Chenglin Li
Ziqun Zhou1
Jieshi Tang- 1Environmental Health Effects and Risk Assessment Key Laboratory of Luzhou, School of Public Health, Southwest Medical University, Luzhou, China
- 2State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing, China
- 3College of Life Sciences, Shihezi University, Shihezi, China
- 4Key Laboratory of Arable Land Quality Monitoring and Evaluation, Ministry of Agriculture and Rural Affairs, Yangzhou University, Yangzhou, China
Land use conversion from flooded paddy fields to upland vegetable systems is becoming increasingly widespread, yet its ecological consequences for soil N2O emissions remain poorly understood. Here, we integrated the potential denitrification-derived N2O flux measurements, microbial community profiling, and network analyses to elucidate how paddy-to-vegetable land conversion reshapes soil microbial interactions and regulates N2O emission dynamics in the Yangtze River Delta region of China. Results showed that N2O emissions increased significantly following the conversion, with fluxes reaching approximately 0.43 and 0.0083 nmol N g−1 h−1 in soils under vegetable cultivation for 4 and 7 years, respectively. In contrast to the trend in N2O emissions, bacterial diversity decreased significantly following the conversion, whereas fungal diversity showed no significant change. Co-occurrence network analysis demonstrated a divergent response of bacterial and fungal communities to land use conversion. In vegetable soils, bacterial networks exhibited enhanced connectivity, with average degrees 1.23 and 1.17 times higher than those in paddy soils after 4 and 7 years of conversion, respectively. Conversely, fungal networks showed markedly reduced connectivity, with average degrees declining by 54.67 and 36.70%, respectively. The number of edges, positive connection edges, negative connection edges, the number of vertices, and average degree in the bacterial network were all significantly positively correlated with N2O emission rates, whereas fungal network connectivity showed opposite trends. Random forest modeling further identified bacterial network features were the most influential determinant of N2O emissions, outperforming traditional soil environmental variables. Altogether, our findings demonstrate that paddy-to-vegetable land conversion alters the architecture, stability, and modularity of soil microbial networks, which may play a pivotal role in enhanced N2O emissions. This study emphasizes the necessity of considering microbial network dynamics in greenhouse gas mitigation strategies.
1 Introduction
Nitrous oxide (N2O) is a powerful greenhouse gas and a significant ozone-depleting compound, exhibiting a global warming potential roughly 273 times greater than that of carbon dioxide (CO2) over a century (Li T. T. et al., 2025). Terrestrial soils, particularly agricultural soils, represent the primary source of N2O emissions to the atmosphere (Zhu G. B. et al., 2025), with its production primarily driven by microbial-mediated nitrogen transformations (Xiong et al., 2025). Land use changes, whether involving the conversion of natural vegetation to cropland or the transition between different agricultural systems, modify various properties of soils (Babur et al., 2025), which in turn influence nitrogen cycling and N2O emissions (Chen et al., 2025). A meta-analysis spanning tropical and subtropical regions revealed that the shift to fertilized agriculture led to the most significant changes in N₂O flux (Van Lent et al., 2015). Additionally, modeling studies indicate that with increasing nitrogen fertilizer use and land use changes, agricultural lands are increasingly contributing to global soil N2O emissions (Ma J. Y. et al., 2025). Given the escalating global shifts in land use practices (Wang et al., 2025), understanding the mechanisms underlying soil N2O emissions in response to land use changes is crucial for effective agricultural ecosystem management and climate change mitigation.
In many regions of Asia and elsewhere, paddy fields (flooded rice systems) have been increasingly converted to upland vegetable cultivation as a result of economic pressures, labor changes, water-management issues and land use planning (Wei et al., 2022). This transformation shifts the soil environment from a permanently or seasonally flooded anaerobic state to a more aerobic and intensively managed upland farming system. When land use practices are replaced by intensive cultivation and management, nitrogen mineralization is often accelerated, increasing the availability of inorganic nitrogen (Zhu K. et al., 2025). These changes typically enhance microbial substrate availability, thereby increasing the potential for N2O emissions (Li C. L. et al., 2022). Furthermore, shifts in hydrological conditions, redox status, root zone processes, fertilization practices, and tillage intensity contribute to the development of distinct soil microbial communities (Li W. H. et al., 2025), which are expected to have profound impacts on N2O production and emission potential.
