Driving forces, Pressures, States, Impacts and
Responses model). In this way, in order to understand the driving forces and patterns behind
the observed changes, we look at the consequences of the observed LCC on biodiversity for
selected transects across Europe. Results show the performance of landscape level
indicators to understand pressures at the biodiversity level. Particular emphasis was placed
on to the analysis of patterns derived from different pressures in order to understand the
relationship between observable patterns and driving forces causing changes that occurred
during the study period (1950’s – 2000).

Methods

The methodology, utilises archived historic and recent aerial photographs (a data
source that has remained consistent over the last 50-60 years) to assess land cover change
around Natura 2000 sites within 15 x 2 km transects from partner countries (United Kingdom,
Spain, Finland, The Netherlands and Slovakia) (See Gerard et al. 2006).
The data consisted of 40 transects from five European countries. Landscape metrics
were calculated for each of the transects using fragstats. From the 53 landscape metrics
calculated only seven were analysed after performing a selection using Pearson’s correlation
coefficient (r > 0.75). The total variance explained was calculated with factorial analysis.

547
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Three input data sets corresponding to the years 1950’s, 2000 and changes were used for
the analysis after eliminating the correlated variables.

Results

Canonical Variate Analysis (CVA), better known as Fisher linear discriminant analysis
(Braak & Smilauer 2002) provided a good separation between observations from 1950’s and
2000 (p<0.002). Changes in the European landscape from the 50’s are important mostly in
terms of pressures from urbanisation and agricultural intensification. The amount of land
cover changes at CORINE level one (e.g. from agriculture to urban) increased with the
distance outside Natura 2000 sites and decreases with the distance inside Natura 2000 sites.
Furthermore, the types of land cover changes are very different inside from outside the
Natura 2000 sites. In all, protection measures within Natura 2000 sites seems to have a
positive influence on the type of land cover changes and their acreage. But at the same time
it must be noted that conservation sites are becoming more isolated. We should therefore be
concerned about changes taking place outside protected areas to maintain a sustainable
habitat in terms of biodiversity quality towards the CBD 2010.

In the overall the landscape matrices studied for Europe showed an important division
within the study period. In particular, Finland, Spain and the Netherlands are in the group
with a very strong association with the Splitting Index (SPLIT), showing a landscape
configuration with an important level of division. This loss of continuity is also evident for the
United Kingdom and Slovakia in a second group that is associated with the increase on edge
density. On the other hand, the previous configuration in the 1950’s represented a more
complex and connected matrix in particular for Spain and Finland.

The methods used are transferable and applicable to a wide range of landscape
studies. The demonstrated methodology could be applied to general monitoring of landscape
change or to more localized areas such as the landscape surrounding environmentally
protected sites, as showed for Natura 2000 in Europe. Historic aerial photographs, once
digitized, could provide an important tool for monitoring when combined with landscape
metrics.

References
ECNC, EEA, UNEP, Council of Europe (2004) Joint meeting on development of plan and guidelines
for indicators and monitoring to achieve the 2010 target for biodiversity in Europe. EEA in
Copenhagen 21-23 April 2004. 26pp
Gerard, F.; Gregor, M.; Luque, S.; Huitu H.; Köhler, R.; Olschofsky, k.; Hazeu, G.; Mücher C.A.;
Halada L.; Bugár G; and Pino J. (2006) Land Cover Change in Europe from the 1950’ies to
2000. Aerial Photo Interpretation And Derived Statistics From 59 Samples Distributed Across
Europe. Edited by Raul Köhler, Konstantin Olschofsky. University of Hamburg, World Forestry
Institute, Germany. 364 pp.
Ter Braak, C.J.F. & Smilauer, P. (2002) CANOCO Reference manual and CanoDraw for Windows
User's guide: software for Canonical Community Ordination (version 4.5). Microcomputer Power
(Ithaca, NY, USA), 500pp

548
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Towards a standardized biodiversity assessment approach in topographically

complex landscapes

G. Hofer1, F. Herzog1, H.H. Wagner2, R.G.H. Bunce3, P.J. Edwards4

1
ART Agroscope Reckenholz, Swiss Federal Research Station for Agroecology and
Agriculture, 8046 Zürich, Switzerland,
e-mail: [email protected]
2
WSL, Swiss Federal Research Station for Forest, Snow and Landscape, 8903
Birmensdorf, Switzerland
3
Alterra, P.O. Box 47, 6700 AA Wageningen, The Netherlands
4
Geobotanical Institute, Swiss Federal Institute of Technology, ETH, 8092 Zürich,
Switzerland

The high plant diversity of mountainous farmland is increasingly threatened by land use
change (MacDonald et al. 2000), and sensitive biodiversity indicators and efficient sampling
designs are needed to monitor these areas. Although measures of habitat heterogeneity are
often used as indicators for species diversity in agricultural landscapes (Benton et al. 2003)
these usually assume a patch-mosaic model of the landscape, with homogenous patches
and clear borders between patches. Mountain areas, however, exhibit gradual shifts within
and between land units and so do not fit the patch-mosaic model well (McGarigal &
Cushman 2005). To estimate species diversity in mountain areas, more appropriate
heterogeneity indicators are needed which allow for gradual shifts and heterogeneous
patches.

Topography is known to affect the variability of abiotic microsite conditions (Swanson et

al.1988), which in turn affect vegetation and landuse patterns. In this study, we investigated
effects of topographic variables from a digital elevation model on the variability of plant
species richness and composition within and between local landscapes to develop a
standardized and generally applicable sampling approach for topographically complex areas.

Within a biogeographically and climatically homogenous dairy farming area of 250 km2 in
the Swiss pre-Alps, 12 local landscapes of 1 km2 were selected and plant species of 40
micro sites were recorded within each. Topography was integrated into the sampling design
at two levels: 1) at landscape level to sample local landscapes randomly along a gradient of
topographic variability within the study area, and 2) within coarse habitat types to maximize
the environmental range among plant species relevées within each local landscape. Linear
regressions were used to analyse the effect of topographic variability of local landscapes on
alpha and beta components of species richness at landscape, habitat and microsite levels.
Multivariate analyses were used to analyse species composition at landscape and microsite
level.

All components of species richness increased with topographic variability of local

landscapes at all levels of analysis (Figure 1). The main shifts in species composition at
landscape level correlated with topographic variability (0.82) and also with mean nitrogen (-
0.91), humidity (0.66) and pH (-0.66) indicator values of the plant communities. In the model
to explain species composition at the microsite level, including topographic variables as well
as habitat types increased the variance explained by 54 percent.

These results indicate that topographic variability is appropriate to estimate potential

relative differences in species richness at landscape level. To assess biotic heterogeneity the
ecological range can be efficiently covered by maximizing the gradient of topographic
variability between samples of local landscapes and between species relevées within local
landscapes. As this approach is based on maximizing relative differences between samples

549
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

within defined units, it is independent of expert knowledge, which is important for an

approach of general applicability.

Figure 1. Effects of topographic variability of the landscape on α and β species richness

components at microsite (squares), habitat (triangles) and landscape level (points). Each
symbol denotes for a species richness component of one local landscape (*=p<0.05,
**=p<0.01, ***=p<0.001).

References
Benton, T.G., Vickery, J.A. & Wilson, J.D. (2003) Farmland Biodiversity: Is Habitat Heterogeneity
the Key? Trends in Ecology and Evolution 18, 182-188.
MacDonald, D., Crabtree, J.R., Wiesinger, G., Dax T., Stamou, N., Fleury, P., Lazpita, J.G. &
Gibon, A. (2000) Agricultural Abandonment in Mountain Areas of Europe: Environmental Con-
sequences and Policy Response. Journal of Environmental Management 59, 47-69.
McGarigal, K. & Cushman, S. (2005) The gradient concept of landscape structure. J. Wiens and M.
Moss (Eds.) Issues and Perspectives in Landscape Ecology. Cambridge University Press,
Cambridge, pp. 112-119.
Swanson, F.J., Kratz, T.K., Caine, N. & Woodmansee, R.G. (1988) Landform Effects on Ecosystem
Patterns and Processes. Bioscience 38, 92-98.

550
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Landscape monitoring of Spain based on air photo interpretation of a stratified

network of land samples

R. Elena-Rosselló1, J. Manuel García del Barrio2, F. Bolaños1, V. Gómez-Sanz1, M.

Ortega2, C. García-Feced1, S. González1
1
Departamento de Silvopascicultura. Escuela Universitaria de Ingenieros Técnicos
Forestales. Universidad Politécnica de Madrid. Ciudad Universitaria s/n. 28040 Madrid.
Spain. e-mail: [email protected]
2
Centro de Investigación Forestal. Instituto Nacional de Investigación y Tecnología Agraria y
Alimentaria. CIFOR-INIA. Carretera de la Coruña KM 7,500. 28040 Madrid, Spain.

Landscape monitoring of Spain has been carried out by developing a system called
SISPARES (SIStema PAisajes Rurales ESpañoles) (Bolaños, et al, 2003). This system is
based on the principles of (1) land stratification for selecting representative samples and (2)
air photo interpretation for measuring and modeling their landscape spatial structure. The
land stratification is based on a previously developed biogeoclimatic land classification
CLATERES (Elena-Rosselló, 1997). The selected landscape samples are objectively
representative of the land ecological structure of Spain and consequently, their evaluation
can be statistically extrapolated for providing consistent national figures.
SISPARES has two main components: (a) REDPARES the Network of Spanish Rural
Landscapes, with 215 samples, and (b) SIGPARES the Geographical Information System,
where the information from surveys is stored and processed. At the moment, three overall
surveys have been carried out: 1956, 1984 and 1998, and a fourth is to be done based on
2004 aerial photos.
Results from the surveys showed the potential of SISPARES for evaluating habitat and
landscape spatial pattern. One of the best values of CLATERES as land stratification system
arose when checking the correlation between habitat and landscape composition and land
biogeoclimatic values (Ortega et al., 2006). Forest habitats declined following the main aridity
gradient, as agricultural habitats increased their own presence. At the same time, landscape
diversity reached the highest value in the middle of the environmental gradient.
Other important results showed the potential of the land classification for monitoring habitat
changes. After three landscape surveys, statistically significant evaluation of the changes
have being carried out, on features such as ecological quality, naturalness, fragility,
accessibility etc.
Among other national European experiences, SISPARES has been considered as a sound
methodological contribution to the BIOHAB project.

551
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Understanding the causes and consequences of landscape change: lessons from

the British Countryside Survey monitoring programme

S.M. Smart1, S. Petit1, D.C. Howard1, L.C. Maskell1, L.R. Norton1, R.C. Stuart1,
R.G.H. Bunce2
1
Centre for Ecology and Hydrology, Lancaster Environment Centre, Bailrigg, LA1 4AP
Lancaster, UK.
e-mail:[email protected]
2
Alterra Green World, PO Box 47, 6700AA Wageningen, The Netherlands

Introduction

Over the last 40 years the Centre for Ecology and Hydrology has developed an integrated
procedure for sampling at the landscape level in order to follow changes in a range of
ecological parameters in space and time as described by Sheail & Bunce (2003) .This long
term monitoring programme is designed to assess change in the extent and change in
ecological parameters at the landscape level. Information is recorded in 1 km square
samples randomly stratified in the country according to 40 Land classes sharing similar
broad environmental characteristics Countryside surveys have taken place in 1978, 1984,
1990, 1998 and the fifth survey is currently underway in 620 1km square in Great Britain but
also simultaneously in Northern Ireland. During the surveys, land use and landscape features
such as hedgerows and ponds are recorded in the field and digitally mapped. The concept is
to obtain an overall picture of what is happening in a range of different landscapes distributed
to be representative of the UK. The vegetation data has been very successful in showing
that changes in quality rather than the extent of habitats, is the recent trend in British
landscapes. One overriding success of the approach, which involves quality control and
assessment at all stages, is that the results are widely accepted by all national agencies.

Changes in vegetation and habitats

The results from the 1990 survey (Barr et al., 1993) showed that the major losses of habitats
that had taken place because of the intensification of agriculture since the war had begun to
slow down. However an exception was the length of hedgerows which are a central feature
of British landscapes which had declined over 20% in the nine years since the previous
survey. The major impact of the survey was however in the significant loss of species from
many habitats - eg over 20% in relatively infertile grasslands that had stayed as the same
habitat since 1978. Even the fields of crops had lost up to 20% of species over that period
despite the intensive use of herbicides for many years. These losses involved not only
botanical but associated biodiversity such as farmland birds and butterflies. Examination of
the ecological consequences of these changes (Bunce et al. (1999) and Firbank et al.
(2000).) showed that there were several dominant drivers such as eutrophication, more
intensive farm practices and lack of management of linear features which had led to these
losses. The survey carried out in 1998 (Haines-Young et al., 2003) confirmed that the rate of
habitat loss had slowed but species losses continued.

Landscape ecological implications

Countryside Surveys indicate that land use is undergoing important changes across Britain
and that although some trends are common, e.g. fewer but larger parcels of arable land,
specific regional patterns can be identified . In particular, the potential for land use
intensification seems to have been almost reached in some areas whilst other regions are
currently experiencing important transformations including the rapid fragmentation of
grasslands. The probability of such changes occurring at the field level clearly depends on
spatial attributes such as the size of individual parcels and the amount of intensive land use
within 1 km squares (Petit & Firbank, 2006). In parallel, landscape heterogeneity within 1 km

552
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

squares seems to have increased in some areas of Britain (Haines-Young et al., 2003).
These could be related to different processes including urbanisation and new woodland
planting (Petit et al (2004a)).In terms of vegetation, regional trends segregate into a non-
random draw of traits related to reproduction, dispersal and establishment (Smart et al.,
2006a). The profile of 'winning' versus 'losing' plant species differs between habitat types
suggesting the conditional effect of land use (Smart et al., 2006b). There is an overall
positive relationship between the diversity of the land and plant species richness. Landscape
structure is also affecting targeted functional groups of plant species, with clear regional
patterns. For example, the size and isolation of woodland patches are strongly constraining
the occurrence of specialist woodland species in the lowlands while no such effect could be
detected in the uplands (Petit et al., 2004)
.
References
Barr C.J., Bunce R.G.H., Clarke R.T., Fuller R.M., Furse M.T., Gillespie M.K., Groom G.B., C.J.,
Hornung M., Howard D.C., Ness M.J. (1993) Countryside Survey 1990: main report, London:
Department of the Environment..
Bunce R.G.H, Smart S.M., van de Poll, H.M., Watkins, J.W, Scott, W.A. (1999) Measuring change
in British vegetation. ECOFACT vol. 2. London: Department of the Environment, Transport and the
Regions.
Firbank, L.G., Smart S.M, Bunce, R.G.H, Hill, M.O., Howard, D.C., Watkins, J.W. and Stark, G.J.
(2003) Causes of change in British vegetation. ECOFACT vol. 3. London: Department of the
Environment, Transport and the regions.
Haines-Young, R., Barr, C. J., L.G., F., Furse, M., Howard, D. C., McGowan, G., Petit, S., S.M., S.
& Watkins, J. W. (2003) Changing landscapes, habitats and vegetation diversity across Great
Britain. Journal of Environmental Management 67:268-281
Petit, S., Howard, D,C., & Stuart R.C. (2004 a) A notional perspective on recent changes in the spatial
characteristics of woodland in the British landscape.Landscape and Urban Planning 69:127-35.
Petit, S., Griffiths, L., Smart, S.S., Smith, G.M., Stuart, R.C.& Wright, S.M. (2004) Effects of area
and isolation of woodland patches on herbaceous plant species richness across Great Britain.
Landscape Ecology 19: 463-471.
Petit, S., Firbank L.G. (2006) Predicting the loss of semi-natural habitat to intensive agriculture
at the national scale. Agriculture, Ecosystems and Environment 115: 277-280.
Sheail J., Bunce R.G.H. (2003) The development and scientific principles of an environmental
classification for strategic ecological survey in the United Kingdom. Environmental Conservation
30: 147-159
Smart, S. M., Clarke, R. C., van de Poll, H. M., Robertson, E. J., Shield, E. R., Bunce, R. G. H. &
Maskell, L. C. (2003) National-scale vegetation change across Britain; an analysis of sample-
based surveillance data from the Countryside Surveys of 1990 and 1998. Journal of
Environmental Management 67: 239-254.
Smart S.M., Thompson K., Marrs R.H., Le Duc M.G., Maskell, L.C. and Firbank, L.G. (2006a)
Biotic homogenization and changes in species diversity across human-modified ecosystems.
Proceedings of the Royal Society B-Biological Sciences 273: 2659-2665
Smart S.M., Marrs R.H. Le Duc M.G., Thompson K., Bunce R.G.H., Firbank L.G. (2006b) Spatial
relationships between intensive land cover and residual plant species diversity in temperate
farmed landscapes J. Appl. Ecol. 43: 1128-1137

553
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

A landscape integrated analysis and classification: application to a river valley in

central Spain

A. Patrono1, A. Saldaña López 2, A. Gómez Sal 2

1
European Union Satellite Centre – Apdo. de Correos 511, 28850 Torrejón de Ardoz
(Madrid), Spain.
e-mail: [email protected]
2
Department of Ecology, Faculty of Sciences, University of Alcalá – Ctra. Madrid-Barcelona
Km 33.6, 28871 Alcalá de Henares (Madrid), Spain

Introduction

A method for landscape analysis and classification is presented. The method takes into
account quantitative techniques to integrate biotic and abiotic variables (transformed into
continuous ranges), minimizes subjective interpretations and avoids the predefinition of
spatial regular units or constraints associated to a specific study area or scale. The purpose
was to obtain a coherent organization of the land in some synthetic terrain units to be used in
landscape analysis.

Material and methods

Study area and data

A set of environmental variables was considered to explore the landscape of the Lozoya
River valley, located on the southern slope of the Guadarrama mountain range (Madrid,
Spain). Topographic variables include the DTM and the slope. Climatic variables include the
precipitation and mean annual temperature range, modelled using the DTM and watershed
information. Irradiance (potential) was calculated modelling the yearly illumination condition.
Soil variables were estimated from profile descriptions. Land-use variables were derived from
a multi-temporal set of Landsat imagery, classified with a maximum likelihood algorithm to
obtain the probability distributions of 10 pre-selected land cover classes.

Ecological classification
The approach sought a compromise between simplifying the spatial variability given by all
possible combinations among variables (20) and the need of delineating terrain units with a
feasible area and degree of homogeneity (Bryan, 2006). The method analysed the loss of
internal variability (estimated by the standard deviation) within the terrain units defined by
segmenting the area into an increasing number of clusters with the Isodata algorithm (Duda
and Hart, 1973). The objective was to define the limit from where any further segmentation
did not produce any remarkable change. The trend analysis was done by fitting a function to
the clustering scores whose first derivative represented the rate of variability loss (Patrono
and Saldaña, 2006). The resulting units were grouped into classes characterised by a unique
combination of variables scores and transformed into a vector database.

Multivariate statistical methods were applied to examine the ecological significance of the
obtained landscape classification. Principal Components Analysis was used to reduce the
size of the database, minimising the loss of information. The relations among the
transformed 20 variables were studied with a cluster analysis whose result was over-
imposed on the vector space generated by the factor coordinates of the variables (Jongman
et al., 1987).

Results

554
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

The result of the classification yielded 40,538 homogeneous terrain units grouped into
962 classes. The derived vector database represented less than 2% of the original data
volume. PCA successfully reduced the size of the database (the first eight principal
components captured over 92% of the variation in the 20 input parameters). The result of the
cluster analysis overlaid on the ordination by the first two principal components (Figure 1)
highlighted the relations of the environmental variables scores that typify the terrain units.
The grouping of the variables defined two main clusters that tended to characterize
mountainous and valley bottom ecosystems respectively. Several subgroups could be
defined in the two clusters. For example one gathered the variables typifying higher elevation
ecosystems (including mountainous shrublands), one grouped more developed soils
(including juniper stands) and one comprised a subgroup with deciduous and evergreen oak
forests, dehesas, grasslands and shrublands (dominated by Spanish lavander).

Figure 1. Ordination by the first two principal components of the transformed 20 variables.
The result of the cluster analysis is overlaid.

Discussion and conclusion

The results confirmed that the proposed method could be used to summarize landscape
complexity. In addition to an important data reduction, a systematic landscape subdivision
into homogeneous and ecologically meaningful elements was obtained, minimising
subjectivity and constraints. The result could be mapped at detailed scale and in a robust
manner which offers environmental specialists a spatial database open to multiple
applications as verified in the analysis of the relations and the ecological significance of the
environmental variables that characterize the terrain units.

References
Duda, R.D & Hart, P.E. (1973) Pattern Classification and Scene Analysis. John Wiley and Sons, New
York.
Bryan, B.A. (2006) Synergistic Techniques for Better Understanding and Classifying the
Environmental Structure of Landscapes. Environmental Management 37(1): 126–140.
Jongman, R.H.G., ter Braak, C.J.F. and van Tongeren, O.F.R (1987) Data Analysis in Community
and Landscape Ecology, PUDOC Scientific Publ., Wageningen.
Patrono, A. & Saldaña, A. (2006) A proposal for a landscape integrated classification. A. Marçal (Ed).
Global Developments in Environmental Earth Observation from Space. Millpress, Rotterdam, pp.
233-240.

555
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Optimising thematic resolution of landscape metrics for biodiversity monitoring

in rural Europe

D. Bailey1, R. Billeter2, S. Aviron1, M. Parracchini3, O. Schweiger4 and F. Herzog1

1
Agroscope Reckenholz-Tänikon Research Station ART – Reckenholzstr. 191, CH-8046
Zurich, Switzerland.
e-mail: [email protected]
2
ETH Swiss Federal Institute of Technology, Geobotanical Institute, Zurichbergstr. 38, CH-
8044 Zurich, Switzerland.
3
Joint Research Centre of the European Commission, Institute for Environment and
Sustainability, Rural, Water and Ecosystem Resources Unit, TP 262, 21020 Ispra, Italy
4
UFZ - Centre for Environmental Research, Department of Community Ecology,
Theodor-Lieser-Strasse 4, D-06120 Halle, Germany

Introduction

There is an urgent requirement for relatively cheap and rapid methods for monitoring the
status of biodiversity in agricultural landscapes. Most direct monitoring methods are
expensive requiring extensive field work and the use of multiple species groups with
contrasted ecological requirements. The potential use of landscape metrics as inexpensive
indicators was proposed by O’Neill et al. (1988). Whilst the identification of a generic set of
metrics is challenging due to map interpretation issues, spatial scale and ecological
appropriateness, it is widely acknowledged that biodiversity depends on landscape
properties. Here we identify landscape pattern metrics that act as indicators for biodiversity in
the context of monitoring Temperate European agricultural landscapes. As the affectivity of
metrics will be highly influenced by the map classification (Bailey et al. 2006a) we explore the
role of thematic resolution and examine a suite of biological and functional groups.

Materials and methods

Twenty four study sites of 16 km2 were located within arable landscapes of seven
temperate European countries. Land cover was digitised using recent orthophotos and
classified using an adaptation of the European EUNIS habitat classification system (Davies
and Moss 1999).To study the response of biological groups to landscape metrics at different
thematic resolutions, the original land cover maps (47 habitats) were reclassified using three
coarser classification systems (2, 3 & 14 habitats). The four resolutions were considered
appropriate to taxonomic and functional biodiversity. Forty-one common landscape-level
metrics representing five main aspects of landscape structure (grain, edge, shape,
configuration, diversity) were then calculated using FRAGSTATS 3.3 (McGarigal et al. 2002).
Manageable subsets of metrics for each thematic resolution were them obtained using
exploratory analysis (Bailey et al. 2006a). Biodiversity of the sites were assessed using
taxonomic groups that have different ecological requirements (plants, birds, wild bees, true
bugs carabid beetles, hoverflies and spiders). The arthropod groups were also divided into
functional groups representing body size. Linear mixed models were used to detect
correlations between metrics and species richness values (Bailey et al. 2006b).

Results - discussion

The exploratory approach to landscape-level metric selection allowed for both the
identification of metrics and the general areas of landscape pattern which correlate with
taxonomic and functional groups at the different levels of thematic resolution. Few
landscape-level metrics correlated significantly with the biodiversity data at the coarser
scales of thematic resolution and grain metrics (patch density, largest patch index) were the

556
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

only measures to correlate with perhaps the more robust biological groups (plants, large
sized arthropods). At the fine scale of thematic resolution there were also few significant
correlations with the biodiversity data and a diversity metric (e.g. Simpson’s diversity index)
was the most appropriate measure when using relatively heterogeneously defined biological
groups. The intermediate thematic resolution offered most promise for biodiversity
monitoring. Many metrics were significantly correlated to most taxonomic and functional
groups. Metrics suitable to monitor biodiversity at this level of thematic resolution include the
largest patch index, edge density, nearest neighbour, the proximity index, circle and
Simpson’s diversity index. The correlation of landscape-level metrics with biodiversity groups
were therefore sensitive to thematic resolution and an intermediate scale proved the most
informative for both taxonomic and functional groups. Our results show how very coarse
parameters of landscape structure calculated at the intermediate scale can indicate whether
landscapes are likely to sustain biodiversity. Higher biodiversity, for example, can be
expected in European agricultural landscapes which have a greater habitat diversity
(Simpson’s diversity index), large variations in patch distribution (nearest neighbour) and
more edge and patchy habitats (edge and patch density).

Conclusions

The thematic scale issue should not be underestimated. If simple landscape

classifications have been used it is perhaps easier to choose metrics using an expert
approach as the main effect of landscape pattern is likely to be a species-area relationship.
High thematic resolution data is unlikely to correlate with generally defined biological groups
and should be used to answer specific research questions related to carefully defined
species data. Our findings suggest that for more general biodiversity monitoring, intermediate
levels of thematic resolution are appropriate. The metrics calculated at this level correlate
well with most species groups and provide information about interactions with landscape
pattern. Two possible applications could be to incorporate the metrics identified in this study
in a monitoring system for rural landscapes and to use them to identify biodiversity hot spots
in European agricultural landscapes.

References
Bailey, D; Herzog, F. Augenstein, I. Aviron, S. Billeter, R. & Baudry, J. (2006a) Thematic resolution
matters: Indicators of landscape pattern for European agro-ecosystems. Ecological Indicators
0.1016/j.ecolind.2006.08.001.
Bailey, D; Billeter, R. Aviron, S. Schweiger, O. & Herzog, F. (2006b) The influence of thematic
resolution on metric selection for biodiversity monitoring in agricultural landscapes. Landscape
Ecology DOI 10.1007/s10980-006-9035-9.
Davies, C. & Moss, D. (1999) EUNIS Habitat Classification. Final Report to the European Topic
Centre on Nature Conservation. European Environment Agency, Paris. 256pp.
McGarigal, K; Cushman, S. Neel, M. & Ene, E. (2002) FRAGSTATS. University of Massachusetts:
www.umass.edu/landeco/research/fragstats/fragstats.html.
O’Neill, R; Krummel, J. Gardner, R. Sugihara, G. Jackson, B. DeAngelis, D. Milne, B. Turner, M.
Zygmunt, B. Christensen, S. Dale, V. & Graham, R. (1988) Indices of landscape pattern.
Landscape Ecology 1: 153-162.

557
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

The use of species compositional dissimilarity to assess the effectiveness of

ecological classification schemes

B.E. Lawson1,3, S. Ferrier2, G. Wardell-Johnson1, D.V. Pullar3, R.J.S. Beeton1

1
School of Natural and Rural Systems Management, The University of Queensland, Gatton
QLD 4343, Australia.
email: [email protected]
2
New South Wales Department of Environment and Conservation, P.O. Box 402, Armidale
NSW 2350, Australia
3
Department of Geographical Sciences and Planning, The University of Queensland,
Brisbane QLD 4072, Australia

Introduction

Ecological classification schemes implicitly assume that map units represent some level
of species similarity within each unit and some level of species dissimilarity between units.
We describe our use of species compositional dissimilarity (SCD) to assess the utility of an
existing ecological classification scheme from Queensland, Australia. The ‘bioregional
hierarchy’, as described by Sattler and Williams (1999), comprises three levels:
• Subregions which represent significant associations of geological, geomorphology
and climate within a bioregion, and are mapped at 1:500 000 scale.
• Land zones delineate major geological types and their associated landforms, and are
mapped at 1:250 000 scale.
• Regional ecosystems (REs) are vegetation communities that are consistently
associated with a particular combination of geology, landform and soil in a bioregion.
They are mapped at 1:100 000 scale and finer.
Although there is significant work relating to the importance of SCD in determining the
efficacy of ecological classification schemes (e.g. Phillippi et al. 1998, Ferrier 2002), its
application to existing classification schemes remains understudied and under-reported.

Methods

Site-based floristic presence-absence data from 418 survey sites in the South-east
Queensland Bioregion were used. Each site included details of its membership to bioregional
hierarchy units. SCD was calculated using the Kulczynski metric, which is considered to
provide the best representation of non-linear ecological gradients (Faith et al. 1987).
Firstly, pair-wise species dissimilarity comparisons were calculated for all within-unit and
between-unit comparisons for every classification unit at each level of the bioregional
hierarchy. This was done to assess the general patterns of SCD for classification units and to
determine the extent to which each hierarchy level portrayed floristic patterns.
The second phase used randomisation to determine the effects of number of sites per unit
on SCD patterns. From the pool of sites belonging to a particular unit, the requisite number of
sites were randomly drawn. Due to the time consuming nature of the analysis, only 10
permutations of each number of sites for each unit were performed. A random site allocation
was also performed to provide a baseline against which units could be compared.

Terminology
Within-unit dissimilarity refers to the per cent floristic dissimilarity between sites within the
same map unit, and between-unit dissimilarity to the per cent dissimilarity between sites in a
map unit with sites outside of that unit. Differential dissimilarity refers to the per cent
difference from within-unit to between-unit dissimilarity for a particular map unit.

Results and Discussion

558
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Comparisons of SCD between units revealed that dissimilarity levels and the variation
increased from broader-scale subregion level to finer-scale regional ecosystems. The five
subregion units examined exhibited an average of 4 % differential dissimilarity, the seven
land zones an average of 10 %, and the 59 REs an average of over 19 %. This shows finer-
scale classifications to be more effective at discriminating SCD than broader classifications.
Figure 1 shows these trends for representative units from each bioregional hierarchy level.
Analysis of the effects of site number on SCD showed consistent trends across all levels
of the hierarchy. For all hierarchy levels, minimal variation in differential dissimilarity was
evident when 10 or more sites are used to derive the information. This suggests that very few
survey sites per classification unit are required to obtain a meaningful estimate of a unit’s
SCD. Results from a representative unit from each hierarchy level are presented in Figure 2.
The results reveal a robust and transparent way to guide future ‘lumping’ or ‘splitting’
decisions for a priori classification schemes, to achieve a more even balance in SCD levels
between classification units as advocated by Ferrier 2002. However, more widespread use
will require the development of software to efficiently derive and display such information.
Our current research is focussed on using this approach to disentangle the effects of sample
size and ecosystem area on SCD, and to gain a better understanding of its relationship to
species richness within units. Additionally, we also aim to test whether similar patterns are
evident for different classification schemes and environments.

Figure 1. The extent and variation of Figure 2. The effect of site number on
within- and between-unit species species compositional dissimilarity at the
compositional dissimilarity at the three three levels of the bioregional hierarchy
levels of the bioregional hierarchy from the from the South-east Queensland
South-east Queensland bioregion. Several bioregion. Several units representative of
units representative of the trends are the trends are shown.
shown.

References
Faith, D.P; Minchin, P.R. & Belbin, L. (1987) Compositional dissimilarity as a robust measure of
ecological distance. Vegetatio 69: 57-68.
Ferrier, S. (2002) Mapping spatial pattern in biodiversity for regional conservation planning: where to
from here? Systematic Biology 51: 331-363.
Philippi, T.E; Dixon, P.M. & Taylor, B.E. (1998) Detecting trends in species composition. Ecological
Applications 8: 300-308.
Sattler, P.S. & Williams, R.D. (1999) The Conservation Status of Queensland’s Bioregional
Ecosystems. Environmental Protection Agency, Brisbane.

559
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Large scale land cover classification systems – a pragmatical appraisal

C.S. Cruz1, J.P. Fernandes2, N. Guiomar2, T. Baptista3, M.J. Mateus3,

1
Universidade de Évora, Rua Romão Ramalho, 59 – 7002-554 Évora, Portugal e-mail:
[email protected]
2
CEEM, Universidade de Évora, Rua Romão Ramalho, 59 – 7002-554 Évora, Portugal
3
AMDE, Rua 24 de Julho, 1 – 7000-673 Évora, Portugal

Introduction

Large Scale Land Cover classification systems are critical instruments for landscape
planning, ecological assessment, environmental evaluation and every landscape assessment
and management process.
Different systems of classification have been proposed and are being currently used such
as the CORINE Land Cover classification - CLC (prepared for a medium scale of 1:100 000)
and the EUNIS habitat classification system (more adapted to the classification of vegetation
formations). Their adaptation to large scales (above 1:25 000) and to complex vegetation
covers (e.g. classifying simultaneously the habitat type, dominant land use and the varying
type of undercover) poses increasing problems because of one hand to the need of using
automatic classification methods for remote sensing data and on the other, the need for a
detailed classification not only of the dominant vegetation layer, but also of the other layers
present.
The development and implementation of an integrated CLC classification adapted to large
scales (detailed to the 5th digit) and its comparison with complex classification systems
combining both information on the habitat, dominant vegetation or land use, the associated
vegetation and the conservation status are presented and discussed.
The use of complex classification systems, combining information on the morphology,
geology and pedology (meaning globally the geocenosis), the vegetation complex
(phytogeocenosis) and the human influence (namely type of vegetation management), is
also discussed in terms of their utility, in comparison with the combination of thematic
cartography. Comparative examples of application of different systems are presented and
their relative advantages compared.

Development of a conceptual framework

At the large scale of landscape zoning the main differences or units derive from
microclimatic variations, soil differences (e.g. water and nutrient availability) and, eventually,
morphology in terms of local variation of spatial dynamic patterns (e.g. flow direction, flow
intensity, erodibility). Land use builds an additional zoning factor in which it conditions many
of these factors and introduces energy factors that determine disturbance particular
disturbance patterns that emerge as circumstantial landscape features (eg. matrix, patches
or corridors).
Stability patterns of some spatial variables are presently prone, due to human influence to
experiment accelerated cycles of variations that determine that man’s management
intervention must be based on the knowledge of the different factors that determine a given
landscape pattern at a given moment. It is also necessary to assess the different stability of
that pattern according to spatial and temporal arrangement of disturbances. This implies that
a good land use classification should be able to characterise and classify simultaneously not
only the present land use mosaic but also the landscape organisation according to its natural
resources and processes and should eventually be linked to disturbance factors.
This classification system implies the simultaneous consideration of the stable
geocenotical patterns (presenting a high stability), related or not to vegetation descriptors
(e.g. potential natural vegetation) and of the antropic land use spatial patterns with its lability.

560
Theme 5: Monitoring and classification
5.1 Symposium 10: Monitoring at the landscape scale

Our classification system is aimed in the first stage at enlarging the CLC classification
until the 5th level of detail:

1. AGRICULTURAL AREAS
1.1 Arable land
1.1.1 Non-irrigated arable land
1.1.1.1 Rainfed herbaceous cultures
1.1.1.1.1 Cereals

This classification system is only aimed at present land uses and is not able to represent
other landscape features, such as geocenosis or natural or potential phytocenosis.
Neverthless, it is possible to build combined classification systems (e.g. aaaaaa bbbbbb
cccccc) where each of the three sets of digits represents one landscape characteristic
(Loureiro & Cruz, 1993, FAO, 2005). Such a combined classification (e.g. CLC (land cover);
EUNIS (natural vegetation); geology, geomormology or morpholitology) would give the
planners and managers an integrated appraisal, not only of the present actual character of a
landscape, but also of the set of factors that determine and characterise the given situation in
terms of stable factors. This information allows simultaneously an identification of eventual
disturbance processes and intensities, a better characterisation of the spatial arrangement of
dynamical processes and the characterisation of the natural arrangement of landscape
features like Matrix, patches and corridors (disturbed and not disturbed).

Figure 1. Comparision between the present land use structure and the natural stable structure

References
Cruz, C. S. (2002) A cartografia das fitogeocenoses aplicada à gestão de áreas protegidas. Tese de
Doutoramento, Universidade de Évora.
Loureiro, N. S.; & Cruz, C. S (1993) Cartografia dos usos do território e dos habitats de Portugal.
Projecto INASP, ICN, Lisboa..
FAO (2005) Land cover classification system: classification concept and user manual (Software
version 2). FAO, Rome.
Fernandes, J. P. (1993) Classificação das Unidades Ecológicas Adoptada para Portugal. Seminário
sobre Avaliação de Impacte Ambiental em Sistemas Ecológicos, CEPGA, SPRCN, Serra da
Estrela.
Guiomar, N.; Fernandes, J. P.; Cruz, C. S.; Baptista, T.; & Mateus, J. (2006) Sistemas de
classificação e caracterização do uso e ocupação do solo para zonamento microescalar.
Proceedings of ESIG 2006, Oeiras.

