Icarus 157, 535–548 (2002)

doi:10.1006/icar.2002.6854

Photosynthesis: Likelihood of Occurrence and Possibility of Detection

on Earth-like Planets
R. D. Wolstencroft
Royal Observatory Edinburgh, Blackford Hill, Edinburgh EH9 3HJ, United Kingdom
E-mail: [email protected]

and

J. A. Raven
Division of Environmental and Applied Biology, School of Life Sciences, University of Dundee, Biological Sciences Institute,
Dundee DD1 4HN, United Kingdom

Received August 4, 2000; revised February 21, 2002

I. INTRODUCTION
Although there are considerable technical challenges to be over-
come during this decade, the prospects for the detection of Earth- The increasing optimism that Earth-like planets (ELPs) will,
like planets (ELPs) orbiting nearby stars are encouraging. If life in a decade or so, be found to be orbiting a significant fraction of
has developed on some of the ELPs that may be discovered by so- the nearby main-sequence stars has widened the scope for remote
phisticated telescope systems, such as the Terrestrial Planet Finder,
sensing of extraterrestrial life from the planets and satellites of
the detection of photosynthesis is an attractive possibility. Here we
discuss the likely preconditions and subsequent events that have
the Solar System to include ELPs in planetary systems around
led to the occurrence of O2 -producing photosynthesis on Earth and such stars (Des Marais and Walker 1999). Plans to detect ELPs
then extend this discussion to how this may have occurred on ELPs orbiting nearby stars using an infrared space interferometer that
orbiting in the habitable zone of a variety of main-sequence stars can null out the light of the central star are being explored by
from spectral type F0V to M0V. We point out how the need for liq- NASA (Terrestrial Planet Finder/TPF) (Beichman et al. 1999)
uid water and the need to avoid UV radiation have influenced the and ESA (Infrared Space Interferometer/IRSI). To assist the
evolution of photosynthesis on Earth, how the absorption spectra strategy for the search for ELPs we would ideally like to have
of the dominant (chlorophyll) photosynthetic pigments may have some statistical information on the frequency of ELPs that might
been determined in natural selection, and how and when the evo- come from transit experiments such as COROT, KEPLER, or
lution of the ability to use water as an electron donor took place. EDDINGTON or from gravitational lensing. Such information
Models for the photosynthetic productivity of ELPs orbiting at the
could guide the priorities in deciding the dwell times and types
inner edge of the habitable zone are discussed both from aquatic and
land-based photosynthesis, making some allowance for global cloud
of stars that will be surveyed by TPF/IRSI. The other aspect is
cover on the ELP. The photosynthetic generation of O2 is greatest on whether life is likely to flourish, if initially present, given the
cloud-free planets with hot (e.g., F0V) parent stars, though the ad- distance of the ELP and the type of parent star. This is not only
vantage over cooler stars depends on the fraction of the planet cov- a question of whether liquid water could exist at the planetary
ered by oceans. The low O2 generation in ELPs orbiting cooler stars surface, as encapsulated in the term “Habitable Zone,” but also
is due to the poor match between the parent star’s spectral energy whether organisms on the ELP could find a reliable and long-
distribution and the assumption of terrestrial pigment properties. lived source of energy that would allow them to survive, multi-
We discuss the possibility that a three- or four-photon mechanism ply, and evolve over planetary time scales. Oxygen-generating
might operate on such planets (as opposed to the two-photon system photosynthesis is such a source of energy which, because it uses
on Earth) and how it could influence the spectral properties of the water as the electron donor, would be able to provide an abundant
ELP. We also emphasize the role of tectonic and other geological pro-
source of energy for microorganisms on the ELP.
cesses as well as biology in determining the O2 level on Earth and on
ELPs. c 2002 Elsevier Science (USA)
The detection of photosynthesis on an ELP is an attractive pos-
Key Words: atmospheres; evolution; Earth; prebiotic environ- sibility for the remote sensing of extraterrestrial life (Des Marais
ments. and Walker 1999). On the assumption (to be analysed below) that
photosynthesis on an ELP would lead to O2 accumulation in its

535

0019-1035/02 $35.00

c 2002 Elsevier Science (USA)
All rights reserved.
536 WOLSTENCROFT AND RAVEN

atmosphere, then the detection of free atmospheric O2 via spec- The calculations of Jakosky and Shock (1998) suggest that life
troscopic detection of O3 (produced photochemically from O2 ) based on the inorganic chemistry of hydrothermal vents could
could be used as an indication of the occurrence of life (Leger produce 0.6 M tonnes of organic C per year worldwide. This
et al. 1993). Such a test is a specific case of the general method contrasts with 100 G tonnes of organic C generated today by
for remote sensing of life on other planets suggested by Lovelock photosynthotrophs and at least 0.2 G tonnes organic C produced
(1975), i.e., the detection of gross departures of some chemical each year by chemolithotrophs on Earth today, mainly using for
component(s) of the planetary atmosphere from thermodynamic their exergonic metabolism inorganic substrates which result ul-
equilibrium from other components. While some O2 can be pro- timately from photosynthesis (Raven 1996). The chemolithotro-
duced by the (abiological) photochemical dissociation of water, phy on the early Earth which was not associated with continued
large quantities of O2 (and hence O3 ) require the activity of input of reductant from vents is less readily quantified. However,
(biological) photosynthesis. A further possible means of detect- it would have converted the useable substrates for chemolithotro-
ing photosynthesis is the spectroscopic remote observation of phy (FeS, H2 S, CO2 ; Wächterhauser 1990) into organic carbon
the pigments which are essential for this process. and the relatively inert FeS2 (Russell and Hall 1997), so that
Before discussing the possibilities of making the required there is little possibility of a chemoorganotrophic back reac-
measurements of O2 and pigments, we briefly discuss the likely tion oxidising organic C and FeS2 as oxidant, regenerating the
preconditions (Section II) and subsequent sequence of events chemolithotrophic substrates.
(Section III) which have resulted in the occurrence of O2 - This depletion of substrates would have restricted the pro-
producing photosynthesis on Earth. This permits us to consider ductivity of early chemolithotrophs. This metabolic bottleneck
(Section IV) the likelihood that life on an ELP would have led to provides both an ecological opportunity for light as an additional
photosynthesis (potentially detectable by spectroscopy of pig- energy source and part of the metabolic infrastructure derived
ments) and an O2 -evolving variant of photosynthesis (potentially from chemolithotrophy, which is used in the various mechanisms
detectable spectroscopically as O3 ). of photosynthetic energy conversion found on Earth today.

II. ORIGIN AND PREPHOTOSYNTHETIC EVOLUTION III. EVOLUTION OF PHOTOSYNTHESIS AND THE
OF LIFE ON EARTH ACCUMULATION OF O2 ON EARTH

The presence of liquid water is a major prerequisite for life For light to be used as the energy source for growth of or-
on Earth (or any other planet). Granted the occurrence of ade- ganisms on Earth requires that they live at the land surface, or
quate water, the requirement that a very significant fraction of at depths of less than 300 m in even the clearest waters, if the
this water remains liquid involves a tectonic cycle which recy- light energy supply is to be adequate (Raven et al. 2000). This
cles the greenhouse gas CO2 . Such a cycle can return CO2 to the requirement for a certain minimum level of photosynthetically
atmosphere even when near-global glaciation restricts the part active radiation (PAR) means that the organisms will be exposed
of the cycle which removes CO2 from the atmosphere. This can to UV-B, exacerbated by the high output of UV-B relative to
potentially reverse glaciation, the likelihood of which would be PAR by the Sun 3.5 billion years ago but, for aquatic organisms,
increased early in the evolution of life as a result of the lower ameliorated by the greater absorption of UV-B than of PAR by
radiant energy output of the Sun 4 billion years ago than today natural waters (Cockell 2000, Kirk 1994a,b).
(Gough 1981, Walker et al. 1981). Biogenic methane was prob- Mulkidjanian and Junge (1997) suggest that the evolution of
ably also a very important greenhouse gas until O2 increased photosynthesis involved UV-screening proteins in or associated
around 2.3 billion years ago (Kasting et al. 2001). with the plasma membrane. All proteins (and nucleic acids) ab-
The two current hypotheses as to the origin of life on Earth sorb, and can be damaged by, UV-B; surface-located proteins
are the chemoorganotrophic and the chemolithotrophic. The could screen out some of the UV-B, limiting damage to pro-
chemoorganotrophic hypothesis has become less popular with teins and DNA within the cell. Excitation energy transfer from
the realisation that the early atmosphere was less reducing than the protein to the tetrapyrrol would allow dissipation of this
was believed in the early experiments on production of organic energy as fluorescence, heat, or photochemistry in the earliest
compounds by atmospheric chemistry (Broda 1975, Chang evolutionary stages of photosynthesis. The evolutionary selec-
1999). The alternative and currently favoured hypothesis is tion value of the photochemical alternative for energy dissipa-
that of a high-temperature chemolithotrophic origin of life tion is that the photochemistry could supplement, and ultimately
(Wächterhauser 1990, Russell and Hall 1997, Pace 1997, Ferris replace, chemolithotrophic reactions as the energy source for
1999), in which the energy input to the synthesis of organic com- growth. This photochemistry would have generated ATP and re-
pounds comes from exergonic inorganic reactions. The evolution ductants of relatively low redox potential which could reduce
of life at hydrothermal vents would also have the advantage of CO2 to produce organic compounds. Such conversion of reduc-
limiting access of UV-B radiation from the young Sun to early tants such as Fe2+ and S2− to Fe3+ and S0 in photosynthetic CO2
life since hundreds of metres depth of seawater would act as a reduction permitted the use of Fe3+ and S0 to act as electron ac-
UV-B screen. ceptors in respiratory reactions of chemoorganotrophic growth,
PHOTOSYNTHESIS ON EARTH-LIKE PLANETS 537