Traditionally, studies of soil N2O emissions have emphasized bulk microbial abundance or key functional genes (Muhammad et al., 2022), such as the nirK/nirS genes encoding nitrite reductase and the nosZ gene encoding N2O reductase (Li et al., 2024). Research has shown that after land use changes, increases in the abundance of nitrifying and denitrifying microbial genes often lead to elevated soil N2O emissions (Zhang et al., 2022). However, emerging evidence underscores the pivotal role of microbial community interactions, such as species co-occurrence, competition, or cooperation can influence functional outcomes (e.g., N2O production or consumption) in ways that extend beyond mere gene abundance (Li et al., 2023). Network-analysis approaches (e.g., co-occurrence networks and molecular ecological networks) provide a framework to characterize microbial community complexity, modularity, connectivity and stability (Baker et al., 2024). This methodology has been demonstrated as an effective tool for exploring microbial community interactions and organization (Faust and Raes, 2012). Research has indicated that the topological features of these networks are influenced by changes in environmental conditions and can serve as indicators of alterations in the ecosystem’s potential functionality (Felipe-Lucia et al., 2020). For instance, Cornell et al. (2023) reported that converting temperate grasslands to conventional croplands increases microbial network complexity (e.g., connectance) as a means of compensating for the ecological functional losses resulting from biodiversity decline. Whereas, in ecosystems under high stress and resource scarcity, microorganisms may survive by employing niche differentiation strategies to promote diverse nutrient utilization, resulting in the formation of additional micro-modules (Yang et al., 2025). Regarding N2O emissions in nitrogen transformation processes, it has been demonstrated that in freshwater lakes, the emission levels are significantly influenced by the complexity of denitrifying community networks (Su et al., 2024). These findings suggest that microbial network properties (topology, keystone taxa, modularity, and negative/positive links) may serve as mechanistic levers linking soil environmental change to greenhouse-gas fluxes. Yet explicit investigations of how land use conversion (especially paddy-to-vegetable) reconfigures microbial networks and thereby regulates N2O emissions are still scarce.
In China’s Yangtze River Delta region, a globally important agricultural and economic hub, traditional flooded rice fields are increasingly being converted into vegetable fields. This region provides an ideal natural laboratory for investigating how paddy-to-vegetable fields conversion affects soil microbial networks and N2O emissions. Previous studies in this region have documented elevated N2O emissions following paddy conversion (Wei et al., 2022), but the microbial network mechanisms underlying these changes remain poorly quantified. Given the complex interactions among soil microorganisms, identifying how bacterial and fungal networks respond to this transformation is critical for elucidating the microbial ecological drivers of N2O dynamics in this region. Specifically, two major knowledge gaps persist: (1) whether microbial diversity and network structural characteristics respond differently to short-term and long-term land use conversion; (2) how significant are network characteristics for N2O emissions under this type of land use change. To address these gaps, this study investigates the responses of bacterial and fungal network characteristics in a rice field and vegetable fields converted from rice field for 4 years and 7 years at the Yangtze River Delta and explores their linkages to soil N2O emissions. We hypothesized that (1) the conversion of rice fields to vegetable cultivation increases the complexity and stability of microbial networks, but such enhancement diminishes with prolonged conversion; (2) microbial network characteristics have a significant impact on N2O emissions during land use change.
2 Materials and methods
2.1 Soil sampling
Soil samples were collected from a rice field (RF) and vegetable fields converted from paddy soils for 4 and 7 years (VE4 and VE7, respectively), near the Changshu Agricultural Ecological Experimental Station of the Institute of Soil Science, Chinese Academy of Sciences, in Jiangsu, China. The region traditionally follows a wheat-rice rotation system with N fertilization rates of 500 kg ha−1. Driven by economic incentives, some rice fields have been converted to vegetable cultivation, with N application rates typically exceeding 800 kg ha−1. At the time of sampling, soil samples were collected from the top 20 cm layer by removing the top 2 cm of soil to eliminate plant residue interference. To minimize the short-term influences of fertilization pulses as well as crop-related disturbances on soil nitrogen transformation processes and microbial communities, soil sampling was uniformly conducted within 1 week after crop harvest. This timing was intended to better capture the background state of soil microbial networks and nitrogen transformation processes across different land use types, thereby enhancing comparability among samples. Three independent sampling locations (≥ 10 m apart) were selected for each land use type, with five random sub-samples taken at each location and thoroughly homogenized to generate one composite sample. Resulting in nine samples in total. After collection, samples were promptly transported to the laboratory and stored at 4 °C for incubation experiments. A portion of the samples was air-dried for analysis of soil properties. The basic characteristics of the soils under different treatments were provided by Li C. L. et al. (2022) and are also presented in Supplementary Table S1.