561
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

5.2 Symposium 25: Advances and Applications of Landscape Character

Mapping

European Landscape Typologies as a Reference Base for Data on Cultural

Heritage and Traditional Knowledge Systems in the Mediterranean

T. Kizos

Department of Geography, University of the Aegean, University Hill, Mytilini 81100, Greece,
0030 22510 36447,
e-mail: [email protected]

Introduction:

Cultural heritage lies in the heart of current debates on conservation, preservation and
sustainable development in Western societies. A way to estimate, manage and evaluate
cultural heritage can be by the knowledge it produces. This ‘traditional knowledge’ stands for
practices, representations, expressions, knowledge, skills as well as instruments, objects,
artefacts and cultural spaces associated with communities, groups and, in some cases,
individuals that recognize them as part of their cultural heritage. This intangible cultural
heritage is constantly recreated by communities and groups in response to their
environment, their interaction with nature and their history (2003 UNESCO Convention).
Landscape, as the footprint of all human activities on nature, can be the medium for
identifying, studying and marking traditional practices and knowledge in an area. Landscapes
can offer the spatial framework for associating bio-environmental, social and management
sciences and provide the reference base for data on cultural heritage and traditional
knowledge systems.
In the Mediterranean, cultural heritage, from the cultural landscapes of rural areas to the
historic town centres, is the expression of its identity, but also part of the everyday
environment of numerous people. Threats of this rich heritage may come from
‘modernization’ of production and society, for which cultural heritage is often viewed as a
constraint. In attitudes, practices and worldviews, it may be viewed as ‘conservative’ and
‘backward’ and for knowledge issues most if not all ‘traditional knowledge’ is considered as
‘outdated’ and ‘obsolete’. But, cultural heritage and the knowledge produced during its
making and reproduction can be used in order to enrich ‘modern’ practices for sustainable
use of resources while continuing to mark local identities.
In this paper, existing European landscape typologies are used as a reference base against
traditional knowledge systems in the Mediterranean. Recent landscape typologies are
examined in three examples of traditional knowledge farming systems to evaluate the
suitability to describe actual landscape differences.

European Landscape Typologies and Landscape Character

Landscape typologies at the European scale are rare, due to environmental conditions and
habitats diversity on one hand and the variety and depth of human interventions on the other.
Existing efforts tend to focus either mostly on natural forces (Mucher et al. 2003), or on
human management systems (Meeus 1995). A recent effort that attempts to combine both
approaches has been a typology developed for the ELCAI (European Landscape Character
Assessment Initiative) initiative that is based upon the use of different layers of human
elements (land cover, land use, special characteristics) upon natural landscape elements
(climate, altitude) (Wascher 2005). This typology is developed to be used for defining and
assessing Landscape Character (LC) that is a distinct, recognisable and consistent pattern of
elements in the landscape that makes one landscape different from another. The emergence
of LC assessments has brought forward ‘new’ spatial units that can be used for landscape
typologies (Wascher 2004).

562
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

Traditional Knowledge Systems in the Mediterranean and Landscape Typologies

The link between landscape typologies and traditional knowledge systems is not as
straightforward as the use of information on the cultural elements for the making of these
typologies might imply. Traditional knowledge systems are complex systems that may be not
is use today and therefore not considered when creating such typologies, although their
marks may be still evident in the landscape. In the Mediterranean, where agriculture has
undergone great changes: replaced by ‘modern’ agriculture or abandoned.
Such an example is terracing. Cultivation terraces are very characteristic of a landscape,
combining value (ecological, economic, symbolic, etc.) and function, if the terraces are not
abandoned. In the Mediterranean, they are found in many places (Grove and Rackham
2002). Existing LC assessments and landscape typologies of the Mediterranean ignore them,
focusing on land uses. Reasons behind this relate with the lack of relevant data.
The examples examined in this paper include three very different systems of traditional
knowledge systems on three different Aegean Islands that involve cultivation terraces (Figure
1): (a) Olive plantations on Lesvos; (b) Vines on Serifos; and (c) Mixed farming (cereals and
trees) on Nisyros Islands. The basic principles of these systems are described briefly to
assess differences and similarities. They are all characterized as “Mediterranean hills of
pasture land or permanent crops”, but the empirical material presented demonstrates that
they are very diverse, including high diversity inside each type, that they clearly should be
assigned with different landscape characterizations as a result of the different traditional
knowledge systems they have resulted from. Therefore, new ways of creating typologies and
characterizing landscapes have to be used that have to consider traditional knowledge
systems in the cultural elements they use. Local research is required for the concretization of
the relationship for specific spatiotemporal settings.
Another issue briefly touched in this paper is the future of such systems. The examples
offered here demonstrate that this approach marks heritage and knowledge as concepts that
should not be preserved as ‘museums’ of a past however ‘golden’, but actively conserved to
enrich contemporary practices and that landscape typologies which consider them can serve
as a reference base for successful and meaningful landscape characterizations in the
Mediterranean.

References
Grove, A.T. & Rackham, O. (2002) The Nature of Mediterranean Europe: An Ecological History, Yale
University Press, New Haven.
Kizos, T. & Koulouri, M. (in press) Same Land Cover, Same Land use, Different Landscapes: Small
Scale Landscape Change in Olive Plantations on Lesvos Island, Greece. Landscape Research.
Meeus, J.H.A. (1995) Pan-European Landscapes. Landscape and Urban Planning 31, 57-79.
Mucher, C.A.; Bunce, R.G.H.; Jongman, R.G.H.; Klijn, J.A.; Koomen, A.J.M.; Metzger, M. &
Wascher, D. (2003) Identification and Characterization of Environments and Landscapes in
Europe, Alterra Rapport 832, Alterra, Wagenigen.
Wascher, D. (2004) Landscape-indicator development: steps towards a European approach. R.
Jongman (Ed.) The New Dimensions of the European Landscape, Springer, Wageningen UR
Frontis Series Nr. 4, Berlin, pp. 237-252.
Wascher, D. (2005) European Landscape Character Areas: Typologies, Cartography and Indicators
for the Assessment of Sustainable Landscapes, Final Project Report of the ELCAI, Project,
available at www.elcai.org.

563
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

From Landscape Character Assessment to Sustainable Design – Tools for

Planning and Policy

D. Wascher, P. Smeets, W. Timmermans, V. Kuypers

Alterra, P.O. Box 47, 6700 AA Wageningen, The Netherlands.

e-mail: [email protected]

Landscape Character Assessment

Landscape Character Assessment (LCA) stands for an assessment technique that is

descriptive without being judgmental regarding the state of a landscape. As a means for
developing landscape profiles and typologies, LCA is recognized as a key reference system
for researchers and planners dealing with integrated land use and spatial planning.

Its integrative and synergetic power links LCA well to a variety of research and policy
objectives in the field of sustainable development. Sustainable land use is considered to be
intrinsically linked to the concept of multi-functionality. Rural, urban and peri-urban
landscapes provide a variety of integrated functions including production, living, regulation,
and information functions. The way and degree to which these functions are being offered
determine the environmental and socio-economic qualities of a landscape – from the
perspective of policy makers, researchers, planners, investors, land users and a variety of
stakeholders. Linking LCA to landscape planning tools can hence be considered as an
essential prerequisite for sustainable design in the wider countryside.

This paper will analyze the rural and urban dimension of introducing sustainable design
principles when developing spatial development perspectives by drawing upon selected
cases in the Netherlands and examining possible applications of above landscape tools for
the wider European context.

Spatial Development in Europe

Europe’s rural landscapes are traditionally perceived and appreciated for their diversity,
traditional character, regional identity and high qualities with regard to standards of living,
transport infrastructure and the level of spatial planning. Another important aspect is the
historically and functionally defined relation between rural and urban areas. However, during
recent decades, uncontrolled urban sprawl, agricultural intensification coupled with land
abandonment have resulted in a clear decline of landscape diversity, biodiversity, ‘green
values’ in and around cities and cultural heritage. Rural areas have continually shrunk as
residential districts, recreational facilities and business estates have been built. In many
places, the rural areas are fragmented due to an increase in scattered construction and new
infrastructure. The spatial contrasts between the city and the countryside are diminishing, as
is the diversity of urban and rural environments. These developments make many regions
more monotonous, less attractive, more exchangeable and less distinct. As a consequence,
the boundaries between rural and urban areas do not reflect the spatial and structural
characteristics of the ‘parent’ landscapes. At the same time, demand from society for a
greater diversity of environments is increasing, not in terms of the traditional, physical
distinction between the city and the countryside, but in terms of attractive, safe city areas in
which to spend time and live - attractive, distinctive landscapes, and accessible rural areas
for recreation and relaxation. The rural areas are taking on the function of public space.

The Rural Dimension

564
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

The task is to analyze and assess the quality of the rural areas in relation to current and
planned policies, especially with regard environmental pollution, biodiversity and amenity
values of cultural landscapes. Of relevancy here are the European and global agricultural
policies, land areas policy, and the environmental, nature and spatial policy. In response to
severe epidemics of Classical Swine Fever (1997) and of critical nitrate loads due to high
livestock densities, Dutch authorities have launched massive reconstruction plans for large
rural regions. The Dutch law on reconstruction for concentration regions (2002) offers a
spatial framework for the restructuring of intensive livestock farming in areas with high
livestock densities. Within the concentration regions, mono-functional ‘development zones’,
multi-functional ‘interwoven zones’ and zones where livestock densities are substantially
reduced, among other things by transfers of farms (‘expansion zones’) have been
established. Reviewing the implementation of these reconstruction plans will form the basis
for examining the question whether the Dutch experience can find application at the
European level with special attention to the role of landscape character assessment. EU
environmental standards, aspects of regional identity and economic efficiency must be
considered as key objectives for sustainable European rural areas. Special emphasis will be
on the role of landscape assessment tools for the identification of regional profiles relevant
for sustainability assessment and of their corresponding thresholds and risks regarding
pressures and driving forces affecting land use change.

The Urban Dimension

As in the Dutch context rural and urban development is not only a matter of fringes, but can
be regarded as a spatial transformation to a metropolitan landscape, these tools are also
highly relevant for the urban dimension. Planning of whole new towns like Almere are based
on historical, cultural and ecological conditions and characteristics. Methodological the actual
landscape is stripped into layers from the past to the future. A former great lake –
Haarlemmermeer - became a polder at the end of the 19th century, because it was a threat to
the development of the city of Amsterdam. It became rationalized farmland and was later
consumed by Schiphol airport. Nowadays the pressure of urban development and mainport
facilities lead to scarcity of space, that poses questions of multiple land use such as
combining flood risk areas with housing (www.bouwenmetwater.nl). For such a development
the basic principles are the characteristic lines in the landscape, the design of the polder and
water system, sweet and salt groundwater patterns, peat and clay layers. These layers have
been analyzed in order to create a 60 hectare sustainable water basin in which up to 1500
houses can be build. The size makes it possible to relieve large parts of the water system
and in the same time respect the cultural and ecological conditions that “made” the
landscape.

This paper reports on ways of applying landscape character assessment tools such as
landscape typologies, indicator-based landscape functions, and landscape change models,
when developing integrated and stakeholder-oriented design concepts for sustainable design
in rural and urban areas.

565
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

Risk and vulnerability of landscape identity

É Konkoly-Gyuró1, S Jombach 2
1
University of West Hungary 1 – H-9400 Sopron, Bajcsy-Zsilinszky ut 4. Hungary
e-mail: [email protected]
2
Corvinus University of Budapest 2 – H-1118. Budapest, Villányi ut 35-43. Hungary

Introduction

Landscape is the tangible and perceivable result of the co-existence of culture and nature.
Landscape identity reflects the way different cultures interact with nature, shows how
adaptive their land-use techniques are, how harmonious or disharmonious settlements,
infrastructures, production surfaces are inserted into the biophysical environment. Man
creates landscape identity; it is the imprint of the human activities driven by culture.
Landscape identity is strongly related to sustainability, as it is the expression of the
interaction between environment, economy and society. If any of these three pillars is
damaged, identity will be destroyed.
Risk and vulnerability are adjunct terms in a cause effect chain. Risk refers to external
impacts, but the factual effect of them depends on the vulnerability of a system. Based on the
ongoing research within the SENSOR project (www.sensor.org), this presentation will give an
overview about the possible assessment of the risk and vulnerability of landscape identity at
a European scale.

Research goal and method

The SENSOR, a EUFP6 Integrated Project (2004-2008), aims to create an ex-ante

assessment tool, for evaluating the policy impacts on sustainability focusing on the regional
multifunctionality of land-use. In addition to the 32 impact issues, listed in the Impact
Assessment Guidelines (European Commission, SEC 2005), a new issue, the landscape
identity was introduced. Landscape identity can be best described by landscape character
areas formed by unique characteristics of the cultural and natural heritage. In the SENSOR
project however our spatial reference framework (Renetzeder, Ch. et al. 2004) consists of
571 regional units of the EU 27+2 at NUTSX (2/3) level. The first task has been to create
adequate indicators for showing the persistence of the valuable landscape identity. Three
crucial questions have arisen: 1) What can be considered as valuable identity of a region,
where the continuity is desirable? 2) What are the most important, land-use sensitive
components of the landscape identity? 3) Which relevant indices can be calculated within a
model for which the data availability is given now and also in the future at NUTSX level for
the whole Europe?
We suppose that identity is valuable if attractivity and the unique sense of place exist and the
landscape are also perceived and appreciated by people. Thus we’ve chosen visual
attractivity, appreciation of existing landscape heritage and unique character as key issues of
the assessment. Accordingly two indicators, reflecting the persistence of the valuable
identity, have been created.

Indicator 1. Change of visual attractivity

Visual attractivity means the scenic value of the landscape and environment that is perceived
and appreciated by people. It depends predominantly on the land cover diversity that can be
measured by the edge density (Kiemstedt, 1971) Edges between the visually different land
cover types are visible manifestation of landscape structure and fundamental elements of
landscape identity. Scale and speed of its change show not only loss or gain in diversity both
visually and biologically but also risk and vulnerability of identity. First results of the indicator

566
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

value calculation affirmed the assumption that the most significant transformations occurred
on the peripheries of Europe (Ireland, Iberian Peninsula, Eastern-Central European
countries).

Indicator 2. Continuity of appreciated landscape heritage

The higher the level of appreciation is the stronger the landscape identity is manifested. The
level of appreciation is calculated in each spatial unit from the designations and legal
protection measures and from the attendance by tourists. In those areas where the
appreciation level is high the continuity of land-use, accordingly the land cover structure, is
highly important as it contributes largely to the preservation of landscape identity. The
transformation risk of the cherished traditional landscapes in the new accession countries,
especially in Bulgaria and Rumania, is extremely high.

Risk and vulnerability

The indicators reflect change or continuity of important landscape identity factors. In the
further step of the research we will define thresholds in NUTSX regions according to the risks
and vulnerability. Risk and vulnerability can be evaluated at scales that are ended by
opposite qualities, like security and stability. So there are counter-impacts and value pairs
like risk/security and vulnerability/stability. The assessment of these can be done by SWOT
analysis. Risk and security are external impacts: threats and opportunities. Risk level
coincides with the lack of protection measures and the extreme growths of tourism. An
important security factor is the designation of heritage. Vulnerability and stability are internal
qualities: weaknesses and strengths. A region is vulnerable if the land-cover is homogenised,
as the landscape diversity and the population is decreasing. Sign of stability is the diverse
and persistent landscape structure as well as moderately changing population density.

SENSOR project provides an important first step toward the methodical development of the
risk and vulnerability assessment of landscape identity. The fact that this topic is considered
as a sustainability impact issue reflects a new approach, the growing acknowledgement of
culture’s importance. Nevertheless the large European scale and the given spatial framework
(regions instead of landscape character areas) of the current research result in some
generalisations and erase certain differences between and within the regional units. Future
research scope should be enlarged toward the accurate assessment of the richness and
diversity of the smaller scales.

References
European Commission, (2005) Impact Assessment Guidelines, SEC(2005) 791 pp. 29-31
Kiemstedt, H., (1971) Harzlandschaft und Freizeit Harzer Ferkehrsverband: Schriftenreihe des Harzer
Verkehrverbandes
Renetzeder, Ch., Eupen, M. van, Mücher, S., Wrbka, T., Wascher, D., Kienast. F., (2004) Report
on methodology and map for integrated Spatial Regional Reference Framework SENSOR
Deliverable 3.1.3

567
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

Comparative review of European national and international landscape

classifications with regard to future applications

G. Groom1, D. Wascher 2, S. Mücher 2

1
NERI – Grenaavej 14, 8410 Roende, DK.
e-mail: [email protected]
2
ALTERRA – Droevendaalsesteeg 3, 6708 PB Wageningen, NL.

Introduction

The last 10 – 15 years have seen significant developments in Landscape Character

Assessment (LCA) work. Today LCA is recognised as an important tool for policy
stakeholders, providing them with quantitative and qualitative evidence to reach a dynamic
management adjustable to new demands of regional identity. LCA has come to be seen
as central to sustainable development and management of land. Many organisations have
developed sophisticated landscape character mapping tools that are scientifically sound,
region-specific and stakeholder-oriented.
Two recent developments in relation to Europe-wide LCA work have been a review of
over 50 examples of regional, national, and pan-European LCA works (Groom et al. 2006)
and production of a pan-Europe landscape typology and map, LANMAP2 (Mücher et al.
2003; Wascher 2005). The former activity focused on the characteristics of LCA as a process
of classification-orientated mapping. The picture from that review was one of both
convergence and differences in terms of the landscape factors considered and the
mapping methods applied. Questions were noted relating to the mapping of landscape
character types, the use of crisp mapping approaches and the representation of landscape
change. The latter activity, LANMAP2, was successful in identifying and mapping
consistently for pan-Europe the major biophysical landscape characteristics and using land
cover information to add interpretation representative of the cultural characteristics of
Europe’s landscapes. However, comparisons of LANMAP2 to regional and national LCA data
revealed mixed results (Wascher 2005).
This paper re-visits the questions raised by the earlier LCA-review work and shortcomings
in the LANMAP2 evaluation work to identify possibilities for clarification of the European LCA
arena and further development of pan-Europe landscape characterisation work.

From review to refinement

The 2005 review of European LCA work was structured around the predominant mode of
the examined LCA works, their structure properties, the landscape factors (e.g. biophysical
environment, cultural) that they considered and the classification/mapping methods used
(Groom et al. 2006). Distinction between landscape character types (i.e. generic, repeated
spatial units) and landscape character areas (i.e. unique spatial units) was noted mainly as
an aspect of the structural properties of an LCA activity. What has become apparent since is
the more fundamental nature of the differences between LC-Type work and LC-Area work.
Through this paper and conference presentation two apparent patterns are investigated:
(a) That overlap between the ranges of significant spatial scales for LC-Type and LC-Area
work breaks down as one focuses upon the more local, micro-scale (i.e. sense of LC-Type is
lost), but less-so as one focuses upon the global, more macro-scale (i.e. sense of LC-Area is
preserved, as in LC-Areas such as The Dordogne, or The Weald).
(b) That “classification” has different roles, forms and degrees of prominence between LC-
Type and LC-Area work. In the former, scientifically formulated classification, in most of its
classic forms including its spatial extensions, has been the engine applied to the job of
systematising a breadth of landscape into a set of boxes. Though a variety of classifications
have yielded LC-Areas, the emphasis in much LC-Area work, particularly at the more local

568
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

scale, moves towards the organisation of information. Systematic classification work itself
has only a sub-ordinate role.
The earlier work (Groom et al. 2006) noted that whilst consideration of a broad range of
landscape factors is recognised and understandable for LC-Area work, the degree to which
landscape factors other from the biophysical ones have a role in LC-Type work is less clear.
This question is also re-examined through re-evaluation of LANMAP2 (see below). The aim
of this paper and presentation is to inquire into the nature of “LCA” as either just one, or else
two, or possibly several, distinct types of activity, and, if it is considered as not being merely
the former of these, then how best to develop clear understanding and communication of the
respective forms.

Stable vs. dynamic aspects of landscape units

The prevailing understanding of European LCA work was the basis for evaluation of
LANMAP2 (Wascher 2005). This had a number of consequences. Firstly, the fundamental
differences between LC-Types and LC-Areas were not, at that time, considered sufficiently,
with LANMAP2 evaluation comparisons made to examples of both. Restriction of comparison
to typologies (LC-Types work) could have provided a more objective evaluation. Secondly,
the comparative evaluation should have been applied to LC-Type work representative of
spatial scales most appropriate to the LANMAP2 mapping scale (1:1,000,000), in particular
landscape issues that are manifest at regional as well as national levels.
Follow-on evaluation exercises upon LANMAP2 that are reported by this paper /
conference presentation have aimed to address these earlier shortcomings. These have
involved comparison of LANMAP2 to two regionally well-documented landscape types, with
particular attention to the nature of the boundaries (e.g. sharp, straight indistinct) and the
respective area profiles. Comparisons are made firstly with respect to the biophysical factor
driven main layer of LANMAP2, and secondly to the land cover information layer of
LANMAP2, in order to investigate to question of the role of non-biophysical factors in LC-
Type work, which was noted above.
The issue of LCA-change with respect to landscape change was another key question
raised by the review of LCAs. The ways that the major national exercises in LCA mapping,
for LC-Types and LC-Areas, can accommodate changes in the patterns of landscape
character, due to environmental, land use or other changes is not clear. In this paper a
demonstration is made of how dynamism, based upon regular updates in the source data
can be incorporated into a temporal series of LANMAP products.

References
Groom, G; Wascher, D., Potschin, M. & Haines-Yong, R. (2006) Landscape character assessments
and fellow travellers across Europe : a review. R.G.H. Bunce & R.H.G. Jongman (Eds) Landscape
Ecology in the Mediterranean: inside and outside approaches. Proceedings of the European IALE
Conference 29March – 2 April 2005 Faro, Portugal. IALE Publication Series 3, pp.221-231.
Mücher, C.A.; Bunce, R.G.H., Jongman, R.H.G., Klijn, J.A., Koomen, A.J.M., Metzger, M.J.,
Wascher, D.M. (2003) Identification and characterisation of environments and landscapes in
Europe. (ALTERRA report 832). ALTERRA, Wageningen (NL)
Wascher, Dirk M. (2005) European Landscape Character Types – Typoloies, Cartography and
Indicators for the Assessment of Sustainable Landscapes. Final Project Report as deliverable
from the EU’s Accompanying Measure project “European Landscape Character Assessment
Initiative (ELCAI). (ALTERRA report 1254). ALTERRA, Wageningen (NL).
Mücher, C.A., Wascher, D.M., Klijn, J.A.,. Koomen, A.J.M, Jongman, R.H.G (2006) A new
European Landscape Map as an integrative framework for landscape character assessment.
R.G.H. Bunce and R.H.G. Jongman (Eds) Landscape Ecology in the Mediterranean: inside and
outside approaches. Proceedings of the European IALE Conference 29 March – 2 April 2005
Faro, Portugal. IALE Publication Series 3, pp. 233-243.

569
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

The role of landscape science in higher education in Europe: the ATLAS

experience.

G.B.M. Pedroli 1, H. Palang 2, M. Bogers1

1
Alterra Wageningen UR, P.O. Box 47, NL 6700 AA Wageningen, The Netherlands.
e-mail: [email protected]
2
Tallinn University, Uus-Sadama 5, EE 10120 Tallinn, Estonia

Landscape education and training: a fragmented provision

The fragmented nature of education and training provision in sustainability impact

assessment for landscape planning is a major barrier to the management of rapid land use
change. The European Spatial Development Perspective (Potsdam 1999) recognises the
challenges posed by European enlargement: while levels of environmental damage in
accession countries are declining due to declining manufacturing industry and successful
environmental protection policies, the future success of such policies depends on economic
prosperity. Large-scale abandonment of agricultural landscapes poses particular questions
for the environmental and social sustainability of rural populations. Sustainability Impact
Assessment Tools (SIAT) are currently being developed worldwide and also under different
nomenclature. In the European Union (EU) and especially in the European Commission the
term SIAT is favoured today and large research projects are developing such tools in
coordination with policy makers. Sustainable impact assessment tools consist of various
indicator systems and effect prediction instruments to determine the contribution to
sustainable development of changes in land use brought about by defined policies.

Inventory of educational provision over Europe: www.ATLAS-EU.org

If one focuses on an environmentally sustainable situation for European citizens, the

planning and management of landscape that contributes to sustainable development
involves the analysis of a number of complex issues. As a result, the skills and knowledge
required are often fragmented, so that the capacity of both professional and non-professional
groups to resolve sustainable land use issues is often limited. The EU funded project ATLAS
(Action for Training in Land use And Sustainability) analysed educational provision in a broad
sense from complete university courses to professional short courses and pays special
attention to training in landscape management taking particular account of the issues raised
by transition. The ATLAS website (www.atlas-eu.org) provides an interactive route planner
about courses and material available all over the 25 EU member states. One aim is to
enhance the demonstration of assessment tools that apply for instance to landscape and
landscape change, not just for the evaluation of major initiatives, but also to examine longer
term effects of predictable changes and changes that are policy driven.

The ATLAS project was designed to:

o take stock of what educational resources exist;
o assess their adequacy; and
o stimulate the development of appropriate strategies and initiatives for the future.

The status quo in relevant educational provision for policy and practice in this area through-
out Europe was analysed:
o through a comprehensive survey of the status of educational provision at
practitioner’s, professional, undergraduate and Master’s levels, within Europe
o followed by a SWOT-analysis (strengths, weaknesses, opportunities, threats) of the
extent to which this provision meets current needs, with clear recommendations for
improvement;

570
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

o leading to an interactive route planner for training in land use sustainability

assessment providing better European organisation of the educational provision
leading to appropriate professional qualifications.

Although ATLAS focuses on the EU, the results of this comprehensive overview of more than
three thousand courses is of relevance as well for the exposure of the European educational
provision to the rest of the world. Firstly, about a fourth of the courses explicitly address
landscape development issues worldwide. Secondly, more than half of the courses focus to a
high degree on methodological issues and on SIAT. Improved co-ordination of their efforts on
education and training would have great benefits for the effectiveness of the policies
currently developed. This will open up a large number of existing educational programmes to
enhance coordinated capacity building for policy makers in the field of land use management
practice, design, planning and research.

Conclusion: need for educational cooperation for proper covering of landscape issues

It is concluded from our analysis that, even though sustainability has been part of
European society already for a few decades, in all EU countries there is still an apparent
need for clearer definitions, practical guidelines and information of which aspects to consider
in implementation, especially related to landscape management and policy.
In Norway, Sweden, Denmark, the Netherlands, Belgium, France and the UK the situation in
sustainability teaching related to landscape issues seems to be better than in the other
countries. Thus, one could say that these countries form the core of educational provision.
The new EU countries, Czech Republic, Estonia, Latvia, Lithuania, Hungary, Poland,
Slovakia and Slovenia are still struggling with the different assessment tools and basic issues
related to sustainable development and landscape. On the whole, practical work and training
should be offered more. Teaching sustainability in landscape development should be carried
out in cooperation of different academic disciplines by professionals that are aware of the
aspects of sustainability in their fields. This is easier said than done, since the current trend
in science is more towards the opposite – fragmentation and specialisation.

Acknowledgement
The project was effectively implemented in strong mutual cooperation by: Thomas
Blaschke, Marion Bogers, Philippe Deuffic, Annelene Kammer, Grete Kukk, Carla Oonk,
Hannes Palang, Bas Pedroli, Marion Potschin, Ingrid Sarlöv-Herlin, Milada Stastná and
Gabor Turcsanyi

571
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

Defining and mapping the landscapes of Italy

C. Blasi1, D. Guida2, V. Siervo2, M. Paolanti3, L Michetti1, G. Capotorti1, D.

Smiraglia1
1
Dept of Plant Biology, University of Rome “La Sapienza”, P.le Aldo Moro 5, Rome Italy.
e-mail: [email protected]
2
Dept of Civil Engineering University of Salerno, Via Ponte Don Melillo 1, Fisciano (SA) Italy
3
CHOROS, Via Pietro Cuppari 33, 00134 Rome - Italy

Ecosystems are the result of a complex interaction of physical, social and economic
factors. Due to the importance of land management and biodiversity conservation,
ecosystems need to be described, characterised and spatially located (Sims et al., 1996).
Recently, ecosystem classification and mapping has received renewed attention, since its
relevance for understanding ecological patterns and processes and addressing
environmental tasks (Urban et al., 1987; Klijn and de Haes, 1994; Zonneveld, 1995; Metzger
et al., 2005; Jongman et al., 2006). Factors controlling pattern can be used for classifying
types as well as for mapping their boundaries (Bailey, 1996).
Spatial structure of landscape elements is recognized as a crucial factor in affecting
ecosystems functions (Forman, 1995; Turner et al., 2001) and its relevance is considered
also in the international legislation. With the Convention on biological diversity (Summit of the
Earth, Rio de Janeiro, 1992), the Habitats Directive (92/43/EEC) and the European
Landscape Convention (Council of Europe, Florence, 20.X.2000), the importance of
landscape diversity in the strategies of nature conservation is fully acknowledged. Moreover,
the Pan European Biodiversity and Landscape Strategy (1996) introduced the need to
protect Europe’s natural heritage and landscape diversity in order to protect and enhance the
natural environment. At present, the European Landscape Character Assessment Initiative
(ELCAI) represents the most important EU project for reviewing and documenting the state-
of-the-art of landscape character assessment techniques in Europe (Wascher, 2005). It can
be applied at a range of scales, from national to regional and local, and its importance to
support decision making about conservation and socio-economic goals is arising.
Within this context, a project for mapping the landscapes of Italy at broad scales
(1:500,000-1:1,000,000) was recently undertaken with the support of the Ministry of the
Environment of Italy, in order to provide a reference model at national scale. Classification
and mapping of landscape types was based on a hierarchical spatial framework developed
by Blasi et al. (2000), that integrates the basic physical aspects of the landscape. The
overlay of different environmental layers enables to define and map homogeneous units of
land characterised on climatic, lithological and geomorphological basis. The Climatic Regions
have been derived from an existing map (Blasi and Michetti, 2005). The proposed approach
employs real monthly data to define and map climatic types. A review of the geological maps
produced in Italy was necessary for the lithological final product and a new geomorphological
map was realised using an innovative method for landform identification, classification and
mapping. This method follows a new proposal of multi-scale geomorphological map legend.
The Lithological Map was produced from a systematic reinterpretation and homogenization
of existing regional and national geological units in term of multi-scale hierarchical lithological
units (IAEG-UNESCO, 1976). The following nested lithological entities was thus recognized
and mapped: Litho-System, adequate at scale 1.000.000 to 1:250.000 and national analysis
level, corresponding to Sequence in the IAEG-UNESCO classification (1976), as rock
succession with lithogenetical homogeneity; Litho-Complex, adequate at scale 1:250.000 to
50.000 and regional analysis level, as rock succession with litho-stratigraphical homogeneity
(i.e. arenaceous-conglomerate complex, coral reef carbonate complex); Litho-Type,
adequate at scale 1:50.000 to 1:25.000 and basin or local analysis level, as rock succession
with litho-technical or litho-pedological homogeneity (i.e. arkose sandstone or conglomerate
with clayey matrix). The Morphological Map, instead, was achieved using a semi-automatic
method from a digital elevation model with 75m resolution and implementing a step-by-step

572
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

procedure: i) basic morphometric analysis, ii) topographic position analysis; iv) grid-based
neighbourhood geomorphometric analysis, iv) object-based geomorphic analysis, and v)
grid-to-vector translation and comparison with sample areas maps carried out with the
traditional geomorphological approach. The following nested morphological entities was thus
recognized and mapped: Morpho-System (i.e. summit, hillslope, piedmont, plain, coastal
morphological systems), adequate at scale 1.000.000 to 1:250.000 and national analysis
level, with physiographic and morphogenetic homogeneity; Morpho-Facet, adequate at scale
1:250.000 to 50.000 and regional analysis level, as part of a Morpho-System with
morphological and morpho-evolutive homogeneity (i.e. talus, alluvial fan and scree in
Piedmont Morpho-System); Morpho-Type, adequate at scale 1:50.000 to 1:25.000 and basin
or local analysis level, as part of a Morpho-Facet with morphometric and morphodynamic
homogeneity (i.e. apex, mid-fan or proximal fan Morpho-Types in the Alluvial Fan Morpho-
Facet). Each Landscape Type, obtained from progressive overlapping of the above themes,
has been successively analyzed and characterised in terms of potential natural vegetation
and land cover, employing the 1:250,000 “Map of the vegetation series of Italy” (Blasi et al.,
2004) and the 1:100,000 CORINE Land Cover Map IV level.
This project highlights the role of physical determinism in characterising landscapes, as
pointed out in the European Landscape Convention. The land classification scheme provides
descriptive land units useful for survey, monitoring, management and for adequate
sustainable development initiatives. Furthermore, it provides a framework for comparing the
actual and potential heterogeneity of landscapes on a structural basis, in order to assess
their state of conservation.

References
Bayley, R.G. (1996) Ecosystem geography. Springer-Verlag, New York.
Blasi, C; Carranza, M.L; Frondoni, R. & Rosati, L. (2000) Ecosystem classification and mapping: a
proposal for Italian Landscape. Applied Vegetation Science 3:233-242.
Blasi, C; Filibeck, G; Frondoni, R; Rosati, L. & Smiraglia, D. (2004) The map of the vegetation
series of Italy. Fitosociologia 41(1) suppl. 1: 21-25.
Blasi, C. & Michetti, L. (2005) Biodiversità e clima C. Blasi; L. Boitani; S. La Posta; F. Manes & M.
Marchetti (Eds). Stato della Biodiversità in Italia. Palombi Editori, Roma, pp.57-66.
Forman, R.T.T. (1995) Land mosaics. The ecology of landscape and regions. Cambridge
University Press, Cambridge.
Jongman, R.H.G; Bunce, R.G.H; Metzger, M.J; Mücher, C.A; Howard, D.C. & Mateus, V.L. (2006)
Objectives and applications of a statistical environmental stratification of Europe. Landscape
Ecology 21: 409-419.
Klijn, F. & Udo de Haes, H.A. (1994) A hierarchical approach to ecosystems and its implications for
ecological land classification. Landscape Ecology 9(2):89-104.
Metzger, M.J; Bunce, R.G.H; Jongman, R.H.G; Mücher, C.A & Watkins, J.W. (2005) A climatic
stratification of the environment of Europe. Global Ecology and Biogeography 14: 549-563.
Sims, R.A; Corns, I.G.W. & Klinka, K. (1996) Global to local: ecological land classification.
Environmental Monitoring and Assessment 39:1-10.
Turner, M.G; Gardner, R.H., O’Neill R.V. (2001) Landscape Ecology in theory and practice. Pattern
and process. Springer, New York.
Urban, D.L; O’Neill, R.V., Shugart, H.H. (1987) Landscape ecology. A hierarchical perspective can
help scientists understand spatial patterns. Bioscience 37(2):119-127.
IAEG - UNESCO (1976) Engineering geological maps. A guide to their preparation. UNESCO Press,
Paris.
Wascher, D.M. (2005) European Landscape Character Areas. Typologies, Cartography and
Indicators for the Assessment of Sustainable Landscapes. Final Project Report as deliverable
from the EU’s Accompanying Measure project European Landscape Character Assessment
Initiative (ELCAI), funded under the 5th framework Programme on Energy, Environment and
Sustainable Development. Landscape Europe – ELCAI, Wageningen.
Zonneveld, I.S. (1995) Land Ecology. SPB Academic Publishing, Amsterdam.

573
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

Developing a science-based approach for delineating natural heritage system in

Southern Ontario

D. Puric-Mladenovic, S. Strobl

Information Management and Spatial Analysis Unit

Southern Science and Information Section, Ontario Ministry of Natural Resources
300 Water St., 4th Floor South Tower, Peterborough, ON, Canada K9J 8M5
e-mail: [email protected]

Introduction

In southern Ontario, natural areas are facing threats from unsustainable urban
development that fragment and degrade natural heritage features and functions. In order to
sustain ecosystems and quality of life in southern Ontario, we need to understand where the
critical natural areas that protect biodiversity, ecological functions and ecological services
are, so we can actively work to conserve them. In Ontario, planning authorities are required
to identify natural heritage systems (NHSs) as part of the land use planning process and
have generally applied Geographic Information System (GIS) analyses based on scoring
patches for several, often related, ecological criteria.
To provide a strategic basis for implementing a voluntary stewardship program to
encourage private landowners in southern Ontario to protect and restore natural areas we
were challenged to develop a science-based methodology to identify and delineate NHSs
that still considers expert input. We developed and tested an iterative spatial analysis
approach that: incorporates empirical targets for biodiversity and habitat protection based on
a priori gap analysis; utilizes the best available data and a mathematical algorithm to
optimize numerous targets; provides a number of NHS scenarios that can be comparatively
assessed and selected by experts; and is adaptable and repeatable over time as new
information becomes available.

Methods

The methodology was piloted in two ecodistricts, 7e5 and 6e6 representing distinct
climatic zones and differing amount of natural cover (14% for 7e5 and 36% for 6e6), land use
and amount of protected area.
Goals for the NHS included protection of the diversity of ecological communities,
sensitive surface water and groundwater features and aquatic and wildlife habitats. For each
ecodistrict, a gap analysis identified the current levels of biodiversity conservation (e.g.,
representation of natural vegetation communities), ecological function (e.g., total amount of
forest cover; amount of interior forest) and ecological services (e.g., amount of natural
vegetation in headwater and riparian areas). On the basis of available standard data layers
for each ecodistrict, empirical targets for over 60 objectives were developed e.g., conserve:
35 % of each vegetation community; natural vegetation within 75 % of the riparian buffer (30
m); 100 % of the forest interior (100 m from edge).
Available spatial information on biodiversity, including current, standard land
cover mapping was prepared, assembled and tessellated to 5-ha hexagon planning units
for the ecodistrict and a 2.5 km surrounding buffer. Since fine-scale vegetation maps
were not available we combined natural cover mapping with soil texture and drainage
mapping to derive a surrogate vegetation type mapping.
A simulated annealing algorithm available in MARXAN (Ball and Possingham 2000)
was used to find a system of spatially cohesive sites that met the suite of targets. The
program was run 100 times for each scenario, and units identified in 60 % of the time were
identified as a possible NHS. Several scenarios were run by varying the levels for targets
and the treatment of planning units (i.e., locked-in, preferred, or available) for each
ecodistrict. All scenarios excluded planning units with > 50 % urban or > 10 % road area.

