Introduction

Sugarcane ratoons are derived from the residual root system of harvested sugarcane and produce a new generation of sugarcane from the emerging lateral buds on the old sugarcane roots. The rhizosphere ecosystem consists of plant-soil-microbe interactions and plays a crucial role in the feedback between plants and the environment, and it is closely associated with the growth status of crops and obstacles to ratoon development1, 2. Examining the trends, interrelationships, and interactive effects of rhizosphere physicochemical factors, biochemical functional factors, enzyme activities, microbial diversity, and endophytic bacterial diversity in different years of sugarcane ratoon growth can provide valuable insights into the mechanisms underlying sugarcane ratoon decline. However, systematic investigations in this area are currently lacking.

Crop growth and development are based on specific nutrients provided by the soil. The ecological factors of untilled rhizospheric crop ratoon soil undergo long-term evolution that leads to deficiencies in essential nutrients and accumulation of allelopathic substances released through root exudation and organic matter decomposition. Under normal ratoon cultivation practices, crop ratoons often experience reduced stress tolerance, compromised growth conditions, and decreased yield and quality3,4,5,6,7,8,9,10. However, few studies have investigated the rhizosphere soil ecological environment of sugarcane ratoons, with only a few reports focusing on variations in the physicochemical properties of the rhizosphere soil11, 12 and alterations in microbial community structure and functionality13, 14. The factors contributing to the decline of crop ratoons are diverse, and the underlying mechanisms vary. Researchers in China and abroad have primarily focused on the long-term accumulation of crop residue decomposition products, secretion of allelopathic substances, deterioration of physicochemical and biochemical factors in the rhizosphere soil, and degradation of the microbial community status in the rhizosphere soil and endophytic microorganisms8, 15,16,17. The cultivation practices of crop ratoons profoundly influence the rhizosphere soil physicochemical properties, nutrient supply, and morphological and physiological activities. In field production, sugarcane yield exhibits a decreasing trend with increasing years of ratoon cultivation, and these changes can also vary depending on the regional climate18. The ratoon nature of sugarcane is influenced by multiple factors, including genetic characteristics and soil conditions, which include various rhizosphere ecological factors that affect sugarcane growth19,20,21. Research has demonstrated that roots, as the main organs for nutrient and water absorption in crops, are closely related to factors in the soil ecological environment that influence crop growth22. Sugarcane ratoon roots derived from old ratoons gradually lose metabolic vitality in soils with accumulated allelopathic substances, high salinity, and harmful secretions, causing diseases, such as yellowing23, 24. Variations in the quantity and morphology of sugarcane ratoon roots directly affect the plant’s water and nutrient absorption capacity, tillering ability, stress resistance, and pest and disease resistance, and these changes alter the source-sink relationship of the plant and influence sugarcane yield and sucrose content, among other quality characteristics25, 26.

Li et al.22 suggested that soil physicochemical factors significantly influence sugarcane root biomass, root morphology, root respiration, water and nutrient absorption capacity, and stem and leaf growth and development. Lan27 discovered that the yellowing of sugarcane ratoon seedlings was related to excessive absorption of lead and manganese elements, an imbalance in iron and manganese, and the accumulation of phenolic acids and other allelopathic toxins, which damage the sugarcane root system and affect plant growth and development. Yang et al.14 found that compared to the soil of sugarcane that is non-susceptible to smut disease, the rhizosphere soil of susceptible plants demonstrated a decreasing trend in bacterial diversity and richness and showed drastic changes in dominant bacterial community structure at the phylum and genus levels. The plant-soil feedback loop is driven by the intricate interplay of complex factors including rhizosphere physicochemical properties, soil biochemical functions, soil biological community structures, endophytic community compositions, and their secondary metabolites28, 29.

The present study focuses on the ecological factors of the rhizosphere of different sugarcane ratoon plots (same variety) planted under the same climatic and soil conditions. By analysing the dynamic variations and correlations among rhizosphere ecological factors at different temporal scales (different ratoon ages) and their associations with the properties of sugarcane growth, we aimed to investigate the relationship between sugarcane ratoon decline and rhizosphere ecological factors. The results of this study may improve the growth and yield of sugarcane ratoons.

Results and analysis

Analysis of the ecological factors in the soil rhizosphere of sugar ratoons of different ages

Variation trends in the ecological factors of the soil rhizosphere of sugar ratoons of different ages