regenerating CO2 and the reductants Fe2+ and S2− . However, Sleep (2001; see also Kump et al. 2001) discusses two hy-
such recycling would have been limited by the sedimentation potheses in which the mantle is involved in restricting net burial
of particulate organic C and of particulate oxidants (Fe3+ , S0 ) of Fe2 O3 in permitting buildup of O2 . The argument is that net
which, with tectonic activity, would remove these substrates for removal of Fe2 O3 from the crust permits the reaction 2FeO +
anaerobic respiration. H2 O → Fe2 O3 + H2 to proceed, with O2 removal from the at-
Such depletions of reductants for photosynthesis would have mosphere by reaction with H2 . Accumulation of Fe2 O3 at the
provided a selective advantage for organisms which use the very surface limits H2 production, allowing O2 to accumulate. Fur-
high potential but abundant oxidant H2 O, with production of O2 . thermore, H2 escape following biogenic H2 production and pho-
The use of the high-potential reductant H2 O as electron donor for tolysis of biogenic CH4 could have caused irreversible oxidation
CO2 reduction involves the use of two photons in photochemistry of early Earth (Catling et al. 2001, Hoehler et al. 2001).
for each electron transferred from H2 O to CO2 (Falkowski and The final aspect of this consideration of the evolution of pho-
Raven 1997). Even with two photons per electron the lower tosynthesis on Earth concerns the factors determining the de-
energy per photon at longer wavelengths imposes a limit for tectability of O2 -producing photosynthesis by remote sensing.
photons used in O2 -evolving photosynthesis, which is usually Spectroscopy is the only method available, and it could theo-
700 nm but can extend to 730 nm in a few organisms (Halldall retically detect the catalysts of photochemistry (photosynthetic
1968, Schiller et al. 1997, Koehne et al. 1999, Mimura et al. pigments) and one of the products of O2 -generating photosyn-
1999). The higher energy per photon at the lower wavelength thesis (O3 , derived photochemically from O2 ). Seasonal changes
limit for photosynthesis (380–400 nm) is always converted, after in photosynthetic pigments on the land surface should be de-
absorption, to the lower energy per photon of longer wavelengths tectable from space (with the Sun nulled out) with their char-
before use in photochemistry. acteristic (for Earth) 700-nm cutoff. Photosynthetic pigments in
The evolution of this O2 -producing photosynthesis could have the ocean are less readily perceived in this way.
occurred as early as 3.56 billion years ago (Golubic and Seong- The next part of this paper (Section IV) considers the likeli-
Joo 1999, Summons et al. 1999, Des Marais 2000), yet the first hood that a similar sequence of events to those described here
evidence for global oxygenation comes from little earlier than for Earth could have occurred on an ELP orbiting a nearby star.
2.5 billion years ago (Holland 1984, Falkowski and Raven 1997, In terms of observability by spectroscopy from Earth, we need
Brocks et al. 1999, Watanabe et al. 2000). O2 accumulation in to investigate the likelihood of (seasonally varying) photosyn-
the atmosphere requires that some of the organic C produced thetic pigments on the land surface in sufficient quantities to be
by photosynthesis is removed from the possibility of biological detectable and the likelihood that O2 would accumulate suffi-
(respiratory) reoxidation by sedimentation. There is evidence ciently to generate enough O3 to be detectable spectroscopically
that sedimentation of organic C, apparently produced by pho- as a proxy for O2 .
tosynthesis with H2 O as reductant, occurred well before global
oxygenation. Sedimentation of organic C also removes other IV. POSSIBLE ORIGIN AND EVOLUTION OF LIFE
nutrient elements (such as phosphorus) from the biosphere, so AND OF PHOTOSYNTHESIS ON ELPS AND ITS
O2 accumulation coupled to organic C sedimentation also de- SPECTROSCOPIC DETECTABILITY
pends on P and other nutrient inputs from weathering processes
and thus on continued exposure of unweathered rock by tectonic We begin by asking how Earth-like an ELP must be and what
activity (Algeo and Scheckler 1998, Berner 1998). Tectonic ac- orbit it must occupy in relation to a particular type of star, if life
tivity is also responsible for returning the carbon sedimented and photosynthesis of anything like the sort found on Earth are
as organic C to the biosphere as CO2 , together with inorganic to occur. This latter requirement is defined in terms of the range
reductants which consume O2 . On the present Earth sedimen- of distances from a star within which an orbiting planet would
tation removes about 0.2 G tonnes of reduced C each year out be habitable over a period of time; for the Earth this period is
of 100 G tonnes of primary productivity (Raven and Falkowski (at a minimum) 3.8 billion years ago to the present. This range
1999). of distances in relation to a specified time span is known as the
The long lag between the probable time of evolution of Continually Habitable Zone (CHZ).
O2 -producing photosynthesis and the global (as opposed to lo- The assumption made hereafter is that our ELP originally had
cal; McKay and Hartman 1991) accumulation of O2 in the bio- a CO2 /H2 O/N2 atmosphere (modifiable by living organisms)
sphere involves “missing O2 ” which was presumably consumed and that habitability requires the presence of liquid water on the
in the oxidation of residual Fe2+ (to Fe3+ ) and S2− (to S0 and planetary surface. We note that even if oceans form on ELPs,
SO2−4 ), including that supplied from hydrothermal vents, possi- they can also disappear (e.g., Yung and De More 1999), as has oc-
bly at higher rates than today, with Fe2+ and S2− from the deep curred on Venus, and that even persistent oceans may not be deep
dark ocean mixed with photosynthetic O2 by ocean circulation enough to act as effective UV screens (see later in this section).
(Berner and Petsch 1998, Canfield et al. 2000). This suggestion As we saw in Section II the control of temperature on the planet’s
is not easily distinguished from the alternative of oxidation of surface by temperature-independent changes in the atmospheric
Fe2+ and S2− by anoxygenic photosynthetic bacteria. CO2 level depends on changes of volcanism and plate tectonic
538 WOLSTENCROFT AND RAVEN

activity, so these are also requirements for our ELP. Franck TABLE I
et al. (2000) have combined volcanism, weathering, and tec- Stellar Properties and Radiation Fluxes at the Inner and Outer
tonics into a general modelling scheme for identification of the Limits of the CHZ a
habitable zone (HZ), i.e., the range of distances from a star within Stellar Teff λmax b λmax c Sind rine
which an ELP would be habitable over a long but unspecified type (K) (µm) (µm) M/M(sun) L/L (sun) (Ssun ) Sout (AU) rout
time span. However, a significant constraint on such models,
pointed out by Whitmire et al. (1991), is that low atmospheric M0V 3700 0.7832 0.9919 0.5 0.06 1.00 0.46 0.25 0.36
temperatures can cause CO2 condensation, increasing cloudi- G2V 5700 0.5084 0.6439 1.0 1.0 1.10 0.53 0.95 1.37
F0V 7200 0.4025 0.5097 1.3 4.3 1.25 0.61 1.85 2.70
ness via the production of CO2 ice clouds and thus increasing
albedo and decreasing the magnitude of the greenhouse effect a From Kasting et al. (1993).
of CO2 via the latent heat release on condensation and its ef- b Wavelength of maximum energy flux [F(λ)] at Teff .
c Wavelength of maximum photon flux [N (λ)] at T .
fect on decreasing the adiabiatic lapse rate. An effect working eff
d Stellar flux relative to present solar constant (1360 W m−2 , 0.24–4.5 µm)
in the opposite direction is the greenhouse effect of clouds of
at the inner and outer limits of the CHZ.
CO2 ice via the scattering of outgoing thermal radiation back up e Distance of planet from parent star at the inner and outer limits of the CHZ.
to the surface (Forget and Pierrehumbert 1997). CO2 condensa-
tion thus provides an outer limit for the CHZ, assuming constant
stellar luminosity with time (Kasting et al. 1993, Kasting 1998). relative to the Sun, and the relative energy inputs at the top of
Tectonic activity can give rise to continental crust and thus to the atmosphere of a planet at the inner and outer limits of the
a habitat for land-dwelling biota (Kasting et al. 1993, Wetherill CHZ (corresponding to “water loss” and “CO2 condensation”
1996). Tectonic activity also permits the preconditions for a respectively) relative to the present solar constant of the Sun
chemolithotrophic route at hydrothermal vents. We assume an (1360 W m−2 , 0.24–4.5 µm). Since photosynthesis depends on
acidic Fe2+ -rich ocean with reducing, H2 S-rich hydrothermal the number of photons in the photosynthetically active range
vent water. We also assume that our ELP does not suffer from rather than their energy we also tabulated the wavelengths of
the climatic extremes which can be associated with chaotically maximum photon flux for the three stellar types. Finally, Table I
variable, and often extreme, obliquities lacking the stabilizing lists the computed orbital distances (AU) for the inner and outer
influence of a large (relative to the planet) satellite (Laskar et al. limits of the CHZ for the Earth and the ELPs orbiting the two
1993, Williams and Kasting 1997). other stars, computed from the stellar luminosities and the in-
The inner limit for the CHZ is determined by the stellar flux verse square law. We shall later use these estimates in consider-
at which water loss occurs. Taking the Sun as an example, over ing the potential for photosynthesis. However, we note that the
the next several billion years the Sun’s luminosity will increase, inner and outer limits of the CHZ are defined by a climate model
resulting in a modest increase in the water vapour content of the in which the main uncertainty is the role of clouds and its effect
troposphere and a much larger increase in the stratospheric water on the planetary albedo.
vapour content (Kasting et al. 1993). The increased cloudiness Evolution of life by the chemolithotrophic route could have
would increase the planetary albedo and a not readily quanti- occurred on an ELP in the CHZ in the manner indicated earlier
fied negative feedback could decrease the surface temperature. in our consideration of evolution of life on Earth. Evolution deep
Neglecting this feedback, when the Sun’s luminosity reaches in the sea at hydrothermal vents would mean that high UV flux
110% of its current level, catastrophic water losses would oc- incident on the planetary atmosphere, especially for F0 stars
cur (in as little as 1 billion years) and habitability would cease among those cited in Table I, would not damage UV-absorbing
∼0.2 billion years later. For low-luminosity M dwarf stars the essential molecules in the early living organisms. This is because
HZ lies quite close to the star (0.25 to 0.47 AV for an F0V star); the organisms are protected by hundreds or thousands of metres
because a planet orbiting close to its parent star is forced to ro- of UV-absorbing water even in the absence of O2 (and hence O3 )
tate synchronously due to tidal forces (Huang 1960), and hence in the atmosphere to absorb the damaging UV-B (and UV-C)
to have one permanently illuminated hemisphere, it has been shorter wavelengths of UV.
widely argued that such planets will not be habitable. However, In assessing the potentially damaging role of UV radiation we
there are arguments that such planets could support life (Kasting must be aware that UV emission has two components, both of
1996, Heath et al. 1999). which vary with stellar age. The photospheric UV component
Kasting et al. (1993) compute the likely CHZs for ELPS as- depends on the stellar surface temperature and increases only
sociated with two different main-sequence stars for comparison slowly over the stellar lifetime (Gough 1981); it is appreciable
with the Earth. The Sun is a G2V star, with an effective pho- for F stars but relatively weak for the cooler G, K, and M stars.
tospheric temperature (Teff ) of ∼5700 K and, from Wien’s Law For these latter stars, the UV emission comes mainly from the
(λmax /µm = 2, 898/Teff ), a peak energy output at 0.508 µm. second UV component, the active chromosphere whose energy
This is contrasted with an F0V star (Teff = 7200, λmax = derives from dynamo-generated magnetic activity. The energy
0.403 µm) and an M0V star (Teff = 3700, λmax = 0.783 µm). source for the dynamo is thought to be stellar rotation, which
Table I shows these data, together with the mass and luminosity slowly declines with age, and hence the chromospheric UV is
PHOTOSYNTHESIS ON EARTH-LIKE PLANETS 539