574
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

Results and Discussion

In Ecodistrict 7e5 where natural cover is low (14 %) and with the more than 60 targets
set at levels identified by the Canadian Wildlife Service (2004), Scenario 7e5-S1 with
provincial parks (PP) locked in and previously evaluated natural heritage (NH) features
preferred resulted in an NHS that overlapped by 85 % (and identified similar total area) with
Scenario 7e5-S2 where all planning units were available.
In Ecodistrict 6e6 where natural cover is higher (36 %), however, scenario runs (6e6-
S1T4) with the levels for targets set the same as those used in Ecodistrict 7e5 demonstrated
greater sensitivity to planning unit treatment. In addition, setting the levels for targets the
same as those used in 7e5 identified an NHS that comprised a much larger proportion of the
ecodistrict (> 80 %) and likely unacceptable to planning authorities. The MARXAN model was
considerably more sensitive to the levels set for targets in Ecodistrict 6e6 since there were
considerably more hexagons that could meet targets (Scenarios 6e6-S1T2, S1T3, S1T5).
Thus, in ecodistricts with higher natural cover, there is both a greater need to carefully
consider target levels and the model’s sensitivity with locally-knowledgeable experts and a
greater need to have fine-scale vegetation composition and structure and key species habitat
mapping in order to set more specific targets for conservation and protection of biodiversity
and ecological functions.

Table 1: Percentage of targets achieved among scenarios for just 5 of the over 60 targets. All
scenarios excluded planning units with > 50 % urban or > 10 % road area.
Ecodistrict-Scenario 7e5-S1 7e5-S2 6e6-S1T2 6e6-S1T3 6e6-S1T4 6e6-S1T5
1
Planning Unit Treatment PP-L All PP-L PP-L PP-L PP-L
NH available NH NH NH NH
areas-P areas-P areas-P areas-P areas-P
% existing forests in NHS 65% 65% 40% 62% 99% 62%
% underrepresented forest in Not Not 29% 53% 96% 43%
NHS Tested Tested
Interior forests (100 m from 86% 88 % 45 % 71% 99% 62%
edge)
Interior forests (200 m from 100% 100% 23% 87% 99% 79%
edge)
Riparian forest (30 m) 67% 71% 26% 70% 100% 71%
Patch size > 200 ha 99% 99% 56% 80% 66% 66%
% of ecodistrict in NHS 39% 39 % 26% 41% 82% 57%
% of NHS capturing lands 24% 23% 11% 12% 17% 18%
intensively cultivated for
agriculture
1
PP - provincial parks; NH- previously evaluated natural heritage features; L= Locked-in; P=Preferred;

References
Ball, I.R., H.P. Possingham. 2000. MARXAN (V1.8.2): Marine Reserve Design Using Spatially
Explicit Annealing, a Manual.
Environment Canada. 2004. How Much Habitat is Enough?: A Framework for Guiding Habitat
Rehabilitation in Great Lakes Areas of Concern (Second Edition). The Canadian Wildlife Service
of Environment Canada. 80 pp.

575
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

Habitat opportunity mapping at the landscape scale in the UK

G.H. Griffiths1, I.N. Vogiatzakis2

1
Department of Geography, The University of Reading, Whiteknights, Reading, UK.
e-mail: [email protected]
2
Centre for Agri-Environmental Research (CAER), The University of Reading, Earley Gate,
Reading, UK

Introduction

The identification of suitable sites for habitat restoration and re-creation given limited
resources and competing land use is a challenging and urgent task given continuing
concerns about the loss of species and habitats. There is a long tradition of site selection for
nature conservation based upon a coherent and well understood set of criteria
(Ratcliffe,1977). However, there is mounting evidence that the protection of statutory sites is
failing to arrest habitat loss and species decline (Robinson & Sutherland, 2002). This,
combined with increasing public awareness of the importance of diverse and species-rich
landscapes has resulted, in recent years, in a shift from a site-centred approach focussed on
protected areas to a landscape scale approach.

This change of direction is being accompanied by increasingly sophisticated tools for

identifying sites for habitat restoration and re-creation, often using GIS (Store & Jokimaki,
2003). However, many of these spatially explicit tools fail to recognise the significance of
the landscape context. The paper suggests ways in which contrasts in the physical and
cultural attributes of the landscape can be accommodated into habitat models using
Landscape Character Assessment (LCA). We present and review a number of case studies
from the U.K. in contrasting landscapes ranging from lowland agricultural to upland.

Habitat potential and condition

In the heavily modified cultural landscapes of the UK and other parts of Europe, small-scale
differences in soil type, geology, landform, climate and cultural patterns need to be
accounted for in setting targets for biodiversity and identifying suitable sites for habitat
restoration and re-creation. Landscape Character Assessment provides a relatively simple
and objective system for dividing the landscape into homogeneous environmental strata or
landscape units. The resulting maps provide a convenient spatial framework within which to
set biodiversity targets based upon habitat potential and ecological condition.

Ecological condition is defined spatially in this context, a set of indices that measure the
extent and distribution of surviving semi-natural habitat within a landscape unit. Survival may
depend upon physical factors alone, such as slope and elevation or, equally importantly, it is
related to the historical evolution of the landscape. Frequently, it is the interaction of these
two forces that is important – physical factors limiting the habitat to sites that may also be
marginal for agriculture.

An important methodological challenge is to measure the difference between the potential of

a landscape unit to support a target habitat and its ecological condition. A unit with a high
potential for a specific habitat but with only a small area surviving, may indicate a significant
opportunity for habitat restoration and re-creation.

GIS habitat Models

Recent developments in ecology have moved the discipline away from the largely
descriptive to quantitative distribution models that formulate the species–habitat link to

576
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

predict where species and habitats should occur as well as improved understanding of the
factors involved (Hobbs & Morton 1999). Matrices indicating the likelihood of the presence of
a habitat (widespread or localised) in relation to landform (slope and elevation) and soil type
(fertility and drainage) were established for selected Annex I Priority Habitats and the results
mapped by landscape unit. In this way, maps of habitat distribution are intersected with
habitat potential to determine the extent to which habitat potential is realised in relation to the
extent of surviving habitat.

However, this landscape scale approach is only useful at a national scale: the identification
of optimal sites for a target habitat at the local scale based upon spatially explicit ecological
decision rules is equally important. A GIS habitat model has been developed that uses a set
of spatially explicit ecological decision rules to determine the potential suitability of a site for a
range of target habitats based upon the need to increase habitat area and reduce isolation.

Summary

The combination of these two approaches: setting biodiversity targets at the broad scale
based upon the ratio of the difference between habitat potential and condition, combined with
local-scale GIS habitat modelling is proving to be a powerful policy tool. In particular the
approach is being developed and tested in Wales with funding from the Countryside Council
for Wales and in other parts of the UK including the lowland heath restoration in the Midlands
and ecological sensitivity analysis in a southern county of England.

References
Hobbs R.J. and Morton, S.R. 1999: Moving from descriptive to predictive ecology. Agroforestry
Systems 45: 43-55
Ratcliffe D, 1977. Nature Conservation Review, Cambridge University Press, Cambridge.
Robinson, R.A. & Sutherland, W.J., 2002. Changes in arable farming and biodiversity in Great
Britain. Journal of Applied Ecology, 39, 157–176.
Store, R. and J. Jokimaki, 2003. A GIS-based multi-scale approach to habitat suitability modeling.
Ecological Modeling, 169: 1-15.

577
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

The assessment of landscape diversity for the purposes of spatial organisation of

tourism: the case study of Brest region

I. I. Schastnaya

Department of Geography, Belarus State University, av. Nezalezhnasti, 4

221050 Minsk, Belarus
e-mail: [email protected]

The concept of landscape diversity (LD) is a new research problem in landscape

science. The concept is meant in different lines, the most explored of these being the
traditional-landscape (classical) and anthropogenic ones (Grodzinski, 1999). The classical
interpretation of LD has been more widely developed and is related to the assessment of
uniqueness and mosaics of Natural Territorial Complexes (NTC).
Anthropogenic line of LD is not widely used concept. It is associated with a diversity of
technogenic and Natural-Anthropogenic Complexes (NAL). NAL dominate in our Republic.
They are the result of use of resources of a natural landscape in the certain types of human
activity and include both elements of natural complexes, and anthropogenic elements. NAL
depend on their development first on natural laws. At the same time attributes of their
functioning and dynamics are closely connected with social and economic conditions
(Schastnaya I.I., 2004). The type of economic use is a basic attribute and classification
feature of NALs. This factor can be described by land use distribution, which can be seen as
a premise for calculation and assessment of LD of NAL.
Classical and anthropogenic lines of the concept of LD are interconnected and mutually
complementary. The creation of various types of natural-technical systems is a necessary
element of the expansion of a society. Therefore, the main research problem is to maintain a
diverse and optimum proportion of NTC and anthropogenic complexes. It can be achieved as
a result of LD assessment, which is integrated parameter giving the information on the
system organization of a landscape. LD is an indirect indicator of quality of landscape
functioning (Martsinkevich, Schastnaya, 2005). There are many landscape indexes for
assessment of LD of natural and “natural-anthropogenic” complexes. The number of species
(vid) landscapes (or number of patches of landscape species) and their area are the basic
parameters of assessment of LD of natural complexes.
Calculation of indexes of a landscape diversity of NAL includes many indexes, first of
all:
1) Area and perimeters of patches,
2) Patch numbers within landscape (indicate landscape fragmentation)
3) Number of land use type
4) Proportion of different land use type within landscape
Si
Pi = , (1)
S
Si – area of particular land use type,
5) Mean area of patches within landscape.
S
So = , (2)
n
n – number of patches.

According to the results of analysis, the following indices have been selected -
Menchinik, Margalef, Shannon and focally an index of complexity. Natural territorial
complexes have been used as a unit for assessment. A map of NTC for Brest region has
been made (scale 1:100 000). Widely spread terracing and fluvioglacial landscapes have the
most complex horizontal structure. Unique flood-plain landscapes locate in the western part
of region. Secondary-moraine landscapes have simple structure and spread in northern part

578
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

of research area. The assessment of LD according to selected factors allows to define the
natural complexes having a high values of LD. Terracing and fluvioglacial landscapes have
the maximal and high values of Menchinik and Margalef indexes. The index of complexity
has defined the maximal degree of diversity of fluvioglacial complexes.
A map of natural-anthropogenic complexes of the study area (scale 1:100 000) has
been made. This map reflects the certain regional features of economic development of
territory (Marctsinkevich et al., 2002). The system of classification units of NAL includes three
levels: class - subclass – species (vid). Each level is based on primary indicators. The
highest unit of classification - a class NAL – is recognized by main direction of economic
activities in different branches of national economy. Agricultural and agro-forestry class of
NAL dominate in our Republic. The relative abundance of various type of land use within
landscape (arable, forest-arable, etc.) is the indicator for determination of NAL subclasses.
The species of a landscape considers a type of economic activities within natural complex
(for example, arable secondary-moraine). Agricultural, agro-forest and forest classes of
landscapes are defined in Brest region. Agricultural and agro-forestry landscapes dominate.
They have diverse horizontal structure and include 5 subclasses and 9 species for
agricultural landscapes and 3 subclasses and 12 species for agro-forest landscapes.
The analysis (according to the indices of Мenchinik, Margalef and the coefficient of
complexity) has shown that the diversity of natural and anthropogenic complexes differs.
Only their estimates on the Shannon index are similar to each other for both types of
landscapes. The assessment of LD helps to choose the sites having similar values of
diversity for both maps. These sites together with a network of settlements and areas with
high recreational potential for the region have formed a basis for the creation of a scheme of
spatial organisation of tourism. The sites of maximal and high landscape diversity, as
elements of a natural framework of territory, are recommended for the primary organisation
of all ecological, scientifically cognitive and partially recreational tourism. Areas with lower
values of LD can be used for organisation of recreational tourism with elements of agro-
tourism, rural and business tourism.
Results of research have confirmed that it is possible and necessary to use more
widely parameters of a landscape diversity of natural and natural-anthropogenic complexes
in spatial planning. Application of landscape indexes allows to recommend alternatives for
development of territory and to predict consequences of economic activities.

References:
Grodzinski M.D. (1999). Diversity of landscape diversity Landscape as an integrated concept of ХХІ,
Kiev State University Press: 34-35.
Schastnaya I.I. (2004). Natural-anthropogenic complexes of Mahilew district Geography of Mahilew
district, Mahilew State University Press. Mahilew: 174-180.
Martsinkevich G.I., Schastnaya I.I. (2005) The landscape diversity assessment of natural and
natural-anthropogenic complexes Environmental Management, 61: 98-205.
Martsinkevich G.I., Klitsunova N.K., Schastnaya I.I. (2002). Principles of natural-antropogenic
complexes classification Dynamic of landscapes and problems and conservation and sustainable
development of biodiversity, Belarusian State Pedagogic University Press: 90-91.

579
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

Landscape characterisation in Belgium: integration of different scale levels and

analysing temporal differences

V. Van Eetvelde, M. Antrop

Department of Geography, Ghent University, Krijgslaan 291 S8, 9000 Ghent, Belgium.
e-mail: [email protected]

Introduction

Stimulated by the European Landscape Convention, Landscape Character Assessment

(LCA) became popular. Most of the initiatives elaborated so far, identify landscapes using
classic methods of land(scape) classification, in particular the ones based on a parametric
approach. In these approaches, landscape character is defined through a combination of
different thematic maps, mostly using GIS overlaying, and is related to certain landscapes
types. The Landscape Convention also emphasises the importance of studying landscape
change, including the forces and pressures causing it (Council of Europe, 2000). Landscape
character is believed to be rather stable, while land use is a more dynamic property of
landscape. The core question is how much change in a specific data layer is needed to make
significant changes in landscape character. This is analysed for the landscape
characterisation in Belgium, using the CORINE land cover from 1990 and 2000.

Characterisation of the contemporary landscapes of Belgium

At the national level of the federal Belgian state, the landscape character typology of the
contemporary landscapes is based on four basic datasets that cover Belgium as a whole: a
digital terrain model, CORINE Land Cover 1990, a soil map and a Landsat 5 TM satellite
image. A hierarchical parametric classification method is used, as well as geostatistics and
GIS (Van Eetvelde, 2006, Van Eetvelde et al., 2006). The typology has two scale levels. At
the first scale, landscape types were assigned to a grid with 31473 square kilometre cells
using 18 variables which were derived from the four datasets. The variables were used in a
k-means cluster analysis to define 48 landscape types. At the second scale level, 222
landscape character areas are delineated by a manual holistic interpretation of the spatial
patterns formed by the landscape types of the first scale level. The spatial patterns in these
areas are described using landscape metrics. Landscape types and pattern characteristics
are subsequently used in a hierarchical cluster analysis to define 54 landscape character
types at the second scale level.

Methods and materials

CORINE Land Cover is the only available dataset in the typology that gives temporal
variation. The land cover changes were analysed between 1990 and 2000. First, a transition
matrix was made for the whole of Belgium between 6 main groups of land cover used in the
typology at a resolution of 10ha. These groups were urban fabric, industrial units, arable
land, pastures, forest and semi-natural areas, and water bodies. In this paper, the focus goes
to the urban land use to analysis the problem.
Second, the land cover changes of these groups were defined for each kilometer cell at
the first scale level of the typology. The relative changes were used to detect cells where
probably change in landscape type could occur.

Results

Looking at the overall changes of land cover in the transition matrix, some differences can
not be explained by real changes in the landscape. For example, according to CORINE,

580
Theme 5: Monitoring and classification
5.2 Symposium 25: Advances and Applications of Landscape Character Mapping

urban fabric is mainly transformed to arable land and pastures, industry changed into forest,
and water bodies changed into arable land or forest. This is illustrated in figure 1A, were
4090 cells (13%) have a loss of urban fabric, which is highly improbably to be a real
landscape change. These differences are due to differences in method and technology used
in the two CORINE inventories. CORINE 2000 is an update of 1990 and is results in a more
accurate and detailed classification, resulting in less generalised and smaller patches.

For the CORINE category of urban fabric, 8513 (27%) of the cells show an increase since
1990. 1264 (4%) of the cells are classified as urban in the first scale level, but will not affect a
change in their landscape type or character. Figure 1B shows 7249 cells (23%) which have a
growth of the urban fabric and are classified as non-urban landscape type in the typology.
Only 18 of these cells, indicated in black, have an increase of more than 33% urban fabric
and could potentially change the landscape character of the cell into the urban type.

Figure 1. (A) cells with loss of urban fabric; (B) possible changes in landscape type, based
on land cover changes of urban fabric between 1990 en 2000.

Conclusions

Land cover is the only data layer in the typology that could identify temporal differences of
the landscape character of the kilometre cells at the first scale level. However, most of the
changes are caused by differences in data quality and not by real landscape changes, but it
is possible to identify cells with potential changes in landscape character.

References
Council of Europe (2000) European Landscape Convention and Explanatory Report. Council of
Europe, Document by the Secretary General established by the General Directorate of Education,
Culture, Sport and Youth, and Environment.
Van Eetvelde, V. (2006) Van geografische strekenkaart tot landschapsdatabank. Gebruik van GIS,
informatietheorie en landschapsmetrieken voor het karakteriseren van landschappen, toegepast
op België. PhD thesis, Universiteit Gent, Vakgroep Geografie, Gent.
Van Eetvelde, V.; Sevenant, M. & Antrop, M. (2006) Trans-regional landscape characterization: the
example of Belgium. R.G.H. Bunce & R.H.G. Jongman (Eds.). Landscape Ecology in the
Mediterranean: inside and outside approaches. Proceedings of the European IALE Conference,
29 March - 2 April 2005, Faro, Portugal. IALE Publication Series 3, pp. 199-212.

581
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

Open Session 2: Landscape modelling and earth observation

Problem of characteristic space scale of landscape processes

A.V. Khoroshev, G.M. Aleshchenko, K.A. Merekalova

Moscow State University, Faculty of Geography, Vorobyovy Gory, 119992, Moscow, Russia.
e-mail: [email protected]
.
Introduction
At the moment there is common understanding of the fact that geographical and
ecological processes are scale-specific. Multidimensional systems, such as landscape are
often referred to, are being involved simultaneously in multiple interactions at a certain
number of levels. The concept of multiscale organization is recognized as a critical idea of
modern landscape study. Most studies of the scale problem in landscape ecology are based
either on remote sensing or on digital elevation models. Hence, the establishment of
hierarchy is realized for only one of the landscape components, plant cover or relief
respectively, assuming that it is perfectly indicative of driving forces that operate in the
landscape. However, since landscape is a holistic entity, its distinguishing properties are
determined by interaction of components. Results of interactions (e.g. between landforms
and plant cover) can vary across a heterogeneous landscape, and occasional events can
cause bifurcations of structural properties. We hypothesize, that two types of landscape
patterns should be distinguished in nature: the first based on uniformity of a set of landscape
properties, the second based on uniformity of relations between components in
heterogeneous environment. New data storage systems show that the concept of one-to-one
correspondence between landforms and plant cover should be considered as a particular
case of the more general regularity, namely multi-structure organization of landscape. The
concept of multi-structure organization of landscape assumes that relatively independent
systems are controlled by factors of different origin that can co-exist in space and through
time. The core of the multi-structure concept is multiplicity of driving factors of spatial
organization at each hierarchical level.

Materials and methods

The case study is situated in the central taiga region of European Russia (the
Arkhangelsk region). We modeled types of relations between moisture-sensitive properties of
plant cover (evaluated from space images) and drainage conditions (calculated from digital
elevation model). To reveal specific scale of landscape processes we developed novel
approach based on multiple regression modeling that enables the comparision of types and
the strength of relations between landscape components across the area. We understand
the concept of “type of relations” as follows: 1) a set of independent variables (e.g. relief
morphometry) that affects the value of the dependent variable (e.g. groundwater level); 2) the
relative contributions of independent variables to spatial variability of dependent variables; 3)
the degree to which the dependent variable is determined by the integrated effect of the
independent ones (r2). The challenge is to assess the diversity of factors that organize the
territory under study at the hierarchical level. The territory is represented as a set of
intersecting squares. The size of squares is user-defined and related to the hypothesized
characteristic space with uniform type of between-component relations.
The purpose is to reveal, whether the spatial diversity of the given area can be explained
by the diversity of a single driving force values. If this holds true, then land use decisions
should be adapted to anthropogenic loadings and landscape pattern taking into account the
gradient of the principal driving force. For example, if bioproductivity is proved to vary
continuously within the territory depending on groundwater level, land reclamation measures
should vary according to the model of relations between plant cover and groundwater.
Our approach aims at “bottom-up” determination of hierarchical levels using multiscale
analysis of between-component relations. We do not identify levels based on a single binding

582
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

factor, but generate levels based on relations between components. The technique includes
multiple regression modeling of relations in a moving square window by means of
consequent examination of various combinations of scale parameters (extent and cell size)
until “resonance” combinations are found. “Resonance effect” means that the agreed spatial
variability of a set of landscape attributes under certain combination of scale parameters
indicated by high r2 value. The Operational Unit (OU) is characterized not only by its own
properties (e.g. species composition, microrelief) but also by properties imposed by the
higher-level unit (e.g. drainage conditions, mesorelief). When we increase moving window
size we consequently evaluate relative importance of higher order units (levels “+1”, “+2”
etc.) for the focus unit (level “0”) properties. A set of OU within a moving window is used to
design a statistical model that relates properties of OU to properties of higher-order unit
(level “+1”). If a set of OU within moving window covers the territory comparable to the
landscape level, then the model answers the question: do the relations between properties of
OU (level “0”) and level “+1” within the landscape (level “+2”) belong to one type or to many
types? Comparing accuracy of models designed for different window sizes one can assess
adequacy of the environment for subordination of between-component relations.
Changing linear dimension of OU (from 30 m to 400 m in this study) we pose the question
as follows. If some law of relations “level “0” – level “+1” within the level “+2” is proved to
exist, whether the same law describes relations in the system “level “+1”-level”+2”? This
analysis clarifies, whether the relations at adjacent levels are self-similar or driven by
different factors. Processes responsible for differentiation at each level can be revealed using
a combination of regression coefficients. To verify the results of multiple regression modeling
Spearman nonparametric correlations, Jacobian determinant and information measures were
calculated following the same routine.

Results and conclusion

Resonance effects are multiple at most of the territory, showing that between-component
relations also operate at multiple hierarchical levels. At the level 2 (OU 400 m) moisture-
sensitive properties of plant cover are related to drainage effects in larger areas as compared
to the level 1 (OU 30 m). At the level 2 the linkage is realized in the most perfect way in
transitional zones between flat watersheds and gentle upper portions of river valley slopes.
Linearity of relations is inherent for space with linear dimension 5 times as long (2000 m) as
OU (400 m). This holds true for more than a half of the territory. Recalculation of the model
for increased moving window (more than 2000 m) detected that linearity disappeared on the
most territory. Normally, in the valleys r2 decreases as extent increases. Thus, diversity of
relations in the “relief – plant cover” system is greater in the valleys in comparison with
watersheds, i.e. characteristic space scale of relations decreases in valleys. At the level 2
relations are revealed distinctly for the square environment with a linear dimension of 2000
m. In larger environments it is replaced by other types of relations, i.e. a shift to another scale
results in change of driving force. However, in certain areas with mean diameter 3-4 km the
type of relations is repeated at various hierarchical levels. We identified two types of
locations: with self-similarity of relations at different hierarchical levels and with scale-
dependent process. While comparing parametric and nonparametric calculations we
revealed a particular scale at which linear relations are replaced by non-linear ones. For the
fragment of the watershed the type of relations remains the same if linear dimension of
moving window increases from 2000 m to 5200 m, but changes abruptly under linear
dimension 6800 m accompanied by critical decrease of determination coefficient r2. This
indicates disappearance of linearity. At the same time, nonparametric Spearman correlations
are significant. This proves that the functional linkage between moisture factors and drainage
conditions is significant, but linearity is inherent for the scale domain 2000-5200 m being
replaced by non-linearity in broader scale.

583
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

Using between-patch boundaries parameters for conservation status assessment

on coastal dune ecosystems

M.L. Carranza¹, S. Feola2, A. Acosta2, A. Stanisci¹

¹Dip. STAT. University of Molise. Contrada Fonte Lappone - 86170 Pesche (IS), Italy.
e-mail: [email protected]
² Dip. of Biology. University of “Roma Tre”, V.le Marconi 446 - 00146 Roma (RM), Italy.

Introduction
Coastal dune systems are natural formations characterized by a strong sea-inland
zonation. In such conditions of landscape sustainability is greatly influenced by the specific
contiguity between natural habitats patches.
In the present work we propose to use an adjacency matrix for landscape analysis and
conservation status assessment on coastal dune ecosystems. Specifically we analyze the
variation in the number and extension of boundaries between adjacent land cover categories
using the Rènyi parametric generalized entropy profile. The study was carried out based on
detailed CORINE land cover maps of the Molise Region coast, central Italy.

Materials and Methods

A detailed land cover map (scale 1:5,000) of the Molise coast (central Italy) was used
covering a 500-m-wide strip starting from the coastline towards inland. Land cover was
manually interpreted on video, with the help of a Geographic Information System (ArcView
3.1. ESRI 2000) and field survey. The legend followed CORINE land cover (Anon. 1993)
expanded to a fourth level of detail for natural and semi-natural areas (Acosta et al. 2005). 22
land cover typologies were identified and mapped.
Then, three 1 km x 350 m. windows with different disturbance regimes were analyzed and
compared. For each window, number and extension of the edges between each pair of land
cover categories were calculated and synthesized in an adjacency matrix.
In order to analyze differences in landscape diversity, the variation in the number and
extension of boundaries between adjacent land cover categories (patches) from the three
windows were evaluated. To do this, for each window a Rènyi parametric generalized
entropy profile based on the number and extension of the boundaries was calculated.
Boundaries were applied because they could be more informative than the use of patches.
Rényi’s generalized entropy represents the cornerstone for a continuum of possible diversity
measures. In fact, for a distribution function characterized by its proportional abundance
pi=(p1, p2, ..., pN) Rényi’s (1970) extended the concept of Shannon’s information (entropy)
N
1
Hα = log ∑ piα
defining a generalized entropy of order (α) as 1−α i =1 Where 0 ≥α≥∞ and pi
denote the relative abundance of the ith element in a system (i=1, 2, ..., N) such that 0≤pi≤1
∑
N
i
i =1
p =1
and . Note that in our context N denotes the total number of boundary types for
each data and ni is the total extension of boundaries belonging to the ith boundary type.

Results and discussion

Rènyi’s diversity profiles vary from α= 0 up to α= 10 in the three compared windows, as

shown in Fig. 1. From Campomarino (W3) to Frantoio di pietrisco (W2), diversity curves
show an increasing number of boundary types with higher richness (α=0), higher diversity
values according to Shannon’s entropy index (α =1), and higher evenness. Note that no
curve cross or touch any other.
Results concerning the increasing richness, Shannon’s diversity values and evenness on
Frantoio di pietrisco area (W2), were related to an increase in landscape fragmentation. This
area is characterized by a diffuse coexistence of artificial and natural areas.

584
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

On the other hand, Lido di campomarino area was very simple (W3), as expressed by
very low richness and diversity and dominance by few contact types related to consistently
human pressure. This window is actually characterized by extensive urban areas in the
foredunes and by high tourist pressure and managed dunes.
In our case, the intermediate values of Ramitelli area indicate a quite good conservation
status. In fact, this area is inside a pSCI because of the integrity of its coastal dune
vegetation zonation.

4,5
Lido di Campomarino
4 Frantoio di pietrisco
Ramitelli
3,5

2,5

1,5
0 1 2 3 4 5 6 7 8 9 10

Figure 1. Rènyi's diversity profiles (Hα) vs α of the edges in the different compared windows
extending from α=0 to α=10. Frantoio di Pietrisco (W1), Ramitelli (W2), Lido di
Campomarino (W3)

Conclusions
In the analyzed areas, results showed a good relation between adjacency matrices
information and conservation status. In natural conditions few adjacency types, most of them
between natural areas, dominated the landscape. Moderately disturbed coastal areas were
characterized by a fragmented landscape with high number of adjacency types. By contrast,
in highly disturbed sites there were a few dominant adjacency types.
The major advantage in applying the Rènyi generalized parametric diversity function to
compare landscape mosaics is that diversity profiles display not just a single index but a
family of indices, many of which currently applied and widely used in landscape ecology. In
this way, profiles allow a complete summarization of trends in landscape richness and
dominance concentration.
Since it was based on standard land cover classification, the proposed method could
represent a good tool in planning issues regarding coastal dune areas in many other
European countries.

References
Anon. (1993). CORINE Land Cover. Guide technique. CECA-CEE-CEEA. Bruxelles.
Acosta A.; Carranza M.L.; Izzi F. (2005) Combining Land cover mapping with coastal dune
vegetation analyses. Applied Vegetation Science. 8: 133-138.
ESRI. (2000). Arc-View 3.1. CA, US.
Rènyi, A. (1970). Probability theory. North Holland Publishing Company. Amsterdam.

585
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

Landscape-dynamical analysis: studies in the North-Western Russia

G.A. Isachenko

University of St. Petersburg,

Research Institute of Geography; 33, 10 line, V.O., 199178, St.-Petersburg, Russia
e-mail: [email protected]

By the end of the 20th century the necessity of synthesis of morphologic («static») and
functional-dynamic approaches in landscape studies has been clearly recognised. Such
studies have been created in the different centers of landscape science and in many
respects were initiated by the development of applications of a landscape approach:
landscape planning and design, environmental assessment etc. The author and his
colleagues the from Laboratory of Landscape Research and Ecological Mapping (University
of St. Petersburg) since 1990 have carried out field studies of landscapes in the north-
western part of the European Russia, within the limits of taiga and hemi-boreal forest zone.
On the basis of a 15-year research programme the concept of the landscape-dynamic
analysis has been developed. The main substantive principles can be formulated as follows:

1. In each landscape (geocomplex, natural territorial complex) different components

and elements have different characteristic time (time of full change of object or time of one
full cycle at cyclic character of changes).

2. Depending on the characteristic of time, in any elementary landscape (geocomplex)

the stable part, or a site, and a dynamic part which is described by a set of states of different
duration (from diurnal up to long-term ones) are recognized. The site is described by the
basic elements as a form of a relief and the upper layer of pedogen (soil-forming) bedrock. It
was established based on extensive of evidence that the basic morphological features of
relief and the characteristics of the upper layer of bedrocks in identical climatic conditions
unequivocally cause a character and degree of moistening regime (or conditions of natural
drainage) of sites and a mode of migration of substances. Characteristics of states
(concerning basically vegetation and some soil properties) change 1-3 orders more slowly,
than attributes of sites. Borders of geocomplexes are drawn on the maps first of all as
borders of sites. The characteristics of environment vary more strongly between sites, than
within the limits of sites (= geocomplexes).

3. The dynamics of a geocomplex is understood as a sequence of all the states of

different duration, and also the set of transitions between the states. Transitions which have
certain duration can be examined as states (using a longer time scale). It is possible to speak
about short-, middle-and long-term dynamics of geocomplexes (correspondingly states
duration is less than 1 year, from 1 year up to 10 years, tens and then hundreds years).

4. Irreversible changes of a relief and substratum (upper layer of pedogen bedrock)

under impact of processes with characteristic time, as a rule, are more than 1000 years and
may be considered as the evolution of geocomplexes. It should be pointed out that a
geocomplex at a local level can transform in other geocomplex as a result of catastrophic
processes (earthquakes, landslides etc) or strong technogenic impacts (e.g. open mining).

5. Processes in landscapes and its components have a different causality. Some

processes have spontaneous character, i.e. occur without any human participation, and
sometimes without an opportunity of such participation. Spontaneous processes can be
caused by exogenic (e.g. the development of bogs under influence of climate changes) or
endogenic factors (the development of bogs due to neo-tectonic lowering of territory). Some
processes are determined by self-development of separate natural bodies: for example,
overgrowing of the surfaces which have been freed from the water. The majority of

586
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

spontaneous processes are caused by simultaneous action of several external factors. An

impact is understood as the event caused by external (natural and/or anthropogenic) factors
and causing relatively fast change of a state of geocomplex. Three basic groups of impacts
on geocomplexes can be determined: point-source, linear and areal impacts. Anthropogenic
influences can be short-term, playing the role of an initial push with subsequent «start» of
spontaneous processes (e.g. natural forest regeneration after clear cutting), or long-term
(ploughing up the territory and its use as an arable land).

6. The character and intensity of human impacts on a landscape in each historical

period are determined by set of the economic, social, political, ethnic factors realized in
regional system of landscape management. To study the modern state and trends of change
of any geocomplex, it is necessary to analyze changes at previous historical periods.

7. Concerning processes it is necessary to note two basic moments. First, the

superposition of different processes always takes place. So, forest regeneration on clearings
is complicated and slowed down by periodic local fires, overgrowing by herbs, partial bogging
etc. Second, processes of different causes can have similar character of manifestation in
landscapes. As an example we can mention forest fires which result in similar consequences
irrespective of the reason of the concrete fire.

8. Each impact can be considered as a starting point of the subsequent dynamic

trajectories of a geocomplex. The trajectory represents a sequence of long-term states. We
can not always indicate with hundred-per-cent probability, in what direction the given
geocomplex will change after the impact. In contrast, the number of possible trajectories of
the given geocomplex as a result of any impact (or during its realization) usually exceeds
one. The set of possible trajectories is less than the connections in a landscape which are
more rigid and than the set of plant and animal species which can occupy the released
ecological niches. In the taiga of North-West Russia the lowest variety of dynamic trajectories
is typical of geocomplexes of extreme sites e.g. granite rocks and oligotrophic bogs.

9. Every superposition of the impacts complicates a dynamic trajectory of a

geocomplex and brings in it various «lateral branches». Our studies show, that the number of
additional branches is not infinite as the amount of influencing factors in nature exceeds the
number of possible reactions to impacts. Any type of a landscape at any period of time is
represented in space by various states (or their modifications) of one or several dynamic
trajectories (e.g. different stages of coniferous forest regeneration after different-time
clearings and fires).

Hence, the key idea of landscape-dynamic analysis is to divide the characteristics of

basic landscapes into two categories: site characteristics and state characteristics. Sites are
relatively stable in time, states are «mobile» and change due to spontaneous processes and
numerous impacts. Both sites and states can be classified and typified. Typology of sites
gives a basis for landscape-dynamic mapping. We have developed a typology of sites for
landscapes of the taiga in North-West European Russia, including 30 main types of sites.
This typology is applied to landscape-dynamic mapping in scales varying from 1: 5000 up to
1: 500 000.

587
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

Spatial modelling of landscape patterns derived from land use and land cover
changes

J. P. Fernandes, N. Guiomar

CEEM, Universidade de Évora, Rua Romão Ramalho, 59 – 7002-554 Évora, Portugal

e-mail: [email protected]

Introduction

Land use is the instantaneous expression of the best solution of the equation operating
environmental functions such as production, regulation and information, social economical
functions such as land use history, available techniques and investment capacity culture and
traditions. This implies that the solutions vary along time and throughout space according to
natural conditions and social, economic and cultural factors.
The representation of this equation through a spatial model is of major importance in
order to evaluate the impact of land use activities on natural and cultural resources. At the
same time, the possibility of forecasting such spatial land use distributions according to
development scenarios allows the evaluation of incremental and cumulative effects of the
diverse activities.
The analysis of environmental and social-economical factors that determine or lead to a
given land use system and therefore to a given landscape, is developed in order to
determine the existence of regular patterns, allowing the development of simulation models
and assessing the relative weight of each factor throughout space and time.
This evaluation makes use of geo-statistical analysis of the eventual dependences of land
use on environmental factors. A detailed analysis of the regional land use systems and
history is also performed in order to identify spatial and temporal patterns of influence or
interdependence as well as its inertia. This process allows the identification of areas with
homogeneous behavior.
On the basis of these homogeneous areas stochastic land use simulation models will be
developed according to socio-economical development scenarios.

Conceptual framework and methodology

Land use patterns can be interpreted and analyzed from many perspectives. For
example, they can, to a greater or lesser degree, provide information about landscape
function, economic opportunity, and environmental amenities (Zube, 1987). To understand
the structure, function, and dynamics of ecosystems it is necessary to integrate both
ecological and human processes (Jennerette & Wu, 2001). The distribution of landscape
patterns is strongly influence by environmental discontinuity, human persecutions and other
social activities (de la Ville et al., 1998).
Applying transition probability models and geostatistical descriptions to generate
simulations of future landscape patterns requires one to assume some degree of temporal
stationarity in the process (i.e., that the types of locations that change the and patterns of
change will be constant over time). Furthermore, applying the models generated in one place
to another place would require the assumption of spatial stationarity (Brown et al., 2002).
Modelling the relationships between land use, landscape structure, land policies and
biophysical environment provides a framework structured fro decision making. This is
particularly important to the evaluation of conservation measures for given targets species
faces the difficulty that those species demand particular spatial habitat arrangements, factor
of particular complexity when we deal with agricultural landscapes and complementary
habitat demands. For example, when considering alternative rotations or different
combinations of production, the resulting spatial patterns have different impacts on the
populations of those species which need to be previously assessed.

588
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

This issue is of particular importance when agriculture land use factors are changing
rapidly. These changes determine that evaluation instruments of foreseeable impacts on
particularly threatened species will constitute a critical instrument for future conservation
policies in agricultural environments. Based on economic simulations of foreseeable land use
changes in the area of Castro Verde (southern Portugal) the agricultural use of individual
parcels is simulated through random allocation of the different components of alternative
rotations (traditional and conservation-aimed). This detailed spatialisation of the land
management patterns allows the evaluation of habitat suitability for different target species,
based on the particular demands of those species to given spatial arrangements. The results
were then analysed in order to explain variations between the occurrence of some of those
species and predictions obtained trough traditional Habitat Suitability Models.

Figure 1. Patterns of land use and land cover change analysis

References
Brown, D. G.; Goovaerts, P.; Burnicky, A.; & Li, M.-Y. (2002) Stochastic simulation of land-cover
change with geostatistics and generalized additive models. Photogrammetric Engineering &
Remote Sensing, 68(10): 1051-1061.
de la Ville, N.; Cousins, S. H.; & Bird, C. (1992) Habitat suitability analysis using logistic regression
and GIS to outline potential areas for conservation of the Grey Wolf (Canis lupus). S. Carver
(Eds). Innovation in GIS, Taylor & Francis.
Jenerette, G. D.; & Wu, J. (2001) Analysis and simulation of land use change in the Central Arizona –
Phoenix Region, USA. Landscape Ecology, 16: 611-626.
Zube, E. H. (1987) Perceived land use patterns and landscape values. Landscape Ecology, 1(1): 37-
45.