Significant differences (P > 0.05) were not observed in the pH, total carbon, total phosphorus, total potassium, lead, chromium, copper, and arsenic content between newly planted sugarcane and the rhizosphere soils from sugarcane of different ratoon ages, as shown in Table 1; Figs. 1, 2. Similarly, significant differences (P > 0.05) were not observed in the soil water content (SWC), pH, total carbon, total nitrogen, total phosphorus, total potassium, lead, chromium, copper, arsenic, respiration rate, acid phosphatase activity, cellulase activity, sucrase activity, and sucrose content among the rhizosphere soils from sugarcane ratoons of different ages. However, with increases in sugarcane ratoon age, a gradual increase in electrical conductivity, organic matter, total phosphorus, ammonium nitrogen/nitrate nitrogen ratio, lead, copper, and mercury levels was observed. This suggests that the rhizosphere plays a role in the accumulation of organic matter, total phosphorus, and certain heavy metal elements. Specifically, in the third-year sugarcane ratoon (TST) plot, only organic matter, the ammonium nitrogen/nitrate nitrogen ratio, and mercury were significantly higher (P < 0.05) than those in the first-year sugarcane ratoon (FST) plot. Total carbon, total potassium, available phosphorus, chromium, and nickel showed an initially increasing and then decreasing trend, with only available phosphorus and nickel showing significant differences (P < 0.05) between different ratoon ages. Furthermore, the SWC, pH, bulk density, total nitrogen, ammonium nitrogen, nitrate nitrogen, available potassium, arsenic, all biochemical processes, all enzyme activities, and all growth properties exhibited a gradually decreasing trend. Thus, these soil ecological factors were positively correlated with sugarcane growth properties. The rhizosphere soils of newly planted sugarcane showed significant differences (P < 0.05) in the SWC, bulk density, total nitrogen, ammonium nitrogen, nitrate nitrogen, available phosphorus, available potassium, ammonium nitrogen/nitrate nitrogen ratio, nickel, all biochemical processes (excluding respiration rate, ammonification, nitrification, and nitrogen fixation), all enzyme activities, and all growth properties when compared to the corresponding ecological factors of the rhizosphere soils from the different ratoon age plots. Similarly, the electrical conductivity, bulk density, organic matter, ammonium nitrogen, nitrate nitrogen, available potassium, ammonium nitrogen/nitrate nitrogen ratio, nickel, mercury, all biochemical processes, enzyme activities (excluding acid phosphatase activity, cellulase activity, and sucrase activity), and all growth properties (excluding sucrose content) of the FST exhibited significant differences (P < 0.05) compared to that of the second-year sugarcane ratoon (SST) or TST plot. These findings indicate that as the age of sugarcane ratoons increased, most of the non-microbial ecological indicators in the rhizosphere soil underwent significant deterioration (P < 0.05).

Table 1 Comparison of the physicochemical factors in the rhizosphere across sugarcane ratoon ages.
Full size table
Fig. 1
figure 1

Comparative analysis of biochemical functions and enzymatic activities in the rhizosphere across sugarcane ratoon ages.

Full size image
Fig. 2
figure 2

Comparative analysis of growth characteristics in sugarcane across ratoon ages.

Full size image

Correlation analysis between the ecological factors in the soil rhizosphere of different ratoon ages and their growth properties

The correlation analysis results shown in Fig. 3 indicate that the physicochemical factors of rhizosphere soils of different ratoon ages were significantly correlated with the soil enzyme activities. Cellulase, sucrase, and urease activities were significantly correlated with almost all physicochemical factors of the soil (except for organic matter, total potassium, and organic matter) (P < 0.05). This suggests an important relationship between physicochemical factors and soil enzyme activities. Peroxidase activity was not significantly correlated with any of the physicochemical factors (P > 0.05) and was highly negatively correlated with organic matter content. The total carbon, organic matter, and total phosphorus primarily exhibited a negative correlation with soil enzyme activities. Total potassium exhibited a low correlation (approximately 0) with soil enzyme activities, whereas the other physicochemical factors were positively correlated. In terms of soil biochemical functions, respiration, ammonification, nitrification, and nitrogen fixation were significantly positively correlated with catalase, cellulase, sucrase, urease, and hydrolase activities (P < 0.05). Soil acid phosphatase activity was significantly positively correlated only with the nitrification process among soil biochemical functions (P < 0.05).

According to Fig. 4, which presents line graphs representing the different enzyme activities in the soil (normalised using the Log10 transformation for uniformity), the peroxidase activity and catalase activity in the rhizosphere of sugarcane demonstrated a relatively gradual decline, whereas the hydrolase activity exhibited the largest gradient of change and the most noticeable decrease. This indicates that variations in ratoon age had the greatest effect on hydrolase activity. All eight soil enzymes showed a decreasing trend as the ratoon age increased, which was consistent with the variations in two key sugarcane yield indicators.

Fig. 3
figure 3

Spearman correlation heatmap of the relationship between soil physicochemical factors and enzyme activities in sugarcane rhizosphere across different ratoon ages. *, **, and *** represent significant Spearman correlations at the 0.05, 0.01, and 0.001 levels, respectively. Blank spaces without * indicate non-significant correlations (P > 0.05). Different colour variations represent the Spearman correlation coefficient (R) between variables, with orange, black, and white representing positive, negative, and no correlation, respectively. The darker the colour, the stronger the correlation.

Full size image
Fig. 4
figure 4

Trends in the variations of enzyme activities in the sugarcane rhizosphere and theoretical yield across different ratoon ages. FST, SST, and TST correspond to rhizosphere soils from the first-year, second-year, and third-year sugarcane ratoons, respectively. Histograms depict the theoretical yield indicators of sugarcane for different ratoon ages: theoretical sugarcane yield (A) and theoretical sugar yield (B). * Significant difference between groups (P < 0.05).

Full size image

Analysis of rhizosphere microbial ecological factors in sugarcane ratoons of different ages

Microbial diversity in the rhizosphere of sugarcane ratoons of different ages

Figures 5, 6 show that the alpha diversity significantly differed between bacteria and fungi in both the rhizosphere soil and endophytic communities of sugarcane roots. In the rhizosphere soil, the alpha diversity of bacteria showed a similar trend, with the Shannon and Chao1 indices indicating an initial increase followed by a decrease with increasing ratoon age. However, the alpha diversity of fungi remained relatively stable and showed minimal variations. Among the different ratoon ages, the highest alpha diversity of bacteria was observed in SST, followed by that in TST, whereas the lowest diversity was observed in FST. In terms of the alpha diversity of endophytic microbial communities in the sugarcane roots, significant differences were observed between the Shannon and Chao1 results, which was possibly due to algorithmic variations. The Chao1 index estimates the community richness by considering the number of undetected species30, whereas the Shannon index considers both the richness and evenness of the community31. The trends of the Shannon and Chao1 indices were similar for the microbial community but showed greater differences in the endophytic microbial community, indicating higher heterogeneity and lower evenness within the groups. The analysis of alpha diversity revealed inconsistent trends between the rhizosphere soil and endophytic communities of sugarcane roots, particularly in the variability of alpha diversity.