2.2 Soil incubation experiment
Soil N2O emission rates were measured using the slurry incubation method (Chen et al., 2019). Soil samples stored in a refrigerator were mixed with He-flushed water at a 1:5 ratio and transferred into 12 mL glass vials (Exetainer; Labco, United Kingdom) in an anaerobic chamber. After sealing, the vials were placed on a vertical shaker and pre-incubated at 25 °C for 1 day to restore soil activity and deplete background oxygen. Then, the slurry was incubated at 25 °C for 7 h, after which 200 μL of saturated ZnCl2 was added to each sample to terminate the incubation. Following incubation, the samples were centrifuged, and 5 mL of supernatant was transferred from each vial into 18.5 mL vacuum serum bottles using a syringe. Environmental air was then introduced into the bottles, and they were immediately equilibrated at 5 °C for 24 h. Three milliliters of headspace gas were extracted from each vial with a syringe and analyzed for N2O concentration using a gas chromatograph (Agilent 7890A, United States). The total N2O concentration in the slurry was estimated based on the measured headspace N2O concentration and the Bunsen solubility coefficient at 5 °C. The N2O emission rate was subsequently derived from the temporal variation in N2O concentrations using linear fitting over the incubation period. It is important to note that since all soil samples were measured under anaerobic conditions, N2O emissions are considered a result of denitrification. The N2O emission rates measured under anaerobic conditions are regarded as the potential for N2O emissions. Therefore, the subsequent N2O emission rates reflect only the potential denitrification-derived N2O production under controlled anaerobic conditions.
2.3 Soil microbial community analysis
Soil samples were freeze-dried, and 0.5 g was used for DNA extraction using the Fast DNA SPIN kit (MP Biomedical, Cleveland, OH, United States). Microbial community composition was analyzed through high-throughput sequencing, with amplicon sequencing performed on an Illumina MiSeq platform (Illumina, San Diego, CA, United States). Amplification of the V4-V5 regions of the universal 16S rRNA gene and the ITS1 region of fungal ITS genes was performed using the corresponding primer sets (Zhu et al., 2024). Following data filtering, 260,892 bacterial sequences and 313,902 fungal sequences were obtained. The bioinformatics analysis process and PCR reaction conditions are described in detail in the Supplementary Table S2. Furthermore, we constructed a co-occurrence network using all samples. After excluding low-abundance groups and performing trimmed mean of M-values (TMM) normalization, spearman correlation coefficients were calculated between taxon (at the genus level). Only pairs exhibiting a correlation coefficient above 0.7 and a p value below 0.05 were maintained (Fan et al., 2019; Ma X. et al., 2025). The network and its primary modules were visualized using Gephi.1 For each module, microbial group relative abundance was derived from the standardized species abundance (z-score). Linear regression was then applied to assess the relationship between taxon abundance within each module and N2O emission rates. Taxa exhibiting high Zi values were classified as keystone nodes (Liu et al., 2025). The Zi values, network robustness, and subnetwork topology across different treatments were calculated using the ggClusterNet package in R (Wen et al., 2022). Microbial network cohesion was quantified using the microeco package in R (Liu et al., 2023).
2.4 Statistical analysis
All statistical analyses, except for the comparison of network robustness differences, were conducted using SPSS (version 27.0). Statistical significance between different land use types was tested using one-way analysis of ANOVA, followed by LSD testing. Differences in microbial network robustness between land use types were compared using a two-sided t-test, performed in R software. Differences in microbial communities across distinct land use types were examined using the ANOSIM method. The Pearson correlation between N2O emission rates and microbial network topology was calculated using the linkET package in R. The habitat niche breadth of taxa was quantified by the niche.width function from the spaa package (Zheng et al., 2025). And key factors influencing N2O emissions were analyzed and assessed using the randomForest package in R. Principal component analysis (PCA) was used to construct composite indices for soil microbial diversity, network characteristics, and soil properties (Chen Q. et al., 2021). The first principal component (the dimension with the highest explained variance) was selected for random forest analysis.