589
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

The application of satellite imagery to identify landscape structure

C.A. Mücher1, C.C. Vos1, C. Renetzeder2, T. Wrbka2, M. Kiers1, M. van Eupen1, R.

Bugter1
1
Alterra, Box 47, 6700 AA Wageningen, The Netherlands.
e-mail: [email protected]
2
University of Vienna, Department of Conservation Biology, vegetation and Landscape
Ecology, Althanstraße 14, 1090 Vienna, Austria.

Introduction

Habitat fragmentation is widely recognized as one of the major causes for the loss of
biodiversity. Spatial cohesion (Opdam et al. 2003) is a sustainability indicator to determine
whether the size and connectivity of ecosystem networks is sufficient for sustainable
biodiversity protection. Spatial cohesion is one of the biodiversity indicators used within the
SENSOR project of the Sixth Framework Programme. The landscape permeability of cultural
landscapes determines for a large part the spatial cohesion for specific species groups. The
landscape permeability is especially related to the land use and landscape elements present
within a specific landscape. The presence (veining) of all kinds of semi-natural (green)
elements within a cultural landscape is also referred to as green veining, and offers refuges
and corridors for a large array of species. The Greenveins project (www.greenveins.nl)
indicated relationships between the amount of green veining, the landscape structure, the
intensity of agricultural land use and biodiversity. However, actual information on landscape
structure and amount of green veining does often not exist for many areas. Many landscape
elements can be measured directly by the use of aerial photographs or very high resolution
satellite data with spatial resolutions below one meter (Mücher et al. 2001). However the
disadvantage of such satellite imagery is that it is mostly used for small areas and certainly
not throughout Europe. Although high resolution satellite data such as Landsat or SPOT, with
spatial resolutions between 10 and 25 meters for multispectral data do cover the globe and
are in general cheaper to obtain, they do have a limitation in that most linear and small
landscape elements cannot be detected directly. The hypothesis is that measuring landscape
structure might be a good substitute for the amount of green veining. The major objective of
this paper is to show the applicability of satellite imagery to identify landscape structure and
to show that there exists a relationship between the satellite derived landscape structure and
the amount of green veining in a landscape. Special attention is given to the segmentation of
Landsat 7 ETM+ satellite imagery of the Image2000 database (image2000.jrc.it) that covers
almost the whole of Europe.

Materials

Image2000 was produced from ETM+ Landsat 7 satellite, providing both multi-spectral (6
bands and 25 m spatial resolution) and panchromatic data (1 band and 12.5 m spatial
resolution) for the reference year 2000. Image2000 products cover Europe almost entirely
(image2000.jrc.it) and are intended to be the main data source for updating the CORINE land
cover database. However, the CORINE land cover database provides very limited
information about the landscape structure and therefore the Landsat satellite images were
segmented to provide additional information about the landscape structure.

Methodology

For the derivation of the SENSOR biodiversity indicator “Spatial Cohesion” at a regional level
we need information about the landscape structure – which we want to derive by
segmentation of Landsat-TM satellite imagery. Although the IMAGE2000 database covers
the EU28 it was not be feasible within the project to derive the landscape structure for the

590
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

whole of Europe. Therefore a sampling strategy was needed. Stratified random sampling is
obvious choice to take representative samples over entire Europe. Since the SSRF (Spatial
Regional Reference Framework) cluster regions are determining the spatial reference
framework within the SENSOR project it seemed to be logical to use the SRRF clusters
(Renetzeder et al, 2007) as the strata for the sampling. Finally, ten cluster regions were
selected randomly. For each cluster regions three samples of 50 by 50 km (using the
INSPIRE spatial grid) were selected randomly to describe the variation within the cluster
region. So, in total we selected 30 samples, each covering an area of 50 by 50 km. For the
segmentation of satellite images to determine the landscape structure in terms of individual
field parcels there is a wide variety of software packages available. At this stage eCognition
is still one of the advanced software packages for segmentation and classification of satellite
images and was used within this study. The selection of the best parameters thresholds in
eCognition is the most difficult task and is to a large degree depending on experience and
“trial & error”.

Results and discussion

The best segmentation results of all thirty European sample sites were found with all six
Landsat bands and a scale factor of 30 and a shape factor of 0 (these are eCognition
parameter settings). Segmentation results are directly exported as shapefiles. The land
cover information was derived from the CORINE land cover database (CLC2000) and was
labelled as an attribute to the shapefiles. These results are now available for all thirty
European samples. Fragstats will be used to calculate specific landscape metrics that
describe the landscape structure in the best manner. The correlation between the specific
landscape metrics and the amount of landscape elements (focussing especially on
hedgerows, lines of trees, and small woodlands) – as available within the 25 Greenvein sites
and the digital topographic map of the Netherlands (Top10-vector) – still has to be
determined. If there exists a significant correlation between specific landscape metrics as
derived by the above mentioned methodology and specific landscape elements for the whole
Europe or for example biogeographical regions, these results can be used to assess the
amount of greenveining within the thirty SENSOR samples to assess finally the spatial
cohesion of these areas for specific species groups.

References
Mücher, C.A; Thunnissen, H.A.M.; de Bont, C.; Clement, J.; Kramer, H. & Koomen, H.J.M.
(2001). Toepassing IKONOS satellietbeelden in het Meetnet Landschap. BCRS rapport 01-40,
Delft.
Opdam, P.; Verboom, J. & Pouwels, R. (2003). Landscape cohesion: an index for the conservation
potential of landscapes for biodiversity. Landscape ecology 18, 113-126.
Renetzder, C.; van Eupen, M.; Mücher, C.A. & Wrbka, T. (2007). Linking landscape characterics
and socio-economic profiles for sustainable impact assessment at the regional level – the spatial
reference framework (SRRF). To be published in the IALE world conference proceedings 2007,
Wageningen, the Netherlands.

591
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

Identification of geosystem formative factors on the basis of field and remote

sensing data

D.N. Kozlov1, M.Yu.Puzachenko2, M.V. Fediaeva1, Yu.G.Puzachenko2

1
Lomonosov Moscow State University Faculty of Geography, 119899, Moscow,
Leninskiye Gory 1, Russia,
e-mail: [email protected]
2
A.N. Severtsov Institute of Ecology and Evolution RAS, 119071, Moscow,
Leninsky prospect 33, Russia

In spite of all inconsistencies of various scientific interpretations, the term “landscape

forming factors” mean in landscape science the properties of components which determine
endogenous character and intensity of interaction. Among these important properties are
climate, geological structure, relief, biotic and soil features. While researching a specific
territory the main task of a researcher is to reveal a role of zonal and azonal factors in
landscape structure differentiation as well as to determine dominant and subdominant factors
for different hierarchical levels. The special task is to reveal individual and integral properties
for each landscape component. Individual ones which determine regularities of structure
organization of particular landscape components are the object of branch sciences of
geography. Integral properties provide wholeness of natural geosystem is studied by
landscape geography.
Decisions of stated tasks differ for various territories and depend of basic perceptions
of researcher and selected methods of material collection and analysis. Development of
methods of an objective estimation of factors which form the complex of environment
conditions is an actual task of modern landscape science.
The stated aim is complicated because we need to extract hypothetical factors from
huge number of properties which determine conditions of landscape components. Each
component has its own characteristic time. Moreover, these properties inevitably need
various methods of description and with various accuracy. The most optimal scheme of
landscape factors search we can offer the way of consequent, step-by-step decreasing
number of properties for each functional part of component. For example, soil can be
described by the following properties: thickness of each genetic horizon, their color and
texture. Similar functional parts for vegetation are separate layers of canopy characterized
via species composition, height etc. Relief as the factor of moisture redistribution on various
scales can be presented through relative elevation, slope and its form. Reflected solar
radiation registered by satellite sensor is in fact momentary measurement of energy balance
components, different spectral indices reflect biophysical properties of vegetation.
Joining this massif of data and using different methods of multidimensional statistical
analysis, we can sequentially present each measured property or combination of properties
through relief and remote sensing data characteristics. In simplest case this operation can be
realized by step-by-step discriminant and factor analysis. Properties which discrimination
statistically is not confident are excluded from further analysis. Eventually, few dozens of field
measured properties are reduced to a few (5 to 6) independent factors. During this
transformation a number of properties excluded from final system. These are the ones which
variation in space is very specific or, only the determined in whole system part of this
property can be determined. The properties that are described by independent factors are
admitted as belonging to unified landscape system. The disadvantage of the offered method
is selection of all initial properties on the basis of relief and remote sensing data and mapping
via factor analysis only linear relations. Its advantage is visualization of any property and
independent factors variation directly in cartographic form. This allows us to semantically
control the obtained results and to make physical interpretation of independent factors
through analysis of their connection with functional properties.
Decreasing of system dimension essentially possible using method non-parametric
scaling. Using this method we do not need to use remote information as unified base. But the

592
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

second way in more bulky although it provides the informative analysis. Results of proposed
method are demonstrated on the example of searching landscape forming factors for
southern taiga landscape of Walday Hill (Central Forest Reserve, 33º E, 56.2º N). It is based
on 1150 field descriptions of 20 properties.
We have stated that all diversity of properties measured in situ, can be described by
eight independent factors which reflect properties of landscape components in more than
30% cases. For example three variables of Munsell soil color charts (HUE, VALUE, CROMA)
reconstructed in 50% cases. First factor determines 21.3% of total variation and leading for
most of landscape components. It reflects redistribution of moisture depending on slope
steepness on various hierarchical levels and determines development of raised peat bogs
and forest bogs, main properties of soils, development of moss cover and pine distribution in
space. Second by its significance factor (16.5%) also reflects moisture redistribution but
through form of surface. It is closely connected with amount of sun radiation used for
transpiration and biological productivity and sort of humus (soft or coarse). Third independent
factor (14.5%) reflects succession stages. Maximum of factor corresponds to forests with
highest biological productivity with spruce and birch domination. Fourth factor (12.5%) again
connected with moisture redistribution and with degree of humus accumulation dependently
of backwater moisture. It determines distribution of alder and elm trees. Fifth factor (9.5%) do
not depend on relief, it determines development of thin podzoloc soil horizon hypothetically in
places where ground water unloaded through carbonated moraine deposits. Here the low
sparse forests with poor productivity appear. Sixth factor (7.9%) is connected with absolute
elevation and determines thickness of humus layer and way of organic decomposition. All
other equal the higher territories characterized by thinning of humus layer and more soft
humus sort. Seventh factor (7.3%) is connected with soil texture and absolute elevation. The
higher elevation is, the more dense soil forming rock. Thickness of humus and podzolic soil
layers increases under grassy spruce-birch forests. At last the weakest eighth factor (5.7%)
is connected with slope steepness at microlevel and determines the type of redox regime
which influences the color of podzolic soil layer.
All factors with different weights determine states of various landscape properties. Their
real state is a result of relatively independent forces. But generally relations resolve to
multidimensional redistribution of moisture by relief, independent vegetation disturbances
and various mineralization of ground water.
Stated relations allow to obtain landscape map with utmost possible portrayed genesis
of landscape properties and to prepare basis for landscape forming processes modeling.
The research is made with support of RFBR projects # 03-05-64280.

593
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

Integrated Remote Sensing Monitoring for Environmental Changes of a Solid

Mineral Deposit Area (Diamond Deposit. Arkhangelsk District. Russia)

T.V.Orlov

Institute of Environmental Geoscience Russian Academy of Sciences (IEG RAS). Ulansky

pereulok 13, building 2, P.O.Box 145, 101000 Moscow, Russia.
e-mail: [email protected]

Introduction

Extreme geoecological conditions and high level environmental impact characterized solid
mineral deposits in northern regions. But solid mineral deposits exploitation is important part
of development modern society. However, mineral resource industry bring about irreversible
environmental changes. Mechanical, geochemical, hydrogeological impacts and their non-
linear interference determine this changes.
Integrated geoecological monitoring system should be an obligatory part of any solid mineral
deposit project. Different orientation of local impacts, their interference and changes, non-
linear finish result are the main problems of integrated geoecological monitoring. Integrated
geoecological monitoring system is number of measures to study and control different nature
component changes.
Remote sensing monitoring is a component of the integrated environmental monitoring
system of the diamond deposit. Field researches show environmental changes, remote
sensing monitoring shows spread of these changes for the whole investigated area.
The purpose of this issue is to investigate environmental conditions of diamond deposit area
using remote sensing data and field researches. The aim of this issue was to develop map of
natural and human-caused ecosystems and map of environmental changes.

Methods and Materials

Landsat 7 images (30m/pix, 2000), QuickBird (2.4, 0.6 m/pix, 2004, 2005, 2006), aerial
images (2000) and field researches of 2004-2006 were the basic issue materials.
Unsupervised and supervised classifications of satellite data were used for developing map
of natural and human-caused ecosystems. Samples for supervised classification were based
on types and features of natural ecosystems and degree of human-caused variations,
observed during field researches. Results of these classifications were verified by check data
set. Maps of natural and human caused ecosystems of different years are compared with
each other. The environmental changes map was developed using difference between
ecosystems maps of various years. Final map are verified by last satellite data and field
researches data.

IEM development problems

Problem of the monitoring point’s location.

A monitoring point is located at a way of influence. The method to solve the problem is
suggested: monitoring points groups are located at main ways of influence, other ways of
influence are controled by remote sensing data..
Problem of the monitoring point type selecting.
Another problem is to select monitoring point type and consider natural component
interaction. One of the methods to solve the problem may be characteristic-indicator search.
These indicators are specified for each region and impact type.
Problem of the observation characteristic set selecting.
Observation characteristic set selection came from prospective ecosystems changes and the
most sensitive environmental elements. Besides, there are monitoring points oriented on the

594
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

environmental changed flow control (e.g. waste water after bog passing). In that case
observation characteristic set should to consider flow changes. In example with waste waters
it is exchange reactions within wasted waters, bog and natural bog composition. Modeling
and field experiments are needed for those reaction describing.
Problem of measurement frequency
IEM system controlled processes have different rates.It is sufficient to find characteristics-
indicators that display this process in most completed way.

Conclusion

The main principles of rational IEM system develop was determined. They are: considering
landscape differentiation, considering of natural components interactions and reaction
integrity, considering primary impact processes, considering geochemical migration flows,
combination of point-contacted filed researched method and areal remote sensing methods,
integration of analyze and processing data, using indication characteristic, adaptivity of
structure and work order.
The IEM system for diamond deposit on the north of Russia was developed, based on thus
principles,. The IEM system consists of approximately 100 monitoring points, 20 routes and
profiles and covers the area of 40 sq.km. Maps of natural (fig. 1a) and man caused
ecosystems (fig. 1b) were developed.

Fig. 1a. Map fragment of natural ecosystems. Fig. 1b Map fragment of man caused
(green colors different bogs, purple colors – ecosystems (purple and green polygons –
different spruce forests, orange colors – pine forest bogging, green circles – man caused
forests, yellow and grey – men changed subsidences, green line - man caused
areas underflooding, blue polygon - water flooding

One of the most spread environmental changes is the land clearing and expansion of waste
piles and dams. Forest ecosystems bogging is the second type of changes. Soil wetness
grows up, water surfaces and streams appear as a result of bogging. Land subsidence
appears on the dams and vegetation cleared surface. Soil denudation and accumulation
present on the grass slopes, covered with 2-5 sm alluvium. Processes of environmental
changes are local at the present day; this investigation hasn’t fetch out significant spatial
changes. But diamond deposit exploitation impact is rather serious, and it will be heavier in
future.

595
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

Landscape mapping: does the scale lie within?

F. Tolle, J.-C. Foltête

ThéMA UMR 6049 CNRS - Université de Franche-Comté, 32 rue Mégevand, 25030

Besançon cedex.
e-mail: [email protected]

Introduction
The question of scale is a major issue in landscape ecology. Li and Wu (2004) do stress
the fact that a distinction should be made between the scale of landscape patterns and the
scale to which ecological processes are sensitive. These scales are not necessarily equal
and they might even not be linked (Wu, 1999). It is therefore critical that the scale at which
the landscape is described coincides with the process of interest. The way species use their
environment is constrained at various scale levels (Thies et al., 2003). It is now the
landscape approach that is privileged in most studies (Kareiva and Wennergren, 1995;
Wiegand et al., 1999). Gehring and Swihart (2003) explored the influence of habitat
fragmentation on mammal predators. Generalist predators and small animals do not have the
same perception of landscape (Vos et al., 2001). Inter-species relationships and the
complexity of prey-predators systems do require a multiscale approach (Brown and Litvaitis,
1995). This study aims at characterizing the functional response of species to landscape
structure without any a priori assumptions on the species behaviour. The influence of
landscape on the presence of the parasite Echinococcus multilocularis in a vector-borne
disease system has been explored. This parasite is responsible for the fatal zoonotic disease
alveolar echinococcosis in humans. The life cycle of the parasite is dependent upon two
animal vectors. Micromammals such as voles do host the larval form of the parasite and
foxes carry its adult form. The main goal of this work was to identify critical scale levels
favourable to the presence of Echinococcus multilocularis in the Doubs department in
eastern France. This area is known as endemic for the parasite and most of French human
cases have been diagnosed in the Doubs (Giraudoux et al., 1996).

Material and methods

In the study area, a database of 175 georeferenced samples has been constituted and
nine samples revealed the presence of the parasite. A classified image of the study area was
derived from IRS remote sensing data and resampled at a 25 m resolution. Landscape
contexts were derived around each sample and the composition of these landscapes was
recorded. The ratio of each land use class for each sample was expressed as a composition
vector (Wharton, 1982). This vector was expressed for each point data P of an image with c
land use classes as XP = (d1,d2,…,dc). Manhattan distances were then used to calculate the
distance between two composition vectors. A method developed by Foltête et al. (2002) was
used to compare the distance between the mean composition vector of all samples and the
composition vector of positive samples, at each analysis radius. This index, noted u, shows
low values when the distance between positive- and all-samples landscape composition
vectors is the greatest. It has been computed at radii ranging from 25 to 10000 meters and
revealed three scale levels of interest. Landscape proved the most different at radii of 700 m,
2200 m and 4775 m. These scales were selected for subsequent analysis.

Results
At the three scales of analysis, the composition of the landscape has been compared.
The first conclusion is that the landscape surrounding positive samples consistently shows a
higher ratio of complex patches (abandoned agricultural areas, hedges, forest margins…)
which can be interpreted as the expression of the need for species to find both food and
shelter. Conversely, cultivated fields are not attractive and are inappropriate to the
development of the parasite’s life cycle. This was expected as fields are not optimal habitat
for the intermediate hosts because of the disturbances caused by ploughing. Other

596
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

parameters such as forest do not seem to have an influence on the epidemiological

processes. Three maps were computed and illustrate the Manhattan distance of the
landscape of each cell of the study area to the mean composition vector of positive samples.
At a 700 m radius, forest edges seem to appear significant whereas important uniform
agricultural areas do not seem to play a role. At a 2200 m radius, the most complex areas
are opposed to the same main field areas. Eventually, at a 4775 m radius, the area of the
first plateau and the high land of the Jura mountains seem to be of value to be parasite. As
shown here, both a precise and spatial description of the specificities of landscapes related
to Echinococcus multilocularis can be established using these methods. The main interest of
this approach lies in the fact that the scale of analysis is expressed by the data and not
based on expert knowledge, therefore excluding potential mistakes in the critical choice of
the scale level at which the ecological phenomenon should be observed. Subsequent
sampling campaigns could be designed based on these results and used to validate the
conclusions of this work.

Acknowledgments
The authors are grateful to Patrick Giraudoux and Francis Raoul from the Laboratoire de Biologie
Environnementale, Déborah Gottschek, Stéphanie Favier and Frantz Catarelli at the ERZ for the
coordination of the monitoring survey. Denis Augot, Franck Boué and Anne Thuault at the AFSSA for
the samples’ analysis work. Financial support from the European EchinoRisk program QLK-2-CT-
2001-01995 “Risk assessment and prevention of Alveolar Echinococcosis” and US National Institutes
of Health and National Science Foundation “Ecology of infectious diseases” (TWO1565-02).

References
Brown, A.L. & Litvaitis, J.A. (1995) Habitat features associated with predation of New England
cottontails: what scale is appropriate? Canadian journal of zoology. 73, 1005-1011.
Foltête, J.C.; Monteil, C. & Deconchat, M. (2002) Habitat animal et image numérique : méthode de
reconnaissance exploratoire appliquée à des occurrences d’espèces. Proceedings of the 6èmes
journées Cassini, 187-206.
Gehring, T. M. & Swihart, R. K. (2003) Body size, niche breadth, and ecologically scaled responses
to habitat fragmentation: mammalian predators in an agricultural landscape. Biological
Conservation. 109, 283-295.
Giraudoux, P.; Vuitton, D.A.; Bresson-Hadni, S.; Craig, P.; Bartholomot, B.; Barnish, G.;
Laplante, J.J.; Zhong, S.D.; Wang, Y.H. & Lenys, D. (1996) Mass screening and epidemiology
of Alveolar echinococcosis in France, Western Europe, and in Gansu, Central China: from
epidemiology towards transmission ecology. In: J. Ito et N. Sato (eds.), Alveolar echinococcosis:
strategy for eradication of alveolar echinococcosis of the liver, Fuji Shoin, Sapporo 060, Japan,
197-211.
Kareiva, P. & Wennergren, U. (1995) Connecting landscape pattern to ecosystem and population
processes. Nature. 373, 299-302.
Li, H. & Wu, J. (2004) Use and misuse of landscape indices. Landscape Ecology. 19, 389-399.
Thies, C.; Steffan-Dewenter, I. & Tscharntke, T. (2003) Effects of landscape context on herbivory
and parasitism at different spatial scales. Oikos. 101, 18-25.
Vos, C.C.; Verboom, J.; Opdam, P.F.M. & Ter Braak, C.J.F. (2001) Toward ecologically scaled
landscape indices. American naturalist. 157, 24-41.
Wharton, S.W. (1982) A contextual classification method for recognizing land use patterns in high
resolution remotely sensed data. Pattern Recognition. 15(4), 317-324.
Wiegand, T.; Molony, K.A.; Naves, J. & Knauer, F. (1999) Finding the missing link between
landscape structure and population dynamics: a spatially explicit perspective. Am. Nat. 154, 605-
627.
Wu, J. (1999) Hierarchy and scaling: extrapolating information along a scaling ladder. Canadian
journal of remote sensing. 25, 367-380.

597
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

Investigation of spatial resolution in remote sensing of fragmented environments

for ecological models

A.M. Lechner1, S.D. Jones 1, S.A. Bekessy 2

1
School of Mathematical and Geospatial Sciences, RMIT University,
GPO Box 2476V, Melbourne 3001, VIC, Australia.
e-mail: [email protected]
2
School of Global Studies, Social Science & Planning, RMIT University,
GPO Box 2476V, Melbourne 3001, VIC Australia.

Introduction

This project investigates the use of high spatial resolution remote sensing for ecological
models, particularly in fragmented environments. Due to the lack of accurate up-to-date
maps of an appropriate scale, fragmented natural habitats such as peri-urban areas are
some of the most difficult to model. Remote sensing is often heralded as a solution to fill this
information void, however objects may be spectrally and spatially similar and may thus
confound mapping - fragmented environments are spatially complex with habitat patches
varying in size from median strips (~10m2) to large vegetation remnants contained within
national parks (100km2).

Correctly mapping habitats is often critical for the development of accurate ecological
models. However, depending on the remote sensing data, classification technique and class
description, large differences in the classification of landcover may occur, and the resulting
map might vary in the extent, patchiness and accuracy of classified areas. This study
compares different sensor resolutions, the influence of changing the extents of the study
area and the final mapped products.

Measuring the effect of changing spatial dependent factors

This study utilizes presence / absence tree cover data produced for the state of Victoria in
Australia that covers an area of approximately 227,416 km². The study area includes
agricultural, urban and wilderness areas with a variety of levels of habitat fragmentation.

10m 50m 100m

Figure 1. Example of a clipped area where the resolution has been degraded from the
original 10m pixel size to the lowest resolution of 100m.

Subsets of this image were randomly clipped at 3000m, 10000m, and 20000m replicating
landscapes of different extents (see figure 1). Other studies have compared scaling effects
on landscape metrics (e.g. Wu et al. 2002), however this study is unusual in that the large
study area allows for multiple replications at the landscapes level of real landscapes.
Simulated landscapes have difficulties in capturing all the characteristics of real landscapes

598
Theme 5: Monitoring and classification
5.3 Open Session 2: Landscape modelling and earth observation

(Li et al. 2004). For each clip the image was degraded from the original 10m pixel size to
100m at 10m intervals. Landscape metrics were then calculated using the fragstats package
(McGarigal et al. 2002).

Results
The study found that while the total area classified remained relatively constant when the
image resolution changed there were large differences in the patchiness (see figure 2). As
image resolution increased the fine scaled levels of patchiness no longer appeared. Small
patches either aggregated into larger patches or disappeared. Furthermore, the relationship
between resolution and patchiness was not linear and changed at different extents.

a) b)
450 0.5
400
Extent 3000
Proportion Present
350 0.4
Mean Patch Area

300 Extent 10000

0.3
250 Extent 20000
200
0.2
150
100 0.1
50
0 0
10 20 30 40 50 60 70 80 90 100 10 20 30 40 50 60 70 80 90 100
Resolution Resolution

Figure 2. a) Mean patch Area. b) Proportion of imagery classified as having vegetation

present.

Conclusion

It can be seen that changes in scale dependent factors affect the patchiness and total area
classified. This study demonstrates that landcover maps are the product of the resolution of
the imagery and study extents.

This paper is part of a larger project that aims to assess multiple scale dependent factors and
compare the magnitude of their influence on accuracy, patchiness and total area classified.

References
Li, X, He, HS, Wang, X, Bu, R, Hu, Y & Chang, Y (2004) Evaluating the effectiveness of neutral
landscape models to represent a real landscape, Landscape and Urban Planning 69(1): 137-
48.
McGarigal, K, Cushman, SA, Neel, MC & Ene, E (2002) FRAGSTATS: Spatial Pattern Analysis
Program for Categorical Maps. Computer software program produced by the authors at the
University of Massachusetts, Amherst., <Available at the following web site:
www.umass.edu/landeco/research/fragstats/fragstats.htm>.
Wu, JG, Shen, WJ, Sun, WZ & Tueller, PT (2002) 'Empirical patterns of the effects of changing scale
on landscape metrics', Landscape Ecology, 17 (8): 761-82.

599
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

5.4 Open Session 15: Landscape metrics and geostatistics

Ecological Aspects of Landscape Fragmentation Measurements

P. Csorba

University of Debrecen, Departement of Landscape Protection and Environmental

Geography, H-4010 Debrecen Egyetem tér 1, Hungary,
e-mail: [email protected]

Landscape metrics deal with characteristics of landscape patterns, like spatial structure of
patches, corridors and barriers. Landscape metrics often examines the fragmentation of
landscapes due to its importance from the aspect of landscape aesthetics, general
landscape and nature protection planning (Farina, 1998).
It is not necessary to take into consideration the ecological features of visual landscape
structure for landscape aesthetic analyses. In such cases aesthetic parameters of landscape
pattern, balance of patches, color, and tone are important factors. (Wöbse, 2002).
In the case of general landscape protection – e.g. a World Heritage cultural landscape –
the aim is the maintenance of a landscape structure that is considered to be valuable by the
society. Ecological aspects are not of primary importance in such cases, because usually,
the main purpose of those efforts is to conserve traditional cultivated plants and land use –
vineyard, pasturing, forestry (dehesa, kampen, bocage) for instance (Wascher and Jongman,
2000).
On the other hand, when data of landscape-metrics analyses are used in nature
protection planning, ecological aspects are stressed (Hawkins and Selman, 2002, Klopatek
and Gardner 1999, Vos et al. 2001).
The degree of fragmentation of 229 microregions of Hungary by roads, railway lines and
settlements was determined (Csorba, 2006). Our datasets were compared to those of Dosh
and Beckmann (1999) for Germany. Results have showed that for estimation of ecological
consequences of fragmentation it is necessary to take into account some other factors in
addition to data on the density of traffic infrastructure.
For landscape metrics based on ecological aspects, important factors are the size, scatter
and shape (pheriphery/area) of patches; ecological contrast of neighboring patches, quality
(width) of ecotones, connectedness of patches and quality of corridors (strip, line, dispersal
corridor). From the aspect of ecological functions traffic loads of roads and railway lines are
also important parameters. Other important factors are: whether there are busy roads and
railroads close to each other, their position relative to the protected areas and whether there
are sections of those traffic routes which cross protected areas.
For the weighting, traffic density of roads was first taken into account. Data for this
parameter is easily accessible from the national database. Traffic density data for Hungarian
road network is divided into eight motor car unit categories: 0-499, 500-999, 1 000-1 499,
1 500-1 999, 2 000-4 999, 5 000-7 999, 8000-19 999, and over 20 0000 car units/day. On
that base five categories were formed by a contraction for ecological weighting (Csorba,
2000a).
In the case of railway lines different weights were given to main lines, branch lines and
narrow gauge lines.
The picture gained this way was refined further by the consideration of consequences of
the topographic situation (Csorba, 2000b). It is quite usual that busy roads and railways run
parallel to each other. In this way they form a dual barrier for the migration of animal and
plant species. The scale of the base map (1 to 250 000) made possible to use a higher
weight in places where a road and a railway line runs closer to each other than 1 km.
In places, where roads or railway lines cross nature protection areas, it is advisable to use
a higher weight again.

600
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Because of lack of data on their traffic density, the fragmentation effect of tourist tracks could
unfortunately not be taken into consideration, although these tracks cross nature protection
areas, and act as strong barriers (mass tourism in the environment of the cities) in many
cases.
It is not easy to quantify the fragmentation effect of settlements. According to Hungarian
landscape protection regulations there must be an at least 200 meters wide migration
corridor left between settlements. In places where degree of agglomeration of settlements
reaches this level it is advisable to also use a higher weight.
It would be reasonable also to take into consideration the ecological migration effect of
the typical scattered settlement type (homesteads) in the Great Hungarian Plain. This
problem has not completely been solved yet, as there are not enough data for the
qualification of the impacts of special land use sites like quarries and airports as well.

References
Csorba, P. (2005) Landscape ecological fragmentation of the small landscape units (microregions) of
Hungary based on the settlement network and traffic infrastructure. Ekológia in press.
Csorba, P. (2005a) Ecological corridors in the foothill area of the Tokaj Mts. Greenways. Conference
Presentations of the Ecological Corridors, Green Corridor, Sopron, pp. 31-44.
Csorba, P. (2005b) Kistájaink tájökológiai felszabdaltsága a településhálózat és a közlekedési
infrastruktúra hatására. Földrajzi Értesítő LIV, 3-4., pp. 243-263.
Dosch, F. & Beckmann, G. (1999) Trends der Landschaftsentwicklung in der Bundesrepublik
Deutschland. Informationen zur Raumentwicklung, H. 5/6. pp. 291-310.
Farina, A. (1998) Principles and Methods in Landscape Ecology. Chapman and Hall, 235 p.
Hawkins, V. & Selman, P. (2002) Landscape scale planing: exploring alternative land use scenarios
Landscape and Urban Planning 60. pp. 211-224.
Klopatek, J.M. & Gardner, R.H. (eds.) (1999) Landscape Ecological Analysis. Issues and
Applications. Springer Verlag, 400 p.
Vos, C.C. & Verboom, J. & Opdam, P.F.M. & Ter Braak, C.J.F. (2001) Toward Ecologically Scaled
landscape Indices. The American Naturalist 183. 1. pp. 24-41.
Wascher, D. & Jongman, R. (2000) European landscapes, Classification, assessment and
conservation. European Environmental Agency, Coppenhagen
Wöbse, H. (2002) Landschaftsästhetik. Ulmer Verlag, 304 p.

601
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Determining patch-level metrics in naturally disturbed and managed forest

landscapes: implications for enhancing ecological attributes and processes

R. Lafortezza1, R.C. Corry 2, R.D. Brown 2, G. Sanesi1

1
greenLab – Department of Scienze delle Produzioni Vegetali, University of Bari, via
Amendola 165/A 7026 Bari, Italy.
e-mail: [email protected]
2
School of Environmental Design & Rural Development, University of Guelph, Guelph,
Ontario, Canada N1G 2W1

Introduction

Relatively biodiversity-rich habitats in highly fragmented landscapes tend to be few and

their characteristics tend to be modified by pressures from the surrounding matrix. The
woodlands, wetlands, and grassland or prairie habitats that are some of the most valued
refuges for biodiversity are valued in part because of their rarity. Beyond the presence and
number of these habitats are other characteristics of their relative quality, like patch size,
shape, and location. Patch shape relates to habitat quality by the ways in which it allows or
enables interactions with adjacent ecosystems. Complex patch shapes are considered to
provide more locations for different foraging behaviours and for encouraging more boundary-
crossing animals through coves and lobes (Dramstad et al. 1996).
Complex shapes lead to microclimatic variability and greater plant diversity along edges
(Forman 1995). Simple patch shapes, conversely, encourage movement along edges (rather
than across) and lead to less interaction between the habitat patch and the matrix (Dramstad
et al. 1996). While complex patch shapes may provide for more interaction among species,
the pervasive effect of human beings has been to divide, settle, and manage land in ways
that simplifies patch shapes into regular linear corridors, rectangles, triangles, and geometric
polygons (Fig. 1) (Corry and Nassauer 2002; Brown et al. 2007).
Beginning with a primeval ecosystem (forest, wetland, grassland/prairie), the perforation
of clearing and settlement gradually converts the intact ecosystem to smaller, fewer, and
more geometrized patch shapes (Forman 1995). Patch shapes would not be simplified only
under circumstances of unusually-high required effort, exorbitant costs, or technical limits
(Corry and Nassauer 2002). Patch sizes and shapes have been shown to have a relationship
in particular locations. Larger patches tend to have more complex shapes, while smaller
patches have more regular shapes. A study in Mississippi (USA) for example noted that
forest patches that were smaller in area were more regularly shaped, while large forest
patches were more complex (Krummel et al. 1987). In another study forest patch shapes
were found to be simpler in Wisconsin (USA) landscapes that were more affected by human
activities. In this paper, we explored the size:shape relationship of forest patches in two
distinctly cultural, yet different landscapes: Southern Ontario and Southern Italy (Apulia
region). The Ontario landscape is relatively young (about 12,000 years since last glaciation)
and its transformation to broad-scale settlement is relatively recent. Southern Ontario has
been surveyed and divided under several different land division systems, creating a
patchwork of towns, fields, and forests with odd angles, gores, and unusual intersections
(Hart 1998). The Southern Italy landscape is older (about 100,000 years since glaciation),
has been settled much longer, and is typified by finer-scale management. The patterns of
land division and settlement have commonly fragmented primeval ecosystems along lines
that coincide with road networks, farm boundaries, and field patterns (Brown et al. 2007).
Cultural artifacts, cropping patterns, and remnants of vegetation (generally) typify the
structure of the landscape and its inherent spatial heterogeneity.

Objectives

602
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Our exploration had the following research objectives:

• to compare similarities and differences between Ontario and Italy forest patch
size:shape relationships
• to learn if patch size and shape were related in ways similar to other studies (i.e.,
larger patches had more complex shapes; smaller patches had more simple
shapes)
• to learn if the duration of settlement (i.e., the length of time that culture has
modified the landscape) is related to more simple patch shapes.
• to estimate – by comparison to spatial attributes – the degree of cultural and
acultural shape effects on patch size classes.
• to suggest how forest patch size:shape relationships can inform landscape
planning, design, and management.

CLC: 311
Size: 42.47 ha
D: 1.29

Figure 1. Aerial photograph of southern Italy landscape showing forest patch size and shape
(D= fractal dimension; SI= shape index).

References
Brown, R.D., Lafortezza, R., Corry, R.C., Leal, D. & Sanesi, G. (2007). Cultural patterns as a
component of environmental planning and design. In: Hong SK, Nakagoshi N, Fu B, Morimoto J,
Wu J.G. (eds.) Landscape Ecological Applications in Man-Influenced Areas: Linking Man and
Nature Systems, Springer.
Corry, R.C. & Nassauer, J.I. (2002) Managing for Small Patch Patterns in Human-dominated
Landscapes: Cultural Factors and Corn Belt Agriculture. In: Liu J, Taylor W (ed.) Integrating
Landscape Ecology into Natural Resource Management, Cambridge University Press. Cambridge,
Massachusetts: pp. 92-113.
Dramstad, W.E., Olson, J.D. & Forman, R.T.T. (1996). Landscape Ecology Principles in Landscape
Architecture and Land-use Planning, Island Press, Washington, DC.
Forman, R.T.T. (1995). Land Mosaics, Cambridge University Press, Cambridge, Massachusetts.
Hart, J.F. (1998). The Rural Landscape, Johns Hopkins University Press, Baltimore, Maryland.
Krummel, J.R., Gardner, R.H., Sugihara, G., O'Neill, R.V. & Coleman, P.R. (1987). Landscape
patterns in a disturbed environment. Oikos 48: 321-324.