also though to decline with age (Ayres et al. 1996, Ayres 1997, ogous) catalysts needed to perform catalysis (Maynard Smith
Simon 1999). and Szathmáry 1997).
The means of energy transformation on an ELP need not nec- Within the constraints indicated, photosynthetic rates on Earth
essarily involve the detailed mechanisms used on Earth, although are limited by resource supply and incident flux density. With
redox reactions are essential. Regardless of the mechanistic de- sufficient nutrients, and hence sufficient pigment per unit area
tails of the evolution of chemolithotrophy, an Earth-like planet to absorb almost all of the incident photons, the lower limit on
would be subject to similar overall energetic constraints on the photon flux density at which gross photosynthesis can occur is
extent of chemilithotrophic primary productivity to those men- set by unavoidable back reactions and, for almost 100% photon
tioned earlier for the Hadean/early Archaean chemolithotrophic absorption and a two photons per electron mechanism, it is some
primary producers on Earth, i.e., to about 0.1% of the present 20 nmole photon m−2 s−1 (Raven et al. 2000).
photosynthetic primary productivity (see the preceding and For a four photons per electron mechanism on an ELP the
Jakosky and Shock 1998). lower limit of 20 nmol photon m−2 s−1 (Raven et al. 2000) would
As on Earth in the late Hadean and Archaean, so at the corre- be 40 nmol photon m−2 s−1 . It is not clear if the more complex
sponding stage on an ELP, there would be a large unfilled niche mechanisms inherent in a four-photon mechanism necessitates
for the catalysed use of light energy to permit CO2 reduction more energy loss by unavoidable back reactions. Furthermore,
using weak reductants at rates higher than can be achieved by a the requirement for four rather than two photochemical reac-
chemolithotrophic mechanism. tions means that irradiation-limited rates of photosynthesis for a
Ultimately, as on Earth, even the very weak reductants dis- mechanism using four photons per electron would only be half
solved in water would be used up, since regeneration of the that of a mechanism with two photochemical reactions, other
reduced form by regeneration from the oxidised form using “res- things being equal. Finally, a greater number of photochemi-
piratory” reactions oxidising organic carbon back to CO2 would cal reactions implies a more complicated mechanism, where the
be incomplete because particulate organic carbon can be sedi- use of more catalytic macromolecules per redox sequence from
mented and thus spatially separated from the oxidant (see also water to CO2 could constrain the rate of radiation-saturated pho-
Walker 1987). This imbalance sets up the scene on the ELP, as tosynthesis (Raven 1984a,b, 1987). We note that the maximal
on Earth, for the evolution of O2 -producing photosynthesis with wavelength for the four-photon mechanism is up to 1500 nm,
water as reductant. which is still well beyond the wavelength of maximum energy
These arguments suggest that the evolution of O2 -producing flux of an M0 star (783 nm; see Table I); the Teff for a star with
photosynthesis is very likely on an ELP. We now consider a λmax of 1460 nm is only 1930 K.
whether the magnitude of such photolithotrophy is likely to yield The upper limit of photon flux density at which photosyn-
sufficient O2 accumulation in the atmosphere of an ELP to yield thesis can occur on Earth is set by the availability of nonpho-
enough O3 in the stratosphere to be potentially detectable for our ton resources used in constructing and using the photosynthetic
Solar System and to act as a UV-B and UV-C screen (Lovelock apparatus and by the occurrence of photodamage at high pho-
1975, Cockell 2000) but also to produce a potentially detectable ton flux densities (Raven 1984a, 1989, 1994, Long et al. 1994,
(from our Solar System) quantity of photosynthetic pigments. Anderson et al. 1997, Maule et al. 1995, Maule and Andrews
A determinant of gross O2 production in photosynthesis is the 1996). These data suggest that some plants on Earth could grow
incident photon flux density, the fraction of the incident photons at a photon flux density of 6000–9000 µmol photon m−2 s−1 ,
which are absorbed, the wavelength range over which photosyn- despite the maximum natural photon flux density (400–700 nm)
thesis is possible, and the number of photons needed to reduce on Earth being 2000 µmol photon m−2 s−1 . We suggest that
one molecule of CO2 to carbohydrate or to evolve one molecule photosynthetic primary production can occur on land on an ELP
of O2 . The two latter factors are closely related, since the ener- with 10,000 µmol photon m−2 s−1 ; aquatic habitats permit the
getics of CO2 reduction and O2 evolution require two photons at screening of supra-optimal irradiance by water, provided the
700–730 nm, three photons at 1050–1095 nm, or four photons water is deep enough (Falkowski and Raven 1997).
at 1400–1460 nm (Hill and Bendall 1960, Hill and Rich 1983, For the application of these considerations on the potential for
Heath et al. 1999). Thus, the greater the long-wavelength limit photosynthesis on an ELP, we deal with an ELP located at the
for photosynthesis, the larger the number of photons needed to inner edge of the habitable zone. Full details of the calculation
fix each CO2 and evolve each O2 . The quantity of photosyn- are given in the Appendix. In Fig. 1 we show the flux of photons
thetic pigment per unit area determines the fraction of incident absorbed by land plants at the surface of a cloud-free ELP or-
radiation absorbed, with a decreasing increment of photons ab- biting at the inner edge of the habitable zone for different parent
sorbed for each doubling of pigment per unit area. The quantity stars: F0V (1.85 AU), G0V (1.02 AU), G2V (0.95 AU), K0V
of pigment per unit area is inter alia not only a function of the (0.67 AU), and M0V (0.25 AU). The dependence of photosyn-
availability of photons to energize synthesis of the pigments but thesis on photon flux rather than energy flux and the decreasing
also a function of the availability of nutrients needed to produce atmospheric attenuation with increasing wavelength provides a
the pigments and the associated catalysts (C, N, P, S, Mg, Fe, bias in favour of cooler stars, but the F0V star still provides
Mn, Cu, Zn, etc.) and of the size of the proteinaceous (or anal- the best parent star environment for an ELP. Integration across
540 WOLSTENCROFT AND RAVEN

diffuse (sky) solar radiation and (b) the cloud cover measured
at meteorological stations across the globe, plus global cloud
cover statistics, to infer the global value of the PAR in terms of
G(clear) the global surface flux of PAR incident on a cloud-free
Earth. In Section A.2 of the Appendix we estimate an annual
time average of G = 0.67 G(clear), but note that this value is
slightly uncertain; this value contrasts with that for a completely
overcast Earth of 0.25 G(clear).
Clearly we have no way of estimating the cloud cover on an
ELP and so we need to calculate models for the global photo-
synthetic productivity for all possible values of the cloud cover,
as well as for values for the fraction, h, of the planet covered in
land as opposed to oceans.
Using the former approach we use Eq. (A5) (in the Appendix),
which gives the global photosynthetic productivity, p (in units
FIG. 1. Flux of photons absorbed by land plants (units of 1017 photons
of molecules of O2 generated globally per second), in terms
m−2 s−1 nm−1 ) at the surface of a cloud-free ELP (direct light from parent star of the fraction of the planet covered in land (h) and oceans
only) for an ELP at the inner edge of the star’s habitable zone for five types of (1 − h) for three values of the fraction of global cloud cover, f .
main-sequence star. Values of p are tabulated in Table A9 for five types of star from
F0V to M0V and illustrated in Fig. 3 for F0V, G2V, and M0V
stars. For given values of f and h the productivity p is greatest
the PAR yields the flux of photosynthetically productive pho- for the hottest F0V stars and least for the coolest M0V stars.
tons (Table A5 in the Appendix). In Fig. 2 the flux of photons Nevertheless, a given level of O2 productivity can be attained
observed by marine algae at 10 m depth has been calculated from a variety of cloud and land fraction values on different
assuming deep ocean seawater transmission, which varies from parent stars.
97% at the blue peak of chlorophyll (450 nm) to 0.3% at the Returning to the question of whether a three- or four-photon
red peak (∼700 nm); the integrated flux of photosynthetically mechanism might have evolved on ELPs orbiting the cooler
active photons is listed in Table A8. In this case the blue bias K or M stars we note that the achieved rate of photosynthesis
favours the hotter F0V star and this bias would increase with an could be as little as two-thirds or half that for the terrestrial two-
increase in the fraction of the ELP covered in oceans. photon mechanism. An example is an M0V ELP (h = 1.0) for
We now tackle the question of clouds. In the Appendix we which the assumed range of PAR is 600–1050 nm and with a
use the empirical relations between (a) the flux of direct plus three-photon mechanism of photosynthesis. We assume that the
pigments absorb the fraction of photons tabulated in Table A3
but at wavelengths of 600 nm (0.95), 675 nm (0.94), etc. The
integrated flux between 600 and 1050 nm PAR yields 14.4
instead of 4.63 for the two-photon mechanisms (in units of
1020 photon m−2 s−1 ; see Table A5). In applying formula (A3) to
obtain global photosynthetic productivity (Table A9) allowance
must be made for the 12 rather than 8 photons needed to gener-
ate one molecule of O2 . Nevertheless, this brings the cloud-free
ELP value of 77 (two-photon mechanism) to 160 (three-photon
mechanism) (in units of 1030 O2 molecules s−1 ) for an M0V
parent star. This advantage could turn into a disadvantage for a
planet largely covered in oceans because of the attenuation by
seawater at wavelengths beyond about 600 nm.
The implicit assumptions about the possible wavelength range
over which photosynthetic pigments absorb photons deserve
rather more attention. We have seen that photosynthesis on Earth
involves chlorin pigments with two absorption peaks, the Soret
band in the blue region of the spectrum (∼430 nm) and the Qy
band in the red (∼670–680 nm). Regardless of whether the pho-
FIG. 2. Flux of photons absorbed by algae (units of 1017 photons m−2 s−1
nm−1 ) at 10 m ocean depth on a cloud-free planet (direct light from the parent
tons are harvested by the Soret Band (second excited singlet) or
star only) for an ELP at the inner edge of the star’s habitable zone for five types the Qy band (first excited singlet) the energy used in photochem-
of main-sequence star. istry is that of the Qy band with a longer lifetime of electronic
PHOTOSYNTHESIS ON EARTH-LIKE PLANETS 541