3 Results
3.1 Changes in N2O emissions and microbial community composition
The conversion of rice fields to vegetable cultivation significantly influenced soil N2O emissions (Supplementary Figure S1). Compared to the RF system, both vegetable fields converted from RF for 4 years and 7 years showed elevated N2O emission rates. Specifically, the N2O emission rates in VE4 were approximately 0.43 nmol N g−1 h−1 higher than those in the RF system, while the emissions in VE7 were nearly 17.86 times higher. The transition from rice fields to vegetable fields also resulted in shifts in microbial community composition (Figure 1; Supplementary Figure S2). In RF soils, Candidatus_Solibacter (2.13%), Knoellia (2.12%), and Arthrobacter (2.05%) were the most abundant bacterial genera (Figure 1A). In contrast, VE4 and VE7 soils showed a marked increase in the abundance of nitrogen-cycling bacteria (Figures 1C,E), such as Bacillus (2.36–5.17%) and Gemmatimonas (1.90–2.40%). Additionally, fungal communities also shifted, with an increase in Lasiobolus, Condenascus, and Fusarium (Figures 1B,D,F).
Figure 1. Relative abundances of dominant bacterial (A,C,E) and fungal (B,D,F) genera in a rice field (RF) and vegetable fields converted from the rice field for 4 years (VE4) and 7 years (VE7).
In terms of microbial diversity, land use conversion significantly altered the β-diversity of bacterial and fungal communities (Supplementary Figure S3), as well as the α-diversity of bacterial communities (Figure 2A). Community α-diversity was evaluated through six commonly used indices, namely Shannon, Simpson, Pielou, Coverage, Heip, and Pd (Supplementary Figure S4). The RF exhibited significantly higher taxonomic diversity compared to the VE4 and VE7 (p < 0.05). However, no significant differences were observed in the α-diversity of fungal communities among the different treatments (Figure 2A). Furthermore, changes in land use did not significantly reduce the ecological niche breadth of bacterial communities, while the fungal community niche breadth in VE4 was significantly reduced by 19.59% compared to RF (Figure 2B).
Figure 2. Microbial community diversity (A) and niche breadth (B) in a rice field (RF) and vegetable fields converted from the rice field for 4 years (VE4) and 7 years (VE7). Different lowercase letters above bars indicate significant differences among treatments for bacterial community, and different uppercase letters indicate significant differences for fungal community (p < 0.05). Error bars represent standard deviations (n = 3). Note: Six indices, Shannon index, Simpson index, Pilou index, coverage index, Heip index, and Pd index were used to reflect the taxonomic diversity in the analysis alpha diversity.
3.2 Variations soil microbial co-occurrence network
The conversion of RF to vegetable cultivation significantly altered the structure of both bacterial and fungal co-occurrence networks (Figures 3A,B), as reflected in key topological characteristics and robustness. We identified three ecological clusters that exhibited significant co-occurrence relationships in the bacterial constructed network (Figure 3A; Supplementary Figure S5A). Within the co-occurrence network, Sphaerobacter and Methylocaldum were identified as core taxa (Supplementary Figure S6A). The network features of bacteria, such as the number of edges, positive connection edges, negative connection edges, number of vertices, average degree, and mean clustering coefficient were significantly lower in RF (p < 0.05). The number of positive connections under VE4 and VE7 land use types increased by 45.49 and 31.39%, respectively, compared to RF, while the number of negative connections increased by 59.57 and 39.39%, respectively (Supplementary Figure S7A). Under VE4 and VE7, the number of vertices was 1.24 and 1.15 times that of RF, respectively (Supplementary Figure S7B). Similarly, the average degree and mean clustering coefficient were 1.23 and 1.17 times, and 1.02 and 1.03 times higher than those under RF (Supplementary Figure S7C,D). But there was no significant difference in centralization betweenness between the land use type (Supplementary Figure S7E). The positive cohesion ranked as RF < VE7 < VE4 in bacterial co-occurrence networks (Supplementary Figure S7F). Moreover, the network robustness decreased significantly (p < 0.05) with increasing years since rice–vegetable conversion, being highest in the RF and lowest in the VF7 (Supplementary Figure S8A).