603
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Mathematical models of landscape patterns

A.S Viktorov

Institute of Environmental Geoscience of Russian Academy of Science

Ulanskii per., 13, P.O. Box 145,101000, Moscow, Russia,
e-mail: [email protected]

Quantitative studies of landscapes spatial patterns face with problems resulting from
unpredictable relationships among different metrics of landscape patterns (Ritters et al,
1995). It is very difficult to determine their sufficiency and superabundance for pattern
analysis and to choose optimal sets of metrics. Mathematical models of landscape patterns
can solve these problems. Our research allowed us to formulate the mathematical models of
landscape patterns. They are specific to areas characterized with a certain set of processes,
such as soil subsidence, fluvial erosion, karst, and other processes. The models are based
on the theory of random processes and describe behavior of patterns’ geometric features.
Mathematical analysis of the models gives us landscape pattern laws such as:
− Poisson distribution for arrangement of subsidence sites,
− Wiener stochastic process describes size logarithm growth for thermokarst depressions,
− Raleigh distribution of khasyrey (drained thermokarst lakes) size.
− Using of the mathematical models of landscape patterns allows us to predict dynamics of
a territory.
Let us examine a good example of nature dynamics at a territory with long time of
diffuse process development. The diffuse nature process is a process generating numerous
often roundish sites randomly located within the area of process development. Processes of
this type include karst, soil subsidence, thermokarst, aeolian processes (wind erosion
dimples), and others.
The model of diffuse processes within a uniform area is based on the following
assumptions (Victorov, 2003):
1. Appearance of a new site during any time within any clear area is an event
independent of other sites and its probability is directly proportional to the time period and
size of the area. The probability that two or more sites occur is negligibly small in comparison
as appearance of only one site.
p1 = λΔsΔt + î (ΔsΔt ) , (1)
where λ is a parameter,
2. A new site cannot appear within another site.
3. Changes of a new site of isometric form are described with an occasional process
F0 ( x, t ) (probability density is f 0 ( x, t ) ) if obstacles are absent. Different sites change
independently of each other.
4. Changes of a new site of isometric form are described with an occasional process
(probability density is F (x ) ) if obstacles are absent. Different sites change independently of
each other.
Thus, the territory can be regarded as a certain flow of developing sites of the diffuse
process. Every site appears at a random moment independently from each other, enlarges
under influence of different factors including random ones and finally reaches a critical value
at a random moment and comes to the stage of degenerate sites. In this very case it possible
to show that under rather general conditions (Victorov, 2003) the territory in question reaches
a certain state close to stationary one with characteristics of dynamic balance. Thus, after a
long period of development a number of quantitative characteristics become stable:
- process impact ( Pd* ) , share of area affected by the dangerous process
- radii distribution for the active sites

604
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

+∞
[1 − F ( x)] ∫ f 0 ( x, u )du
f ( x, ∞ ) = + ∞+ ∞
0
(2)

∫∫
0 0
[1 − F ( x)] f 0 ( x, u )dxdu

- average area density of the active sites

+∞+∞
γ a ∞ = λ[1 − P ] ∫ ∫ [1 − F ( x)] f 0 ( x, u )dxdu
d
*
(3)
0 0
- radii distribution for the degenerate sites
f dk ( x, ∞) = f ( x) (4)
In other words, new active sites continuously appear and disappear within the observed
area, their dimensions change, new degenerate sites appear but the main quantitative
statistic characteristics, such as area density of the active sites, the dimensions of
degenerate and active sites remain stable.
For example, It shows that under certain weak conditions after a long period of time the
термокарстовая plain reaches dynamic equilibrium with constant area and constant spatial
density of thermokarst lakes and constant average khasyrey area.

The mathematical models of landscape patterns give us analytical decision for

estimating damage risk for engineering constructions under exogenous geological hazards.
Using the described above mathematical models of landscape patterns for diffuse
processes one can get impact probability for different types of constructions.
The roundish construction (a cycle with radius I):
Pds (l ) = 1 − exp[−πγ (t )[(r (t ) + l ) 2 + σ 2 (t )] − πγ dg (t )[(rdg (t ) + l ) 2 + σ dg
2
(t )]] (5)
The linear construction (with length L)
Pdl ( L) = 1 − exp[−2[γ (t )r (t ) + γ dg (t )r dg (t )]L] (6)
where γ (t ) , r (t ) , σ (t ) , γ dg (t ), rdg (t ),σ dg (t ) are average area density, average radius, and
radius standard. The equations of impact probability for little, large, and linear constructions
at any given time were developed basing on remote sensing data as initial information. The
equations were verified at key areas of Western and Eastern Siberia, Caspian Lowland, and
piedmont plains of Central Asia.
The mathematical models of landscape patterns give us analytical decision for
interrelations between different pattern metrics and the problem of sufficient metrics including
general information about landscape patterns. For example we can obtain a dependence
between average ( a ) and standard ( σ ) of location density of karst depressions
σ = a (7)
The research branch dealing with the mathematical models of landscape patterns is named
“the Mathematical Morphology of Landscape” (Victorov, 1998).

References
Ritters K.H., O’Neill R.V., Hunsaker C.T. et al. (1995) A factor of landscape pattern and structure
metrics. Landscape Ecology, v. 10, №1, pp.23-39.
Victorov A. S. (2003) An integrated mathematical model for diffuse exogenous geological processes.
Proceedings of the 9th Annual Conference of the IAMG, Portsmouth, GB, pp. 1600-1620.
Victorov A. S. (1998) Matematicheskaya morfologiya landshafta (Mathematical Morphology of
Landscape). Tratek, Moscow

605
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Statistical challenges integrating Landsat time series data with forest inventory
data for characterizing forest disturbance and regrowth in the U.S.

G.G. Moisen1, S.P. Healey 1, R.E. Kennedy 2, S.L. Powell 2, W.B. Cohen 2, S.N.
Goward3, J.G. Masek4, C. Huang3
1
US Forest Service – 507 25th Street, Ogden, UT 84401,USA.
e-mail: [email protected]
2
U.S. Forest Service – Pacific Northwest Research Station, Corvallis, OR 97331, USA.
3
Department of Geography, University of Maryland, College Park, MD 20742, USA.
4
NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA.

Traditional forest inventories

The Forest Inventory and Analysis (FIA) program of the USDA Forest Service collects
data annually on the status and trends in forested ecosystems across the U.S. (FIA, 2005.)
Inventory is conducted via a network of ground-based plots which are located with an
intensity of about one plot per 2,400 ha. Although the program historically collected data
periodically, it recently shifted to an annual rotating panel system which samples 10 to 20%
of each state’s plot network annually (Reams et al., 2005.) These inventory data have
traditionally been used to support estimates of forest population totals over large geographic
areas.

Changing information needs

Recent emphasis has been placed on producing broad-scale maps of numerous forest
characteristics to make these extensive forest resource data more accessible and useful to a
larger and more diverse audience. Important applications of such maps include broad-scale
mapping and assessment of wildlife habitat; documenting forest resources affected by fire,
fragmentation, and urbanization; identifying land suitable for timber production; and locating
areas at high risk for plant invasions, or insect or disease outbreaks. A variety of RS
products are relied upon to improve precision in estimates of forest population parameters,
construct meaningful small area estimates, enable rapid response to catastrophic change,
etc.
An important component of monitoring is gaining a clear understanding of what has
happened in the past. Reconstructing historical trends in forest disturbance using FIA data is
hampered by inconsistent sampling schemes and plot designs, varying definitions, gaps in
plot distributions, irregular and sometimes non-existent temporal sampling, and the list
continues. FIA data alone cannot adequately tell the forest disturbance and recovery history.
In contrast, disturbance maps generated from historic Landsat imagery communicate spatial
and temporal disturbance trends in a straightforward manner (eg., Cohen et al., 2002.)

Collaboration under the North American Carbon Program

FIA is collaborating with NASA, the University of Maryland, and other Forest Service
researchers to map historical disturbances across the country (Healey et al., In press.)
Dense temporal stacks of 23 Landsat scenes have been sampled across the U.S. Forest
biomass available on FIA plots is currently being modeled as empirical functions of Landsat
spectral and ancillary data. These models are then being applied across the temporal stacks
to develop a 30+ year historical record of forest disturbance and regrowth dynamics for North
America.

Importance of local analyses

606
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Five focal scenes are targeted for detailed analyses by FIA scientists, analysts, and clients
for a variety of disturbance issues. Maps of disturbance can be used to identify harvest
trends over time and across ownership boundaries; to track insect outbreaks; to support
purely spatial analyses such as the estimation of edge effects or the revision of fuel or habitat
maps; and to prioritise salvage and recovery efforts. Maps of biomass loss may be of use in
any future carbon accounting framework. Re-growth trajectories can also be produced for
each disturbance, enabling study of factors that affect forest recovery (Healey et al, 2006.)
Leveraging of FIA data with satellite imagery in this way may open up open up new
perspectives of how disturbance operates in our forests over both time and space. However,
the statistical issues abound and here we describe the challenges in integrating design- and
model-based paradigms to construct statistically defensible estimates of forest change
parameters for both local and national analyses.

References
Cohen, W.B.; Spies, T.A.; Alig, R.J.; Oetter, D.R.; Maiersperger, T.K. & Fiorella, M. (2002.)
Characterizing 23 years (1972-1995) of stand-replacing disturbance in western Oregon forest with
Landsat imagery. Ecosystems 5: 122-137.
FIA, 2005. What is Forest Inventory and Analysis. FIA Fact Sheet Series, 3 February, 2005.
Retrieved from http://fia.fs.fed.us/library/fact-sheets/, on Feb 1, 2007.
Healey, S.P.; Moisen, G., Masek; J., Cohen, W.; Goward, S.; Powell, S.; Nelson, M.; Jacobs, D.;
Lister, A.; Kennedy, R. & Shaw, J. (In press.) Measurement of forest disturbance and re-
growth with Landsat and FIA data: anticipated benefits from FIA’s collaboration with NASA and
University partners. In: McRoberts, R. (Ed.) Proceeding of the 7th FIA Science Symposium,
Portland, ME.
Healey, S.P.; Yang, Z., Cohen; W.B.& Pierce, D.J. (2006.) Application of two regression-based
methods to estimate the effects of partial harvest on forest structure using Landsat data. Remote
Sensing of Environment 101: 115-126.
Reams, G.A.; Smith, W.D.; Hansen, M.H.; Bechtold, W.A.; Roesch, F.A.& Moisen, G.G. (2005.)
The Forest Inventory and Analysis sampling frame. In W.A. Bechtold and P.L. Patterson, eds. The
Enhanced Forest Inventory and Analysis Program – National Sampling Design and Estimation
Procedures. US Department of Agriculture, Forest Service, Southern Research Station General
Technical Report GTR-SRS-80. pp. 11-26.

607
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Display of the basic functional properties of landscape cover on the basis of the
remote information for maintenance of initial landscape planning stages

A.N. Krenke, Y.G.Puzachenko

Lomonosov Moscow State University Faculty of Geography, 119899, Moscow,

Leninskiye Gory 1, Russia,
e-mail: [email protected]

Landscape cover - the concept meaning display biophysical properties of a terrestrial

surface. This concept is historically connected to the development of remote sensing by
multispectral measurement of the condition of the land surface and eventually to the tasks
connected to landscape planning.
Field work especially in northern regions incur high expenses; it is therefore necessary in the
initial stages of design and exploration work to use to the maximum remote information
available, in order to get a detailed overview of the current and past condition of the major
properties of landscape cover.
For achievement of this purpose development of special algorithms for the analysis of the
remote information are required, that provide not only allocation of the basic types of
landscape cover, but display of spatial variation of its major properties such as: a
hydrothermal mode, biological efficiency, character and a degree natural and
anthropogenous disturbance.
In the present paper the algorithm of classification is used as a start, given both own
values of brightness spectral channels is examined. These indexes are considered as a
basis, which reliably reflect the certain physical properties of a terrestrial surface. The method
used of dichotomizing hierarchical classification "from the general to the particular" accounts
for both territorial position of selected types and values of indexes. They allow consistently
defining their physical sense of selected classes of a landscape cover. On the basis of the
classification and its discriminant analysis the main factors forming landscape structure are
provided as well as a cartographic display of their values. The constructed maps of modern
landscape cover, as a first approximation allow the estimation of problems which can enable
the designer to estimate the vulnerability of territories, to allocate potential objects of
protection and to develop the most economical and to make an effective avoidance of field
research.
The basic factors are described which allow the construction of qualitative level models of
vulnerability of territories to various impacts. In the study the estimation of fire danger of
territory is shown. It is also possible to estimate engineering works which identify land which
has increased risk of failures of non-production constructions. The most valuable land covers,
the most productive large forests and pastures can be protected.
At the same time these results can be considered only as preliminary outcomes. Field
research should be organized to check up the display of hypothetical landscapes and the
definition of the geographical contents of the allocated types. The received set of optimum
allocated field descriptions will then allow the application of multivariate methods to
interpolate statistical models of landscape cover of territory.
Linking the remote information with a digital elevation model expands the opportunities of
the analysis of structure of landscape cover and creates a basis for the preliminary design of
a network of supervision of monitoring projects.

608
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

The landscape disparity index: an ecologically weighted measure of the

landscape diversity.

P.K. Roche

CEMAGREF, Mediterranean Ecosystems Unit, 3275 Route Cezanne, 12182 Aix-en-

Provence.
e-mail: [email protected]

Introduction

Typically, the analysis of landscape patterns is based on a quantitative analysis of the

composition or/and the spatial arrangements of categorical units mapped from aerial
photographs, satellite images or land use maps using so-called landscape metrics
(Gustafson, 1998).
Usually in landscape studies, authors assume that different landscape habitats or
elements categorised in the mapping process are well defined, and equally different from
each other (Forman 1995 and references therein). But, in most cases the landscape element
types are often not really distinct or at least are not equally distinct from each other and the
degree of dissimilarity between the mapped landscape elements is not usually considered.
In order to improve the ecological relevance of landscape diversity metrics some
ecologically important features should be considered. An index should integrate the habitat
number and their spatial arrangements (O’Neill et al., 1988). It should also have a low
sensitivity to categorical resolution (Frohn, 1998) and/or should be able to handle degrees of
thematic resolution (Loehle and Wein 1994). The two last points could be addressed by
integrating a measure of the ecological contrast between the thematic categories – i.e. the
disparity.

Methods
I propose a new landscape diversity metric: the Landscape Disparity Index (LDI) that
takes into account both the landscape elements adjacencies frequencies and the patches
ecological dissimilarities (1)

LDI =
(
− ∑∑ ln p(i,2 j) ⋅ d (i, j) ) (1)
ln (2 /(m ⋅ (m + 1)))
The LDI was tested using landscape maps from two sites in southern France.
Environmental variables were used to compute the dissimilarity index between and within
landscape element types. Highly autocorrelated variables like altitude and slope were
eliminated. The remaining variables were tested using Kruskal-Wallis tests for significant
differences between landscape elements. Non significant variables were removed. The
remaining variables were used to compute the dissimilarity matrices.
The LDI was compared with eight other widely used landscape metrics using the
Fragstats 3.1 software (McGarical and Marks, 1994).
Scale issues were analysed by computing the LDI and the eight other landscape metrics
on our test area at 4 scales (30, 50, 70 and 90 m). Thematic resolution issues were
assessed by computing the LDI and the other landscape metric on the 6 and 11 theme
maps. In addition, for the LDI, the sensibility to the dissimilarity measure component using
the two variants of the distance matrix for both 6 and 11 theme maps.
As a biological variable of interest we used the species richness at the landscape level
using both modelled and field data.
If species richness is considered to increase when landscape heterogeneity increases
(Duelli 1997), then landscape metrics measuring landscape heterogeneity indexes should be
positively correlated with species richness at the landscape level.

609
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Results and Discussion

The Landscape Disparity Index appeared to belong to the group of landscape diversity
metrics with some interesting properties not shared by other diversity metrics: low sensibility
to thematic resolution and ecologically significant weighting The Shannon index (SHDI)
appears to be the more sensitive to thematic resolution effects with Diff% higher than 40%
and four indexes have Diff% values under 10% (PLADJ, LDI, COHESION and AI). Two
groups of metrics could be defined. The first one comprised the IJI, RPR, SHDI, LDI that
could be defined as diversity metrics. The second group comprised the AI, PLADJ,
COHESION, LPI and CONTAG indexes that are landscape compaction-aggregation metrics.

140

120
# Species Observed

100

80

60

40 Fig. 1. Biplot of species richness and LDI. Each dot

represents a landscape window of 90x90 m. The full
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Landscape Disparity line represents a GLM regression using a log link.

According to our results high LDI values are also related to high species richness at the
landscape level (Fig1). High LDI values indicate that ecologically contrasted habitats are
clustered together on the analysis scale. Traditionally used landscape diversity indexes such
as the Shannon Index are not as coherent, because high values can be associated with low
species richness. Thus, taking into account the degree of ecological contrast between
landscape elements weights the spatial heterogeneity and improves correlation with
biological patterns. These preliminary results indicate that the Landscape Disparity Index
could be used to identify high diversity spots at the landscape level.

References
Duelli, P. 1997. Biodiversity evaluation in agricultural landscapes : an approach at two different
scales. Agri. Ecos. Env., 62: 81-91.
Forman, R.T.T. 1995. Land mosaics. The ecology of landscapes and regions. Cambridge University
Press, Cambridge, 622 pp.
Frohn, R. 1998. Remote sensing for landscape ecology. New metrics indicators for monitoring,
modelling and assessment of ecosystems. Lewis publishers, Boca Raton.
Gustafson, E.J. 1998. Quantifying landscape spatial pattern: what is the state of the art ?
Ecosystems, 1: 143-156.
McGarigal, K. and Marks, B.J. 1995. FRAGSTATS: spatial pattern analysis program for quantifying
landscape structure. General Technical Resport PNW-GTR-351. U.S. department of Agriculture,
Forest service, Pacific Northwest Research station. Portland, OR. 122 pp.
O’Neill, R.V., Krummel, J.R., Gardner, R.H., Sugihara, G., Jackson, B., DeAngelis, D.L., Milne,
B.T., Turner, M.G., Zigmunt, B., Christensen, S.W., Dayle, V.H. and Graham, R.L. 1988. Indices
of landscape pattern. Land. Ecol., 1: 153-162.
Turner, M.G. 1989. Landscape ecology: the effect of pattern on process. Ann Rev Ecol Syst, 20: 171-
197.

610
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Are scaling functions for landscape pattern metrics really accurate and useful?

S. Saura and S. Castro

Departament d’Enginyeria Agroforestal. ETSEA. Universitat de Lleida. Av. Alcalde Rovira

Roure, 191. 25198. Lleida. Spain.
e-mail: [email protected]

Introduction

The availability of a wide variety of landscape data and remote sensors at multiple spatial
scales urges the need for practical scaling techniques that allow comparing and transferring
pattern estimates across different spatial resolutions. Different studies have found scaling
functions (such as power laws) accurately describing the variations of several landscape
pattern metrics with spatial resolution (Frohn, 1998; Frohn and Hao, 2006; Saura, 2001;
Saura, 2004; Wu et al., 2002; Wu, 2004). Considering the good apparent fit of these scaling
functions to a wide set of landscape data, it has been suggested that the extrapolation and
interpolation of the values of these metrics across different pixel sizes can be done simply
and accurately (Wu et al., 2002; Wu, 2004). However, Saura (2004) noted that the
coefficients of the scaling functions (which are needed for the scaling process itself) cannot
be known a priori for a certain image or landscape; rather they have to be determined
empirically by previously fitting the scaling function to a set of metric values computed on the
aggregated image at different spatial resolutions. For this reason, scaling functions seem to
be of little aid for upscaling pattern estimates (i.e. obtaining metrics values at coarser spatial
resolutions), since in fact they require as an input those metric values at broader pixel sizes.
The major interest of these scaling functions may be obtaining subpixel estimates of pattern
metrics, and in fact this may represent the only operational procedure to downscale spatial
pattern characteristics (Saura, 2004), although no quantitative results have been provided
yet on this respect, apart from a single forest class and range of spatial resolutions reported
by García-Gigorro and Saura (2005).

Material and methods

To provide further insights into this scaling problem, we analysed a wide set of landscape
data derived from remotely sensed images covering different study areas, sensor spatial
resolutions, and classification approaches (pixel-based and object-based), which were
aggregated to coarser spatial resolutions through majority filters. We considered a set of
eight landscape pattern metrics (number of patches, mean patch size, patch size standard
deviation, largest patch index, edge length, landscape shape index, area-weighted mean
shape index, area-weighted mean patch fractal dimension) for which stable and predictable
scaling functions at the class level had been previously reported (Frohn, 1998; Frohn and
Hao, 2006; Saura, 2001; Saura, 2004; Wu et al., 2002; Wu, 2004). We fitted these functions
to the metrics values at different spatial resolutions (broader than the target resolution),
comparing the resultant downscaled estimates (obtained through the scaling functions) with
the true value of the pattern metrics at that target resolution.

Results and discussion

Even when all the eight metrics (and the corresponding scaling functions) have been
previously reported as predictable in terms of their variations with spatial resolution, we found
large differences in the actual accuracy of the resultant subpixel estimates among pattern
metrics. For the mean patch size, the landscape shape index or the edge length, quite
accurate subpixel estimates were achieved in all the datasets, while other metrics could not
estimated accurately at finer resolutions through available scaling functions.

611
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

The most accurate subpixel estimates were obtained when only a narrow range of spatial
resolutions (closest to the target resolution) was used to fit the scaling function, instead of the
full range of spatial resolutions that can be obtained through aggregation, which agrees with
García-Gigorro and Saura (2005). This suggests that the scale behaviour of the pattern
metrics (as described by available scaling functions) is not really invariant across the full
range of spatial resolutions; the rate of variation of the pattern metrics at increasingly broader
scales diverges from the characteristic variation at subpixel resolutions.

We also found that the performance of available scaling functions was much lower in
object-based data than in per-pixel classified data, particularly when scaling at spatial
resolutions below the characteristic minimum mapping unit of the interpreted or segmented
image. All the scaling functions for landscape pattern metrics reported so far have been
based only on per-pixel classifications of remotely sensed data, and we suggest that other
different scaling functions may be necessary to provide more accurate results when dealing
with object-based data, which are increasingly common due to the important advantages that
this classification approach presents for landscape pattern analysis.

Conclusions

Most of the previous research has not gone beyond a merely descriptive analysis of scale
effects on pattern metrics. In this study we intended to go further by quantitatively assessing
and validating the true accuracy and practical utility of available scaling functions for
obtaining subpixel estimates of pattern metrics. We conclude that scaling functions may be
useful and reasonably accurate for estimating the values of some pattern metrics at the
subpixel level but only if the specific scaling recommendations and limitations reported in this
study are conveniently taken into account.

References
Frohn, R.C. (1998) Remote sensing for landscape ecology: new metric indicators for monitoring,
modeling and assessment of ecosystems. CRC-Lewis Publishers, Boca Raton, Florida, USA.
Frohn, R.C. & Hao, Y. (2006) Landscape metric performance in analyzing two decades of
deforestation in the Amazon Basin of Rondonia, Brazil. Remote Sensing of Environment 100: 237-
251.
García-Gigorro, S. & Saura, S. (2005) Forest fragmentation estimated from remotely sensed data: is
comparison across scales posible? Forest Science 51: 51-63.
Saura, S. (2001) Influencia de la escala en la configuración del paisaje: estudio mediante un nuevo
método de simulación espacial, imágenes de satélite y cartografías temáticas. Ph. D. Thesis,
Universidad Politécnica de Madrid, Spain.
Saura, S. (2004) Effects of remote sensor spatial resolution and data aggregation on selected
fragmentation indices. Landscape Ecology 19: 197-209.
Wu, J.; Shen, W.; Sun, W. & Tueller, P.T. (2002) Empirical patterns of the effects of changing scale
on landscape metrics. Landscape Ecology 17: 761-782.
Wu, J. (2004) Effects of changing scale on landscape pattern analysis: scaling relations. Landscape
Ecology 19: 125-138.

612
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

AMOEBA algorithm to identify natural clusters of mixed forest in Poland using

CLC2000 data

T. Edman

Swedish University of Agricultural Sciences, Forest Faculty, Department of South Swedish

Forest Research, SE-230 53 Alnarp, Sweden.
e-mail: [email protected]

Connectivity of habitats within and between landscapes and regions is of great importance
for the persistence of viable populations in the landscapes and regions. There are several
methods to establish the connectivity between separated habitats. These methods include
grid based algorithms averaging the amount of habitats in an area, calculation of landscape
metrics of different areas, k-mean clusters amongst others. Most of the methods are
computing and data intensive.

The long history of land utilisation in Europe affects the potential for biodeversity and there
are differences between regions due to variance in economic history, recent economical
development and bio geographical conditions. These differences are recognised as an east-
west gradient with low biodiversity in the Benelux area and high in the Baltic-states and in the
Carpathian Mountains (Edman et al. manuscript). Poland (Figure 1) is a good arena for
investigations of the relationship between landscape characteristics and biodiversity as it
contains both high and low diversity areas. Poland consists mainly of agricultural plains with
small forest patches and of forest landscapes in the northeast, southeast and in the
northwest.

Mixed forests are of great importance for biodiversity as they often are less intensively
managed and thus contain more suitable habitat for forest specialised species. Woodpeckers
are connected with habitats that are common in the mixed forests of Europe (Martikainen et
al., 1998) and the area of mixed forests are important for the functional diversity of forest
vertebrates in temperate Europe (Edman and Angelstam, manuscript).

I used the AMOEBA algorithm proposed by Estivill-Castro and Lee (2000) as a less
computing intensive cluster algorithm, to establish clusters of mixed forests in Poland.
CORINE 2000 with resolution of 100m was used as land cover data (EEA, 2004). The
geographic modifications of the CORINE 2000 raster was made with ArcWorcstation 8.2
Each delineable polygon was attributed with centre position, area, perimeter, land cover type,
minor and major axes of the smallest ellipse that can be fitted over the polygon as well as the
angle of the axis in relation to the x- and y-axis of the EUREF 99 coordinate system.
According to the AMOEBA algorithm a Delaney triangulation table was established, the
distances between all ellipses that are used as proxies for the actual land cover polygons
were calculated and mean local (Lm) and global (Gm) segment length as well as global
standard deviation (Gstd)of segment length. If a segment is shorter than Gm+Gstd*(Gm/Lm)
the segment is considered to connect two patches that are part of a cluster, see Estivill-
Castro and Lee (1999) for further elaboration of the algorithm. This procedure is repeated to
obtain sub clusters of the first level clusters. All calsulations was made in Matlab 14.

613
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Figure 1 Map over Poland displaying the clusters of mixed forest,

identified with the AMOEBA algorithm.

AMOEBA clustering proved to be a good way to explore networks of connected forest

patches in Polish landscapes. As the cut off level for cluster belonging is adjustable and
might be replaced with an ecologically based figure the AMOEBA clustering seems to have a
good potential for further development and use in landscape ecological models, where
species movements and connectivity between habitat patches is concerned.

References
EEA (2004) CORINE Land Cover 2000. pp.Landcover in Europe. EEA, Copenhagen.
Estivill-Castro, V. & Lee, I. (1999) AMOEBA: HIERARCHICAL CLUSTERING BASED ON SPATIAL
PROXIMITY USING DELAUNATY DIAGRAM. Proceedings of the 9th International Symposium on
Spatial Data Handling (SDH2000).
Martikainen, P., Kaila, L. & Haila, Y. (1998) Threatened Beetles in White-Backed Woodpecker
Habitats. Conservation Biology, 12, 293-301.

614
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

How important is the third dimension for the analysis of spatial patterns of
landscape structure?

U. Walz, S. Hoechstetter, N.X. Thinh

Leibniz Institute of Ecological and Regional Development (IOER),

Weberplatz 1; D-01217 Dresden;
e-mail: [email protected]

Introduction

The analysis and quantification of patterns of landscape structure are crucial for
understanding ecological dynamics. In the Patch model ecologically meaningful 3D-features
like terrain shape or elevation are not taken into account and valuable information is lost for
the analysis. In a study financed by the German research foundation (DFG), several methods
to include the third dimension in the analysis of landscape structure are developed, tested
and customised.
3D-aspects can be integrated into structural analysis by the computation of “true”
perimeter and area on the basis of digital elevation models (DEM) or by the consideration of
actual surfaces on the basis of digital surface models (including vegetation and buildings).
The latter includes the investigation of height differences between the landuse structures of
individual patches or the characterisation of the surface texture within the landuse classes
and/or individual patches.

Integration of “true” surface area and perimeter into landscape metrics

Most Geographic Information Systems (GIS) systematically underestimate the areas and
perimeters of patches in a land mosaic and the distances between them because of the
planimetric projection of land elements. As most of the commonly used landscape metrics
utilise information about patch geometries as input parameters, this circumstance can lead to
systematic errors in resulting index values. To improve this, we tested a method to
approximate the realistic surface area from digital elevation models developed by Jenness
(2004) and customised this method in order to calculate the true surface perimeters of each
patch. For the analysis, we used a DEM from Airborne Laser Scanning and the digital vector
data of the German ATKIS landuse data. The results of the computation of area and shape
metrics for the planimetric and the 3D-situation with different spatial resolutions shows that
there are distinct differences for the area metrics “Landscape Area” and “Patch Area”,
especially in steep terrain. Values tend to get closer to each other for the coarser resolution
level, as differences in relief are levelled out when the elevation model is resampled to a
larger grid size. On the other hand, metrics like Perimeter/Area-Ratio, Fractal Dimension
Index and Shape-Index, which are based on fractions between patch perimeters and areas,
exhibit only weak or no response at all to the inclusion of surface geometries into the
calculation, especially for the coarse-grained landscape model. High-quality input data with a
high spatial and thematic resolution appear to be a precondition for a meaningful
implementation of 3D-approaches.

Three-dimensional landscape patterns

For the characterisation of surface patterns, modifications of fractal approaches like the
determination of the three-dimensional box dimension and a new technique based on the
lacunarity analysis of quantitative data, were developed. Besides that, parameters of surface
metrology were used. Particularly the parameter „Average Surface Roughness“ proved to be
suitable as a straightforward approximation of the relief variability within individual landscape
elements. The analysis of surface roughness may serve as a valuable instrument to provide

615
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

highly condensed information about the topographic characteristics of patches. As the basic
surface metrology-indices can easily be calculated and may be integrated into the patch
mosaic model of landscapes, they appear to be a good extension of common landscape
metrics towards the third spatial dimension.

Possible applications in landscape ecology

Our research showed that there is a need for the enhancement of landscape structure
indices in terms of an integration of the third dimension. Especially in environmental
planning, models and indicators are needed which can supply valid statements for the
condition and the change of ecological systems caused by anthropogenic interferences. In
this context, the landuse structure has a key position for the analysis and characterisation of
functional correlations within the ecosystem. Furthermore, elevation and topography in reality
act as “gradients” rather than “discrete structures”. Natural boundaries in the relief can rarely
be detected. The attempt to integrate and to assess three-dimensional structures appears to
be a promising way to take such ecological gradients into account (MCGARIGAL AND
CUSHMAN 2005).
The employment of 3D-indices in the context of environmental modelling or spatial
planning leads to more precise results and may thus help to improve the basis of information.
In particular in the following fields the application of 3D-indices of landscape structure
appears to be relevant:
Methods of habitat connectivity and modelling of biodiversity (determination of isolation, size
and neighbourhood relations; delineation of transition areas (ecotones));
Measures of landscape fragmentation. Currently used measures (e.g. effective mesh size)
are based on surface computations. Including “real” area can thus lead to closer-to-reality
results.
Distribution models of species frequently use structural descriptors as explanatory variables.
One can expect that the consideration of three-dimensional surface properties leads to the
improvement of such models.
Methods and models in forest sciences, in particular concerning height structures and
surfaces of forest stands (e.g. information for forest stocktaking or estimation of horizontal
and vertical diversity).
As a result, we can conclude that integrating elevation into the assessment of landscape
structure principally seems to be an encouraging approach. Results indicate that 3D-analysis
provides a great deal of additional information about the environment. Especially the
comparison between standard landscape metrics and their corrected 3D-equivalents reveals
significant differences between the index values. In combination with promising techniques
such as lacunarity or surface metrology indices, which allow to capture ecological gradients
and structure analysis over a range of scales, the proposed framework allows for a more
realistic and accurate representation of landscapes. In particular when high-resolution data
like for example digital terrain models from airborne laser scanning are available, the
appliance of 3D- and gradient analysis techniques is appropriate.

References
Jenness, J.S. (2004): Calculating Landscape Surface Area from Digital Elevation Models. Wildlife
Society Bulletin 32 (3): 829-839.
McGarigal, K. & Cushman, S.A. (2005): The Gradient Concept of Landscape Structure. In: Wiens, J.,
Moss, M. (eds.): Issues and Perspectives in Landscape Ecology. Cambridge University Press,
Cambridge: 112-119.

616
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Approaches to revealing landscape spatial pattern based on analysis of

landscape between-component relations

K.A. Merekalova

Moscow Lomonosov State University, Geographical Faculty, Vorobyovy Gory, 119992,

Moscow, Russia,
e-mail: [email protected]

Introduction

A specific character of landscape approach to nature investigation is the priority given to

the study of relations between landscape components. The Russian landscape science
school has traditionally relied on the concept of strictly determined interconnections between
all landscape components (genetic model) when mapping natural landscapes (Vidina, 1962)
(fig.1). Later, the idea of landscape multi-structural organization emerged, which assumed
the simultaneous existence of spatially superposed but relatively independent partial
landscape units generated by interaction between different geophysical fields (Solntsev,
1981). According to this concept, landscape spatial pattern at a certain scale level is an
aggregate consisting of partial landscape units that are mostly evident in this scale. Lastly,
nowadays there is a problem of translating geographical information from one scale level to
another (upscaling and downscaling) (Marceau, 1999). Such a translation is possible only if
the relationships type between landscape components are scale-independent and similar at
all hierarchical levels in the area under study (Meentemeyer, 1989). Correspondingly, there
is a need to reveal spatial landscape units with unified type of components interrelations at
each scale level.
The purpose of this research is to apply the above approaches to reveal landscape spatial
patterns within the boundaries of the focus region.

Figure 1. Concepts of landscape components interconnections

Region and material

617
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

The study area with an area of about 10 km2 is located in middle taiga region of North
European Russia. Research material includes: 1) digital elevation model (DEM) with
resolution 30 m in pixel derived from topographic map with scale 1:10000; 2) space image
Landsat 7 with resolution 30 m in pixel; 3) field landscape descriptions (more then 300
sample plots).

Results

The problems solved in the research are as follows:

1) Based on genetic approach a landscape map was created (scale 1:20000). The study
area was divided into 3 localities that included 82 patch types.
2) Based on conception of multi-structure organization probabilistic approach was applied.
Mutual adaptation of landscape components was evaluated and a map of partial landscape
units was created (Khoroshev, Merekalova, 2006).
3) Linear multiple regression equations were computed for the system “landform-plant cover”
in a moving window. As dependent variables we used factors of plant cover differentiation
calculated using principal components analysis of Landsat 7 image. Relief characteristics
based on DEM were chosen as independent variables (slope gradient, second derivative of
elevation (Laplacian), distance to the closest waterway, vertical ruggedness (elevation
variance in the area of certain radius) and horizontal ruggedness (total length of waterways
in certain area)). Classification of the territory by regression coefficients revealed landscape
units with unified type of components interrelations, with eight types in total. Comparison of
the results achieved by using a different size of moving window made it possible to reveal the
characteristic scales of landscape components interrelationships.

References
Khoroshev, A.V. & Merekalova, K.A. (2006) Uncertainty of relations between landscape components
– a tool for modeling evolution of spatial pattern. Ecology (Bratislava), Vol. 25, Supplement
1/2006, pp. 122-130.
Marceau, D.J. (1999) The scale issue in social and natural sciences. Canadian Journal of Remote
Sensing, Vol. 25, No. 4, pp. 347-356.
Meentemeyer, V. (1989) Geographical perspectives of space, time, scale. Landscape Ecology, Vol.
3, No. 3/4, pp.163-173.
Solntsev, V.N. (1981) Structural organization of landscapes (in Russian). Science Press, Moscow.
Vidina, A.A. (1962) Methodic of field large-scale landscape research (in Russian). Moscow University
Press, Moscow.

618
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

Use of a bioclimatic classification for a large-scale application of a

biogeochemistry model

A.D. Shkaruba1, V.V. Kireyeu2, I.P. Usava3

1
Central Research Institute for Integrative Use of Water Resources (CNIIKIWR),
Slavinskaga 1/2, 220086 Minsk, Belarus.
e-mail: [email protected]
2
Central European University, Nador u. 9, 1051 Budapest, Hungary
3
Belarus’ State University, Ave. Skaryny 4, 220026 Minsk, Belarus

Knowledge of the spatial distribution of NPP, carbon and nitrogen fluxes is of increasing
importance for a number of policy areas. As a result, large-scale applications of
biogeochemical/geophysical models have become more common, and are also stimulated by
open access to high-resolution spatial datasets and remote-sensed images. Accordingly, the
quality and availability of field data have become a bottleneck for large-scale modelling
applications. This is particularly true for process-based models, which are often preferred
because of their broader applicability and robust extrapolations.

These problems have been faced when setting an application of one-dimensional

CoupModel (coupled heat and mass transfer model for soil-plant-atmosphere systems
(Jansson & Karlsberg, 2004)) for the Eastern Belarus Landscape Province (c.a. 22,000 km2,
120-230 m, dominated by elevated moraine plains). The objective was to simulate actual and
potential carbon stocks and sinks and make predictions in relation with baseline scenarios of
climate change. For the reasons of data availability and computational efficiency, it was
decided to run the model for environmental classes instead of grid meshes which would have
been too coarse even at the highest resolution supported. Comparing to the grid meshes,
classes of an environmental classification, believed ecologically homogeneous within their
boundaries, can reproduce existing patterns of environmental gradients with a greater
fidelity, while keeping within a reasonably low number of clusters. Hence, the specific
objectives of this study were (1) to develop an environmental classification and (2) to
propose procedures for preparing model inputs at a scale of these blocks (classes).

The underlying assumption of the classification is that the spatial diversity of ecological
conditions and habitats primarily depends on the local distribution and re-distribution of the
heat and water with significant contribution from land-use/cover. Other considerations were
related to the input data requirements and design of the CoupModel, very much identical to
those of many other process-based models. For this setup the model needed three kinds of
input data: soil properties (mostly hydrothermic and biogeochemical), plant properties
(biometry, physiology, phenology) and climate data series. Accordingly, the classification
should be bioclimatic by its nature, which would allow the modeller to construct climate
datasets for individual strata. Soil types should coincide with the bioclimatic units (this can be
enhanced by adding to the classification the data on the conditions of saturation), which can
be checked against local soil classifications (though quite impossible to validate by field
data). Plant properties are mostly derived from the generis data and literature (though there
are measured validating data available), but ideally they are also supposed to fit into the
bioclimatic classes.