Fig. 5
figure 5

Alpha diversity analysis of bacteria and fungi in the rhizosphere soil. (A) and (B) Shannon Chao1 alpha diversity indices of bacteria and fungi in the rhizosphere of sugarcane from different ratoon ages, respectively.

Full size image
Fig. 6
figure 6

Alpha diversity analysis of endophytic bacteria and fungi in the root system. (A) and (B) Shannon Chao1 alpha diversity indices of endophytic bacteria and fungi in the rhizosphere of sugarcane from different ratoon ages, respectively.

Full size image

NMDS analysis of microbial community in rhizosphere of sugarcane ratoons in different years

Figure 7 shows that the microbial communities in the rhizosphere soil exhibited significant differences between groups (A, C), indicating relatively large inter-group distances. However, the within-group replicates showed good homogeneity. This overall pattern was also observed in the microbial community analysis (E). Notably, the fungal community in the rhizosphere soil of the FST displayed relatively high heterogeneity (C). The significance tests for microbial community structure in Table 2 further support these findings. The Bray-Curtis distance of microbial communities showed highly significant results (P < 0.001) for all three tested methods: permutational multivariate analysis of variance (ADONIS), analysis of similarity (ANOSIM), and multi-response permutation procedure (MRPP). These results suggest substantial differences in rhizosphere soil microbial communities among the different years, particularly for bacterial communities (A).

The endophytic bacterial communities (B) within the root system displayed higher similarity in another ecological niche, i.e., the root endophyte, across different years compared to the rhizosphere bacterial communities (A). The inter-group beta diversity of endophytic bacteria was lower, indicating greater homogeneity within groups. In contrast, the endophytic fungal communities (D) exhibited no significant variation in beta diversity with respect to sugarcane ratoon root age. These fungal communities showed higher heterogeneity and lower evenness within group replicates. In fact, the Bray-Curtis distances between the five replicates were greater than the distances between different years. The overall microbial community analysis across different ecological niches (E) also demonstrated similar patterns, with notable differences in beta diversity between the two niches. The beta diversity of rhizosphere soil microbial communities in sugarcane ratoons exhibited relatively consistent patterns across different years (blue confidence intervals in E). Conversely, the beta diversity of endophytic bacterial communities in sugarcane ratoon root systems exhibited higher complexity and heterogeneity among groups. These findings align with the results in Fig. 6 showing high alpha diversity heterogeneity among the endophytic microbial communities. The significance tests for microbial community structure (Table 2) further support these conclusions because significant differences (P > 0.05) were not observed in the Bray-Curtis distances of endophytic bacterial communities for all three tested methods: ADONIS, ANOSIM, and MRPP. This suggests relatively small variations in endophytic bacterial communities among different years, especially for fungal communities (D).

Fig. 7
figure 7

NMDS analysis of microbial communities based on the Bray-Curtis distance for different ecological niches. (A) Non-metric multidimensional scaling (NMDS) analysis of bacterial communities in the rhizosphere soil of sugarcane ratoons across different years and ecological niches. (B) NMDS analysis of bacterial communities inhabiting the rhizosphere of sugarcane ratoons within different ecological niches over various years/ages. (C) NMDS analysis of fungal communities in the rhizosphere soil of sugarcane ratoons across different years/ages and ecological niches. (D) NMDS analysis of fungal communities residing within the rhizosphere of sugarcane ratoons across various years/ages and ecological niches. (E) NMDS analysis of microbial communities (including both fungi and bacteria) in the rhizosphere soil and sugarcane ratoon roots according to different ecological niches over different years/ages.

Full size image
Table 2 Significance tests of microbial community structures in rhizosphere soil and endophytic root.
Full size table

Modular analysis of microbial ecological networks in the rhizosphere of sugarcane ratoons in different years

According to Fig. 8; Table 3, the cross-domain ecological network of microbial communities in the rhizosphere soil of sugarcane ratoons contained 1013 nodes, 68,017 edges, and 5 module clusters, with a modularity level of 0.291. The cross-domain ecological network of endophytic fungal communities in sugarcane ratoon root systems included 257 nodes, 762 edges, and 7 module clusters, with a modularity level of 0.385. The number of nodes and edges in the rhizosphere soil microbial network of sugarcane far exceeded that of the endophytic fungal network in the root system. However, in terms of module quantity and modularity level, the cross-domain ecological network of endophytic fungal communities in the root system was more complex, with a richer functional community and a more stable microbial community structure. This may be one of the reasons for the significant differences in alpha and beta diversity between the endophytic microbial communities and the rhizosphere soil microbial community of sugarcane.