FIG. 3. Global photosynthetic productivity, P, of molecular oxygen (units of 1030 molecules s−1 ) for an ELP at the inner edge of the habitable zone, as a
function of the fractional cloud cover, f , and for various fractions of the ELP covered by land and ocean for three types of parent star: (a) F0V, (b) G2V, and
(c) M0V.

excitation, increasing the chance of productive photochemistry on land after the production of an atmospheric UV screen on
leading to stable redox products. In the context of evolution, such a planet would be maximized if the pigments were also
the early occurrence of photosynthesis under the UV screen of active at longer wavelengths, a requirement met by chlorin pig-
liquid water flow would bias the incoming radiation hitting the ments. We note that the evolution of photosynthesis at depth in
ocean surface to the shorter visible wavelengths and the longer the ocean would not have precognition of the wavelength range
wavelengths of UV-A (320–400 nm) by the time a deep-growing to which organisms would be exposed in shallow water or on
photosynthetic organism is reached. Even with greater UV ab- land. While a chlorin pigment using only its Soret band for ab-
sorption relative to absorption at longer wavelengths in the early sorption would not have an energetic requirement to use more
ocean due to the presence of Fe2+ in anoxic waters (Olson and than two photons per electron for O2 -evolving photosynthesis,
Piersen 1986, Garcia-Pichel 1998) there would still be an em- very significant changes to the photosynthetic apparatus would
phasis on shorter wavelengths of visible radiation at the depths be required if more photons per electron are to be used to permit
at which UV-B has been screened out to nonlethal levels. This use of infrared photons and thus use the full wavelength range
would favour the evolution of pigments which have absorption on the planetary surface. It is evolutionarily relatively easy to
in the blue region. If the pigments only absorbed in the blue, then cause large in vivo shifts to longer wavelengths for the Qy band
the depth at which photosynthesis could occur in oceans in the (e.g., to between 760 and 900 nm in photosynthetic bacteria);
CHZ of an ELP associated with an M0 star is restricted since the altering the number of photoreactions is less readily envisioned.
blue (and, perhaps, UV) emission from such stars is relatively For F0 parent stars, the chlorin pigments used on Earth would
small. The occurrence of photosynthesis at shallower depths or be appropriate for UV screening at depth in the ocean (e.g., in
542 WOLSTENCROFT AND RAVEN

early evolution) as well as later for life near the ocean surface and nonlinearities in the feedback effects in O3 photochemistry
or on land on an oxic ELP. The greater depth of ocean needed mean that the ozone layer is only 18% thinner than that of Earth
to screen out the higher UV flux from an F0 star would be per- with the same O2 partial pressure (Kasting et al. 1997). Thus,
mitted by a chlorin pigment due to the higher blue radiation flux the lower flux of longer wavelength UV relative to that of wave-
from the F0 star as a fraction of its higher output. lengths beyond 400 nm from the late-type star is not offset by the
Attempting now to integrate the various evolutionary and even lower flux of short-wavelength UV reducing the O3 screen-
astronomical threads to see what could be the constraints on ing effect in the oxygenated atmosphere. We conclude that the
photosynthesis by ELPs, we deal first with the inner limit of ratio of UV-B to flux above 400 nm is lower on an ELP orbiting
the CHZ. For Earth, orbiting our (G2V) Sun, at the inner limit a late-type planet than it is on Earth.
(Table I) the flux would have 1.10 times the photon input on the Turning to the possible rates of net O2 accumulation on ELPs,
Earth at the moment. This would not threaten the occurrence and the steady-state O2 levels, we summarize the possible rates
of photosynthesis due to photoinhibition even on a planet with of photosynthesis for planets with 50% ocean and 50% land in
a lower atmospheric albedo than the present Earth, nor would Table A9. An ELP at the outer edge of the CHZ intercepts only
the occurrence of photosynthesis be threatened by screening by half as many photons as a similar sized ELP at the inner edge.
any plausible atmospheric albedo. For a cloud-free ELP orbiting For an ELP within the HZ, the range relative to the present Earth
an F0V star, the photon flux at the inner limit is greater but by is as small as 5 : 1 for an M0V ELP at the outer edge of the HZ.
less than a fraction of 2 than that on the present Earth (see Fig. 1), If the global cloud cover, f , were an entirely free parameter
although with more blue and UV-B and less red and infrared. then the range could be significantly greater. These values sug-
Even with a higher albedo (Kasting et al. 1993) such a star would gest that the O2 -producing primary productivity of ELPs of the
permit global photosynthesis at a rate not much greater (∼20%) type considered here are, in the CHZ, between about one-fifth
than for the present Earth (Table A9), albeit with a higher UV and twice that on the present Earth. To translate the capacity for
flux and hence the need for UV screening by compounds pro- photosynthetic O2 production into net O2 accumulation in the
duced by the organism. Kasting et al. (1997) discuss the UV flux atmosphere, we need to consider the processes which consume
from an F2V star and suggest that the 200–300 nm UV flux on a O2 on the ELPs. An ELP with one-fifth the rate of primary pro-
climatically habitable O2 -free planet would be 5–40 times that duction on the present Earth (e.g., an outer limit M0V) might
for the primitive Earth. Kasting et al. (1997) also point out that have an O2 accumulation rate which is much lower than one-
the evolution of photosynthetic life on such planets could (as on fifth that of the Earth. One reason is that biological processes
Earth) involve living under water, forming mats, and develop- other than photosynthesis determine O2 accumulation; these are
ing intra- and extracellular UV-protection mechanisms. Another organic C burial processes rather than reoxidation and intercon-
important point made by Kasting et al. (1997) is that once O2 versions of Fe2+ and Fe3+ and S2− , S0 , and SO2− 4 . To the extent
(from photosynthesis) has accumulated, the high UV flux below that organic C burial is limited by P (and other nutrient) inputs to
200 nm which is incident on the top of the atmosphere of a ELP the biosphere, if weathering is limited by photosynthetic activity
orbiting an F2V star would produce an O3 layer twice as thick (Algeo and Scheckler 1998) then organic C burial may comprise
as on Earth for the same atmospheric O2 partial pressure. This a smaller fraction of primary production than on Earth. How-
would serve to reduce the UV flux incident on the surface of an ever, if phosphorus supply is not constrained in this way, then a
ELP orbiting an F2V star to less than the value of 5–40 times greater fraction of primary production may be buried on an outer
the value incident on the primitive Earth. limit M0V ELP. However, the abiological processes consuming
For M0V parent stars we have already suggested that, be- O2 , i.e., reductant supply from the effectively infinite reductant
cause of the longer wavelength of radiation from such stars, it stores deeper in the crust by volcanic and hydrothermal activi-
is possible that a three-or four-photon mechanism might bring ties (Berner and Petsch 1998, Paytan et al. 1998), are likely to
the photosynthetic rate closer to that on the present Earth, even consume a greater fraction of photosynthetic O2 when primary
granted that absorption by atmospheric CO2 and H2 O vapour productivity is lower.
diminishes the potential for (especially land surface) photosyn- This comparison involves planets very similar in size to Earth.
thesis at very long wavelengths. A second effect of the longer Larger planets with tectonic activity, which are continually bri-
wavelength of radiation is that the (probably) lower UV flux nging fresh reductant from the larger volume (proportional to ra-
from such a later-type star relative to the flux of longer wave- dius cubed) of crust to the area (proportional to radius squared)
length radiation means that the early evolution of life before any of planetary surface, are predicted to have a larger ratio of O2
O2 accumulation in the atmosphere would be less threatened by consumption by inorganic reducing agents to photosynthetic
high UV flux than in the other cases considered here. A counter- O2 production than on smaller tectonically active planets
argument by Sheldon (quoted by Kasting et al. 1997) suggests (McKay 1998).
a minimal UV screening by O3 once O2 has accumulated on an More massive, tectonically active ELPs near the outer limit of
ELP in the CHZ of a late-type star, due to the very strong de- the CHZ with relatively low (radiation-limited) rates of photo-
pletion of the shortest UV wavelengths which convert O2 to O3 . synthesis might produce O2 from photosynthesis at so low a rate
However, detailed calculations do not bear out this expectation, as to not exceed the rate at which O2 is consumed by inorganic
PHOTOSYNTHESIS ON EARTH-LIKE PLANETS 543