The method taken to produce the bioclimatic classification was a statistical stratification
operating on quantitative data (Jongman et al, 1995). The technique is partly based on that
employed in the Environmental Classification of Europe (Metzger et al, 2005). For the
stratification we have selected variables (1) related to climate to account for the distribution
of geophysical factors over the area and (2) related to topography to account for their local
re-distribution. The spatial climate datasets have been downloaded from the CRU of the

619
Theme 5: Monitoring and classification
5.4 Open Session 15: Landscape metrics and geostatistics

University of East Anglia and included mean values for 1961-1990. The topographic
variables have been extracted from USGS 3 arc-second Digital Terrain Elevation Data. They
are elevation and TOPMODEL’s topographical index (TopoIndex) used to account for
topographically determined hydrological conditions (Kirkby & Weyman, 1974) and identify the
meso-forms of local topography. The TopoIndex calculates as ln(α/tgβ), where α is the area
draining through a point from upslope and β is the local slope angle; high index values are
likely to be more saturated. The stratification included (1) consequent Principal Component
Analyses (PCA) to to remove redundancy and to reduce dimensionality within each category
of environmental variables, (2) statistical classification and (3) post-processing. Several
classifications with different number of classes has been generated and tested for the
variability of input values within classes; 30-class stratification fit better the selection criteria.

As expected, the bioclimatic classification showed a good correspondence with the local
soil classification, which is explained by the fact that the area’s major topographic features
coincide with the class boundaries, but often differ in the geological origin and, hence, in the
underlying rocks. The main extent of the strata are always smaller in size than polygons of
the soil classification, therefore, there was no need to subdivide strata massifs (separate
groups of meshes) into smaller units with regard to soil classes. In contrast, land-use
classification showed poor correlation with the strata boundaries, except the land-use
classes strongly dependent on the physical environment, e.g. meadows, peat bogs or
flooded forests. If requested by scenarios simulated, e.g. based on the current land-use, the
strata can be subdivided by land-use boundaries into smaller parcels; if the issue is the
potential vegetation, the strata can be subdivided by the classes of vegetation, which,
however, again are likely to agree closely.

As discussed above, setting climate and vegetation input data to the CoupModel within
this classification can be done in a straightforward manner. It is more difficult with soil data.
The datasets available, coming from the National Land Survey, are not consistent, irregularly
spaced, and have not been based on a accurate sampling design. It was concluded that the
soil data available often are not statistically representative for particular landscapes, land-
uses or strata and, therefore, cannot be subjected to aggregation. In the study the problem
was solved by choosing representative soil samples within each block applying the expert
knowledge of the area. Partly following the methodology suggested by Brus (1994), we tried
to find an appropriate sample for each stratum, but also we took into consideration other
factors of spatial variability (e.g. land-use) by taking more samples when the variability is
higher. Additionally, the options and parameters of the CoupModel were analysed against
abiotic ecological conditions (e.g. high saturation) and biocenoses expected within strata.

References
Brus D.J. (1994). Improving design-based estimation of spatial means by soil map stratification. A
case study of phosphate saturation. Geoderma 62: 233-246.
Jongman R.H.G.; Ter Braak C.J.F. & Van Tongeren O.F.M. (1995). Data Analysis in Community
and Landscape Ecology, Cambridge University Press, Cambridge.
Kirkby, M. J. & Weyman, D. R. (1974). `Measurements of contributing area in very small drainage
basins', Seminar Series B, No. 3, Department of Geography, University of Bristol, Bristol.
Metzger M.J.; Bunce R.G.H., Jongman R.H.G., Mücher C.A. & Watkins J.W. (2005). A climatic
stratification of the environment of Europe. Global Ecology and Biogeography, 14: 549-563.

620
5.5 Posters

5.5 Posters

How to recognize a landscape boundary? – Finding appropriate scale for the

ecosystem management –

N. Tokiwa, J. Morimoto, F. Nakamura

Graduate School of Agriculture, Hokkaido University, JAPAN.

e-mail: [email protected]

Introduction
Ecosystem functions can be maintained by keeping various ecological linkages between
neighboring ecosystems, not by conserving an “isolated” target ecosystem from others
(Meyer 1997). Landscape, a heterogeneous land mosaics composed of a cluster of
ecosystems, and appropriate spatial scales for ecosystem management should be
scientifically recognized. We developed a method to identify a landscape boundary by
analogically applying species – area curves (Mueller-Dombois & Ellenberg 1974) to the
relationships between the number of vegetation categories as ecosystems and management
area in SHIRETOKO World Natural Heritage (WNH).

Methods
[Database] Actual Vegetation Map of 5th survey (MOE, 1999) and Environmental GIS
database of latitude, elevation, relief, snow depth, degree of vegetation naturalness, and
Warmth Index. [Software] ArcGIS9.0 (ESRI) and PC-ORD (MjM Software Design).
[Analyses] 1. Ecosystem – Radius Curves: All data were converted into grid data. Assuming
the northernmost SHIRETOKO Cape the center, concentric 80 circles with radius of 1,000m
to 80,000m were made. The number of vegetation categories included in every circle was
counted. With these data, a curve showing relations with the number of vegetation categories
and concentric radius was delineated. 2. Ordination and Correlation Analysis: for
understanding this curve, 20 sampling areas (5,000m in radius) that cover all SHIRETOKO
WNH were arranged by Non-metric Multidimensional Scaling (NMS).

Results:
1. Ecosystem – Radius Curves: The range of a radius of 80,000m from SHIRETOKO Cape
had two types of curves, i.e., intensively increasing part (fig.1-A,C,E) and relatively flat part
(fig.1-B,D).
2. Ordination by NMS: Sampling areas in zone A and B were distributed over the
approximately first quadrant in a 2 dimensional coordinate plane. In contrast the sampling
areas in zone B, C and D were distributed widely. Vegetation categories appeared in zone A
were distributed concentratedly near the origin. Whereas those appeared in zone C were
distributed over the third quadrant. Vegetation categories which appeared in zone E
distributed far from the origin in the third quadrant.
3. Correlation test: The relationships
between the 1st axis score and 10
environmental variables revealed that 1st
axis expresses steepness to the positive
direction, and low level of naturalness to the
negative direction. The relationships
between the 2nd axis score and 10
environmental variables revealed that high
altitudes to the positive direction, and low
level of naturalness to the negative direction. Fig1. Ecosystem - Radius Curves

621
5.5 Posters

Using graphs and graph theory to describe and analyze connectivity in linear
habitat networks

A.Jellema12, P.F.M. Opdam 2, W.A.H. Rossing 1

1
Biological Farming Systems Group, Marijkeweg 22, 6709 PG Wageningen,
The Netherlands.
e-mail: [email protected]
3
Alterra Green World Research, P.O. Box 47, 6700 AA Wageningen, The Netherlands.

Introduction

Connectivity in a network of linear habitat elements, as found in agricultural or river

landscapes, can not be described or analyzed adequately using traditional landscape
ecological methods. We propose an alternative approach using mathematical graphs, as an
efficient means to describe and analyze linear landscapes (Urban and Keitt 2001). A linear
habitat network is described by a graph by breaking the linear elements in equal sized
sections and representing each section by a node (point). The connections between the
sections can be represented by a set of edges (lines). This methodology has been
demonstrated in a case study on the hedgerow network of the Frisian Woodlands. A series of
figures present the results in the poster.

Methodology

The methodology presented is based on two concepts from graph theory, the minimum
spanning tree, the set of edges in a graph that connects all nodes of the graph at the least
possible cost and the minimum cut, the set of edges with the minimum sum of edge weights
that needs to be removed to break a graph into two sub graphs (Bondy and Murty 1977). We
have used these concepts into a three step approach:
1. The definition of species dispersal groups based on the network configuration.
2. Analysis of the network structure per species group, revealing the backbone, the
bottleneck and the core of the network.
3. Determination of the robustness of the network for habitat destruction.
Using the methodology a quick scan of the connectivity relations within a habitat network can
be obtained for conservationists or spatial planners with limited effort. The concepts and
method as presented can be generalized to patch networks and other ecological networks
types.

Results

Case study results are presented on the poster in two figures.

References
Urban, D. & Keitt, T. (2001) Landscape connectivity: a graph-theoretic perspective. Ecology 82:
1205-1218.
Bondy, J.A. & Murty, U.S.R. (1977) Graph theory with applications. American Elsevier Publishing
Co., New York

622
5.5 Posters

ATLAS-Action for training in Land Use and Sustainability,

B. Pedroli, M. Bogers

Alterra Wageningen UR, Postbus 47 Wageninegn, The Netherlands

[email protected]

The ATLAS project is an EU-funded Co-ordination Action (2005-2007) aimed to provide

insights into delivering adequate types of education and training to decision-makers in the
area of sustainable development and land use. They should be able to judge on the basis of
sound policy analysis in a competent way about the effectiveness of different strategies.

ATLAS has brought together the expertise of the leading European research, education and
training institutions in the area of land use and sustainability, combining innovative research
efforts and practical experiences. It has led among others to a database of about 3000
courses related to sustainability and land use across Europe.

The ATLAS project has designed an Interactive Roadmap to assist users in obtaining
information regarding Sustainable Impact Assessment and Land use that is tailor-made for
their respective needs. The Road-map is based on an analysis of strengths, weaknesses,
opportunities and threats (SWOT) of the educational provision in this area in Europe and is
tuned to the needs identified by the different target groups (students, policy makers,
scientists) in sustainability impact assessment of land use policies. It aims to link the
information needs of the different target groups to the required information. It not only
includes the higher level education provision in this field but also informs about policies on
National and EU level and provides links to the current state of the art within other EU
projects (such as SENSOR, SEAMLESS, and others) and further reading and documents
related to Sustainable Impact Assessment and Land use.

The poster session will be set up interactively2 to show participants the function of the on-line
roadmap and will ask for their feedback for future incorporation. This should provide better
European organisation of the educational provision and lead to appropriate professional
qualifications.

2
for this poster session internet access and projection facilities are necessary

623
5.5 Posters

Landscape character assessment using region growing in GIS

A. Jellema1,3, D.J. Stobbelaar1,2,4, J.C.J. Groot1,2, W.A.H. Rossing 1

1
Wageningen University & Research Centre, Biological Farming Systems Group, Marijkeweg
22, 6709 PG Wageningen, The Netherlands.
e-mail: [email protected].
2
Plant Production Systems Group, 3 Alterra Green World Research 4 Rural Sociology Group.

Introduction
The character of a landscape can be defined as the arrangement, variety and intensity of
different landscape features. Examples of such landscape features are the field pattern,
types of land-use, the presence ponds, tree lines and hedgerows etc. The pattern of
landscape features gives a landscape a specific quality and makes it stand out from its
surrounding areas. We present a new methodology to delineate and analyze such
landscapes using a region growing algorithm in spatial data. The methodology is
demonstrated and evaluated in a case study in the Frisian Woodlands, using an expert map
as a independent reference.

Methodology
Region growing is an iterative bottom-up optimization process, where iteratively the most
similar neighbouring data elements are grouped creating spatially continuous clusters or
regions (Baatz and Schaepe 2000). Each region in the spatial dataset represents a
landscape characterized a data pattern of landscape features. The methodology consists of
4 steps: 1 building a spatial database, 2 Creating a regional adjacency graph representing
the spatial elements (polygons or raster cells) as nodes and their neighbourhood relationship
as edges, 3 delineating landscapes on the basis of the data patterns of the landscape
features, 4 evaluation of the regions.

Results
Based on 7 landscape features, 132 landscapes are distinguished in the study area, using
the region growing algorithm. When overlaying the original data with the region growing
result, all regions and the exact location of most of the borders between regions can be
explained. Also the resemblance with expert map is large; differences between the maps can
be explained by the absence or presence of data. In comparison to the expert map, the
region growing technique shows more consistent delineation of landscapes throughout the
study area. The results are presented in figures in the poster.

Conclusion
Region growing provides a powerful means to delineate landscapes on the basis of a
spatial data set. These landscapes can be used for planning discussions or further research.
For example the quantitative description of the landscape pattern can be used to classify the
landscapes or evaluate the landscape quality in comparison to a reference landscape. The
methodology can be generalized to other scientific disciplines.

References
Baatz, M. & Schaepe, A. (2000). In: J. Strobel and T. Blaschke (Eds.). Multiresolution Segmentation:
an Optimization approach for high quality image segmentation, Angewandte Geographische
Informationsverarbeitung 12. Wichmann-Verlag, Heidelberg, pp.12-23.

624
5.5 Posters

The use of remote sensing data for modeling differing levels of grassland
improvement within the Welsh landscape

J. Breyer1, K. Medcalf2
1
Institute of Geography and Earth Science, University of Wales, Aberystwyth, Ceredigion,
UK.
e-mail: [email protected]
2
Environment Systems, 8G Cefn Llan Science Park, Aberystwyth, Ceredigion, UK.

Unimproved and semi-improved grassland habitats are an important repository of

biodiversity within the Welsh landscape. This poster examines the use of space and airborne
remote sensing data to identify areas of species rich unimproved and semi-improved
grassland. The technique potentially represents a readily available tool for monitoring and
evaluating biodiversity networks relating to these important grassland habitats.
The method was developed using 1 m spatial resolution hyperspectral Compact Airborne
Spectrographic Imager (CASI) data acquired over two upland and lowland long term
experimental sites. Each site contained a number of grassland plots representing a range of
improved to unimproved sward types that are typical to the Welsh landscape, which has
been under traditional agricultural long term management. Fieldwork undertaken
coincidentally with the CASI imagery acquisition indicated that the spectral properties
(specifically mean Near Infrared (NIR) reflectance and NIR variance) of fields at varying
levels of improvement differ and can be related to ecological factors such as species
composition, species richness and biomass.
These relationships were then utilized within a rule-based classification developed within
eCogniton software to classify 10 m spatial resolution multispectral SPOT 5 HRVIR imagery.
The NEXTMap DTM was used to evaluate altitude and slope within the landscape,
supporting conclusions on particularly the drainage conditions in the uplands, which can
determine the prevalent grassland type. To further aid the classification, linear spectral
unmixing techniques within ENVI imaging processing software were utilized to generate
endmember fraction images for shade/moisture, photosynthetic and non-photosynthetic
vegetation.
To reduce error during classification the landscape was initially segmented into objects
(segments) within eCognition software using the Land Parcel Identification System (LPIS)
unit boundaries as a thematic layer, which provides all field boundaries within the lowlands.
The upland was defined as areas above the upper limit of agricultural enclosure again using
the LPIS boundary vector layer. To further increase accuracy, all non grassland habitat areas
were identified and excluded from the classification process. Prior to image classification,
accurate registration of the satellite data was essential, as well as reliable atmospheric
correction to surface reflectance to be able to directly compare between different images.
The resulting map of semi-improved and unimproved grasslands areas within selected
areas of the Welsh landscape is presented in this poster. Maps like these, generated from
remote sensing data, would be a valuable aid to conservation efforts such as UK Biodiversity
Action Plans, particularly in view of maintaining and restoring biodiversity networks and
preventing further isolation of valuable grassland species populations and communities. They
would also support policy decision processes particularly in relation to the effects of the
European Common Agricultural Policy (CAP) reform.

625
5.5 Posters

Application of System Approach towards Small-Scale Mapping and Classification

of Geographical Landscapes for the Purposes of Melioration and Agriculture

A.A. Nikiforova

Moscow Lomonosov State University, Soil Science Department, Moscow, Leninskiye Gory, ,
Russian Federation.
e-mail: [email protected]

It is evident that efficient and environmentally safe agricultural soil use is impossible
without a system approach. In its turn successful practical realization of a system approach
under agricultural ecosystem exploitation is impossible without small-scale mapping.
“Agroecological Soil-Reclamative Map of the Nechernozemnaya Zone of the European
Russia”* in scale 1:1 500 000 is an attempt to apply a system approach towards small-scale
mapping and classification of geographical landscapes for the purposes of agriculture.

Since a system approach in geography is a landscape approach, the map was compiled
using a landscape base. Geographical landscapes are regarded as hierarchically
superordinate natural territorial complexes of different levels composed of interdependent
and associated natural components. Thus soils are only one of the natural components of
geographical landscapes and cannot be considered out of context. The map does not give
integral cartographical representation of geographical landscapes, i.e. it is not “a map-
conclusion”. Due to the layer by layer cartographical representation of the landscapes
components and their characteristics, the map is, in fact, a Geographical Information System
(GIS). It contains: 1) spatial distribution of natural and agricultural geographical landscapes
(agrolandscapes) of different levels and characteristics of their natural components (in other
words, climatic and natural conditions); 2) agroecological alternatives of soil use; 3)
reclamative and erosion control measures; 4) factors complicating the carrying out of
reclamative and erosion control measures; 5) spatial distribution of land areas inadaptable
and restrictedly inadaptable for agricultural use, derelict lands and forest fund.

There are seven hierarchical levels of the geographical landscapes on the map, the bases
of division of which are characteristics of their natural components. Notations of geographical
landscapes of concrete hierarchical level are formed by banding of notations of
characteristics of their natural components which are the bases of division of this and all
higher levels.

Application of a system approach towards small-scale mapping and classification of

geographical landscapes for the purposes of melioration and agriculture has enhanced the
following conclusions to be draw:
1. The validity of landscape classifications cannot be checked up with using a small-scale
landscape “maps-conclusions” with one type boundaries for all subdivisions.
2. Small-scale agroecological maps cannot be “maps-conclusions” because of
insufficiency of graphic arts for cartographical representation of the information held on them.
3. Subdivisions of “intrazonal” geographical landscapes are of no classification value and
do not fill well into an integrated hierarchically superordinate classification system.
4. Morphologic (horizontal) structure of geographical landscapes cannot be used as the
basis of their division because of subjectiveness and ambiguousness in the way they were
constructed.
* The map covers ~2.5 m sq. km (27 constituent entities of the Russian Federation) and has been
prepared for publication in two versions – in the form of bound album and as a wall chart.

626
5.5 Posters

Biophysical surveys integration and multifunctional gis management for running

sustainability-conscious agro-based projects of the millennium challenge
account programs.

E. Amamoo-Otchere1, F. Mensah2
1
Centre for Remote Sensing and Geographic Information Services (CERSGIS), University of
Ghana, Legon;
e-mail: [email protected]
2
Principal Applications Specialist, Centre for Remote Sensing and Geographic Information
Services (CERSGIS), University of Ghana, Legon

Reportedly, Ghana is to benefit from United States Government special grant “MILLENNIUM
CHALLENGE ACCOUNT (MCA), which is basically for implementation of rural-based
agriculture, transportation and community-based agro-related social infrastructure. It is
intended to benefit some of the poorest rural districts where poverty rates range between 40
and 90% of the national standard. It is supposed to raise the income potentials of farmers
through increased production of high-value cash crops along side basic food crops, improve
transportation network and integrate food processing industries and produce export handling
facilities. Twenty-three districts are to benefit from the programme. The components
enhancement of profitability of commercial agriculture among small-scale farmers,
improvement of land tenure to facilitate security of small-scale holdings, improvement of the
Volta Lake transport in order to open the Afram Basin to the outside market, and
improvement of small-scale irrigation.
One of the pre-project challenging tasks is development of comprehensive, community-level
databases for the project. The current data sets on the Biophysical Environment, Land
use/Land ownership, Community’s Social, Economic and Cultural Profiles are not ready for
use. Collection and processing of essential data will take some considerable time. Timeliness
will be essential in data provision for critical decision in the project planning stages. Robust
geographic databases on all the spatial and non-spatial data sets are basic requirements that
the initial phase should give priority to.
Some community-based social infrastructure facilities are being compiled for each
administrative district. However, the biophysical and land use/land land tenure databases are
not yet tackled. The surveys for establishing the requisite databases need to be commenced
in time to facilitate availability of the essential land-related data in user-friendly formats. The
existing datasets do not have the quality and quantity required for the project. An initial
integrated survey should target high resolution database building for soils, hydro-geology,
agro-meteorology, vegetation (at plant species level including ethno-botany), and habitats of
endemic parasitic animal/human disease vectors.
Remote Sensing tools will be used as part of the surveillance technology. Past, current and
future aerospace images and GPS-based point-data acquisition will form part of the GIS-
database building processes. The database will require that all data sets are geo-referenced
to the common base map, the 1/50,000 scale topographic map with coordinates compatible
with GPS values, in turn referenced to the satellite images. All other existing spatial data
(soil, vegetation, water, hydrology/ hydrogeology, agro-climate, among others) will be a set of
baseline data processed and linked to the base map. The database organization will have to
recognize the administrative structures of the district assemblies since the assemblies will
play the implementing role of the projects.The paper utilizes satellite images, other existing
environmental data sets converted to GIS-format to illustrate aspects of the survey
integration GIS-database building processes.

627
5.5 Posters

Structural thresholds at landscapes, across different habitat amount and

aggregation levels

M.C. Ribeiro, J.P. Metzger, A.C. Martensen

Laboratório de Ecologia da Paisagem e Conservação – LEPAC, Instituto de Biociências,

Universidade de São Paulo, Rua do Matão - travessa 14 no 321, São Paulo, Brazil.
e-mail: [email protected]

Thresholds can be understood as a state-and-transition point or zone below or above, which

abrupt changes can be observed in the response variables. Despite some studies
investigating landscape metrics and their relationship with gradients of area and aggregation
(Neel et al, 2004), structural thresholds above which cover and configuration could vary
independently are still unknown. We have investigated the behavior of landscape metrics
across habitat amount and configuration gradients, in order to identify cover thresholds about
which the spatial arrangement of habitat patches presents little or none importance. Neutral
habitat and non-habitat landscapes were generated using the RULE program. Nineteen
levels of habitat amount (P= 5% to 95%, with steps of 5%) and ten levels of aggregation (H=
0.1 to 1, with steps of 0.1) were used as input parameters. Neutral landscapes were
simulated combining P and H, with 100 replicates, in a total of 19,000 binary maps.
Landscape metrics (n= 59) were computed at class-level for the following metric groups:
area, shape, core area, isolation, proximity, contagious and connectivity. As many of class-
metrics are correlated, we select only those that present complimentary information, i.e.
poorly correlated. Linear, Logarithm and Piece-wise regressions were applied to habitat
amount and class-level landscape metrics, for each aggregation levels. Metric selection was
carried out using Principal Component Analysis (PC). Akima´s method for bivariate
interpolation and smooth surface fitting was used to generate 3-D surfaces for selected
metrics (as Z-axis), having P and H levels as X and Y-axis. We accounted 86.2% of total
variance in the first five components (PC1 to PC5). PC scores for these five PC´s helped the
selection of following landscape metrics: PROX_MN (mean proximity index), DCORE_MN
(mean disjunct core distribution), MESH (relative measure of patch structure) and SPLIT
(cumulative patch area distribution). Piece-wise regression between PROX_MN and P levels
showed highest r2 values (from 0.74 to 0.95) when compared to linear and logarithm models.
It suggests a structural threshold between PROX_MN and P. It was also noticed that each
aggregation (H) levels changed the amount of habitat thresholds (P_th), with negative slope.
The thresholds for PROX_MN ranged from P_th=55.6% for H=0.1 (disperse) to P_th=45.6%
for H=1 (aggregated). Piece-wise models were also better for MESH (mean r2 = 0.989),
when compared to linear (mean r2=0.901) and logarithm (mean r2=0.868). Again we observe
that exit structural thresholds were between P and MESH index. The P_th for MESH were
47.3% for H=0.1 and ranging between 50.3% and 51.1% for H>0.2 to 1. When analyzing
DCORE_MN and SPLIT, log models were better than Piece-wise, so no thresholds were
identified. The results found in this study emphasise the existence of structural thresholds on
landscape metric behaviors, and that depending on H levels, these thresholds can change
slightly. Almost for PROX_MN and MESH metrics, the P_th appear to be between 45% and
56% of the amount of habitat. Our findings could be helpful in species persistence studies.

Reference
Neel, M.C., McGarigal, K. & Cushman, S.A. 2004. Behavior of class-level landscape
metrics across gradients of class aggregation and area. Landscape Ecology 19: 435-455.

628
5.5 Posters

Incorporating Local Landscape Knowledge into the Cultural Heritage Mapping

Process

E.A. Moylan

School of Geosciences, University of Sydney, Sydney, 2006, Australia

e-mail: [email protected]

Introduction
A challenge faced by managers of heritage landscapes is the incorporation of local
knowledge into the cultural heritage mapping process. This poster presents cultural heritage
mapping at the World Heritage site of Angkor, in Cambodia. Angkor is best known for its
historically significant temples. However, management of the site involves more than the
conservation of archaeological structures because it is also home to over 100,000 people.
The history of the people living at the site constitutes a cultural landscape, but one on a
different level to that associated with the temples.

Using Map Biographies to identify Local Knowledge

Development of community maps that document cultural heritage in the context of the
landscape is an important step towards the inclusion of the local people in the management
of that community (Fox et al 2005). This poster presents community mapping activities at
Angkor that combine landscape ecology principles with ethno-cartographic techniques. The
map biography approach was used in a series interviews with the local community.

The Significance of Local Knowledge in the Identification of Cultural Heritage

Incorporating a traditional Asian landscape knowledge system into a fundamentally
scientific approach, such as landscape ecology, required a reappraisal of standard mapping
techniques. Some of the issues raised from the reappraisal include the development of local
landscape classification systems, the spatial representation of intangible heritage, and the
significance of place names as indicators of land use history. Local descriptions of landscape
types were encouraged, thereby reducing the influence of preconceived classification
systems. This approach resulted in the identification of subtle differences in landscape types,
an improved understanding of landscape changes, and information to analysis patterns due
to current land use restrictions. The inclusion of intangible heritage into cultural heritage
management is a relatively new approach to heritage management in Asia. The nature of
intangible heritage makes it difficult to identify and define. Mapping spatial attachment to the
landscape from a cultural perspective offers a way to identify and represent heritage features
in it (Byrne & Nugent, 2005). The adoption of this approach resulted in some cases where
archaeologically significant temples were demoted in value over local contemporary religious
structures. Additionally, the significance of place names was found to vary between formal
administrative names, and historical local names. Research indicates that local gazetteers
are an important part of cultural heritage mapping.

References
Byrne, D, Nugent, M. (2005) Mapping Attachment: A spatial approach to Aboriginal post-contact
heritage.Department of Environment and Conservation (NSW), Sydney
Fox, J., Suryanata, K & Hershock, P. Eds (2005) Mapping Communities: Ethics, Values, Practice.
East-West Center, Hawaii.

629
5.5 Posters

Identification of Functional Landscapes in Catalonia (NE Spain). A case-study for

the High Pyrenees Natural Park.

J.M. Serra1, M. Ninyerola2, J. Cristóbal1

1
Autonomous University of Barcelona. Department of Geography, 08193 Barcelona, Spain.
e-mail: [email protected]
2
Autonomous University of Barcelona. Department of Animal Biology, Plant Biology and
Ecology, 08193 Barcelona, Spain.

Introduction
Many attempts have been made in order to define ecological boundaries (Bryan, 2006;
Mora and Iverson, 2002) by defining regions with similar ecological characteristics. In the
present research, Functional Landscapes (FL) are identified for the High Pyrenees Natural
Park. The concept of FL relates to the physical forces driving the ecological behavior of the
study area through topographic and climatic variables. Those FL are areas that have the
same characteristics, hence the same potential to develop similar ecological regions.

Metodology

Topoclimatic data
Topographic and climatic variables have been obtained from Catalonian Digital Climatic
Atlas (Ninyerola et al., 2000) in a 180 meters spatial resolution, wisely transformed and
aggregated.

Cluster analysis
A Principal Component Analysis (PCA) has been first performed among climatic and
topographic variables in order to reduce redundancy. Subsequently, a semi-automatic non-
supervised iterative clustering method using IsoData algorithm (Duda & Hart,1973)
created 10 clusters, which divided the Natural Park in homogeneous environmental areas. A
second cluster analysis using Ward’s method has been performed for the centroid of each
cluster, establishing similarities between each FL and therefore enabling a creation of a
hierarchical legend that groups FLs at a different scale.

Functional landscapes map evaluation

Map evaluation has been performed using former landscape cartography (Chevalier,
1929), actual vegetation maps, forestry suitability maps and field surveys.

References
Bryan, B.A. (2006) Synergistic techniques for better understanding and classifying the environmental
structure of landscapes. Environmental Management 37: 126-140.
Chevalier, M. (1929) Les Paysages catalans: leurs aspects, leur structure et leur évolution, Librairie
scientifique Albert Blanchard, Paris.
Duda, R.D. & Hart, P.E. (1973) Pattern classification and scene analysis, John Wiley & Sons, New
York.
Mora, F. & Iverson, L. (2002) A spatially constrained ecological classification: rationale, methodology
and implementation. Plant Ecology 158: 153-169.
Ninyerola, M; Pons, X. & Roure, J.M. (2000) A methodological approach of climatological modelling
of air temperature and precipitation through GIS techniques. International Journal of Climatology
20: 1823-1841.

630
5.5 Posters

Application of BioHab in the Israeli mediterranean zone: a test of the habitat

sampling method

L. Olsvig-Whittaker1, M. Walczak1, D. Rotem1, Y. Magal1, S. Amir1, A.

Perevolotsky2, I. Tauber3, R. Einav4, A. Zahavi5, Y. Carmel6, M. Sternberg7, R.
Frumkin8, T. Achiron-Frumkin8

1
Science Division, Israel Nature and Parks Authority, 3 Am Ve Olamo Street, Jerusalem
95463, Israel.
email: [email protected].
2
Dept. of Agronomy and Natural Resources, ARO, the Volcani Center, P.O. Box 6, Bet
Dagan 50250, Israel. 3 Keren Kayemit Leumi, Forest Dept, PO Box 45, Kiryat Hayim, 26103,
Israel. 4Blue-Ecosystems Environmental Consulting, Hagat 177, Zichron Yaakov, 30900,
Israel. 5 Beit Berl College, Doar Beit Berl, 44905, Israel. 6 Faculty of Civil and Environmental
Engineering,Technion -- Israel Institute of Technology, Haifa 32000 Israel. 7 Department of
Plant Sciences, Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv 69978, Israel. 8
Environmental Consultants. P.O. Box 2444, Mevasseret Zion 90805, Israel.

Introduction
BioHab is a landscape oriented, species-independent method for habitat description and
classification which has been developed for use in Europe (Bunce et al. 2005, see also
www.biohab.alterra.nl). Several organizations in Israel with responsibility for land
management have expressed an interest in application of this method in our country. In
March 2006, we held a workshop on the subject in Israel with instructors Bob Bunce and
Marc Metzger from Alterra, home base of BioHab in the Netherlands. The result of this
workshop was an agreement to continue trials and assessment of the method. We translated
the coding forms to Hebrew, eliminated those codes irrelevant to Israel, and added some
terms peculiar to our country, in preparation for a second field experiment.

Field trial at Park Britannia

In this stage, we explored the possible use of this method in a field trial conducted in
February 2007. Our study site, Park Britannia, was a 4,000 ha area with a mixture of planted
forest and natural shrub land, in Israel's Mediterranean-Arid transition zone. Participants from
several interest groups (conservation, rangeland, forestry, academic ecological research,
regional planning, environmental impact assessment, and education) did joint sampling and
assessed the field sampling from their professional perspectives.

Conclusion
It seems that BioHab can be used outside the habitat range of Europe, with real advantages
where the taxonomic description used in classic habitat classification is weak or irrelevant.
However, adaptations of the method must be made to incorporate new conditions not found
in Europe. A simplification of the method needs to be developed for use by beginners and
groups with limited time and resources. Analytical approaches after data collection will also
become very important.

References
Bunce, R.G. H., G.B. Groom, R.H.G. Jongman, and E. Padoa-Schioppa (eds.) 2005 Handbook for
Surveillance and Monitoring of European Habitats. Alterra report 1219, ISSN 1566-7197. EU FP5
project EVK2-CT-2002-20018. Alterra, Wageningen, the Netherlands.

631
5.5 Posters

Landscape Types of the Czech Republic

D. Romportl, T. Chuman, Z. Lipský

1
Department of Physical Geography and Geoecology, Faculty of Science, Charles University
in Prague, Albertov 6, Praha 2, 128 43,
e-mail: [email protected]

Landscape classification and typology represents one the most important subject of study for
landscape sciences in the Czech Republic. Although different landscape typologies were
developed in the Czech Republic in the past, none of them is applicable to the recent
situation. The present poster introduces a new methodological system of complex landscape
typology. The basic difference is that the typology presented here is based on exact, easily
quantified data covering both natural and cultural landscape conditions, which can be
classified in GIS.

In a first step, critical review of available environmental data has been undertaken in
order to select the following suitable data sources for the delineation of the major natural
landscape units:
• climate
• relief
• parent material

All data inputs were firstly generalized and converted to raster datasets with the same
pixel size. These raster datasets were analyzed by statistical functions implemented in a
Spatial Analyst extension for ArcGIS. In next step, these three layers were stacked into one
RGB colour composite as a TIFF file, which was segmented in a specific way using the
eCognition software. This object-oriented image classification software allows delineation of
polygons represented natural landscape types in various scale levels.
Additionally, the following data about anthropogenic influences on the landscape were
examined in a Spatial Analyst extension for ArcGIS:
1. land use/land cover
2. secondary landscape heterogeneity

By combining chosen raster datasets a new unique raster dataset was created
where each pixel has a unique combination of selected characteristics. All pixels of the same
summary characteristics represent a particular landscape type. The unique landscape types
were generalized and combined with similar ones in polygons defined by eCognition
segmentation process. The output of the used methodology is a map representing a
preliminary landscape typology of the contemporary Czech landscape that will be the subject
of further modifications and interpretations.

632
5.5 Posters

Habitat Guide of the Landscape Ecological Habitat Mapping of Hungary

J. Bölöni1, A. Kun 2, Z. Molnar 1, E. Illyés 1

1
Institute of Ecology and Botany of HAS – 2163, Vácrátót, Alkotmány 2-4., Hungary
e-mail: [email protected]
2
1037, Budapest, Kolostor 2. Hungary

Introduction

To conserve our natural heritage and for developing sustainable landscape

management strategies, it is essential to know on country level what kind of landscapes do
we have with what kind of habitats and in what state are they in. To fulfill these demands
actual data on the distribution and conservation status of the Hungarian habitats were
needed, since so far there was no map of the actual vegetation of Hungary. Therefore a
satellite image supported field mapping of (semi)natural habitats of all Hungary (MÉTA) was
carried out in a hexagonal grid of 35 hectares. List and area proportions of habitats in each
hexagon, and 17 other attributes including naturalness, threats, presence of invasive
species, land use and landscape-ecological attributes were documented. The Habitat Guide
presented here was compiled especially for the purposes of this survey.

The aim and structure of the Habitat Guide

The aim of the Habitat Guide was to serve as a comprehensive basis for the habitat
survey and to ensure the equal quality of the mapping, since nearly 200 mappers were
involved in this project. However, the Habitat Guide is very suitable for other habitat surveys
beyond the MÉTA project as well.
The Habitat Guide is generally based on the Hungarian National Habitat Classification
System (Á-NÉR) developed in 1997 and used successfully in many vegetation surveys so
far, but it was modified and extended considerably. The habitat classification system of the
Habitat Guide reflects the traditional phytosociological views while the emphasis is put more
on the physiognomical approach and the site conditions. It contains 86 habitat types, with
detailed descriptions of the 81 semi-natural habitats and 5 degraded habitat complexes.
The Habitat Guide describes each habitat type in 1500-2000 words in the following
chapters: definition, site conditions, typical stand structure, characteristic species, vegetation
context in which the habitat type occurs, subtypes (with short descriptions), types not
belonging here (the correct category is given), patterns on the satellite image characteristic
for the certain habitat type, naturalness-based habitat quality and regeneration potential.
There are many examples and explanation given on characteristic and specific situations
seen on the field by experianced mappers.

Uniqueness of the Habitat Guide

The uniqueness of the guide stands in the fact that besides the 21 authors, all of the 200
MÉTA mappers contributed to its compilation in two steps. Firstly, they had to review at least
10 habitat types from the Habitat Giude and report in writing before they started the mapping,
and secondly, during the discussions on the compulsury three-day-long field-trips as
preparation for the mapping they raised their questions and added their comments, which
were inbedded in the final version of the Habitat Guide.

633
5.5 Posters

Programme for planned biodiversity studies: ecologically robust and cost-

effective site-based ecological data for multi-purpose research and management

J-M Hero 1, W.E. Magnusson 2 , B.E. Lawson 3

1
Centre for Innovative Conservation Strategies, School of Environment, Griffith University,
Gold Coast Qld 9726, Australia.
e-mail: [email protected]
2
Instituto Nacional de Pesquisas da Amazonia (INPA), 2.936, Petrópolis, Manaus,
Amazonas, CEP 69060-001, Brazil.
3
School of Natural and Rural Systems Management, The University of Queensland, Gatton
Qld 4343, Australia.

Nationally and internationally, long-term site-based datasets are integral to natural

resource management for purposes including condition assessment, impact assessment,
population estimation, planning and modelling. The Programme for Planned Biodiversity
(PPBio) project involves integrated, standardised long-term ecological survey grid based on
a meso-scale grid of survey trails (5km x 5km), with 30 permanent terrestrial plots (Fig. 1)
and permanent aquatic plots as required.

Figure 1. The layout and design of a standard PPBio long-term ecological survey grid with
30 plots of 250m length and varying in width depending on the taxon measured.

The PPBio programme was developed in Amazonas Province, Brazil, for baseline
ecological monitoring and research. The resultant ecological understanding and cost-
effective design has seen its rapid expansion across Brazil, with substantial interest from
conservation agencies across the globe.

One such project is the PPBio Australasia consortium which has recently established a
survey grid in south-east Queensland, Australia. This consortium includes local government
and natural resource management bodies which are eager for a survey methodology that
can efficiently and effectively integrate ecological research at the site-level to achieve long-
term ecological monitoring in addition to one-off ecological studies by demand. The PPBio
design offers a unique program for multi-purpose and multi-scale biodiversity research, and
excellent opportunities for applied management outcomes.