Figure 3. Network diagram with nodes colored according to each of the three main ecological clusters (A,B). The modules in the networks are clusters of closely interconnected nodes. The size of each node reflects the number of connections (degree). The scatter diagram reflect regressions between the N2O emission rate and the relative abundance of the main bacteria in ecological clusters (C–E).
For the fungal network, four ecological clusters exhibiting significant co-occurrence relationships were identified (Figure 3B; Supplementary Figure S5B). Gymnascella and five other fungal genera were recognized as key nodes (Supplementary Figure S6B). The fungal network in the rice field (RF) exhibited significantly higher numbers of total edges, positive connection edges, and average degree as well as a greater mean clustering coefficient (p < 0.05), compared with those in the vegetable field soils (Supplementary Figures S9A,C,D). Conversely, the number of negative connection edges declined markedly with increasing years since conversion (p < 0.05). The number of vertices peaked in VF7, but declined by 11.44% and 27.49% in RF and VE4, respectively (Supplementary Figure S9B). In contrast, no significant variation was detected in the network’s centralization betweenness or positive cohesion across different land use types. Although VE4 and VE7 did not differ significantly in network robustness, both exhibited significantly greater robustness than the RF.
3.3 Key factors influencing N2O emissions
The relationship between microbial network characteristics and N2O emissions was evaluated by examining the topological characteristics of microbial (bacteria and fungi) networks. Notably, the number of edges, positive connection edges, negative connection edges, number of vertices, average degree and positive cohesion exhibited significant positive correlations with N2O emission rates in bacterial network (p < 0.05). Conversely, the ratio of positive/negative edges showed a significant negative correlation with N2O emission rates (p < 0.05; Figure 4A). Additionally, the number of edges, positive connection edges, number of vertices, and average degree in fungal network were significantly negatively correlated with N2O emission rates (p < 0.05; Figure 4B). Beyond network topological structure, the relative abundance of genera in different modules of the microbial network was also correlated with soil N2O emissions. Specifically, the relative abundance of genus in module 1 of the bacterial network and module 1 of the fungal network was significantly negatively correlated with N2O emissions (p < 0.05; Figures 3C,E). While the relative abundance of genus in module 2 of the bacterial network showed a significant positive correlation with N2O emissions (p < 0.05) (Figure 3D). The relative abundance of genus in other microbial network modules did not exhibit a significant correlation with soil N2O emissions (Supplementary Figure S10).
Figure 4. Pearson correlations between the N2O emission rate, bacterial network topological properties (A) and fungal network topological properties (B) in a rice field (RF) and vegetable fields converted from the rice field for 4 years (VE4) and 7 years (VE7). The color gradient and square denote Pearson’s correlation coefficients. Asterisks denote statistically significant differences at significance levels of *p < 0.05, **p < 0.01, and ***p < 0.001.
To further identify the key factors influencing N2O emissions, a Random Forest analysis was conducted using environmental variables (Supplementary Figure S11), taxonomic diversity, the relative abundance of genus in different modules, and network characteristics as predictors (Supplementary Figure S12). The explanatory power of the model reached 96% (p < 0.05). The analysis revealed that bacterial network characteristics and fungal network characteristics were the most significant predictors of N2O emissions (Figure 5). Furthermore, the analysis indicated that the diversity characteristics of the bacterial community and the relative abundance of genera in module 1 of the fungal network play a relatively important role in regulating N2O emissions, although this did not reach statistical significance. Among these factors, soil properties were identified as the weakest driver.
Figure 5. Potential driving factors influencing soil N2O emission rate variations during land-use conversion. The increase in node purity for each variable is used to estimate predictor importance, with higher values indicating greater importance. Asterisks denote statistically significant differences at significance levels of *p < 0.05. R2 represents the explanatory power of the overall predictive model. The p value indicates the overall significance of the model.