reductants (see McKay 1998). This would mean that even an imaging and spectroscopy with NASA’s Planet Imager would
ELP with long-term (≥2 × 109 years) occurrence of photosyn- finally permit direct evidence for the presence of photosynthetic
thetic organisms generating O2 might not have enough O2 in its pigments.
atmosphere, either to provide sufficient O3 as a UV screen or In a similar vein it is worth recalling our earlier suggestion that
for this low level of O3 to be remotely detected. Further, if the for cooler parent stars, especially M0V stars, the evolutionary
parent star is both cool (say M0V) and old, so that it is rotating pressure for the photosynthetic pigments to match the spectral
slowly, then the combined photospheric and chromospheric UV energy distribution (SED) of the star could in principle lead to
emission could be too low to provide a UV screen, irrespec- three- or four-photon systems that would have a long-wavelength
tive of the rate of O2 generation, as a result of the low level of limit to the pigment absorption system at 1050 or 1400 nm; the
UV flux. Paradoxically, this means that, despite the minimal UV corresponding reflection discontinuity for such ELPs would be
flux incident on the upper atmosphere, sufficient UV could reach such that observations across this boundary would be needed.
the ELP surface to limit the photosynthetic generation of O2 to Ironically if this has in fact happened in some cases there would
organisms in the oceans. These considerations mean that the ac- probably be no great improvement in the star/planet brightness
cumulation of significant O2 on an outer limit ELP may take contrast ratio at these longer wavelengths (if the SED and pig-
several billion years. Thus, most ELPs on which life developed ments remained well matched) and hence in the ability to search
would become oxygenated within a few billion years, but some for evidence of pigments, unless the evolutionary process was
outer limit planets might not. Accordingly, the spectroscopic O2 slow or ineffective.
(via O3 ) assay seems like a reasonable but not all-encompassing
one for life on ELPs. APPENDIX: CALCULATION OF
Finally, what are the prospects for sensing photosynthetic pig- PHOTOSYNTHETIC PRODUCTIVITY
ments and O3 (and hence O2 ) on ELPs? The Earth intercepts
4.6 × 10−10 of the solar radiation and thus, if viewed from a The rate at which O2 is generated on a remote ELP depends on the rate at
which photons are absorbed by the pigments of photosynthetic organisms at
distance of 10 parsecs, the Earth, with its visual albedo of about
the planetary surface. To calculate this rate we need to know the flux, F(λ), of
0.3, would contribute just one in 7.3 × 109 of the photons that radiation from the parent star at the top of the ELP atmosphere that falls in the
might be detected in an alien astronomer’s telescope aperture photosynthetically active range (PAR); the fraction, j(λ), of this radiation that
focussed on the Sun and inner planets. If this astronomer had an reaches the planetary surface; and the fraction, g(λ), of these photons that are
instrument capable of completely separating the solar and terres- absorbed and are used in the photosynthetic process.
trial optical photons entering it—the Earth being just 0.1 arcsec
from the Sun—it would only take 20 min to accumulate 100 ter- A.1. Photon Flux at the Top of the ELP Atmosphere
restrial photons in a 50 bandpass at 550 nm with a collecting area Given the distance of the ELP from its parent star and the type of star, we
equivalent to that of a 5-m telescope with 50% throughput and know F(λ) but have no a priori knowledge of either j(λ) or g(λ). In the simplest
case (Section IV) we have assumed that the planet is identical to the Earth in
thus to allow this observer to start searching for the reflection dis-
terms of its atmospheric j(λ), the properties of the pigments defined by g(λ),
continuity at 700 nm associated with photosynthetic pigments. and its planetary diameter. The energy flux of a star with a V magnitude of zero
However, the technical difficulties of achieving this discrimina- [λ(eff) = 550 nm] is tabulated by Allen (1973) for main-sequence spectral types
tion between the optical photons from the ELP and its parent A0V, F0V, G0V, K0V, and M0V. Since photosynthetic productivity depends on
star are unlikely to be overcome soon. While an 8-m telescope the photon flux and not on the energy flux, we need to calculate the photon flux,
which at the top of the ELP atmosphere will depend on the absolute V magnitude
with wavefront control of λ/5000 ∼ 0.1 nm and an advanced
of the class of parent star and the actual distance between ELP and parent star.
coronagraph could in principle achieve this (Malbet et al. 1995, We adopt the values of the absolute V magnitude tabulated by Lang (1992) and
Woolf and Angel 1998), this approach is unlikely to be realised use them to determine the distance, d, at which each class of star would have
in practice in time for the Next Generation Space Telescope. magnitude V = 0. The photon flux for an ELP 1 AU distant from its parent star
Almost certainly the first step will be to detect mid-IR ELP ra- is then obtained by multiplying by [d/(1 AU)]2 and is tabulated in Table A1.
diation using a nulling interferometer in space (see, e.g., Woolf
and Angel (1998) and the Terrestrial Planet Finder Report by A.2. Net Photon Flux at the ELP Surface
Beichman et al. (1999)) and to search for mid-IR absorption in The net flux of photons arriving at any given location of the Earth’s surface
the ELP atmosphere by O3 (as well as CO2 , H2 O, and CH4 ) comprises the direct sunlight after transmission through the atmosphere and
the diffuse component scattered in the atmosphere (Fleagle and Businger 1963).
as evidence for O2 -generating photosynthesis. The next step
Both of these components depend on the cloud cover at this location. To estimate
would be to infer seasonal changes in the level of photosynthesis the global PAR flux, G, integrated over the planet we need to be able to assess
which could be associated with the relative ocean and land cov- the global cloud cover, which varies with time of year and latitude, as does the
erage in the two hemispheres. Such changes in the atmospheric type, vertical extent, and opacity of “cloud,” and how the diffuse and direct
O3 parameters measured on a pale blue dot would perhaps be components differ at clear and cloudy sky sites. This information comes from
both satellite studies and surface observations from meteorological stations but
globally smoothed out and difficult to interpret for a number of
is not well determined.
factors including seasonal cloud cover and O3 /O2 chemistry. It Measurements of the spectral reflectivity of the Earth viewed towards the nadir
remains to be seen what further progress could be achieved with from the GOME (Global Ozone Monitoring Experiment) spectrometer across
a single pale blue dot to work with, until the advent of optical the PAR range (Burrows et al. 1999) show that for 100% cloud cover the albedo
544 WOLSTENCROFT AND RAVEN

TABLE A1 TABLE A2
Photon Flux a at the Top of the Atmosphere of an ELP 1 AU Direct Component of Photon Flux a at the Surface of a Cloud-Free
from Its Parent Star Planet 1 AU from Its Parent Star (Earth-like Atmosphere)

λ (nm) A0V F0V G0V K0V M0V λ (nm) A0V F0V G0V G2V K0V M0V

350 1474 155 25.9 — — 350 693 73.0 12.2 16.6 — —

400 4001 374 55.6 8.74 0.219 400 2520 236 35.0 27.9 5.51 0.138
450 3346 409 71.9 14.9 0.658 450 2440 299 52.5 44.4 10.9 0.480
500 2831 378 75.6 16.9 0.890 500 2240 298 59.7 48.1 13.4 0.703
550 2363 357 74.8 18.8 1.30 550 1960 296 62.1 50.7 15.6 1.08
600 1942 314 71.1 21.4 1.82 600 1630 264 59.8 52.3 18.0 1.53
800 1193 226 62.2 22.7 2.37 650b 1460 245 59.4 51.3 19.9 1.80
1000 889 177 51.9 19.9 2.15 700b 1320 231 59.0 50.2 20.7 1.97
800 1120 212 58.5 45.9 21.3 2.23
a Units of 1017 photons m−2 s−1 nm−1 . 1000 854 169 49.8 36.4 19.1 2.78

a Units of 1017 photons m−2 s−1 nm−1 .

b Interpolated values.
is between 55 and 60% and approximately wavelength independent; the type of
cloud is not known and so this cannot be assumed to be a typical value for cloud
albedo; however, the neutral colour of the reflected light is probably typical and
bluer than solar but somewhat less in the broken cloud case. However, given
related to the scattering of light by water droplets in the clouds. The diffuse
the overall uncertainties we have ignored these spectral differences, which are
light of the sky seen below the clouds is also approximately neutral in colour
relatively small, and assumed that all components have the same SED. The zenith
relative to sunlight; i.e., it has the same colour as the direct sunlight (see, e.g.,
sea level transmission, T , of the Earth’s clear atmosphere (see Allen 1973) across
Bonhomme 1993) since scattering dominates and little light is truly absorbed in
the photosynthetically active range (400–700 nm) is due mainly to molecular
the clouds in the 400–700 nm range. As expected the GOME spectrum above
scattering and shows a redward bias with T ranging from 63% at 400 nm to 91%
a clear part of the Atlantic Ocean (which has low surface albedo) is dominated
at 700 nm. The wavelength dependence of the photon flux reaching the surface
by Rayleigh scattering from the clear atmosphere.
of a cloud-free planet whose atmospheric transmission is identical to that of the
The dependence of the flux, Q, of diffuse plus direct solar radiation at surface
Earth is given in Table A2 for a planet which is assumed to be 1 AU from its
stations on the cloud cover has been examined by a number of authors, and
parent star. The quantity tabulated is the direct photon flux: the global value,
these studies have been discussed by Coulson (1975). Probably the most useful
G(clear), including the diffuse sky contribution is assumed to be 1.25 times the
relation is Ångstrom’s formula (1924),
value listed in Table A2.