Table 4 presents the major modules and their respective proportions in the cross-domain ecological networks at the phylum level for rhizosphere soil and endophytic fungal communities associated with sugarcane root systems. In the rhizosphere soil microbial network, the three modules (module 1, module 2, and module 3) collectively accounted for over 98% of the overall microbial community. Similarly, in the endophytic fungal network, the three modules (module 1, module 3, and module 6) accounted for over 65% of the entire network and represented the major modules in the ecological network. The stacked bar charts at the phylum level for rhizosphere soil microbes (top 7) showed that the fungal phyla Ascomycota and Basidiomycota were relatively abundant in module 1, module 2, and module 3. Particularly, Ascomycota and Basidiomycota were dominant in module 2 and module 3, making them the primary organisms in these modules. Additionally, the composition of the top seven species was similar across all three modules. In the endophytic fungal network, the stacked bar chart at the phylum level for the top seven species revealed differences between module 3 and modules 1 and 6. module 3 was mainly composed of Ascomycota, Cyanobacteria, Proteobacteria, and Actinobacteriota, which accounted for over 98% of the relative abundance, whereas other species had lower proportions in module 3. In contrast, modules 1 and 6 exhibited similar structural compositions. Ascomycota is the only fungal phylum among the top seven phyla that showed a higher relative abundance in module 3 and module 6. Thus, this phylum may play a significant role in influencing sugarcane yield.

Fig. 8
figure 8

SparCC cross-domain ecological network and composition of major module species. (A) Modular analysis of the cross-domain ecological network 362 for microbial communities in the rhizospheric soil of sugarcane ratoons. (B) Taxonomic composition of module 1, module 2, and module 3, which are the main modules in the cross-domain ecological network of rhizosphere soil microbial communities. (C) Modular analysis of the cross-domain ecological network for endophytic microbial communities inhabiting the root systems of ratoon sugarcane. (D) Taxonomic composition at the phylum level for module 1, module 3, and module 6, which are the main modules in the cross-domain ecological network of endophytic fungal communities in sugarcane ratoon root systems.

Full size image
Table 3 Overview of the structures of the two ecological networks.
Full size table
Table 4 Major modules and their relative abundances in the cross-domain ecological networks of rhizosphere soil and endophytic fungal communities at the phylum level in sugarcane root systems.
Full size table

Correlation analysis between the microbial community of the sugarcane ratoon rhizosphere and soil ecological factors

In Fig. 9A, only module 3 in the rhizosphere soil microbial network was significantly positively correlated with soil catalase activity (P < 0.05). The P-values of the Mantel test for the other two network modules and soil enzyme activities were not significant (P > 0.05). Figure 9B shows that the p-values of the Mantel test between the three modules (module 1, module 3, and module 6) in the endophytic fungal network and soil enzyme activities are not significant (P > 0.05).

Bioenv analysis was conducted to assess the correlation between non-microbial soil ecological factors and the bacterial and fungal communities. The factors with high correlations were selected, and the final set of factors included in Fig. 10 was determined based on their variance inflation factor (VIF). Figure 10 shows that the rhizosphere soil bacterial community was influenced by different ecological factors in different years. Figure 10A clearly shows that total phosphorus was the main driving factor for the bacterial community in the FST plot while total carbon, total potassium, and available phosphorus were the main driving factors for the bacterial community in the SST. Nitrate nitrogen and nitrification were the main driving factors for the bacterial community in the TST plot, with nitrate nitrogen having a stronger effect compared to the other ecological factors. Figure 10B revealed that available phosphorus was the primary driving factor for the fungal community in the SST plot while ammonium nitrogen, nitrification, and nitrate nitrogen were the primary driving factors for the fungal community in the TST plot. Available phosphorus and nitrate nitrogen were high-intensity factors, and the fungal community in the FST plot was less influenced by soil ecological factors. Both bacterial and fungal communities were influenced by nitrate nitrogen, nitrification, and available phosphorus. For both the bacterial and fungal communities, nitrate nitrogen and nitrification were the primary drivers in the TST plot.

Fig. 9
figure 9

Composition of major module microbial species and soil enzyme activity in two ecological niche networks as evaluated by Mantel tests. (A) Mantel test between the taxonomic composition of major modules in the rhizosphere soil ecological network and soil enzyme activities. (B) Mantel test between the taxonomic composition of major modules in the endophytic fungal ecological network and soil enzyme activities. The depth of the red colour represents the Pearson correlation coefficient between soil enzyme activities. The lines indicate the correlation coefficient (R) of the Mantel test between the taxonomic composition of modules and soil enzyme activities, and the thickness of the lines represents the strength of the correlation. Different colours of the lines represent different P-values of the Mantel test. The significance of the correlation between soil enzyme activities is indicated as follows: * P < 0.05, ** P < 0.01, and *** P < 0.001.

Full size image
Fig. 10
figure 10

Redundancy analysis (RDA) of the rhizospheric soil microbial community in relation to soil ecological factors. (A) Results of the redundancy analysis (RDA) of the rhizosphere soil bacterial community in relation to soil ecological factors. (B) Results of the RDA of the rhizosphere soil fungal community.