4 Discussion
4.1 Land use change alters soil microbial community composition
Increasing land use intensity and N management input generally enhance N2O emission rates (Uwiragiye et al., 2024). For instance, the conversion of natural forests to croplands typically stimulates soil N2O release (Zheng et al., 2020; Xiao et al., 2025). Regions showing pronounced increases in N2O emissions following land use change are often characterized by high levels of reactive nitrogen inputs (Ma J. Y. et al., 2025). In our specific land use conversion system, we likewise observed a significant increase in soil N2O emissions following the conversion of rice fields to vegetable cultivation. This pattern is consistent with a study conducted in Jiangxi Province, China, where the conversion of double-cropping rice fields to vegetable cultivation increased the soil N2O emission rate by 14.12 times (Yuan et al., 2016). It is noteworthy that although the elevated trend of N2O fluxes in vegetable fields relative to the rice field persisted across different cultivation durations, the N2O flux in the VE7 soil was significantly lower than that in VE4. This suggests that under long-term vegetable cultivation, the microbial processes in VE7 soil may have reached a new steady or adaptive state (Ji et al., 2021). Such a state could be attributed to shifts in microbial community composition (Shen et al., 2021), changes in nutrient availability (Li L. Z. et al., 2022), strengthened microbial interactions (Lin et al., 2025; see also Section 4.3 for further discussion on their role in N2O dynamics), or altered niche breadth.
The conversion from rice to vegetable fields alters microbial community composition. In the RF soil, the three most abundant bacterial genera were mainly decomposers with strong capacities for organic carbon degradation but showed a complete or partial absence of denitrification-related functional genes (Zhang et al., 2021; Yessentayeva et al., 2024). As a representative, Candidatus_Solibacter encodes diverse carbohydrate-active enzymes and phenol oxidases, which enable it to mineralize plant residues and humic substances effectively (Mastny et al., 2021). By contrast, continuous vegetable cultivation supported a bacterial community dominated by Gemmatimonas and Bacillus. These microorganisms are key participants in denitrification (Liu et al., 2022; Huang et al., 2025), likely contributing to the increased N2O emissions detected in vegetable soils. Moreover, Mortierella was the most abundant fungal genus in RF soil. Jiang et al. (2021) similarly reported that during the milk stage of rice, the peak relative abundance of Mortierella could reach 50%. Analogous to the bacterial communities, the dominant fungal genera in VF4 and VF7 were largely similar after rice-to-vegetable conversion. Although they have been rarely studied, these fungi are presumed to contribute to soil organic matter decomposition (Piontelli et al., 1981), highlighting the need for further research. As for community diversity, microbial community diversity exhibited consistent patterns for both bacteria and fungi. It was highest in RF soils, decreased after 4 years of conversion to vegetable fields, and increased after 7 years. This indicates that rice-to-vegetable conversion may initially disrupt microbial community function, but over time, communities tend to stabilize into a new steady state. These trends align with the temporal pattern of soil N2O emissions, suggesting that changes in microbial diversity may play a key role in regulating N2O fluxes (Shen et al., 2021). In addition, after 4 years of rice-to-vegetable conversion, the fungal niche breadth significantly decreased, likely due to enhanced cultivation intensity that imposed stronger environmental filtering (Zhou and Wu, 2021). The narrowing of fungal niche breadth may have altered the balance between C decomposition and N transformation (Ren et al., 2023). Specifically, the loss of fungal taxa with high capacities for organic matter turnover could have reduced the C substrates available to denitrifying bacteria, thereby promoting incomplete denitrification and indirectly enhancing N2O emissions (Chen Z. et al., 2021). The findings highlight the pivotal role of fungi in maintaining N cycling efficiency and mitigating N2O emissions during land use transitions.
4.2 Soil microbial co-occurrence network variations
The characteristic of the microbial co-occurrence network was significantly influenced by the conversion of rice to vegetable fields. In the bacterial network, three ecological clusters with significant co-occurrence relationships were identified, while the fungal network contained four clusters. Notably, not all microbial groups were significantly correlated with soil N2O emissions, but rather, specific microbial groups within certain modules. In agricultural ecosystems, specific environmental selection pressures may lead to the emergence of subsets of beneficial and detrimental microbial groups within the network (Fan et al., 2019), which have differential impacts on nitrogen cycling processes (Yu et al., 2024). As a result, only microbial groups within specific modules showed significant correlations with soil N2O emissions (Jones et al., 2022). However, the keystone taxa within the networks exhibited weak direct associations with soil N2O emissions (Supplementary Figure S13), suggesting that they may exert their effects by driving module-level metabolic processes rather than by directly regulating N2O fluxes. For example, Sphaerobacter is capable of efficiently degrading cellulose (Hou et al., 2025), whereas Methylocaldum has been shown to influence energy metabolism and thereby play a key role in N transformation processes (Chu et al., 2025).