Q = Q(clear) [k + (1 − k)(n/N )], (A1)

A.3. Photosynthetic Rate at the Planetary Surface (Land)
where Q is the flux of diffuse plus direct solar radiation measured on a horizontal Having determined the photon flux at the surface of a planet identical to
surface during a day when n/N is the fraction of the day with bright sunshine, and the Earth at 1 AU from its parent star we can now estimate the photosynthetic
Q(clear) is the value on a cloudless day (n/N = 1); k is a constant that depends rate for a planet within the habitable zone of its parent star, assuming that the
on the type of cloud and varies from about 0.25 for stratus cloud to 0.82 for photosynthetic pigments are also identical to those on Earth. To evolve one
cirrus (Haurwitz 1948; see also Table 4.1 of Coulson 1975). This is measured O2 molecule requires eight PAR photons, which are absorbed by the pigments:
over a wider wavelength range than that of the PAR, but it should provide a the fraction of photons absorbed by the leaf of a typical land plant is listed in
satisfactory estimate for our purposes. Global cloud cover statistics based on Table A3 (taken from Gates 1980).
surface observations at meteorological stations for the period June 1995 to May Thus photosynthesis due to “typical” plants such as the oleander (Nerium
1996 as a function of latitude (Breon, private communication, 1998) yield an oleander) would generate O2 molecules globally at the photosynthetic rate,
estimate of the global average of 13% clear sky, 65% broken cloud component, p(λ) (units of O2 molecules/s), where
and 22% overcast. As a working model we adopt (k = 0.25, n/N = 0) for the
overcast component and (k = 0.75, n/N = 0) for the broken cloud component, p(land) = 1.25 × (1/8) × h Bq1 × (G/G(clear)) × T1 × [A2 · A3] (A3)
leading to global values G (as opposed to Q values at surface stations) of
and where the factor 1.25 allows for the diffuse sky radiation; (1/8) accounts
G = G(clear) [0.13(1.0) + 0.65(0.75) + 0.22(0.25)] = 0.67 G(clear), (A2) for the number of absorbed photons needed to generate one molecule of O2 ;
B is the cross-sectional area of the solid planet, h is the fraction of the Earth
where G(clear) is the global value of the PAR flux incident at the surface of
a cloudless Earth. This compares to 0.25 G(clear) for a totally cloud-covered
Earth. Note that the appropriate k value to be used here is somewhat uncertain.
Perhaps just as valid would be to assume that the broken cloud component was TABLE A3
half clear and half overcast (k = 0.25), which would yield 0.59 G(clear); i.e., Fraction of Photons Absorbed by the Photosynthetic Pigments
overall a value G/G(clear) = 0.63 ± 0.04 is more realistic. In the preceding in a Leaf of Oleander (Nerium oleander)
we have assumed that these three components have the same SED, i.e., solar,
as modified by the transmission of the atmosphere. This is true for the direct λ (nm) 400 450 500 550 600 650 700
components in all three cases but less so for the diffuse (scattered) components
for clear and broken cloud. For the clear sky case about 20% of the Q(clear) Fraction absorbed 0.95 0.94 0.93 0.78 0.86 0.90 0.82
value comes from the diffuse, Rayleigh scattered component, which is hence
PHOTOSYNTHESIS ON EARTH-LIKE PLANETS 545

TABLE A4 amount of cloud is implicity included by assuming a high surface albedo (0.22),
Direct Component of Flux of Photons Absorbed by Land Plantsa which allows their climate model to be consistent with the current mean global
at the Surface of a Cloud-Free Planet for Unit Air Mass for an ELP terrestrial surface temperature at 288 K. As KWR acknowledge, their assumption
1 AU from Its Parent Star implies a fixed amount of cloud cover with no cloud feedback. There are still
very considerable difficulties in putting clouds (at various optical thicknesses
λ (nm) A0V F0V G0V G2V K0V M0V and atmospheric heights) into realistic climate models. While the KWR models
provide a reliable guide to the location of the habitable zones, they cannot be
400 2394 224 33.3 26.5 5.23 0.131 used to assess the effects of variable cloud cover and some other approach is
450 2294 281 49.4 41.7 10.2 0.451 needed.
500 2083 277 55.5 44.7 12.5 0.654 To deduce the flux of PAR radiation received at the surface of an ELP we
550 1529 231 48.4 39.5 12.2 0.842 clearly need to estimate the cloud cover factor G/G(clear). For the Earth we
600 1402 227 51.4 45.0 15.5 1.316 can consider a simple approach, namely the use of the global version of Eq. (A1),
650b 1314 221 53.5 46.2 17.9 1.620
700b 1082 189 48.4 41.2 17.0 1.615 Q/Q(clear) = k + (1 − k)(n/N ) → G/G(clear) = (1 − f ) + k f, (A1)
a Units of 1017 photons m−2 s−1 nm−1 . where f is the fraction of global cloud cover. The fraction of the day with bright
b Interpolated values. sunshine, n/N , transforms to the fraction of the planet with no cloud, (1 − f ), for
which n/N = 1 and hence G/G(clear) = 1; and the cloudy fraction of the day,
for which n/N = 0, transforms to the cloudy fraction of the planet and hence
G/G(clear) = k. For simplicity we assume either all clear or all overcast (stra-
where surface photosynthesis as opposed to aquatic photosynthesis takes place;
tus: k = 0.25) conditions: hence G/G(clear) = 1 − f + 0.25 f = 1 − 0.75 f .
q1 is the effective fraction of that surface covered in such plants (allowing for
In the case of the Earth we could in principle produce a more sophisticated
the density of plants and nutrient availability); G/G(clear) allows for cloud
estimate of G/G(clear) and attempt to apply it to other ELPs, since we have
cover; T1 is an atmospheric transmission factor that takes account of the fact
detailed knowledge of how the visible solar radiation incident at the top of the
that Table A2 is calculated for a zenith air mass; and [A2 · A3], the product of
Earth’s atmosphere is distributed: namely 31% is reflected into space (global
Tables A2 and A3, is proportional to the flux of photons absorbed by the plant
albedo, A), 23% is absorbed by the atmosphere and clouds, and 46% is ab-
pigments. For planets identical in all respects to the Earth all the factors save
sorbed at the ground (see, e.g., Schlesinger 1997). Of the 31% albedo, about
[A2 · A3] would be the same: [A2 · A3] is listed in Table A4.
6% is attributable to scattering at the surface and by the atmosphere, and 25% is
Atmospheric models by Kasting et al. (1993; hereafter KWR) indicate that
attributable to cloud: recent studies of the global albedo by means of Earthshine
an ELP will not be able to retain its surface water due to catastrophic water loss
observations (Goode et al. 2001) have noted that the albedo varies over time by a
within a certain distance of its parent star, namely the inner edge of the habitable
few percent with the expected correlation between albedo and cloud cover being
zone. Using the figures in Table A4, and for the moment ignoring the other
confirmed. One difficulty in trying to infer useful facts about the G/G(clear) for
factors in Eq. (A3), we list in Table A5 the integrated flux of photosynthetically
the PAR radiation from these values is that the albedo measurements are gener-
productive photons across the PAR range (400–700 nm) for ELPs at the inner
ally made not only at visible wavelengths but also at wavelengths beyond the PAR
edge of the HZ for the various parent stars. As expected this shows that for a
range (>700 nm) where atmospheric absorption plays an important role, so that
cloudless planet the photosynthetic productivity is much greater for the hotter
the figures used here on absorption by clouds and atmosphere (23%) and at the
F0V stars than for the cooler K0V and M0V stars.
surface (46%) do not necessarily apply in the PAR range.
The models of KWR, which were developed primarily to deduce the location
Any attempt to model the relation between albedo and cloud cover, even for
of the planetary habitable zones, also lead to values of planetary albedo that
Earth, is at best a difficult task because it involves a very wide range of albedos
differ for planets orbiting different types of star. These albedos are higher (and
of natural surfaces and of various cloud types (see, e.g., Miranova 1973) and
the PAR flux available at the surface is hence lower) for planets orbiting F
because the radiation transfer processes involved are complex. A simple relation
stars and the albedos are lower for planets orbiting M stars: this results from
between albedo and cloud cover for the Earth might be written
the differing amount of Rayleigh scattering in the ELP atmospheres (which is
greatest where there is the most incident blue parent starlight) and from the
greater atmospheric absorption by water vapour (see Appendix B of Kasting A = (R1 )f + (R2 )(1 − f ),
1988) at near-infrared as opposed to visible wavelengths. However, the KWR
models do not allow for changes in cloud cover: instead the presence of a fixed where the cloud reflectivity is R1 and the reflection into space by air and surface
is R2 . Using the cloud cover statistics of F. M. Breon (private communica-
tion, 1998) mentioned earlier we have f = 0.55, which, given (R1 ) f = 0.25 and
(R2 ) (1 − f ) = 0.6, yields R1 (cloud) = 0.45 and R2 (surface) = 0.13. This im-
TABLE A5 plies that the radiation retained by the Earth, 1 − A = 0.87 − 0.32 f , is 73%,
Integrated Flux of Photosynthetically Productive Photons 69%, and 66% for f = 45%, 55%, and 65%. These figures are not too far from
at the Inner Edge of the HZ the corresponding numbers based on G/G(clear) = 1 − 0.75 f , namely 66%,
59%, and 51%, and the differences between the two sets of numbers perhaps
Spectral Stellar Inner edge of Absorbed photon flux represents a realistic estimate of their errors. It is clear that in estimating the effect
type temperature (K) HZ (distance, AU) (1020 photons m−2 s−1 ) on cloud cover for other ELPs we have no choice but to use the equation (A1)
above for G/G(clear) for all ELPs.
F0V 7200 1.85 21.2
G0V 6030 1.02 14.4
A.4. Photosynthetic Rate in the Planetary Oceans
G2V 5860 0.95 14.0
K0V 5250 0.67 8.89 In the case of the Earth 71% of the planet is covered by oceans and a consid-
M0V 3850 0.25 4.63a erable amount of photosynthesis takes place in the surface layers of the oceans.
However, the photosynthetic productivity of the oceans is relatively poor com-
a 14.4 for photosynthesis using 3 instead of 2 photons/electron. pared to the land: each year the oceans fix about 40 G tonnes of carbon and the
546 WOLSTENCROFT AND RAVEN

TABLE A6 TABLE A8
Aquatic Photosynthesis Loss Factor Integrated Flux of Photosynthetically Productive Photons at the
Inner Edge of the Habitable Zone at 10 m Ocean Depth in the ELP
Seawater Fraction, c, of photons Ocean (Cloud-Free Planet)
transmission(s) absorbed by the green Aquatic loss
λ (nm) at 10 m depth macroalga Codium fragile factor (0.9sc) Inner edge of HZ Absorbed photon flux
Spectral type (distance, AU) (1020 photons m−2 s−1 )
400 0.94 0.91 0.77
450 0.97 0.92 0.80 F0V 1.85 11.4
500 0.94 0.92 0.78 G0V 1.02 7.2
550 0.70 0.86 0.54 G2V (Sun) 0.95 6.8
600 0.14 0.85 0.11 K0V 0.67 3.8
650 0.055 0.90 0.045 M0V 0.25 1.5
700 0.0025 0.73 0.0016