This poster describes the history, design and applications of the PPBio approach, with
specific focus on the Brazilian and Australian programmes.

634
5.5 Posters

Evaluation of the vulnerability to floods on the state of Tabasco, Mexico

L. Gama1, C. Zequeira-Larios1, L. Giddings2, M. Soto-Esparza2, A. Galindo-

Alcantara1, A. Hernadez-Moralez1
1
Universidad Juarez Autonoma de Tabasco Km. 0.5 Carretera Villahermosa-Cardenas,
Entronque a Bosques de Saloya, 86030 Villahermosa, Tabasco, Mexico
e-mail: [email protected]
2
Instituto de Ecologia, A. C. Km. 2.5 Carretera Antigua a Coatepec 351, Congregación El
Haya, 91000 Xalapa, Veracruz, Mexico

According to the information recorded by the Mexican Secretary of Environment and

Natural Resources, floods constitute the most important type of disaster in Southern Mexico.
The damages resulting from floods have the largest economic, social, and environmental
consequences. Although it is impossible to eliminate floods; it is possible to minimize their
effects via projects, and activities based on data regarding vulnerability areas that will protect
the economic and social infrastructure. Tabasco has undergo important land cover changes
due to agricultural, cattle raising activities and petroleum extraction, that affected natural
landscapes. Climate change is also expected to be a cause of important changes because of
the geographical conditions of the state. An analysis of the vulnerability to floods due to the
physical condition of the area, the changes on the coastal line and the increase on
precipitations was done to configure a map.
Precipitation data from the Global Historical Climatology Network (GHCN) and additional
Mexican data from Eric II (compact disk produced by the Mexican National Water
Commission), data were assembled from stations in southern Mexico. Monthly precipitation
data were converted to SPI and interpolated to form monthly contour maps of SPI of the
region. The Standardized Precipitation Index (SPI) developed by McKee et al. (1993)
categorizes the observed rainfall as a standardized departure with respect to a rainfall
probability distribution function. that compares the precipitation of one month with the entire
record of precipitation was applied to individual months of a reporting station as well as to
longer periods.

Aerial photos and a historical review were performed to evaluate changes on the coast
line during the last 20 years (Ortiz Perez, et al. 2006). A great Coastal lost of between 5 to
20 meters each year as well as damage to infrastructure shows that a monitoring system
has to be implemented to evaluate the future impact on the rest of the state. Both factor
increase on precipitation and changes on coastal lines are risk factors for an important
amount of population of the state that have to be monitored.

References
McKee, T. B., Doesken, N. J. & Kleist, J. (1993). Drought monitoring with multiple timescales. Paper
presented at the Preprints, Eighth Conference on Applied Climatology, Anaheim, California.
Ortiz Perez, M. A., A. P., Mendez Linares & J. R., Hernandez Santana, (2006), Sea level rise and
vulnerability of coastal low-land in the Mexican area of the Gulf of México and the Caribbean Sea
(Chapter 14). In: Ecosystem-Based Management in the Gulf Of Mexico (eds. J. W. Day & A.
Yañez-Arencibia), multi-volume series “The Gulf of Mexico: Its Origins, Waters, Biota Human
Impacts”. Texas A & M University Pres.

635
5.5 Posters

Evaluation of farm woodlands as agri-environmental schemes in rural landscapes

M. Sigura, A. Marelli

University of Udine. Department of Agricultural and Environmental Sciences, via delle

Scienze 208, 33100 Udine, Italy.
e-mail: [email protected]

Introduction

Several agri-environmental schemes funded by the European Union have been aimed at
increasing the amount of woodland on formerly agricultural land (EEC Regulation 2080/92
and 1257/99) and the evaluation of their effects on biodiversity is required.
The aims of this paper were to assess the value of that new woodland in preserving
biodiversity, using bird species richness as an indicator, and to evaluate the relations
between landscape pattern and bird species richness

Methods and Results

The study was carried out in lowlands of Friuli Venezia Giulia, North-eastern Italy. Bird
species richness in 28 farm woodland was compared with the one in 26 control habitats, the
latter being patches of semi-natural woods and shrubs within 1 kilometre. Bird survey was
based on one point count (10 minutes, no fixed radius) carried out monthly in each farm
woodland and control habitat between February and June 2005. The habitat was analysed
by the Index of Vegetation Complexity (MacArthur & Horn, 1969, Bibby et al, 2000 and
Radtke & Bolstad, 2001 modified) that describes structural complexity of vegetation for both
farm woodlands and control habitats.
The analysis of landscape pattern was based on the land use of a 2 km buffer around the
habitats, as landscape metrics we consider: at the patch level (farm woodland and control
habitats) the patch area, patch perimeters, shape factor, at different spatial scales (500 m,
1000 m, 2000 m) the relative presence of land use classes. The overall list of species
recorded in farm woodlands doesn’t differ greatly from that in control habitats, but frequency
of each species is usually lower in farm woodlands. On average you are likely to encounter
2.6 species in a given farm woodland vs. 8.7 species in a control habitat. Best regression
model (best sub-set method) includes 8 variable, fits in well with observed data and shows a
high degree of significance (R2 = 0.745, F8.42 =15.35, p<<0.001). The most important
contributions to regression model are by vegetation complexity index, perimeter and
presence of farm woodland in surrounding landscape. Important are also the presence of
edges, woodland and arboreal cultivation but with a different weight at the different spatial
scales. The results show that the variable which best explains low species richness is the
poor vegetation structure of farm woods: vegetation complexity index is strongly correlated
with species richness and is systematically lower in farm woods. The trend of landscape
variables at different spatial scale could indicate that in fragmented landscapes, like the
agricultural landscape, the birds need landscape diversity as number of different patches in
order to satisfy their needs and maintaining one vital population.

References
Bibby C. J., Burgess N. D., Hill A. H., Mustoe S. (2000) Bird census techniques – Second edition.
Academic Press, London.
MacArthur R.H. & Horn H.S. (1969) Foliage profile by vertical measurements. Ecology, 50 (5) : 802-
804.
Radtke P.J. & Bolstad P.V. (2001) Laser point-quadrat sampling for estimating foliage-height profile
in broad-leaved forest. Can. J. For. Res., 31: 410-418.

636
5.5 Posters

Characterisation of the rural landscape. Application in a case study of the plain of

North Eastern Italy

M. Sigura, A. Treleani, P. Bonfanti

University of Udine, Department of Agricultural and Environmental Sciences, Via delle

Scienze, 208_33100_ Udine_ ITALY
e-mail: [email protected]

Introduction

In the modern age the changes in land use and land cover modified quickly the rural
landscape with sudden and wide effects that produced visual uniformity and anonymity with
threats to social identity and quality of life, loss in environmental heterogeneity and
biodiversity as well as cultural heritage. It is necessary to study the proprieties of landscape
changes at different levels (Bastian O. et al, 2006). This paper wants to analyse the changing
of visual characters of the rural landscape in different conditions of land use with the focus on
the rural landscape as perceived by people. The aim was to develop and test a methodology
for the application of visual character analysis, not only qualitative, applied with a multiscale
approach to a plain of a single municipality (Friuli Venezia Giulia, Italy) that presents inside
two specific landscape patterns. From the analysis it was revealed that the first one is
correlated to a land consolidation of the second half of XX century with strong simplification
effects of landscape and environment, while the second keeps clear marks of the
stratification of land use of the past with a larger level of landscape and environmental
heterogeneity.

Methods and results

In this study, indicators (related to diversity and complexity) in the two different areas at
two distinct scales were measured and, subsequently, dates achieved were integrated with a
common derivate index. At the macro scale the photograph relief was developed playing
visual shots related to vantage points (bell towers) and separating the whole diorama in 4
view cones in correspondence with cardinal points. With a common process the photos were
standardized and analyzed referring to 7 classes of visual factors elements (intensive
agriculture, meadows, edges, private green belt, streets/paths, dwellings, factories). At the
detail scale for every macro view cones several foreground points (rural streets) were
selected. They are matched with the global character of the area. The indices calculated
were: the MSIEI at the macro scale, the Normalized Index of Linear Complexity (NILC) at
detail scale, as the medium value of linear structures perceived (normalized to scale 0-1),
and the Derivate Diversity Index (DDI) as the sum of MSIDI and NILC to integrate both
scales. In the DDI the NILC is an incremental factor that adjusts the diversity value calculated
at the macro scale with the complexity value surveyed at the detail scale of the perceived
landscape. MSIEI shows a general simplification of the pattern of both considered areas, but
the values of NILC show different levels of complexity derived from the perception of linear
structures at the detail scale that are higher in the area whole land consolidation has no
interest. We can notice that at the increasing of the IDD values the visual diversity of the
landscape pattern and the linear complexity of the visual units grows. This scale integration
method has integrated the landscape diversity of the two areas providing a further visual
survey instrument to the landscape diagnosis.

References
Bastian O., Krönert R., Lipsky Zdenĕk (2006), Landscape diagnosis on different space and time
scales_ a challenge for landscape planning, Landscape Ecology: 374 -pp.359-374

637
5.5 Posters

Analysis of spatial structure of landscape cover classes of sub-basin Balsas,

south of Maranhão state, Brazil

L. Barreto1, M.C. Ribeiro2, E. Pontes3, A. Veldkamp4

1
Federal University of Maranhão, Oceanography and Limnology Department, Portugueses
Avenue, s/n., 65080-040, São Luís, MA-Brazil.
e-mail: [email protected]
2
University of São Paulo, Landscape Ecology and Conservation Laboratory, São Paulo, SP-
Brazil
3
State University of Maranhão, Lansat Laboratory, 65055-970, São Luís, MA-Brazil
4
Wageningen Agricultural University, Lab. Soil Science and Geology, The Netherlands

Recent studies suggest that land use and land cover change dynamics are also
relevant for site selection when one is are aiming for conservation and restoration of natural
ecosystems (Veldkamp et al., 2001; Verburg et al., 2004). Several metrics such as spatial
configuration, connectivity, can be computed for patches, classes and landscapes, helping
the analysis task. Combination of some of them can express the fragmentation status of an
interest region. Also multi-scale analyses are essential to improve de understanding of the
role of regional scale on the local scale processes.
We analyzed the spatial structure of remaining patches of Balsa’s sub-basin, south
of Maranhão state of Brazil. As a consequence of arable land expansion, natural habitats
suffered severely, causing significant impacts on regional flora and fauna. Understanding
how the remaining habitats are spatial distributed and configured is essential for developing
an effective Conservation Area Network definition. In this work we derived a cover map from
Landsat/TM 5 images acquired in 2000, with the classes: cerrado (brazilian savanna) and
riparian forests; open fields (campo cerrado) and pasture; human activities (crops and
buildings) and d) water body.
The whole region was subdivided into 322 hexagonal cells (planning unit - PU) of
analyses, each with 10,000 hectares. Several structural landscape metrics were computed
for the cells (MacGarigal & Marks 1995), only the metric percentage of native cerrado and
riparian forests (PLAND) are discussed here. The metric was computed for two spatial
scales: a) local scale which included the amount of native forests within the cells; b) regional
scale, computed the amount of native forest in a virtual hexagonal cell of about 50,000
hectares, centered on the midpoint of each local (10,000 hectare) cells.
Our preliminary results shows that when considering only a local scale, about 22%
of the total number of P.U. presented good PLAND values (>60%), 18% presented
intermediate PLAND (>40-60%) and 60% of the regional have low PLAND (<40%). But when
we analyzed the regional scale over the local scale, we noticed that some PU´s shift for both
lower and higher PLAND values. Taking in account our two scales of analyses, local and
regional, we classified the PU´s and suggested some preliminary management actions. We
believe that a multi-scale approach of analyses could provide good information to help the
first steps of the definition of a CAN.

References
Mcgarigal, K. & Marks, B. J. (1995) FRAGSTATS: spatial pattern analysis program for quantify
landscape strucure. USDA. Forest Service Pacific Northhwest Research Station, Portland, OR.
Veldkamp, A.; Verburg, P. H.; Kok, K.; de Koning, G. H. J.; Priess, J. & Bergsma, A. R. (2001)
The need for scale sensitive approaches in spatially explicit land use change modeling.
Environmental Modelinmg and Assesment 6: 111 –121.
Velburg, P. H.; Schot, P. P., Dijst, M. J. & Veldkamp, A. (2004) Land use change modelling: current
practice and research priorities. Geojournal 61: 309-324

638
5.5 Posters

Assessment of a land change model using a three-dimensional matrix at multiple

scales

S. Peethambaram, .G. Pontius Jr

Clark University, 950 Main Street, Worcester MA 01610, USA.

e-mail: [email protected]

This paper proposes a novel statistical method to assess the mapped output from a
spatially-explicit model that simulates change over time among land categories. This paper
allows each pixel in the map to have partial membership to more than one category.
Therefore it is possible to compute the statistics at multiple scales. The technique considers
three maps: 1) a reference map of time 1, 2) a reference map of time 2, and 3) a prediction
map of time 2 from a model that predicts the land change between times 1 and 2. The three
maps derive from an application of the SAMBA modelling project in Vietnam (Castella et al.
2005). Figure 1 shows that there is more error than correctly predicted change at the 30-
meter resolution of the raw data, which is common for land change models (Pontius et al. in
press). Two of the three types of errors shrink substantially as resolution becomes coarser.

Figure 1. Percent of landscape categorized as correct and erroneous at multiple scales.

References
Castella, J.-C., Boissau, S., Trung, T.N., & Quang, D.D. (2005) Agrarian transition and lowland-
upland interactions in mountain areas in northern Vietnam: Application of a multi-agent simulation
model. Agricultural Systems 86(3): 312-332.
Pontius Jr, R.G., Boersma, W., Castella, J.-C., Clarke, K., de Nijs, T., Dietzel, C., Zengqiang, D.,
Fotsing, E., Goldstein, N., Kok, K., Koomen, E., Lippitt, C.D., McConnell, W., Pijanowski, B.,
Pithadia, S., Mohd Sood, A., Sweeney, S, Ngoc Trung, T., Veldkamp, A.T. & Verburg, P.H..
2007. Comparing the input, output, and validation maps for several models of land change. Annals
of Regional Science (available at www.clarku.edu/~rpontius).

639
5.5 Posters

Delineation of landscape units at diverse scales using moving windows for

heterogeneity analysis

E. Diaz-Varela1, R. Crecente-Maseda 2
1
LaboraTe. Escola Politécnica Superior- Campus Universitario s/n. 27002-Lugo (Spain). Er-
mail: [email protected]
2
LaboraTe. Escola Politécnica Superior- Campus Universitario s/n. 27002-Lugo (Spain).

Delineation of landscape units should reflect the spatial variation of the heterogeneity of
landscapes and their hierarchical organization in spatial levels, according to their functional
and structural characteristics. In this work, a method involving the use of moving window
analysis at multiple scales was developed, as a simple but straight-forward approach to the
delineation of landscape units at several scales. The use of moving windows for calculation
of landscape metrics (see e.g. Eiden et al., 2000; Riitters et al., 2000) can provide useful
information on the spatial distribution of landscape heterogeneity, whereas the variation on
window size can be used to emulate different scale levels. The different domains of scale
thus identified can be interpreted as building up a hierarchy in the complexity of space
(Zonneveld, 1995), and each of them can be considered as landscape units.

As landscape heterogeneity is conditioned at intermediate levels by the spatial distribution

of vegetation, animal populations, and land forms (Meisel & Turner, 1998), digital data on
distribution of land cover (raster map, 10x10m pixel resolution, obtained from photo-
interpretation) and physiographic characteristics (slope map derived from 1:5.000 digital
cartography) were used for the analysis. Analyses were made using window capabilities of
FRAGSTATS software (McGarigal et al., 2002), using windows of 250, 500, 750, 1000, 1250
and 1500 m of diameter. Then, majority filtering was done on the slope map using ArcGIS,
following a similar array of scales of analysis than in the land cover map.

Results obtained from the analysis of the land cover map allowed distinguishing areas in
which the level of heterogeneity was constant against other with scale-dependent changes.
Such areas were afterwards overlaid with the results for the slope map allowing the
differentiation among several landscape units. Such units were then organized in a nested-
hierarchical structure, obtaining a final land-unit, multiscale defined map. A subsequent study
of the composition and configuration of the landscape structure in the delineated units
verified differences among their characteristics.

References
Eiden, G.; Kayadjanian, M.; Vidal, C. (2000): Quantifying Landscape Structures: spatial and
temporal dimensions. Retrieved on date 12-18-2006, from:
http://ec.europa.eu/agriculture/publi/landscape/ch2.htm#2.
Meisel, J. E. & Turner, M. G. (1998) Scale detection in real and artificial landscapes using
semivariance analysis. Landscape ecology 13: 347-362.
Riitters, K.H.; Wickham, J.D.; Vogelmann, J.E.; Jones, K.B. (2000): National land-cover pattern
data. Ecology, 81(2): 604-604. Retrieved on date 12-18-2006, from:
http://www.esapubs.org/archive/ecol/E081/004/metadata.htm
Zonneveld, I.S. (1995) Land Ecology: an introduction of landscape ecology as a base for land
evaluation, land management and conservation. SPB Academic Publishing, Amsterdam, The
Netherlands
McGarigal, K.; Cushman, S.A.; Neel, M.C.; Ene, E. (2002) FRAGSTATS: Spatial Pattern Analysis
Program for Categorical Maps. Retrieved on date 12-18-2006, from:
www.umass.edu/landeco/research/fragstats/fragstats.html.

640
5.5 Posters

Prediction of soil properties in paddy rice landscapes using terrain data and
satellite information as indicators

K. Sumfleth1, R. Duttmann1
1
Institute of Geography, Department of Landscape Ecology and Geoinformation Sciences,
University of Kiel, Ludewig-Meyn-Str. 14, 24098 Kiel, Germany.
e-mail: [email protected]

Abstract
Sustainable land management and land use planning require reliable information about
the spatial distribution of physical and chemical soil properties affecting landscape processes
and services as well. Although many studies have been conducted to identify the spatial
patterns of soil property distribution on various scales and landscapes, only less is known
about the relationships underlying the spatial distribution of soil properties in intensively used
paddy soil landscapes in southeast of China. In order to provide adequate soil information
needed for the modelling of landscape processes, such as soil water movement, nutrient
leaching, soil erosion and plant growth, this study investigates to what extend the use of
cheap and readily available ancillary information, derived from digital elevation models and
remote sensing data is suited to support soil mapping and to indicate soil characteristics on
the landscape scale. This investigation focuses on the spatial prediction of the total carbon
content as well as to soil physical soil properties such as topsoil silt, sand and clay content,
topsoil depth and plough pan thickness. Correlation analyses indicate that the distribution of
carbon and silt contents is quite well related to the NDVI (SPOT 5, ASTER and Landsat
ETM+) of vegetated surfaces on the one hand side and corresponds significantly to terrain
attributes such as relative elevation, elevation above nearest drainage channel and
topographical wetness index on the other. Geostatistical analyses furthermore reflect a
moderately structured spatial correlation of these soil variables. The combined use of the
above mentioned terrain variables and the NDVI in a multiple linear regression accounted for
29% (silt) to 41% (total C) of the variance of these soil properties. The spatial distribution of
variables as topsoil depth and plough pan thickness is mainly affected by the long term
history of land use and recent land and soil management practices (e.g. intensity of
ploughing, puddling, rice straw management), so that no and less narrow relationships and
structured spatial dependencies could be recognized here. In order to select the best
prediction method to accurately map soil property distribution, we compared the performance
of different regionalization techniques, as multi linear regression, simple kriging, inverse
distance to a power, ordinary kriging and regression kriging. Except the prediction of topsoil
clay content, in all other cases regression kriging model “C” performed best. The regression
kriging model “C” is assumed to be suited to reduce soil sampling density and to contribute to
a time and cost saving soil mapping on the landscape scale even in intensively used paddy
soil landscapes.

References
Herbst, M.; Diekkrüger, B. & Vereecken, H. (2006) Geostatistical co-regionalization of soil hydraulic
properties in a micro-scale catchement using terrain attributes Geoderma 132: 206-221
Odeh, I.O.A.; McBratney, A.B. & Chittleborough, D.J. (1995) Further results on prediction of soil
properties from terrain attributes: heterotopic cokriging and regression-kriging. Geoderma 67:
215-226
Simbahan, G.C.; Dobermann, A.; Goovaerts, P.; Ping, J.; & Haddix, M.L. (2006) Fine-resolution
mapping of soil organic carbon based on multivariate secondary data. Geoderma 132: 417-489.
Yanai, J; Lee, C.; Kaho, T.; Iida, M.; Matsui, T.; Umeda, M. & Kosaki, T. (2001) Geostatistical
analysis of soil chemical properties and rice yield in a paddy field and application to the analysis of
yield-determining factors. Soil Science and Plant Nutrition 42: 291-301.

641
5.5 Posters

Object-based classification of rural landscapes using remotely sensed data of

various resolutions

K. Hara1, N. Kamagata2, T. Ishitsuka3, M. Tomita4

1,4
Faculty of Information Sciences, Tokyo University of Information Sciences, 1200-2 Yatoh-
cho, Wakaba-ku, Chiba 265-8501, Japan.
e-mail: [email protected]
2,3
Graduate School of Informatics, Tokyo University of Information Sciences, 1200-2 Yatoh-
cho, Wakaba-ku, Chiba 265-8501, Japan.

Introduction

Effectiveness of object-based classification of land-cover using very high resolution (VHR)

data such as IKONOS has been demonstrated in recent research (eg., Burnett & Blaschke,
2003, Kamagata et al., 2006, etc.). We examined the effectiveness of object-based
classification for rural landscapes in central Japan, where small-scale agricultural practices
and complex topography produce a complicated pattern of landscape. In object-based
classification, the target area is divided by segmentation processing, and each segment
identified is considered to be one image object. The segmentation of an image allows for
groups of pixels to be considered as a single unit, or an object, within the image.

Study Sites and Methods

This research was conducted at experimental plots located in agricultural areas of Sakura
City and Sosa City, located in Chiba Prefecture, central Japan. These areas consist of
narrow, highly-branched alluvial valleys, called ‘Yatsu’, which are cut deeply into flat-topped,
plateau-like uplands. Definiens Ver.5 was employed for the object-based classification, and
classification results based on IKONOS (spatial resolution (SR): 4m), ASTER (SR: 15m) and
multi-temporal TM/ETM+ (SR: 30m) images, as well as results from aerial photograph (SR:
0.6m) were compared. Initial segmentation was a multi-resolution, bottom-up system based
on the method of Baatz and Schape (2000).

Results and Discussion

The research results showed that aerial photograph and IKONOS data were capable of
extracting small patches such as fragmented forests and small irrigation ponds, and narrow
corridors such as roads and streams. On the other hand, the lower resolution satellite data,
especially the multi-temporal TM/ETM+ data, proved suitable for identifying and mapping
basic landscape types (such as irrigated paddies and dry fields) over a wider region. These
results indicate that object-based classifications using remotely sensed data of various
resolutions can be effective tools in classifying and mapping rural landscapes.

References
Baatz, M. & Schape, A. (2000) Multiresolution Segmentation – an optimization approach for high
quality multi-scale image segmentation. Strobl, J. & Blaschke, T. (Eds). Angewandte
Geographische Informations Verarbeitung XII. Wichmann-Verlag, Heidelberg, pp.12-23.
Burnett, C. & Blaschke, T. (2003) A multi-scale segmentation/object relationship modelling
methodology for landscape analysis. Ecological Modelling 168: pp233-249.
Kamagata, N., Hara, K., Mori, M., Akamatsu, Y., Li, Y. & Hoshino, Y. (2006) A new method of
vegetation mapping by object-based classification using high resolution satellite data. Proc. 1st Int.
Conf. Object-based Image Analysis, 5pp. CD-ROM, Salzburg, Austria.

642
5.5 Posters

Stewardship and monitoring of conservation lands

P. August1, J. Coit2, D. Gregg3, R. Friday4

1
Department of Natural Resources Science, University of Rhode Island, Kingston, Rhode
Island, USA.
e-mail [email protected]
2
The Nature Conservancy, Rhode Island Field Office, 159 Waterman Street, Providence,
Rhode Island USA
3
The Rhode Island Natural History Survey, Coastal Institute in Kingston, Kingston, 1 USA
4
The Land Trust Council, The Nature Conservancy, Rhode Island Field Office, 159
Waterman Street, Providence, Rhode Island USA

Over 400 square kilometers of the state of Rhode Island have been acquired to protect
fauna, flora, and ecosystem services. The major institutional owners of conservation land
are the State of Rhode Island, United States Department of the Interior Fish and Wildlife
Service, the Audubon Society of Rhode Island, The Nature Conservancy, and small land
trusts and conservancies. The viability of any single parcel of land to support healthy
populations of native plants and animals is a function of the condition of the parcel (site-
scale) and the condition of the land around it (landscape-scale). Threats to the integrity of
land that has been purchased for conservation include invasion by pests, pathogens, and
non-native organisms; inappropriate human activities (dumping, cutting vegetation) within or
near a property; degradation of landscape context such as loss of critical dispersal corridors
to/from a property; ecological changes resulting from declines in water quality or quantity;
and succession into habitat states that do not support target species or ecological
communities. Few of the owners of conservation land in Rhode Island have the technical or
staff capacity to monitor and steward their properties. Moreover, there are no standards,
guidelines, or protocols that are universally followed in conservation land management in RI.
The consequences of not monitoring and stewarding conservation lands are the loss of their
ability to support target habitats and perform desired ecosystem services.

We are developing a framework for conservation land stewardship that consists of the
following steps: 1) Site assessment, 2) Development of stewardship goals and plan, 3)
Implementation of a stewardship plan for each property, 4) Monitoring, and 5) Synthesis,
reflection, and adaptive stewardship. Stewardship and monitoring activities draw from and
contribute to a common database on the biota of Rhode Island. Local land trusts are an
essential participant in the program as they are the portal to volunteer land stewards who
have the capacity to perform site surveys and monitoring. Landscape ecological assessment
is an integral element of the framework as it provides a regional context for conservation and
provides insight in gains/losses of dispersal corridors, upwatershed threats that would
jeopardize site-level habitats or species, and land use changes on nearby properties that
might enhance or diminish the conservation value of a particular parcel.

643
5.5 Posters

Multi-scale landscape pattern analysis for h-resolution imagery of tropical dry

forest: integration of object-oriented approach and topographic data

J.A. Gallardo-Cruz, J.A. Meave

Departamento de Ecología y Recursos Naturales, Facultad de Ciencias, Universidad

Nacional Autónoma de México, Circuito Exterior s/n, Ciudad Universitaria, México, D.F.,
Mexico.
e-mail: [email protected]

Introduction

Landscape complexity must be addressed at multiple scales when evaluating the various
levels of heterogeneity in a system (Burnett and Blaschke, 2003). This evaluation may be
based on the use of landscape indices computed for the elements recognized in a satellite
image (Turner and Gardner, 1991). The information enclosed in very high resolution imagery
sets the basis for the identification of landscape elements. However, to fully exploit this
information the spatial characteristics of the image must be used by the adequate
classification methodologies.
In this study we aimed at evaluating landscape pattern at different scales in a tropical dry
forest region. We used an object oriented classification approach to produce categorical
maps from which landscape indices were calculated. The final goal was to explore the
relationship between landscape pattern and scale in the system.

Methods

The data used come from an 8 × 8 km Quickbird image acquired in December 2005. The
image represents a region of low elevation hills (< 500 m) located in the southern part of the
Isthmus of Tehuantepec (Oaxaca), México. Land cover combines vast portions of seasonally
dry tropical forest and savanna intermingled with other plant communities and human
influenced areas (Pérez-García et al., 2001). The orthorectified and pansharpened image
was classified using an object-oriented approach. This technique allows obtaining different
size objects through an image segmentation procedure (Blaschke and Strobl, 2001). The
objects where then classified according to their spectral, textural, spatial and topographical
properties, to generate three categorical maps of the area representing landscape elements
of different scales. A set of landscape metrics were computed and compared between
scales.

Conclusions

This pattern analysis increased our understanding and improved the characterization of
related processes through a wide range of scales that constitute a landscape, and allow
perceiving a complexity of this landscape and its changes across scales that would otherwise
remain unnoticed.

References
Blaschke, T. & Strobl, J. (2001) What’s wrong with pixels? Some recent developments interfacing
remote sensing and GIS. GIS – Zeitschrift für Geoinformationssysteme 6: 12-17.
Burnett, C. & Blaschke, T. (2003) A multi-scale segmentation/object relationship modelling
methodology for landscape analysis. Ecological Modelling 168: 233-249.
Pérez- García, E.A; Meave, J. & Gallardo, C. (2001) Vegetación y flora de la región de Nizanda,
Istmo de Tehuantepec, Oaxaca, México. Acta Botanica Mexicana 56: 19-88.
Turner, M.G. & Gardner R.H. (1991) Quantitative Methods in Landscape Ecology: The Analysis and
Interpretation of Landscape Heterogeneity. Springer-Verlag, New York.

644
5.5 Posters

Evaluation of landscape characteristics derived from digital elevation models

M.V. Ravibabu, P. Negi, K. Jain

Geomatics Engineering Section, Department of Civil Engineering,

Indian Institute of Technology (IIT), Roorkee-24766, Uttaranchal, India.
e-mail: [email protected]

In Geographic Information Sciences (GIS), digital elevation model (DEM) is one of the
important input parameter for studying landform changes. DEM derives are frequently
employed throughout physical geography for applications ranging from geomorphometry
to hydrological modeling and the physiographic correction of digital satellite images. In
this present paper, DEM generated from various data sources such as topographic
maps, surveyed data and stereo satellite images (ASTER) are utilized to derived
topographic parameters like slope, aspects and shaded map to analyze landscape
variations in study area. The main objective of this paper is to evaluate landscape/terrain
changes from various DEMs. Also we will address the effects of interpolation techniques,
scaling and quality of the terrain in the study area.

645
5.5 Posters

Using conditional transit cost to identify multiple dispersal routes: an example

from the Brazilian Atlantic forest

N. Pinto

Section of Integrative Biology, University of Texas at Austin, Austin, TX, 78712, USA.
email: [email protected]

Introduction

The identification of dispersal routes using friction layers and shortest-path algorithms
provides the opportunity for scaling up from individual behavior to landscape pattern.
Shortest-path algorithms are most often used in landscape ecology to obtain the optimum
route between two points, but in practice the establishment of corridors profits from the
identification of redundant routes. Here, I use a new graph-theoretical approach named
Conditional Transit Cost (CTC), which uses the shortest-path algorithm to output all possible
paths between two points along with their relative quality.

Methods

The method is illustrated with data from a 2000-Km2 region in the Brazilian Atlantic forest
(-24.3, -47.3). The area of interest is a complex landscape mosaic composed of urban
centers, farms, roads, pastures, and isolated forest patches. The goal was to model animal
movement between two conservation units, assuming species display a preference for
primary forest but can also move through secondary forest. Thus, the quality for each 50 x 50
m pixel was a function of the percent tree cover weighted by a value that was highest for
primary forest, intermediate for secondary forest, and lowest for non-forested areas.
I explored the use of two friction layers: F1 and F2, which assumed different relative
quality values for secondary forest. Each friction layer is used to build four graphs: G1, G2,
G3, and G4. For G1, I used a 8-cell neighbourhood, to obtain results similar to commercial
GIS packages such as Spatial Analyst within ArcGIS. For G2, G3, and G4, I increased the
neighbourhood to cover a radius of 100, 200 and 400 m respectively. This was done to
simulate the behavior of species that can take large jumps over habitat patches. Last, the
Conditional Transit Cost was calculated for all eight scenarios. The CTC for pixel C
considering a route between A and B is the cost to move from A to B conditional on passing
through C, using Dijktra's shortest path algorithm. The analyses were done using the Java
graph library Jung 1.7.4.

Results

As predicted, the choice of different neighbourhoods gave different results. When the
species is assumed to take large steps and the graph is built with a larger neighbourhood,
the movement routes are almost straight-lines. But in all eight cases, it was possible to
identify more than one route between the two conservation units.

Applications

By using the CTC model on different scenarios, it is possible to explore how one's
assumptions about species' movement can affect the number and quality of dispersal routes
between two points. The CTC model can also be used to produce hypotheses about animal
movement that can be tested in the field.

646
5.5 Posters

Classification of spatial unites for the forested Korean mountainous landscape

using DEM: Comparison of GIS analyzed data with landscape and memory of
people

J. Kwon, J.H. Oh, J.H. Shin, Y.K. Kim, C.Y. Park

Korea Forest Research Institute, Division of Forest Ecology, Cheongyangni 2-dong,

Dongdaemun-gu, Seoul, 130-712, Korea.
e-mail: [email protected]

Introduction

The character of Korean landscape is a mountain-based forested landscape which

covers 65% of Korean land. In history, there were traditional ecological approaches including
the Korean peninsula into a topography-based structure. The traditional approach would be
described as ‘The land is not flat, it is fluctuating (Ch'oe, 1994). The shape of surrounding was
a key factor to understand Korean nature. However, since the experience of major landscape
change during last half of the century due to war and rapid urbanization, the traditional
character has been weakened, and it is now necessary for a reconsideration of the traditional
ecological approaches to improve the landscape for the future.

What is the spatial units & how it developed?

Nine prototype landscapes

Illustrated concepts of Korean landscape patterns have been tested since 2002 (Kwon,
2002), and nine types of prototype landscape have been defined both using GIS tools and
people’s memory of Korean landscape. Aerial photographs, DEM and ArcScene with a
modified AML have been applied to classify the nine landscapes. The key issues are; can the
prototypes be classified by DEM? How big are they? Can they be useful as a vehicle which
contains various ecological data of Korean peninsula? The hypotheses are; land-use
patterns, forest types and habitats of wildlife can be influenced by the nine types, because
people use the land based on a traditional landscape concept. For example, some types are
strongly related to farming in history and topographical elements which could cause regional
micro-climate differences.

Prime spatial units & traditional forests and landscape

In previous research (2002, Kwon), the nine types were summarized as follows;
Cultivated, Basin, Basin Holy, Corridor Cultivated, Cul-de-sac Holy, Crescent Valley,
Enclosed Cultivation, Large Scale-Open Cultivation, Mountainous Settlement and Landmark
Landscape. As the results of classification using GIS tools, the three types of landscape
which are the most readily identified are Corridor Cultivated, Cul-de-sac Holy and Enclosed
Cultivation Landscape. Locations and shapes of Korean traditional village groves
(MAEULSOOP) have different characters based on the nine types. In a national park area
(Mt. Jiri), the landscape types are, in order: Cul-de-sac Holy and Enclosed Cultivation
landscape.

References
Ch'oe, C.J. (1994) HAN'GUK ŬI P'UNGSU SASANG (A Conception of 'P'UNGSU' in Korea).
Minumsa, Seoul.
Kwon, J. (2002) Sense of place – A concept of Korean prototype landscape with reference to a new
policy of urban fringe forest. Unpublished Ph.D. thesis of Department of Landscape. The university
of Sheffield. UK. pp174-186.

647
5.5 Posters

Decision tree and vegetation indices assisting land use classification in a

heterogeneous landscape, a case study in Brazil -

S. Weel1, M.E. Schaepman 2, J.G.P.W. Clevers2

1
EarthCollective – FSD, PO box 570, 6700AN Wageningen, The Netherlands
e-mail: [email protected]
2
Centre for Geo-Information, Wageningen University, P.O. Box 47, 6700 AA Wageningen,
The Netherlands

Introduction
The depiction of landscape indicators, through the assessment of land cover patterns and
processes, provides spatial information to assist a more balanced approach in the planning
and sustainable management of natural resources. Remote sensing data are the basis for
the depiction of these indicators, through classification techniques and conversion analysis.
However, the traditionally used Maximum Likelihood (ML) rule can misestimate landscape
units due to the assumption of normal distribution of the data (not true for heterogeneous
scenes). The potential of multivariate methods and the integration of spectral vegetation
indices in the recognition of vegetation units may be assessed (DeFries and Townshend,
1994). The objective of this study was to evaluate the Decision Tree (DT) algorithm and
vegetation indices in terms of their contribution to landscape assessment, seeking for more
reliable classification techniques of heterogeneous conditions in the south-east of Brazil.

Methodology
Campinas is a municipality requiring a suitable and strategic environmental policy, due to
intense anthropogenic influences on the original landscape composition. Classes of land
cover and plant functional types were depicted from Landsat TM 5 images from 1988 and
2004, through ML and DT algorithms, in order to derive the landscape pattern. Different band
compositions were considered for each classifier, where the ML used original bands and the
DT took several vegetation indices (NDVI, MVI and TCT). A conversion analysis was based
on post-classification comparison of the classified images; deriving indicators of human
induced changes and natural processes, such as expansion of human activities, degradation
and regeneration of natural vegetation.

Results and Conclusion

The results of the classification procedure indicated that the ML and DT had similar
performance, considering the bands used in each model. However, some classes had a low
accuracy due to the functions of the algorithm, biophysical variations and qualities of the
sensor. The result of the DT presents the MVI as the most important index in the data
splitting process. However, statistical analysis may be performed through a method
presented by Xu et al. (2005). The conversion matrix proved to be very useful in delivering
information for understanding the overall processes occurring within the landscape as well as
providing an insight on the best technique and variables to be used for deriving more specific
processes. The planning and management perspectives of this landscape analysis are
dependent on the goal and scale of the approach of each administrative institution acting at
the regional, municipal, and local level.

References
DeFries R. and Townshend J.R.G. (1994). NDVI-derived land cover classification at a global scale.
International Journal of Remote Sensing, 15:3567-3586.
Xu M., Watanachaturaporn P., Varshney P.K., Arora M.J. (2005). Decision tree regression for soft
classification of remote sensing data. Remote Sensing of Environment, 97:322-336.