Following the conversion of rice fields to vegetable cultivation, management practices were profoundly altered, leading to greater fluctuations in soil physicochemical properties (Wu et al., 2023). The bacterial community responded rapidly to these environmental perturbations (Wang et al., 2024), reorganizing its community composition and forming new interspecific associations. This dynamic adjustment increased the number of correlations among nodes within the network, resulting in a temporary rise in overall network complexity (Huang et al., 2024). However, this apparent complexity largely reflects short-term “stress-induced linkages” rather than stable, functionally coherent relationships. As environmental conditions continue to change, bacterial communities may establish new species interactions through rapid adaptive mechanisms (Sun et al., 2025). Such interactions are often fragile and lack sufficient functional redundancy. Low redundancy implies that when the system experiences external disturbances, there are limited alternative pathways or compensatory functions available, thereby reducing the overall robustness and resilience of the bacterial network (Allison and Martiny, 2008). Compared with bacteria, fungal communities generally exhibit a higher degree of functional redundancy (Zhou et al., 2022), meaning that even if certain taxa are functionally impaired, others can compensate to maintain essential ecological functions (Zeng et al., 2023; Moreira et al., 2025). Land use changes may reduce the complexity of fungal networks, but this does not necessarily imply a loss of function. Instead, fungi may optimize resource use efficiency and adaptive capacity by simplifying their interaction networks (Yang et al., 2022). In dynamic environments, simplified fungal networks can be more efficient in resource allocation than highly complex ones. By reducing unnecessary functional redundancy, fungal communities may direct their activity toward essential ecosystem functions, including the decomposition of organic substrates and nitrogen cycling (Bao et al., 2024). This adaptive strategy enhances the robustness of fungal networks, enabling them to sustain ecosystem functionality under environmental disturbances (Xiong et al., 2024). Our results are not fully consistent with previous findings. Cornell et al. (2023) reported that within 19 months of converting grassland to wheat cropland, both the microbial network’s stability and complexity increased simultaneously. The discrepancy between our results and theirs may stem from differences in land use conversion pathways. In the previous study, the system transitioned from natural grassland to cultivated land, whereas in our study, the conversion was from a flooded environment to upland conditions.
4.3 Mechanistic integration: linking land use change, microbial networks, and N2O emissions
The conversion of rice fields to vegetable cultivation may fundamentally alter bacterial co-occurrence patterns, thereby reshaping soil N transformation pathways and leading to increased N2O emissions. As bacterial networks become more complex yet fragile, the stability of bacterial cooperation may decline (Yonatan et al., 2022), potentially disrupting the delicate balance of N transformation processes. In particular, alterations in the connectivity of denitrifying bacteria may hinder the complete reduction of N2O to N2, resulting in the accumulation of the intermediate product (Carr et al., 2025). Furthermore, under fluctuating environmental conditions, rapid community turnover may favor opportunistic bacterial species with high N2O producing but limited reducing capacities, thereby further enhancing N2O emissions (Shakoor et al., 2025).
Unlike bacterial networks, this study found that a decline in fungal network complexity accompanied by increased robustness was associated with a significant rise in soil N2O emissions. This finding suggests that during the conversion of rice fields to vegetable cultivation, the ecological “integration” of fungal networks may enhance system stability, yet concurrently favor functional groups that promote N2O production rather than its reduction. Such simplified but more robust fungal networks are typically dominated by a few stress-tolerant or resource-efficient taxa (Zhang et al., 2025). These fungi are mostly saprotrophic or opportunistic species that can accelerate organic matter decomposition and N mineralization (Camenzind et al., 2024), thereby increasing the availability of inorganic N and providing a continuous substrate supply for bacterial denitrification. From a functional perspective, FUNGuild-based annotation revealed that saprotrophic fungi were dominant in VE4 (Supplementary Figure S14), supporting the view that fungal networks may primarily influence denitrification by regulating organic matter decomposition.
The disruption of the C-N balance under such conditions may lead to sustained N2O emissions (Zhu et al., 2023). Moreover, the decline in fungal network complexity weakened its regulatory role in the coupling of carbon and nitrogen cycles (Booth et al., 2019). In complex networks, diverse fungal taxa interact with bacteria through competitive or mutualistic relationships, forming multilayered linkages that buffer N loss and maintain soil redox balance (Wagg et al., 2019). Once the network is simplified, such ecological regulation diminishes—fungal N immobilization becomes less effective, while incomplete bacterial denitrification dominates, leading to enhanced N2O accumulation. Thus, the observed “robustness” may, in fact, represent a stable state characterized by persistently high N2O emissions.