The integrated flux over the PAR wavelength range of Eqs. (A3) (land, see
land fixes about 60 G tonnes (Raven and Falkowski 1999). Photosynthesis in Table A5) and (A4) (ocean, see Table A8), i.e., the number of O2 molecules/s
coastal waters is relatively efficient and produces in each year about 2 to 3 G generated globally, is
tonnes of carbon, but the bulk of the photosynthesis occurs in the oceans beyond
the continental shelf. Light loss as the PAR radiation enters the ocean depends P = P(land) + P(ocean)
on the wave height and the solar zenith distance: we shall assume an overall light
= 0.156B × T1 × (G/G(clear))[hq1 [A5] + (1 − h)q2 T2 × [A8]]. (A5)
loss of 10% across the PAR range. The depth at which the great majority of PAR
photons are absorbed varies greatly and it is difficult to estimate the effective
ocean depth at which the bulk of photosynthesis takes place: we adopt a depth A.5. Selected Planetary Models
of 10 m.
The transmission, s, of seawater at 10 m depth, which we list in Table A6, is Table A9 lists the global photosynthetic productivity, P, for a representative
much higher in the blue than in the red, which favours the hotter stars. In this set of planet models. For simplicity we assume all the model ELPs have the
table we also list the fraction, c, of PAR photons absorbed by the photosynthetic same collecting area, B, as the Earth (i.e., 1.28 × 1018 cm2 ) and the same value
pigments in the marine green alga Codium fragile, which is typical of marine of T1 . We ignore the slight dependence of T1 on stellar spectral type and adopt
algae (Lüning and Dring 1985). Multiplication of these factors together with the T1 = 0.83 based on a mean air mass of 2.0. For the efficiency factors we adopt
factor of 0.9 for light loss at the surface yields the aquatic loss factor (0.9sc) q1 = 0.01 and q2 × T2 = 0.27q1 based on the relative terrestrial productivity
to be applied to the photon flux in Table A2 to give (in Table A7) the aquatic of land and oceans. We assume, as before, that G/G(clear) = 1.0 (0.25) for the
equivalent of Table A4. cloud-free (cloud-covered) parts of the planet respectively.
The equation for oceans analogous to Eq. (A3) is

p(ocean) = 1.25 × (1/8) × [(1 − h)Bq2 × (G/G(clear)) TABLE A9

× T1 × T2 × [A2.A6]], (A4) Global Photosynthetic Productivity of O2 a at Inner Edge
of Habitable Zone
where 1 − h is the fraction of the Earth where oceanic photosynthesis could h = 1.0 0.0
potentially be taking place; q2 is the effective fraction of this area where it is Spectral type r (AU) (land only) 0.5 (ocean only)
taking place (allowing for density of algae, etc. and nutrient availability); T2
is analogous to the term T1 for atmospheric transmission; and [A2.A6] is the I. Cloud-free planet
product of Tables A2 and A6 (listed in Table A7). F0V 1.85 351 200 51
G0V 1.02 239 136 32
G2V 0.95 232 131 30
TABLE A7 K0V 0.67 147 81 17
M0V 0.25 77 41 6.7
Direct Component of Flux of Photons Absorbed by Algaea at 10 m
Ocean Depth on a Cloud-Free Planet for Unit Air mass for an ELP II. 50% cloud cover, 50% cloud-free
F0V 1.85 219 125 32
1 AU from Its Parent Star
G0V 1.02 149 85 20
λ (nm) A0V F0V G0V G2V K0V M0V G2V 0.95 145 82 19
K0V 0.67 92 51 11
400 1940 182 27.0 21.5 4.24 .106 M0V 0.25 48 26 4.2
450 1952 239 42.0 35.5 8.72 0.384 III. 100% cloud cover
500 1747 232 46.6 37.5 10.5 0.548 F0V 1.85 88 50 13
550 1058 160 33.5 27.4 8.42 0.583 G0V 1.02 60 34 8.0
600 179 29.0 6.58 5.75 1.98 0.168 G2V 0.95 58 33 7.5
650 65.7 11.0 2.67 2.31 0.90 0.081 K0V 0.67 37 20 4.3
700 2.1 0.4 0.09 0.08 0.03 0.003 M0V 0.25 19 10 1.7

a Units of 1017 photons m−2 s−1 nm−1 . a Units of 1030 molecules s−1 .
PHOTOSYNTHESIS ON EARTH-LIKE PLANETS 547

ACKNOWLEDGMENTS Falkowski, P. G., and J. A. Raven 1997. Aquatic Photosynthesis. Blackwell

Science, Malden, MA.
J.A.R. acknowledges all past and present colleagues who have stimulated Ferris, J. P. 1999. Prebiotic synthesis on minerals: Bridging the prebiotic and
thought and curbed some excesses. We are very grateful to Professor J. L. RNA worlds. Biol. Bull. 196, 311–314.
Monteith for our discussions on clouds.
Fleagle, R. G., and J. A. Businger 1963. An Introduction to Atmospheric Physics.
Academic Press, New York.
Forget, F., and R. T. Pierrehumbert 1997. Warming early Mars with carbon
REFERENCES dioxide clouds that scatter infrared radiation. Science 278, 1273–1276.
Franck, S., W. von Bloh, C. Bounama, M. Steffen, D. Schönberner, and H.-J.
Algeo, T. J., and S. E. Sheckler 1998. Terrestrial–marine teleconnections in the
Schellnhuber 2000. Determination of the habitable zones in extrasolar plan-
Devonian: Links between the evolution of land, plants, weathering processes,
etary systems: Where are Gaia’s sisters? J. Geophys. Res. 105, 1651–1658.
and marine anoxic events. Philos. Trans. R. Soc. London B 353, 113–130.
Garcia-Pichel, F. 1998. Origins of life and evolution of the biosphere. Origins
Allen, C. W. 1973. Astrophysical Quantities, 3rd ed. Athens Press, Univ. of
Life Evolut. Biosphere 28, 321–347.
London, London.
Gates, D. M. 1980. Biophysical Ecology. Springer-Verlag, New York.
Anderson, J. M., Y. I. Park, and W. S. Chow 1997. Photoinactivation and pho-
toprotection of photosystem II in nature. Physiol. Plant. 100, 214–273. Golubic, S., and L. Seong-Joo 1999. Early cyanobacterial fossil record: Preser-
Ångstrom, A. K. 1924. Solar and terrestrial radiation. Q. J. R. Meteorol. Soc. vation, palaeoenvironments and identification. Eur. J. Phycol. 34, 339–348.
50, 121–126. Goode, P. R., J. Qiu, V. Yurchyshyn, J. Hickey, M.-C. Chu, E. Kolbe, C. T. Brown,
Ayres, T. R. 1997. Evolution of the solar ionizing flux. J. Geophys. Res. 102(E1), and S. E. Koonin 2001. Earthshine observations of the Earth’s reflectance.
1641–1651. Geophys. Res. Lett. 28, 1671–1674.
Ayres, T. R., T. Simon, J. R. Stauffer, R. A. Stern, J. P. Pye, and A. Brown 1996. Gough, D. O. 1981. Solar interior structure and luminosity variations. Sol. Phys.
Sleuthing the dynamo: Hubble Space Telescope faint object spectrograph 74, 21–34.
observations of solar-type dwarfs in young galactic clusters. Astrophys. J. Halldall, P. 1968. Photosynthetic capacities and photosynthetic action spectra
473, 279–293. of endozoic algae of the massive coral Favia. Biol. Bull. 134, 411–424.
Beichman, C. A., N. J. Woolf, and C. A. Lindensmith 1999. The Terrestrial Haurwitz, B. 1948. Insolation in relation to cloud type. J. Meteorol. 5, 110–113.
Planet Finder: A NASA Origins Program to Search for Habitable Planets. Heath, M. J., L. R. Doyle, M. M. Joshi, and R. M. Haberbe 1999. Habitability
JPL Publ. 99-3. of planets around red dwarf stars. Origins Life Evolut. Biosphere 29, 405–424.
Berner, R. A. 1998. The carbon cycle and CO2 over Phanerozoic time: The role Hill, R., and F. L. Bendall 1960. Function of the two cytochrome components
of land plants. Philos Trans. R. Soc. London B 353, 75–82. in chloroplasts: A working hypothesis. Nature 186, 136–137.
Berner, R. A., and S. T. Petsch, 1998. The sulfur cycle and atmospheric oxygen. Hill, R., and P. R. Rich 1983. A physical interpretation for the natural photosyn-
Science 282, 1426–1427. thetic process. Proc. Nat. Acad. Sci. 80, 978–982.
Bonhomme, R. 1993. The solar radiation: Characterization and distribution in Hoehler, T. M., B. M. Bebout, and D. J. Des Marais 2001. The role of microbial
the canopy. In Crop Structure and Light Microclimate (C. Varlet-Grancher, mats in the production of reduced gases on the early Earth. Nature 412,
R. Bonhomme, and H. Sinoquet, Eds.), pp. 17–28. Inst. National de la 324–327.
Recherche Agronomique, Paris.
Holland, H. D. 1984. The Chemical Evolution of the Atmosphere and Oceans.
Brocks, J. J., G. A. Logan, R. Buick, and R. E. Summons 1999. Archean Princeton Univ. Press, Princeton, NJ.
molecular biology and the early evolution of eukaryotes. Science 285, 1033–
Huang, S.-S. 1960. Life outside the Solar System. Sci. Am. 202(4), 53–65.
1036.
Jakosky, B. M., and E. L. Shock 1998. The biological potential of Mars, the
Broda, E. 1975. The Evolution of Bioenergetic Processes, revised reprint of the
early Earth and Europa. J. Geophys. Res. 103, 19359–19364.
1st ed. Pergamon, Oxford.
Burrows, J. P., M. Weber, M. Buchwitz, V. Rozanon, A. Ladstatter- Kasting, J. F. 1988. Runaway and moist greenhouse atmospheres and the evo-
Weissenmayer, A. Richter, R. DeBeek, R. Hoogen, K. Bramstedt, K. V. lution of Earth and Venus. Icarus 74, 472–494.
Eichmann, and M. Eisinger 1999. The Global Ozone Monitoring Experi- Kasting, J. F. 1996. Planetary atmosphere evolution: Do other habitable planets
ment (GOME): Mission concept and first scientific results. J. Atmos. Sci. 56, exist and can we detect them? Astrophys. Space Sci. 241, 3–24.
151–175. Kasting, J. F. 1998. Habitable zones around stars and the search for extraterres-
Canfield, D. E., K. S. Hobicht, and B. Thamdrup 2000. The Archean sulfur cycle trial life. Bull. Am. Astron. Soc. 30, 1328.
and early history of atmospheric oxygen. Science 288, 658–668. Kasting, J. F., D. P. Whitmire, and R. T. Reynolds 1993. Habitable zones around
Catling, D. C., K. J. Zahnle, and C. P. McKay 2001. Biogenic methane, hydro- main sequence stars. Icarus 101, 108–128.
gen escape, and the irreversible oxidation of early Earth. Science 293, 839– Kasting, J. F., D. C. B. Whittet, and W. R. Sheldon 1997. Ultraviolet radiation
843. from F and K stars and implications for planetary habitability. Origins Life
Chang, S. 1999. Planetary environments and the origin of life. Biol. Bull. 196, Evolut. Biosphere 27, 413–420.
308–310. Kasting, J. F., A. A. Pavlov, and J. L. Siefert 2001. A coupled ecosystem–
Cockell, C. S. 2000. The ultraviolet history of the terrestrial planets— climate model for predicting the methane concentration in the Archaean at-
Implications for biological evolution. Planet. Space Sci. 48, 203–214. mosphere. Origins Life Evolut. Biosphere 31, 271–285.
Coulson, K. L. 1975. Solar and Terrestrial Radiation. Academic Press, Kirk, J. T. O. 1994a. Optics of UV-radiation in natural waters. Arch. Hydrobiol.
New York. Beih. 43, 1–16.
Des Marais, D. J. 2000. When did photosynthesis emerge in Earth? Science 289, Kirk, J. T. O. 1994b. Light and Photosynthesis in Aquatic Systems, 2nd ed.
1703–1705. Cambridge Univ. Press, Cambridge, UK.
Des Marais, D. J., and M. R. Walker 1999. Astrobiology: Exploring the origins, Koehne, B., G. Elli, R. C. Jennings, C. Wilhelm, and H.-W. Trissl 1999.
evolution and distribution of life in the Universe. Annu. Rev. Ecol. Syst. 30, Spectroscopic and molecular characterization of a long wavelength absorbing
397–420. antenna of Ostreobium sp. Biochim. Biophys. Acta 1412, 94–07.
548 WOLSTENCROFT AND RAVEN