Full size image

Correlation analysis between microbial community and sugarcane yield

The relationship between bacterial and fungal alpha diversity (Shannon index) in the rhizospheric soil and endophytic root-associated microbial communities and the theoretical sugarcane and sugar yield were examined using the generalised additive model (GAM). According to Table 5, the alpha diversity of bacteria and fungi in the rhizosphere ecological niche of sugarcane was significantly positively correlated with theoretical sugarcane yield (R > 0.00, P < 0.05) and positively correlated with the theoretical sugar yield, although the correlation was not statistically significant (P > 0.05). According to Table 5, the alpha diversity of bacteria in the ecological niche of sugarcane root endophytic fungi was significantly positively correlated with theoretical sugar yield (P < 0.05) and positively correlated with the theoretical sugar yield, although the correlation was not statistically significant (P > 0.05). The trends in bacterial alpha diversity in these two distinct ecological niches consistently influenced the sugarcane yield indicators. The fungal alpha diversity in the rhizospheric soil ecological niche was significantly positively correlated with the theoretical sugarcane yield (P < 0.05) and positively correlated with the theoretical sugar yield (P > 0.05). However, the fungal alpha diversity in the endophytic root-associated microbial niche exhibited a negative correlation with both sugarcane yield indicators (R < 0.00, P > 0.05).

Table 5 GAM analysis of the microbial community and sugarcane yield.
Full size table

Discussion and conclusion

The present study suggests that almost all non-microbial ecological factor indicators in the newly planted sugarcane rhizospheric soil exhibited significant differences (P < 0.05) compared to the corresponding ecological factors in the rhizospheric soils from the FST, SST, and TST sugarcane ratoons. In particular, most non-microbial ecological factor indicators in FST showed significant differences (P < 0.05) when compared to the corresponding factors in TST. The significant increase (P < 0.05) in organic matter could be related to the accumulation of organic compounds in the rhizospheric soil due to the decomposition of old sugarcane roots, including residues of external fauna and flora. Urease activity promotes the production of urea, ammonia, CO2, and water, thereby enhancing soil nitrogen cycling and influencing soil biochemical processes such as ammonification and nitrification32,33. The increase in total phosphorus in the soil may follow a similar pattern as organic matter accumulation because external contributions from the surrounding environment allowed for its accumulation in the soil. Variations in available phosphorus could be associated with a decrease in acid phosphatase activity. Acid phosphatase hydrolyses glycerophosphates and sugar phosphates under acidic conditions, thereby generating orthophosphates and increasing phosphorus availability in the soil. Soil functions such as respiration, ammonification, nitrification, and nitrogen fixation directly reflect soil characteristics8. Soil respiration encompasses root respiration, microbial respiration, and soil organism respiration, and it represents the chemical oxidation of carbonaceous materials and reflects the carbon cycling efficiency of soil. Carbon is an essential element for plant growth. Ammonification, nitrification, and nitrogen fixation are indicators of soil nitrogen cycling34,35. Main conclusion 1: With increasing ratoon age, a small proportion of non-microbial ecological factor indicators exhibited a gradually increasing trend. This was potentially due to an enrichment effect from consecutive ratoon cropping. However, most indicators showed a significant decline (P < 0.05). Notably, reductions in soil enzymatic activity and biochemical function may have reduced the availability of soil nutrients, such as N, P, and K, consequently impacting sugarcane yield. These findings indicate that the significant degradation (P < 0.05) of most non-microbial ecological factors in the rhizospheric soil with increasing ratoon age is a potentially crucial contributor to ratoon decline.

The present study suggests that only the activities of peroxidase and polyphenol oxidase in the soil were not significantly correlated with the physicochemical factors. Both enzymes exhibited a negative correlation with organic matter content, and the correlation coefficients were relatively high. Peroxidase in the soil belongs to the oxidoreductase class of enzymes, which can enzymatically hydrolyse peroxides, thereby reducing the harmful accumulation of peroxides on crop roots and promoting crop growth36. Polyphenol oxidase in the soil participates in the cycling of aromatic compounds and other phenolic substances37. The main yield indicators of sugarcane showed similar trends of soil enzyme activity across different ratoon ages. The sources of various enzymes in soil are primarily soil microbial activity and root exudates, and their activity levels are directly related to soil microbial and physicochemical properties, thereby significantly influencing plant growth38,39. A strong correlation was observed between soil physicochemical properties and soil enzyme activity. Soil enzymes participate in all biochemical processes in the soil and play a crucial role in soil and vegetation ecosystems. They are also closely related to the environment and can be influenced by physical, chemical, and biological factors in the environment. The activity of soil enzymes can be used to assess soil fertility32,40. Main conclusion 2: Studies have indicated that oxidases in the soil are important redox enzyme systems involved in the synthesis of humus and alleviation of oxidative damage. Soil enzymes can decompose substances such as straw in the soil41, reflecting the level of organic matter accumulation in the soil42. Peroxidase promotes the decomposition of organic substances in the soil43, possibly by facilitating the oxidation of organic matter (such as phenols and amines) into quinones. Organic matter is degraded under the influence of peroxidase. Consistent with the findings of the present study, a negative correlation was observed between peroxidase and polyphenol oxidase activities and organic matter content.