648
5.5 Posters

Classification of the landscape of Huelva (Andalusia, Spain) using multivariate

methods

J. Alcántara Manzanares1, J. M. Muñoz Álvarez1, J. Quijada Muñoz2, J. M. Moreira

Madueño2
1
Department of Botany, Ecology and Plant Physiology. University of Córdoba. Campus
Rabanales. Ed. Celestino Mutis. 14014. Córdoba. Spain. [email protected]
2
Environmental Department. Andalusian Regional Council. C/ Manuel Siurot, 50. 41013.
Sevilla. Spain

Introduction
This study was conducted in the province of Huelva (Andalucía, Spain), which has a
surface area of 10,128 Km2. It comprises four structural sectors: (1) the Sierra, in the North of
the province (up to 1000 m),; (2) the Campiña, in the middle of the province, with low hills,
loamy soils and rainfed crops; to the South lie (3) the Coast, with dunes, expanses of sand,
forestry (pine and eucalyptus groves), crops and housing developments, and (4) the
Marshes, flat areas with limestone soils and characteristic vegetation. The climate is
Mediterranean Oceanic, with mild winters and hot summers;.
The aim of the study was to classify the landscape of Huelva using multivariate methods
and GIS tools.

Method
The province was divided into grid squares of 1Km x 1Km to which information was
associated on visually-perceivable variables: soil use, plant cover, lithology and relief.
Multivariate Classification Analysis.TWINSPAN: This approach offers an advantage over
other classification techniques: it allows grouping of elements and simultaneously provision
of an ecological interpretation of how groups differ (McGarigal et al., 2000).
Multivariate ordination analysis. DCA: Enables visual analysis of the position occupied by
squares in the ordination space, corroboration of classification analysis groups and
establishment of degrees of difference between groups according to the distance between
them (Kent and Coker, 1992).
Multivariate Classification-Validation Analysis. Discriminant Analysis: For a meaningful
validation, the set of elements classified must be different from that subjected to Discriminant
Analysis (Legendre and Legendre, 1998).

Results
Analysis of results of TWINSPAN classification yielded 8 major landscape types, from
North to South: High Sierras; Low Sierras; Peneplains and Piedmonts; Slopes and Hills;
Campiñas; Coastal and Pre-coastal Dunes; Sands; Marshes. DCA results for the first two
axes reveal that the distance between groups in the ordination space is is lower for groups
belonging to the same hierarchical division (except for Campiñas); the greatest distance was
that between High Sierras and Dunes. The classification was checked by Discriminant
Analysis, which yielded an 80% match with the TWINSPAN estimate.

References
Kent, M. & Coker, M. (1992). Vegetation Description and Analysis. A Practical Approach. CRC Press,
Inc., Corporate Blvd., N.W., Boca Raton, Florida.
Legendre, P. & Legendre, L. (1998). Numerical Ecology. Second English Edition Elsevier Science
B.V., Amsterdam.
McGarigal, K; Cushman, S & Stafford, S. (2000). Multivariate Statistics for Wildlife and Ecology
Research. Springer-Verlag. New York.

649
5.5 Posters

Environmental Impact Monitoring Based on Ecological Carrying Capacity

W.C. Su, C.Y. Chang, K.C. Sung

Landscape Architecture Group, Department of Horticulture, National Taiwan University, 138

Keelung Rd. 10673 Taipei, Taiwan.
e-mail: [email protected]

Introduction
Taiwan has faced a critical problem for several decades. Tourists are seriously impacting
natural conservation areas, especially by the traffic problem (Reijnen, et al., 1995; Forman &
Deblinger, 2000). Once the environment is damaged, it can only be regenerated by time, if at
all. Some protected areas limit the number of tourists allowed to ensure environmental
quality. However, the restriction of quantity is more likely based on facility capacity. In other
words, a regular volume of traffic enters these protected areas. In fact, Bowles (1995) has
pointed out that each “normal” sound which affects natural ecology is noise. This study aims
to measure the current effect on the environment and avoid future damage.

Method and material

This study was carried out in the highland farm of National Taiwan University. We
monitored the birds and measured the noise in 12 sites near the highland farm in holiday and
non-holiday areas. In order to understand the correlation between noise and bird species, we
carried out Spearman Analysis with SPSS 13.0.

Result and discussion

Table 1 shows that when the Lmax increases, the bird volumes near highland farm have a
dramatic decrease. It means the effect is not different on all species. Out of the holiday
period, it is more significant because the surroundings are very quiet. This result agrees with
research showing that recreation activities affect ecology directly (Riffell et al., 1996).
Therefore, the more recreational activities there are, the more impacts they cause. Limiting
the facilities or the social carrying capacity does not work to control the damage. Therefore
the buffer zones should be regulated in land use plans to improve ideal habitats for wild
birds.

Table 1. The Spearman Analysis between measured noise and bird species indices
species volume diversity evenness
Spearman -0.282 -0.622* 0.301 0.301
Lmax(dB)
p 0.375 0.031 0.342 0.342
Spearman -0.211 -0.224 -0.049 -0.049
Lmin(dB)
p 0.510 0.484 0.880 0.880
Spearman -0.176 -0.049 -0.350 -0.350
Leq(dB)
p 0.584 0.880 0.265 0.265

References
Bowles, A. E. (1995) Response of wildlife to noise. Wildlife and Recreationists: Coexistence through
Management and Research. Island Press, Washington, D.C.
Forman R. T. T. & Deblinger, R. D. (2000) The Ecological Road-effect Zone of a Massachusetts
(U.S.A) Suburban Highway. Conservation Biology 14: 36-46.
Reijnen, R., Foppen, R., Braak, C. T. & Thissen, J. (1995) The Effects of Car Traffic on Breeding
Bird Populations in Woodland. III. Reduction of Density in Relation to the Proximity of Main Roads.
The Journal of Applied Ecology 1: 187-202.
Riffell, S. K., Gutzwiller, K. J. & Anderson, S.H. (1996) Does Repeated Human Intrusion Cause
Cumulative Declines in Avian Richness and Abundance? Ecological Applications 2: 492–505.

650
5.5 Posters

Satellite remote sensing for monitoring ecological integrity of Canada’s national

parks

D. Pouliot1, I. Olthof1, R. Fraser*1, A. Clouston2, S. Wang1, W. Chen1, J. Poitevin3,

D. McLennan3, J. Kerr4, M. Sawada5
1
Canada Centre for Remote Sensing, Earth Sciences Sector, Natural Resources Canada,
588 Booth St, Ottawa, ON, Canada K1A 0Y7.
e-mail: [email protected]
2
Noetix Research Inc., on contract with the Canada Centre for Remote Sensing, 588 Booth
St, Ottawa, ON, Canada K1A 0Y7
3
Parks Canada Agency, National Parks Directorate, Ottawa, ON, Canada K1A 0M5
4
University of Ottawa, Department of Biology, Ottawa, ON, Canada K1N 6N5
5
University of Ottawa, Department of Geography, Ottawa, ON, Canada K1N 6N5

Canada’s national parks system includes 42 parks covering 3% (276,275 sq km) of the
country’s landmass and representing the full diversity of natural regions. The Panel on the
Ecological Integrity of Canada’s National Parks concluded in 2000 that the Ecological
Integrity (EI) of virtually all of Canada’s national parks is threatened from a variety of internal
and external stresses. In response, the Canada National Parks Act was amended to make
EI the ‘first priority’ in parks management and for reporting the state of Canada’s national
parks to all Canadians.

Considering the vast and often remote areas under protection, Parks Canada Agency
(PCA) considers Earth Observation (EO) technology to be an integral component of a
national park EI monitoring program. A multi-agency initiative is underway to develop
operational EO based methods that use Landsat class imagery to monitor and report the EI
of parks and their greater park ecosystems. This poster describes four types of landscape
scale EI measures that correspond to Landscape Pattern, Succession and Retrogression,
Net Primary Productivity (NPP), and Focal Species Distributions. The measures and
processing methodologies are demonstrated for three pilot study parks (La Mauricie NP,
Quebec; St. Lawrence Islands NP, Ontario; and Pacific Rim NP, British Columbia) that
represent a range of ecological characteristics, using time series of Landsat TM and ETM+
imagery ranging from 1985-2005.

Landscape pattern analyses are discussed in relation to landscape metric selection,

temporal stability, and spatial scaling. Major vegetation disturbances impacting Succession
and Retrogression patterns were identified using a hybrid change detection technique that
combines vegetation index differencing and constrained spectral signature extension from a
baseline land cover product. Ecosystem productivity, as an integrated measure of
ecosystem health, was tracked using a remote sensing-based modeling approach known as
EALCO (Ecological Assimilation of Land and Climate Observations). Finally, changes in the
distribution of focal species were modeled using Genetic Algorithm for Rule Set Production
(GARP) with the temporal remote sensing products and field observations of species
presences serving as inputs.

651
Author list
Baudry, J, 23,
Baumel, A, 307,
Baveco, J.M, 390,
Beaujouan, V, 213,
Beeton, R.J.S, 558,
Acebes, P, 371 Bekessy, S.A, 598,
Achiron-Frumkin, T, 631, Bélanger, L, 394,
Acosta, A, 584, Bennett, A.F, 48, 365
Affre, L, 307, Benzler, A, 447,
Ahern, J., 287, Bertaudière-Montes, V, 236,
Akbarzadeh, M, 132, Bianchi, F.J.J.A, 86,
Alberti, M, 207, Biggs, J, 64,
Alcántara Manzanares, J, 649, Billeter, R, 556,
Alderman, J, 398, Black, B , 319,
Aleshchenko, G.M, 582, Blair, R.B, 199,
Allard, A, 539, Blank, L, 419,
Allemand, P, 384, Blasi, C, 122, 264, 267, 572
Almeida, R.C.G, 526, Boada Juncà, M, 527,
Alvarez, E, 394, Bogers, M, 570, 623
Álvarez, D, 297, Bohnet, I, 116,
Amamoo-Otchere, E, 627, Boitani, L, 325,
Amir, S, 631, Bolaños, F, 551,
Andersen, B.J, 102, Bolli, J.C, 441,
Angermeier, P.L, 517, Bölöni, J, 121, 633
Angold, P.G, 203, Bommarco, R, 147,
Antrop, M, 110, 580 Bonfanti, P, 522, 637
Anwar, M.M, 231, Bontadina, F, 446,
Apan, A, 353, Borghi, C.E, 371,
Arcangelo, C., 499, Borin, M, 412,
Archambault, L, 394, Boscolo, D, 347, 420
Arens, P, 448, Botta-Dukát, Z, 121,
Arlettaz, R, 446, Bottoni, L, 171, 274
Arrignon, F, 117, Bouwma, I.M, 450,
Attorre, F, 100, Bouyjou, B, 117,
Attwood, S.J, 62, 377 Bowyer, R.T, 317,
Aubad, J, 413, Brannom, R, 455,
Auger, P, 339, Bregt, A, 56,
August, P, 643, Bresciani, S, 248,
Avelar, A.S, 526, Breuste, J.H, 197, 189
Aviron, S, 74, 556 Breyer, J, 625,
Awade, M, 420, Brierley, G, 497,
Azari-Dehkordi, F, 58, Brouwers, N.C, 425,
Baas, P, 485, Brown, D, 260,
Babaei Kafaki, S, 132, Brown, R.D, 602,
Bach, H, 181, Brown, S.D, 86,
Bailey, D, 556, Bruckner, A, 381,
Bailey, N, 124, 417 Brus, D.J, 543,
Baker, M.E, 483, Buen, A.A, 373,
Balkenhol, N, 309, Bugmann, H, 441,
Banks, C, 345, Bugter, R, 145, 590
Baptista, T, 560, Buitenzorgy, M, 143,
Barendregt, A, 133, Bunce, R.G.H, 545, 549, 552,
Barnard, F.J, 205, Burel, F, 141, 339, 355,
Barreto, L, 638, Burgess, P.J, 76,
Bartelt, P.E, 313, Burt, T.P, 493,
Basnou, C, 547, Buschmann, A, 434,
Batáry, P, 379, Butet, A, 215, 355
Bateman, D.S, 315, Byomkesh, T, 524,

652
Cabrera, L, 134, Crutsinger, G.M, 299,
Cadenasso, M.L, 293, Cruz, C.S, 560,
Campagne, P, 307, Csorba, P, 600,
Candia-Gallardo, C., 420, Cullum, C.J, 503,
Capotorti, G, 264, 572 Cunningham, M.A, 363,
Caravello, G, 435, Cushman, S.A, 305,
Cardoso, Y, 232, Da, L, 258,
Carey, P.D, 70, 80 Dale, V.H, 299,
Carmel, Y, 419, 631 Daniel, H, 213,
Carranza, M.L, 584, Danielsdottir, M, 455,
Carré, G, 147, David, A, 436,
Cassar, L.F, 135, Davies, B.R, 64,
Castro, S, 611, Davies, D.H, 221,
Catchpole, R.D.J, 72, 415 de Bie, T, 520,
Caycedo-Rosales, P, 90, de Hoog , J.C.J , 525,
Čekstere, G, 273, de Jong, J, 351,
Cerezo, A, 392, Deconchat, M, 117,
Cesarini, S, 163, Delacourt, C, 384,
Chang, C.P, 136, 444 Delettre, Y, 355,
Chang, C.Y, 240, 650 Delmas, V, 250,
Chardon, P, 390, Demeter, L, 367,
Chen, Y.N, 511, Deschamps-Cottin, M, 236,
Chen, M.H, 507, Di Pietro, F, 148,
Chen, P.C, 532, Dias, A.M, 529,
Chen, W, 651, Diaz-Varela, E, 640,
Chen, Y.C, 136, Diekötter, T, 414,
Chen, Y.J, 461, Digiovinazzo, A.B, 274,
Chen, Y.P, 511, Dilly, O, 181,
Cheng, J, 444, Dixo, M, 423,
Chifflet, R, 147, Domon, G, 115, 427
Choda, M, 297, Drapela, T, 381,
Choudhury, M, 78, Drechsler, M., 513,
Christman, M, 481, Duelli, P, 446,
Chuman, T, 632, Dumas, E, 242,
Ciancaglini, A, 88, Dunn, P.K, 353,
Cilliers, S.S, 205, Dunnett, N, 244,
Ciocheti, G, 422, 438 Durand, P, 489,
Ciontescu Camargo, N.A, 94, Dutoit, T, 236,
Classen, A, 299, Duttmann, R, 641,
Clergeau, P, 215, Eastman, J.R, 331, 333
Clevers, J.G.P.W, 648, Edlich, B, 238,
Clouston, A, 651, Edman, T, 613,
Cobben, M.M.P, 448, Edwards, P.J, 441, 549
Coelho Netto, A.L, 526, Eigenbrod, F., 153,
Cohen, W.B, 606, Einav, R, 631,
Coit, J, 643, Elbersen, B.S, 80,
Collard, S.J, 62, Elena-Rosselló, R, 551,
Comor, V, 355, Ennos, R, 185,
Cooper, A, 537, Espartosa, K.D, 426,
Cormont, A, 442, Esseen, P-A, 539,
Corry, R.C, 602, Etherington, T.R, 424,
Corstanje, R, 481, Etter, A, 118,
Corugati, C, 100, Euler, U, 447,
Cosson, J.F, 303, Eycott, A.E, 388,
Coulon, A, 303, Faggi, A.M, 275,
Cranston, J, 369, Fahrig, L, 153, 361
Crecente-Maseda, R, 640, Falcetta, M.F, 100,
Crist, T.O, 414, Fall, A, 440,
Cristóbal, J, 630, Favero, L, 475,
Croci, S, 215, Fediaeva, M.V, 592,

653
Feist, B.E, 455, 465 Gresswell, R.E, 315,
Feola, S, 584, Griffiths, G.H, 135, 576
Fernandes, J.P, 84, 560, 588, Grigg, G., 499,
Ferrier, S, 558, Groom, G, 568,
Fessey, M.C, 124, 417 Groot, J.C.J, 624,
Fialho, A.P, 530, Grove, J.M, 293,
Fievet, C, 123, Guida, D, 572,
Findlay, S, 167, Guiomar, N, 84, 560, 588,
Florgård, C, 209, Gulbinas, Z, 106,
Foltête, J.-C, 596, Gulinck, H, 256, 265, 413,
Forchhammer, M.C, 329, Gyllin, M, 223,
Forman, R.T.T, 175, 479 Tian, H.Y, 284,
Fortin, M.-J, 440, Haase, D, 179, 189, 227,
Foster, D, 123, Hadar, L, 127,
Francis, C, 361, Hadzi Pecova, S, 142,
Franco, D, 98, 475 Hahs, A, 163, 191
Frank, K, 347, Håkansson, N, 432,
Frank, T, 381, Halada, L, 547,
Franklin, S.E, 369, Hall, K, 430, 443
Fraser, R, 431, 651 Hall-Beyer, M, 369,
Friday, R, 643, Han, X.L., 402,
Fröde, A, 541, Handley, J, 185,
Frondoni, R, 267, Hao, X.M, 511,
Frumkin, R, 631, Hara, K, 96, 258, 642,
Fujihara, M, 258, Harrison, P.A, 145,
Fukushima, M, 471, Hartel, T, 130, 367
Galindo-Alcantara, A, 635, Hashimoto, H, 349,
Gallant, A.L, 313, Haslem, A, 48, 365
Gallardo-Cruz, J.A, 644, Hasui, E, 357, 421
Gallo, C, 435, Haynes, K.J, 414,
Gama, L, 635, Hazeu, G, 547,
Garay, I, 418, Healey, S.P, 606,
García, S, 297, Hecnar, S., 153,
García-Albarado, J.C, 244, Hefting, M.M, 495,
García-Feced, C, 551, Helming, K, 181,
García-Salgado, G, 375, Henkin, Z, 127,
Garnier, E, 430, Hepinstall, J.A, 207,
Geniaux, G, 242, Herlihy, A.T, 469,
Gerard, F, 547, Hernadez-Moralez, A, 635,
Ghetti, P, 475, Hernández-Jiménez, V, 282,
Giannico, G, 467, Hero, J-M , 634,
Giannoni, S.M, 371, Herranz, J, 169,
Giddings, L, 635, Herrmann, M, 161,
Gill, S, 185, Hervás, I, 169,
Gil-Tena, A, 54, 359 Herzog, F, 66, 74, 76, 549, 556,
Girard, P, 521, Hewison, A.J.M, 303,
Gledhill, D.G, 221, Hidding, M.C., 38,
Glimskär, A, 539, Hilborn, R, 473,
Gomes, L.C, 530, Hinsley, S.A, 398,
Gómez Sal, A, 554, Hladnik, D, 46,
Gómez-Sanz, V, 551, Hockman-Wert, D.P, 315,
González, S, 551, Hoechstetter, S, 615,
Goodwin, B.J, 327, Hofer, G, 549,
Gottschalk, T.K, 137, Hollinshead, J, 451, 520
Goward, S.N, 606, Holzhauer, S.I.J, 428,
Goyal, S.P, 436, Hommel, P.W.F.M, 501,
Gragson, T., 123, Hong, N.M, 461, 505
Grashof-Bokdam, C.J, 390, Honsell, G, 522,
Graves, A.R, 76, Hope, N., 124,
Gregg, D, 643, Horváth, F, 121,

654
House, A.P.N, 86, 377 Khazaee, N, 58,
Howard, D.C, 552, Khoroshev, A.V, 582,
Howorth, R.T, 138, Kie, J.G, 317,
Hrnčiarová, T, 128, 276 Kiers, M, 590,
Huang, C, 606, Kim, Y.K, 647,
Huettmann, F, 311, Kinzig, A, 123,
Hughes, R.M, 469, Kirby, K.J, 25, 140
Huili, G, 534, Kireyeu, V.V, 619,
Hull, A.P, 451, 520 Kizos, T, 562,
Humphrey, J.W, 68, Klar, N, 161,
Illyés, E, 121, 633 Klaver, R.W, 313,
Imai, K, 515, Kleinn, C, 434,
Imanishi, A, 125, 515 Klingemann, A, 452,
Imanishi, J, 125, 269, 349, Kloehn, K, 473,
Ingegnoli, V, 248, Klopfer, S.D, 517,
Inghe, O, 539, Klotz, S, 266,
Isachenko, G.A, 586, Knapp, S, 266,
Ishitsuka, T, 642, Koh, J, 294,
Iverson Nassauer, J, 27, 35, 260, König, B, 181,
Jaarsma, C.F, 177, Konkoly-Gyuró , É , 566,
Jaeger, J.A.G, 159, Koomen, A.J.M, 543,
Jain, K, 645, Koponen, P.K, 131, 143
Jalink, M.H, 525, Kőrösi, Á, 379,
James, P, 221, 225 Kost, S, 108,
Janin, A, 384, Kozlov, D.N, 323, 592
Jeanneret, P, 74, Kramer-Schadt, S, 161,
Jellema, A., 622, 624 Krenke, A.N, 608,
Jensen, D, 455, Kruse, A, 146,
Jenson, D, 465, Kuby, L, 123,
Jessel, B, 531, Küchler, M, 272,
Jesus, F.M, 410, Kuhlman, T, 181,
Jim, C, Y, 284, Kühn, I, 266,
Jiménez, E, 90, Kun, A, 633,
Jiménez, L, 119, Kuo, M, 252,
Jochem, R, 390, Küsel, K, 487,
Johnsingh, A.J.T, 436, Kuypers, V., 201, 564
Johnson, D.H, 363, Kwon, J, 647,
Joly, P, 384, Lafortezza, R, 602,
Jombach , S , 566, Landman, K, 232, 277
Jones, J.E, 278, Lange, I.A, 455,
Jones, S.D, 598, Lavorel, S, 430,
Jongman, R.H.G, 392, 450, 518, 521, 545, Lawson, B.E, 558, 634
Jönsson, L, 430, 443 Lechner, A.M, 598,
Jordan, T.E, 483, Lee, Y.C, 532,
Juang, D.F, 532, Leeuwestein, J.M, 521,
Kamagata, N, 642, Léna, J.P, 384,
Kameyama, S, 471, Leronni, V, 439,
Kaneko, M, 471, Li, D.H, 402,
Kang, E.S, 509, Li, D, 295,
Kareiva, P, 123, Li, H.W, 467,
Kasmo, B, 533, Li, Z., 404,
Kassahun, D, 139, Li, W.H, 511,
Kati, V, 337, Li, X, 29, 509
Kato, S, 254, 287 Ligtenberg, A, 56,
Kattan, G.H, 351, Lin, Y.P, 461, 505, 507,
Kazmierczak, A.E, 225, Lindström, T, 432,
Kehm, W, 232, Linke, J, 369,
Kennedy, R.E, 606, Lipský, Z, 632,
Kerr, J, 651, Liu, H.L, 402,
Keunen, L.J, 543, Liu, H.Y, 404,

655
Livingston, M, 278, McDermid, G, 369,
Lögdberg, F, 433, McDonald, R.I, 123,
Lombardi, G, 88, McDonnell, M.J, 191,
Lönn, M, 443, McGrath, B.P, 293,
Loos, G.H, 268, McKenzie, S.J.P, 537,
Lorenzen, D, 155, McLennan, D, 651,
Lorini, M.L, 418, McSwain, M.D, 315,
Lorrillière, R, 250, Meave, J.A, 644,
Lotfi, A, 141, Medcalf, K, 625,
Lourival, R, 499, 513 Medoza, J.E, 90, 351
Lozano-Zambrano, F.H, 90, Meeus, S, 256, 265
Lu, L, 509, Meeuwsen, H.A.M, 390,
Lucas, E, 427, Mehaffy, M.H, 279,
Ludwig, J, 335, Meirelles, S.T, 410,
Lunt, ID , 408, Melcher, A, 457,
Luque, S, 547, Menezes, H, 129,
Lyra-Jorge, M.C, 120, 422, 438, Mensah, F, 627,
Maas, G.J, 501, 543 Merekalova, K.A, 582, 617
Mac Nally, R, 365, Merot, P, 489,
Machado, R, 499, Metzger, J.P, 343, 345, 347, 357, 373, 382,
Madriñán, L.F, 467, 386, 410, 420, 421, 423, 628
Magal, Y, 631, Meurk, C.D, 229,
Magnusson, W.E, 634, Michel, N, 355,
Mahoney, S.P, 341, Michetti, L, 572,
Main, M.B, 165, Midha, N, 519,
Mairota, P , 439, Millard, A, 211,
Makaske, B, 501, 518 Minai-Tehrani, D, 58,
Malinowska, H, 442, Mininni, M, 439,
Malo, J.E, 169, 173, 371, Minunno, F, 439,
Manuel García del Barrio, J, 551, Moeller, M, 123,
Marchese, M, 264, Moilanen, A, 396,
Marchetti, M, 122, Moisen, G.G, 606,
Marco, A, 236, Mollo, B, 267,
Marelli, A, 636, Molnar, Z, 633,
Maron, M, 62, 353, 377, Molnár, Zs, 121,
Marta, M, 264, Monteil, C, 117,
Martensen, A.C, 343, 421, 628, Moreira Madueño, J.M, 649,
Martín, B, 119, Moretti, M, 446,
Martinez, C, 413, Morimoto, J, 621,
Marzluff, J, 207, Morimoto, Y, 125, 269, 349,
Masek, J.G, 606, Morison, N, 147,
Mashayekhi-Kerahroodi, N, 58, Morse, S, 135,
Maskell, L.C, 552, Mörtberg, U.M, 195,
Masri, Z, 533, Mortelliti, A, 325,
Mata, C, 169, 173 Moseley, D G , 319,
Matarán Ruiz, A, 60, 82 Moser, D, 381,
Mateus, M.J, 560, Moylan, E.A, 629,
Mathey, J, 238, Mücher, C.A, 547, 568, 590,
Mathur, P.K, 321, 519 Muhar, S, 457,
Mathur, V.B, 449, Muklada, H, 127,
Mattiuzzo, E, 475, Müller, K, 272,
Mattson, K.M, 517, Muñoz, C, 119,
Mauffrey, J-F, 236, Muñoz Álvarez, J.M, 649,
Maund, S, 64, Nabe-Nielsen, J, 329,
Mayor, S.J, 341, Nagamatsu, D, 126,
Mazeffa, D, 414, Nakagoshi, N, 271, 524, 534,
McAlpine, C, 335, Nakamura, F, 621,
McCall, S, 163, Nakamura, N, 112,
McCallum, H, 499, Nakao, K, 471,
McCann, T, 537, Napoleone, C, 242,

656
Nasi, R, 143, Pellegrino, P, 291,
Natuhara, Y, 125, 515 Penev, P, 142,
Naxara, L, 400, Peregovits, L, 379,
Negi, P, 645, Pereira, D, 282,
Nemes, S, 367, Perepelizin, P.V, 275,
Newton, A, 425, Perevolotsky, A, 631,
Nicieza, A.G, 297, Pérez Campaña, R, 60,
Niemelä, J, 193, Perez-Soba, M, 181,
Nieuwenhuizen, W, 543, Perovic, P, 392,
Nijhof, B.S.J, 501, Persson, V.G, 418,
Nikiforova, A.A, 626, Pess, G.R, 465, 473
Nikodemus, O, 273, Peterson, C.R, 313,
Nilon, C.H, 217, Petit, S, 552,
Ninyerola, M, 630, Pfund, J-L, 131,
Nordén, B, 396, Pham , U.D , 271,
Norton, L.R, 552, Phinn, S, 335,
Noss, F.R, 165, Piani, G, 88,
Nowak, D, 219, Pichancourt, J.B, 339,
Noy-Meir, I, 127, Pickett, S.T.A, 293,
Nuissl, H, 179, Pilliod, D.S, 313,
Obrist, M.K, 446, Pimentel, R.G, 357, 421
Ocón, B, 282, Pinay, G, 493,
Odede-Oremo, D.W., 114, Pino, J, 547,
Odle, A, 455, Pinto, N, 646,
Ogram, A, 481, Pinto-Correia, T., 42, 129
Oh, J.H, 647, Pirnat, J, 46,
Oliveau, S, 236, Pivello, P.R, 410,
Öllerer, K, 130, 367 Pivello, V.R, 120, 422, 438,
Olsvig-Whittaker, L, 631, Pleijte , M., 38,
Olthof, I, 431, 651 Poggesi, M.C, 171,
Onufrenya, I.A, 323, Poggio, S.L, 23,
Opdam, P.F.M, 622, Poirazidis, K, 337,
Oriolo, G, 522, Poitevin, J, 651,
Orlov, T.V., 594, Pontes, E, 638,
Ortega, M, 551, Pontius Jr, G, 639,
Örvössy, N, 379, Poppe, M, 457,
Osvalde, A, 273, Porter, W.P, 313,
Ouin, A, 117, Possingham, H.P, 513,
Padoa-Schioppa, E, 171, 274 Possingham , H., 499,
Padovani, C.R, 518, Potter, C.A.Potter, 33,
Palang, H, 570, Potts, S.G, 147,
Paliwal, A, 449, Pouliot, D, 431, 651
Palma, J.H.N, 76, Poulsen, J.G, 143,
Paltto, H, 396, Pouyat, R.V, 219,
Pándi, I, 121, Powell, S.L, 606,
Paolanti, M, 572, Prentice, H.C, 430, 443
Pape, A, 369, Prévot-Julliard, A.C, 250,
Pardini, R, 373, 382, 400, 426 Price, B, 335,
Park, C.Y, 647, Primdahl, J, 36,
Park, S.E, 62, Pullar, D, 335, 558
Parracchini, M, 556, Pungetti, G, 146,
Parris, K.M, 280, Puric-Mladenovic, D, 574,
Parsons, B.C, 50, Puzachenko, M.Y, 323, 592
Pascual-Hortal, L, 406, Puzachenko, Y.G, 323, 592, 608,
Patrono, A, 554, Pywell, R.F, 70,
Patton, J.L, 278, Querner, P, 381,
Pauleit, S, 185, Quijada Muñoz, J, 649,
Pazete de Oliveira, M, 528, Quinn, T, 473,
Pedroli, G.B.M, 570, 623 Qureshi, Q, 436,
Peethambaram, S, 639, Rains, M.C, 485,

657
Rakhimova , N.R , 262, Scheidegger, Ch, 441,
Ramesh, K, 436, Schellhorn, N.A, 86,
Ramírez, C, 531, Schetke, S, 189,
Ramos, I.L, 129, Schindler, S, 337,
Rantier, Y, 355, Schippers, P, 390,
Rashid, M.S, 524, Schlaepfer, R, 452,
Raulier, F, 394, Schmid, H, 429,
Ravibabu, M.V, 645, Schmidt, B.R, 452,
Rawat, G.S, 436, Schmutz, S, 463,
Ray , D , 319, Schneider, D.C, 341,
Ray, N, 303, Schouten, M, 133,
Rayfield, B, 440, Schrautzer, J, 155,
Reardon-Smith, K, 62, Schröder, B, 429,
Rebollo, S, 375, Schüpbach, B, 74,
Reck, H, 155, Schwartz, M.K, 305,
Redman, C, 123, Schwarz, K, 219,
Reisner, Y, 76, Schweiger, O, 556,
Reitalu, T, 430, 443 Serra, J.M, 630,
Remolina, F, 437, Servais, V, 250,
Renetzeder, C, 590, Settele, J, 147,
Revermann, R, 429, Sevenant, M, 110,
Ribeiro, M.C, 120, 357, 421, 422, 438, 628, Sgouros, D, 283,
638, Shahbazyan, E, 445,
Richards, W.H, 416, Sheer, M.B, 455,
Richner, W, 66, Shin, J.H, 647,
Roberts, J.D, 50, Shkaruba, A.D, 619,
Roberts, S.P.M, 147, Short, J.C, 50,
Robiglio, V, 144, Short, K, 96,
Roblas, N, 119, Sibly, R, 329,
Roche, P, 147, 307, 609, Sieber, S, 181,
Rodet, G, 147, Siedentop, S, 183,
Rodriguez, M.A, 413, Siervo, V, 572,
Roedenbeck, I.A, 157, Sifneos, J.C, 469,
Rojas, G, 531, Sigura, M, 522, 636, 637,
Romportl, D, 632, Šimonides, I, 128,
Rossing, W.A.H, 622, 624 Sinclair, F.L, 144,
Rotem, D, 631, Smaniotto Costa, C, 238,
Roullé, N, 427, Smart, S.M, 552,
Rozas, C, 119, Smeets, P, 564,
Rozé, F, 213, Smiraglia, D, 122, 572
Rubilar, H, 531, Smith, A.C, 361,
Ruiz, J, 115, 427 Smith, D.J, 165,
Rusmantoro, W, 143, Smith, E.R, 279,
Sadler, J.P, 203, Smith, M, 68,
Saeed, A, 533, Smulders, M.J.M, 448,
Saldaña, A, 375, Snep, R, 201,
Saldaña López, A, 554, Somma, D, 392,
Sànchez Mateo, S, 527, Soomers, H, 52,
Sanesi, G, 602, Soons, M.B, 491,
Sangermano, F, 331, 333 Soto-Esparza, M, 635,
Sarthou, J.P, 117, Spaar, R, 429,
Sasaki, T, 269, Spooner, PG , 408,
Sato, A.M, 526, Ståhl, G, 539,
Sattler, C, 88, Stanisci, A, 584,
Sattler, T, 446, Steel, E.A, 455, 465
Saura, S, 54, 359, 406, 611 Steffan-Dewenter, I, 147,
Sawada, M, 651, Steinmeier, C, 272,
Schaefer, J.A, 341, Stenhouse, G.B, 369,
Schaepman, M.E, 648, Sternberg, M, 631,
Schastnaya, I, 578, Stewart, C, 163,

658
Stoate, C, 104, van der Veen, M, 390,
Stobbelaar, D.J, 624, van der Ven, P.J.M, 485,
Stokman, A, 234, van der Zande, A.N, 21,
Stringer, C.E, 485, van Doorn, A, 450,
Strobl, S, 574, Van Eetvelde, V, 580,
Stuart, R.C, 552, van Eupen, M, 518, 590
Su, W.C, 240, 650 van Lammeren, R, 392,
Suárez, F, 169, 173, 371, van Langevelde, F, 177,
Suárez-Rubio, M, 246, van Lier, H, 392,
Sudharsan, K, 329, van Noordwijk, M, 131,
Sumfleth, K, 641, van Staalduinen, M.A, 92,
Sundquist, S, 539, Vandewalle, M.O, 430,
Sung, K.C, 650, Vargas, W, 90,
Swaffield, S R , 44, Veldkamp, A, 638,
Swetnam, R, 547, Verboom, J, 151, 390, 448,
Sykes, M.T, 430, 443 Verburg, P.H, 56,
Szentgyörgyi, H, 147, Verhoeven, J.T.A, 485, 491
Tabbush, P, 181, Verweij, P.A, 52,
Takahashi, K, 96, Vianna, L.G.G, 526,
Taleska, K, 142, Viktorov, A.S , 604,
Tambosi, L.R, 422, 438 Vogiatzakis, I.N, 576,
Tarantino, C, 439, von dem Bussche, J, 429,
Tatoni, T, 242, 307 Vos, C.C, 390, 590
Tauber, I, 631, Arntzen, W.J, 301,
Teillac-Deschamps, P, 250, Wagner, H.H, 441, 549
Tejerina Garro, F.L, 477, 528, 529, 530 Waits, L.P, 309,
Terada, T, 289, Walczak, M, 631,
Thinh, N.X, 615, Walter, T, 66,
Thomlinson, J.R, 246, Walz, A, 270,
Thompson, S, 64, 124 Walz, U, 615,
Timmermans, W, 201, 564 Wan, H.F , 444,
Tocchetto, D, 412, Wang, S, 651,
Tokiwa, N, 621, Wang, T, 509,
Tolle, F, 596, Wang, Z, 260,
Tominaga, S, 126, Wardell-Johnson, G, 558,
Tomita, M, 258, 642 Warren, P.S, 217,
Topping, C.J, 329, Wascher, D.M, 181, 568
Torgersen, C.E, 315, Wassen, M.J, 52,
Torras, O, 54, 359 Wasserman, T.N, 305,
Traba, J, 371, Watts, K, 388,
Trautwein , C., 457, 463 Weel, S, 648,
Treleani, A, 637, Weijschede, T.J, 543,
Tscherning, K, 181, Weller, D.E, 483,
Tung, C.P, 461, Wennergren, U, 432, 433
Turkelboom, F, 533, Werger, M.J.A, 92,
Uezu, A, 386, Westerberg, L, 432,
Umetsu, F, 382, Westerveld Cardoso, Y, 277,
Usava, I.P, 619, Westphal, C, 147,
Uthes, S, 88, Whigham, D.F, 485,
Vaishar, A, 187, Whitbread, T, 138,
Vaissière, B.E, 147, White, S, 467,
Valbuena, D, 56, Wiggering, H, 181,
Valenzuela Montes, L.M, 60, 82 Williams, N.S.G, 205,
Vallet, J, 213, Williams, P.J, 64,
van Adrichem, M, 390, Wittmer, H, 179,
van de Berg, M, 390, Wolfert, H.P, 501,
van den Heuvel, R.N, 495, Wolters, V, 137, 428
van der Grift, E.A, 151, Worrell, R, 319,
van der Maat, T.N.M, 543, Woyciechowski, M, 147,
van der Ree, R, 163, Wrbka, T, 590,

659
Wu, P.J, 505,
Wu, Z.F, 136, 444
Xavier-da-Silva, J, 418,
Xiao, D, 523,
Xu, C.C, 511,
Yamashita, S, 112,
Yang, Y, 258,
Yankee, D.H, 279,
Yen, N, 281,
Yesilonis, I.D, 219,
Yokohari, M, 40, 289
Yu, F.C, 281,
Yu, K, 295,
Yuan-Farrell, C, 123,
Yun, P, 534,
Zahavi, A, 631,
Zaiping, X, 534,
Zaller, J.G, 381,
Zammit, C, 377,
Zanetto, G, 475,
Zanini, F, 452,
Zavattero, L, 122,
Zbinden, N, 429,
Zequeira-Larios, C, 635,
Zessner, M, 459,
Zhao, Y, 258,
Zhou, H, 523,
Zhou, K, 523,
Zipperer, W.C , 219,
Zorza, R, 522,
Zuin, A, 475,

660