Kump, L. R., J. F. Kasting, and M. E. Barley 2001. Rise of atmospheric oxy- Raven, J. A. 1987. Limits to growth. In Microalgal Biotechnology (M. A.
gen and the “upside-down” Archean mantle. Geochem. Geophys. Geosyst. 2, Borowitzka and L. J. Borowitzka, Eds.), pp. 331–356. Cambridge Univ. Press,
U1–U10. Cambridge.
Lang, K. R. 1992. Astrophysical Data: Planets and Stars. Springer-Verlag, Raven, J. A. 1989. Fight or flight: The economies of repair and avoidance of
New York. photoinhibition of photosynthesis. Funct. Ecol. 3, 5–19.
Laskar, J., F. Joutel, and P. Robutel 1993. Stabilization of the Earth’s obliquity Raven, J. A. 1994. The cost of photoinhibition to plant communities. In Photoin-
by the Moon. Nature 361, 615–617. hibition of Photosynthesis from Molecular Mechanisms to the Field (N. R.
Leger, A., M. Pirre, and F. J. Marceau 1993. Search for primitive life on a Baker and J. R. Bowyer, Eds.), pp. 449–464. Bios Scientific Publications,
distant planet: Relevance of O2 and O3 detections. Astron. Astrophys. 277, Oxford.
309–313. Raven, J. A. 1996. The role of autotrophs in global CO2 cycling. In Microbial
Long, S. P., S. Humphries, and P. G. Falkowski 1994. Photoinhibition of photo- Growth on C1 compounds. Proceedings of the 8th International Symposium
synthesis in Nature. Annu. Rev. Plant Physiol. Plant Mol. Biol. 45, 633–662. on Microbial Growth on C1 Compounds (M. E. Lidstrom and F. R. Tabita,
Lovelock, J. E. 1975. Thermodynamics and the recognition of alien biospheres. Eds.), pp. 351–358. Kluwer, Dordrecht.
Proc. Soc. London B 189, 167–181. Raven, J. A., and P. G. Falkowski 1999. Oceanic sinks for atmospheric CO2 .
Lüning, K., and M. J. Dring 1985. Action spectra and spectral quantum yield Plant Cell Environ. 22, 741–755.
of photosynthesis in marine algae with thin and thick thalli. Mar. Biol. 87, Raven, J. A., J. E. Kübler, and J. Beardall 2000. Put out the light, and then put
119–129. out the light. J. Mar. Biol. Assoc. UK 80, 1–27.
Malbet, F., J. W. Yu, and M. Shao 1995. High-dynamic-range imaging using Russell, M. J., and A. J. Hall 1997. The emergence of life from iron monosul-
a deformable mirror for space coronagraphy. Publ. Astron. Soc. Pac. 107, phide bubbles at a submarine hydrothermal redox and pH front. J. Geol. Soc.
386–398. London 154, 377–402.
Maule, H. G., and M. Andrews 1996. Relationship between shoot to root ra- Schlesinger, W. H. 1997. Biogeochemistry, 2nd ed. Academic Press, San Diego.
tio of leaf protein content of Tradescantia fluminensis exposed to different Schiller, H., H. Senger, H. Miyashita, S. Miyachi, and H. Dau 1997. Light-
irradiance levels. J. Exp. Bot. 47, 1317–1318. harvesting in Acaryochloris marina—Spectroscopic characterization of a
Maule, H. G., M. Andrews, J. D. Morton, A. V. Jones, and G. T. Daly 1995. chlorophyll d-dominated photosynthetic antenna system. FEBS Lett. 410,
Sun–shade acclimation and nitrogen nutrition of Tradescantia fluminensis, 433–436.
a problem weed in New Zealand native forest remnants. N. Z. J. Ecol. 19,
Simon, T. 2001. The connection between rotation and activity. In Proceedings
35–46.
of the 11th Cambridge Workshop on Cool Stars, Stellar Systems and the Sun
Maynard Smith, J., and E. Szathmáry 1997. The Major Transitions in Evolution. (R. J. G. Lopez, R. Rebolo, and M. R. Z. Osorio, Eds.), ASP Conference
Oxford Univ. Press, London. Series, Vol. 233, pp. 235–249. Astron. Soc. Pacific, San Francisco.
McKay, C. P. 1998. Life in the planetary concept. In Origins. (C. E. Woodward, Sleep, N. H. 2001. Oxygenating the atmosphere. Nature 410, 317–319.
J. M. Schull, and M. A. Thronson, Eds.), ASP Conference Series, Vol. 148,
Summons, R. E., L. C. Jahuke, J. H. Hope, and G. A. Logan 1999. 2-Methyl-
pp. 449–455. Astron. Soc. Pac., San Francisco.
hopanoids as biomarkers for cyanobacterial oxygenic photosynthesis. Nature
McKay, C. P., and H. Hartman 1991. Hydrogen peroxide and the evolution of 400, 554–557.
oxygenic photosynthesis. Origins Life Evolut. Biosphere 21, 157–163.
Wächterhäuser, G. 1990. Evolution of the first metabolic cycles. Proc. Natl.
Mimura, M., S. Akimoto, I. Yamazaki, H. Miyashita, and S. Miyachi 1999. Acad. Sci. USA. 87, 200–204.
Fluorescence properties of chlorophyll d-containing prokaryotic alga, Acary-
ochloris minima: Studies using time-resolved fluorescence spectroscopy on Walker, J. C., P. B. Hays, and J. F. Kasting 1981. A negative feedback mechanism
intact cells. Biochim. Biophys. Acta 1412, 37–46. for the long-term stabilization of Earth’s surface temperature. J. Geophys. Res.
86, 9776–9782.
Miranova, Z. F. 1973. Albedo of Earth’s surface and clouds. In Radiation Char-
acteristics of the Atmosphere and the Earth’s Surface. (K. Ya. Kondratev, Walker, J. C. G. 1987. Was the Archaean biosphere upside down? Nature 329,
Ed.), NASA TT F-678. Asmerind Publishing Co. Ltd., New Delhi, 192–239. 710–712.
Mulkidjanian, A. Y., and W. Junge 1997. On the origin of photosynthesis as Watanabe, Y., J. E. E. Martini, and H. Ohmoto 2000. Geochemical evidence for
inferred from sequence analysis. A primordial UV-protection as common terrestrial ecosystems 2.6 billion years ago. Nature 408, 574–578.
ancestor of reaction centers and antenna proteins. Photosynthesis Res. 51, Wetherill, G. W. 1996. The formation and habitability of extra-solar planets.
27–42. Icarus 119, 219–238.
Olson, J. M., and B. K. Pierson 1986. Photosynthesis 3.5 thousand million years Whitmire, D. P., R. T. Reynolds, and J. F. Kasting 1991. Habitable zones for
ago. Photosynthesis Res. 9, 251–259. Earth-like planets around main sequence stars. In Bioastronomy: The Search
Pace, N. R. 1997. A molecular view of microbial diversity and the biosphere. for Extraterrestrial Life (J. Heidmann and M. J. Klein, Eds.), pp. 173–178.
Science 276, 734–740. Springer-Verlag, Berlin.
Paytan, A., M. Kastner, D. Campbell, and M. H. Thiemens 1998. Sulfur isotopic Williams, D. M., and J. F. Kasting 1997. Habitable planets with high obliquities.
composition of cenozoic seawater sulfate. Science 282, 1459–1462. Icarus 129, 254–267.
Raven, J. A. 1984a. Energetics and Transport in Aquatic Plants. A. R. Liss, Woolf, N., and J. R. Angel 1998. Astronomical searches for Earth-like planets
New York. and signs of life. Annu. Rev. Astron. Astrophys. 35, 507–537.
Raven, J. A. 1984b. A cost–benefit analysis of photon absorption by photosyn- Yung, Y. L., and W. B. De More 1999. Photochemistry of Planetary Atmo-
thetic unicells. New Phytol. 98, 593–625. spheres, Chap. 8. Oxford Univ. Press, New York.