The SparCC cross-domain ecological network and composition and proportion of species in the main modules revealed similarities in the top seven species across the three modules. The relative abundance of Acidobacteria showed a gradually decreasing trend with age, whereas the relative abundance of Basidiomycota exhibited an increasing trend with age. Acidobacteria showed the highest relative abundance in the FST plot, and the value decreased in the following two years. The relative abundance of Ascomycota reached its maximum in the second-year sugarcane ratoon roots (GSSTs). The remaining top seven phyla demonstrated relatively small variations. Saprophytic fungi contribute significantly to soil enzyme activity, and their alpha diversity exhibited minor variations across different years. However, both enzyme activity and yield exhibited a decreasing trend over time, and the succession of fungi, specifically Ascomycota and Basidiomycota, may have been a major contributing factor44, 45. The stacked bar chart of the top seven species at the phylum level in the sugarcane rhizosphere reveals that Ascomycota had a relatively high abundance in modules 3 and 6. It is the only fungus among the top seven species that exhibited a higher relative abundance in the GSST and third-year sugarcane ratoon roots (GTST), which may have contributed to the decline in sugarcane yield indicators. Main conclusion 3: Network modular analysis revealed the taxonomic compositions of the primary modules in rhizosphere soil and endophytic root microbial communities. Mantel tests against soil enzymatic activities indicated that module 3 of the rhizosphere soil microbial ecological network exhibited a significant positive correlation (P < 0.05) with certain soil enzyme activities, potentially driven by the phylum Ascomycota.

The composition of major species in the rhizosphere soil microbiota and endophytic fungal community and the activity of soil enzymes were investigated using Mantel tests. The results revealed a significant positive correlation (P < 0.05) between module 3 and soil catalase activity, which may be attributed to the notable differences in the relative abundance of Basidiomycota compared to that in the other modules. However, the Mantel test results for the other two network modules and soil enzyme activity did not identify significant correlations. Similarly, Mantel tests examining the microbial species composition within the endophytic fungal community network modules (modules 1, 3, and 6) and soil enzyme activity did not identify significant correlations (P > 0.05). Hence, the present study suggests that the endophytic fungal community may not exert a direct impact on soil enzyme activity. The RDA of the rhizosphere soil microbiota in relation to soil ecological factors revealed that nitrate nitrogen and nitrification played a crucial role in shaping the microbial community in TST. In contrast, other ecological factors drove the microbial community in the sugarcane rhizosphere during different years. The declining trend of soil enzyme activity and sugarcane yield with age is depicted in Figs. 1, 2, 3, and the trends observed in the physicochemical factors are presented in Table 1. The findings indicated a weakened nitration process, as evidenced by decreased levels of nitrate nitrogen and ammonium nitrogen in TST. Main conclusion 4: The fungal community within the sugarcane root system may be associated with sugarcane diseases, which could potentially cause a decline in sugarcane yield. Moreover, sugarcane diseases induced by the endophytic fungal community could lead to decreased crop productivity. In contrast, the fungal community in the rhizosphere soil, which does not directly interact with the sugarcane roots, may affect sugarcane yield by influencing soil enzyme activity. For instance, saprophytic fungi are major contributors to enzyme activity and soil functionality46,47. The alpha diversity analysis did not identify significant differences in the alpha diversity of the fungal community among different years in the sugarcane rhizosphere soil (as illustrated in Fig. 6). However, soil enzyme activity exhibited a decreasing trend with increasing age. This suggests that despite the lack of significant variations in fungal alpha diversity between different years, the species composition and structure of the fungal community may have undergone alterations, indicating a possible succession of dominant fungal species.

RDA analysis revealed that various ecological factors exerted differential influences on the rhizosphere microbial communities. Nitrogen was identified as the primary driver shaping the microbial composition in the third ratoon year. In contrast, distinct ecological factors drove the microbiome composition during other ratoon years. This could be attributed to a potential decline in soil nitrogen cycling functions during the third ratoon year, which altered the microbial community structure. Enzyme activity in the rhizosphere soil of sugarcane ratoons was significantly correlated (P < 0.05) with the physicochemical factors of the soil. The decrease in soil enzyme activity and biochemical functions may reduce the effective supply of elements, such as N, P, and K, which ultimately affects sugarcane yield. The structure of the microbial community in the rhizosphere soil varied significantly (P < 0.05) among the different age plots. The microbial community in the rhizosphere soil indirectly influenced sugarcane yield by affecting soil enzyme activity, with fungi having a particularly significant impact (saprophytic fungi being the primary contributors to soil enzyme activity). Main conclusion 5: The presence of endophytic bacteria and soil bacteria was positively correlated with sugarcane yield (P < 0.05), which was likely due to their interaction with non-microbial ecological factors in the soil and potential role in promoting the cycling of elements, such as N, P, and K. This synergistically shaped the rhizosphere ecological environment and influenced sugarcane growth through interactions with the root system. The alpha diversity of endophytic fungi was negatively correlated with sugarcane yield (P > 0.05). Notably, Ascomycota species, which had a relatively high abundance in the sugarcane ratoon root system and were among the top seven species, may be a causal factor underlying sugarcane diseases. Several diseases affecting sugarcane, such as pineapple disease and red rot disease, are fungal diseases. Therefore, endophytic fungi, especially Ascomycota species, could be the primary factors contributing to sugarcane diseases48,49.

Materials and methods

Sampling area overview

The sampling area is located in the National Modern Agricultural Industrial Park in Laibin City (N 23°48′, E109°12′), Guangxi Province, which is situated in central Guangxi downstream of the Hongshui River. The research site is located in the transitional zone from the south subtropical to the central subtropical climate. The average annual temperature is approximately 21 ℃, and the area has relatively dry conditions in winter and spring and experiences abundant rainfall during summer and autumn. The average annual precipitation is approximately 1300 mm, with an average of 1500 h of sunshine per year50. The research site covers an area of 10 hectares and is characterised by relatively flat terrain. The experiment was conducted at a site characterized by sandy soil51. A four-year crop rotation scheme was employed, and it consisted of one year of newly planted sugarcane followed by three consecutive ratoon years (2018–2021). Within this fixed location, 5 plots measuring 10 m x 10 m were established, resulting in a total of 20 sample plots. Investigations and sampling were conducted during the sugarcane harvest period (mid-November annually). This provided samples and data on the rhizosphere soil of the FST, SST, TST, first-year sugarcane ratoon roots (GFST), GSST, and GTST plots as well as indicators of sugarcane growth. The fertilization regime52 involved two applications: a base fertilizer of compound fertilizer (N: P:K = 15:15:15; total nutrients ≥ 45%) at 375 kg/hm2 and urea at 225 kg/hm2, followed by a mid-tillage application (early July annually) of compound fertilizer (375 kg/hm2), urea (525 kg/hm2), and potassium chloride (225 kg/hm2).

Sample collection

Soil sample collection: Soil samples were collected within the plots using the “S-shaped” five-point composite sampling method. After removing surface weeds and scraping off the topsoil, the rhizosphere soil from a depth of 10–20 cm below the ground was collected. After the sugarcane roots were excavated, the roots were gently shaken to remove large clumps of soil, loose soil, and debris. A sterile brush was then used to collect residual soil from the root surface. Soil samples from multiple points within the same plot were combined in equal amounts, mixed thoroughly, and placed in sterile bags. These bags were then transported to the laboratory on dry ice or ice packs for further processing. A portion of the samples was stored in a refrigerator at 4 ℃ for approximately one week to conduct tests on soil physicochemical properties, heavy metal content, biochemical activity, and soil enzyme activity. Another portion was rapidly frozen with liquid nitrogen and stored in a −80 ℃ freezer to test the microbial diversity (16 S/ITS) of the rhizosphere soil53.

Sugarcane root sample collection: The sugarcane root samples were collected within the plots using the “S-shaped” five-point composite sampling method. We randomly selected 3–5 healthy root tissues of varying lengths (approximately the length of fingers) as one sample and ensured that the soil adhering to the root surface was removed. The collected samples were then placed in sterile bags and transported to the laboratory on dry ice or ice packs. The root samples were first rinsed with sterile water for 30 s, followed by immersion in 70% ethanol for 2 min. Finally, the roots were washed three times with sterile water to ensure the surface of the roots was sterilised54.

Analytical parameters and methods

Determination of physicochemical properties of the sugarcane rhizosphere soil: The collected soil samples were air-dried and then ground to pass through a 1 mm sieve. The samples were mixed thoroughly and kept for further analysis. The methods described by Beijing Forestry University55 and the National Agro-Tech Extension and Service Center56 were followed to measure the physicochemical properties of the sugarcane rhizosphere soil. SWC was determined using the drying method, and pH was measured using a pH meter. Electrical conductivity was measured using a conductivity meter, and the bulk density was determined using the ring knife method. Total carbon (TC) was measured using the ethanol-ethanolamine absorption titration method, and organic matter (OM) was determined using the potassium dichromate digestion method with a heating furnace. Total nitrogen (TN) was measured using the Kjeldahl method, and total phosphorus (TP) was determined using the sodium hydroxide fusion-molybdenum antimony anti-spectrophotometric method. Total potassium (TK) was measured using the sodium hydroxide fusion-flame photometry method, and ammonium nitrogen (AN) was determined using the potassium chloride extraction method with indophenol blue colourimetry. Nitrate nitrogen (NN) was measured using the phenol disulfonic acid colorimetric method, and available phosphorus was determined using the NH4F-HCl extraction method. Available potassium was measured using the neutral ammonium acetate extraction method.

Determination of heavy metals in the sugarcane rhizosphere soil: Lead (Pb) was analysed using graphite furnace atomic absorption spectrophotometry according to the “Soil Quality-Determination of Lead, Cadmium” (GB/T 17141 − 1997). Chromium (Cr VI) was analysed according to the “Determination of Hexavalent Chromium in Solid Waste” (HJ 687–2014) using alkaline digestion/flame atomic absorption spectrophotometry. Nickel (Ni) and copper (Cu) were analysed using flame atomic absorption spectrophotometry according to the “Determination of Copper, Zinc, Lead, Nickel, and Chromium in Soil and Sediment” (HJ 491–2019). Arsenic (As) was analysed using atomic fluorescence spectrometry according to “Determination of Total Arsenic in Soil,” which is Part 2 of the “Determination of Total Mercury, Total Arsenic, and Total Lead in Soil” (GB/T 22105.2–2008). Mercury (Hg) was analysed using atomic fluorescence spectrometry according to “Determination of Total Mercury in Soil,” which is Part 1 of the “Determination of Total Mercury, Total Arsenic, and Total Lead in Soil” (GB/T 22105.1–2008).

Determination of biochemical functional activities in the sugarcane rhizosphere soil: Ammonification, nitrification, and nitrogen fixation activities were measured following the methods described by Lin et al.57 and Xu et al.58. Soil respiration was determined using a soil carbon flux measurement system to assess the respiratory activity.

Determination of enzyme activities in the sugarcane rhizosphere soil: The activity of polyphenol oxidase (PPO) was measured using the pyrogallol colourimetric method and an enzyme assay kit (Shanghai Youxuan Biotechnology Co., Ltd, China), where the production of 1 mg of pyrogallol per gram of sample per day was defined as one unit (U) of enzyme activity. The activity of catalase (CAT) was determined using the micro quartz colourimetric method and an enzyme assay kit (Shanghai Youxuan Biotechnology Co., Ltd), where the degradation of 1 µmol of H2O2 per gram of sample per day was defined as one unit (U) of enzyme activity. The activity of peroxidase (POD) was measured using the micro glass colourimetric method (purpurogallin) and an enzyme assay kit (Shanghai Youxuan Biotechnology Co., Ltd), where the production of 1 mg of purpurogallin per gram of sample per day was defined as one unit (U) of enzyme activity. The activity of acid phosphatase (ACP) was determined using the disodium phenyl phosphate method and an enzyme assay kit (Shanghai Youxuan Biotechnology Co., Ltd), where the production of 1 nmol of phenol per gram of sample per day was defined as one unit (U) of enzyme activity. The activity of cellulase (CL) was measured using the 3-dinitrosalicylic acid colourimetric method and an enzyme assay kit (Shanghai Youxuan Biotechnology Co., Ltd), where the production of 1 mg of glucose per gram of sample per day was defined as one unit (U) of enzyme activity. The activity of sucrase (SC) was determined using the 3,5-dinitrosalicylic acid colourimetric method and an enzyme assay kit (Shanghai Youxuan Biotechnology Co., Ltd), where the production of 1 mg of reducing sugar per gram of sample per day was defined as one unit (U) of enzyme activity. The activity of urease (UE) was measured using the indophenol blue colourimetric method and an enzyme assay kit (Shanghai Youxuan Biotechnology Co., Ltd), where the production of 1 µg of NH3-N per gram of sample per day was defined as one unit (U) of enzyme activity. The activity of fluorescein diacetate hydrolase (FDA) was determined using the micro quartz colourimetric dish method and an enzyme assay kit (Shanghai Youxuan Biotechnology Co., Ltd), where the production of 1 µmol of fluorescein per gram of sample per day was defined as one unit (U) of enzyme activity.

Determination of agronomic traits and yield quality of sugarcane: Data on the agronomic traits and yield quality of sugarcane were collected in the designated sampling area. During the harvest period, the data collection method involved randomly selecting 10 sugarcane plants from each replicate plot for sucrose content analysis, determined using the double polarization method59. Additionally, for each replicate plot, three rows of sugarcane were randomly selected, with each row measuring 3 m in length, to estimate the field yield of sugarcane. The collected data included the number of green leaves, plant height (measured 30 cm below the thickest part of the sugarcane stalk), stem thickness, and effective stem count. The average values of the collected data within each group were used for further analysis. The theoretical sugarcane yield (t/hm2) and theoretical sugar yield (t/hm2) were calculated as follows: single stalk weight (kg) = (stem thickness (cm))2 × plant height (cm) × 0.7854 ÷ 1000; theoretical sugarcane yield (t/hm2) = single stalk weight (kg) × effective stem count (stems/hm2) ÷ 1000; and theoretical sugar yield (t/hm2) = sucrose content (%) × theoretical sugarcane yield (t/hm2).

Microbial diversity sequencing

The extracted DNA from the root-zone soil and root samples was stored at −80 ℃ and then subjected to DNA extraction, PCR amplification, and Illumina MiSeq sequencing. DNA extraction and PCR amplification were performed according to the instructions provided with the E.Z.N.A.® soil DNA kit (Omega Bio-tek, Norcross, GA, USA). The quality of DNA extraction was verified using 1% agarose gel electrophoresis, and the DNA concentration and purity were measured using a NanoDrop 2000 spectrophotometer. For the 16 S rRNA gene region, the 338 F (5’-ACTCCTACGGGAGGCAGCAG-3’) and 806R (5’-GGACTACHVGGGTWTCTAAT-3’) primer pairs were used for PCR amplification. For the ITS1-ITS2 region, the ITS1F (CTTGGTCATTTAGAGGAAGTAA) and ITS2R (GCTGCGTTCTTCATCGATGC) primer pairs were used for PCR amplification. The PCR products from the same sample were pooled and recovered using a 2% agarose gel, purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA), and checked for quality using 2% agarose gel electrophoresis. The recovered products were quantified using a Quantus™ Fluorometer (Promega, USA). Library preparation was performed using the NEXTFLEX® Rapid DNA-Seq Kit, and sequencing was carried out using the Illumina MiSeq PE300 platform. The sequencing work was conducted by Shanghai Majorbio Bio-pharm Technology Co., Ltd.

Data processing methods

Data statistics were recorded using WPS Office 2016 XLSX spreadsheets. Minitab 16 was used to perform the ANOVA (analysis of variance) (including standard deviation and p-value analysis for significance of differences). Graphs were generated using Origin 2018 software. Analysis of variance (ANOVA) was carried out using SPSS 22.0 statistical software. Experimental data are presented as the “mean ± standard deviation.” Microbial communities were normalised based on the R package “Phyloseq.” Alpha and beta diversity analyses of microbial communities, correlation analyses between environmental factors and microbial communities, and Spearman correlation analyses were performed using the R language (version 4.2.1) and “vegan” package. Microbial community composition analyses were conducted using the R package “iCAMP.” The significance of community composition disparities was analysed using the R package “vegan.” The GAM analysis was conducted using the R package “mgcv.” Species composition stacked plots were generated using the Python library “Pandas” for “Groupby” analysis, and GraphPad Prism 8 was used for plotting and significance calculations (based on Student’s t-test and ANOVA). Network construction was performed using the SparCC procedure in the iNAP platform, with the majority set to 10 out of the 15 sample points.