It is noteworthy that in both bacterial and fungal networks, the ratio of positive to negative connections exhibited a strong negative association with N2O emissions. Networks characterized by lower positive/negative connection ratios typically exhibit intensified internal competition (Ma et al., 2020). This finding further suggests that weakened microbial cooperation and enhanced competitive interactions disrupt the functional coordination among N-cycling guilds, thereby increasing N2O emissions. The recovery of the positive-to-negative connection ratio in VE7 compared with their lower levels in VE4 suggests that microbial processes in VE7 soils may be, or are in the process of, reaching a new steady state. Finally, the random forest analysis reaffirmed that microbial network features were the most influential predictors of N2O fluxes, surpassing traditional soil physicochemical variables. Consistently, studies in various ecosystems have also highlighted the pivotal role of microbial network architecture in regulating soil N2O emissions (Jameson et al., 2023; Qiu et al., 2023). Furthermore, no significant associations were detected between integrated soil properties and microbial network characteristics (Supplementary Figure S15). Consequently, the indirect regulatory effect of integrated soil physicochemical properties on N2O emissions mediated through microbial network features appears to be relatively weak.
Taken together, a potential mechanistic framework emerges: the conversion of rice fields to vegetable cultivation reshapes the composition and co-occurrence architecture of microbial communities. Within this context, microbial networks act as key mediators that translate environmental perturbations into changes in ecosystem process rates. The contrasting responses of bacterial and fungal networks highlight their complementary ecological roles—bacterial networks are closely linked to denitrification, whereas fungal networks may indirectly regulate N2O dynamics by influencing soil carbon turnover and microhabitat accessibility. Nevertheless, it should be emphasized that our findings are mainly indicative of microbial responses to site-specific differences in the potential denitrification-derived N2O emissions, rather than direct responses to actual field emissions.
4.4 Limitations and future directions
Field N2O emissions originate from multiple processes, including nitrification, denitrification, and their coupled pathways under fluctuating redox conditions (e.g., nitrifier denitrification) (Zhu G. B. et al., 2025). The measurement approach used in this study represents only the potential N2O production associated with denitrification under controlled anaerobic conditions. Consequently, the N2O emission results presented here should be interpreted as a process-based indicator that aids in explaining the spatial variability of field N2O emissions and provides mechanistic insight into the role of denitrification in shaping field-scale N2O emission patterns, rather than as a direct quantitative prediction of in situ N2O fluxes. Besides, the reshaping of microbial networks induced by land use change underscores the urgent need for integrated management strategies that account for microbial dynamics when mitigating N2O emissions. Emerging evidence suggests that interventions such as crop diversification and the application of organic amendments can enhance nutrient cycling efficiency in agricultural soils while simultaneously reducing greenhouse gas emissions (Li et al., 2024). A deeper understanding of how these practices influence microbial interactions and network architecture in regulating N2O fluxes will enable more precise soil management strategies that balance agricultural productivity with environmental sustainability. Although our study reveals strong correlations between microbial network properties and N2O emissions, the limited sample size may reduce the statistical power and thus increase the risk of type II errors, especially in microbial co-occurrence network analyses. Therefore, the results of this study should be interpreted primarily as indicative of potential ecological patterns. Future studies could verify the robustness of our findings by increasing the sample size or incorporating multi-scale sampling approaches. Manipulative experiments targeting specific network modules or taxa would clarify their functional roles. Long-term monitoring across diverse soil types and climatic regions is also needed to generalize the observed patterns. Furthermore, integrating metagenomic or transcriptomic analyses could link network structure to gene expression, providing deeper mechanistic insight into microbial contributions to N2O dynamics.
5 Conclusion
Overall, our findings demonstrate that converting rice fields to vegetable cultivation markedly alters microbial community composition and stimulates the potential denitrification-derived N2O emissions. Specifically, compared with the original rice soils, bacterial co-occurrence networks in vegetable soils exhibited a greater number of edges, higher connectivity, and lower robustness. In contrast, fungal networks showed fewer edges, reduced connectivity, and enhanced robustness. With increasing cultivation duration, the number of edges and the degree of connectivity in bacterial networks significantly decreased, whereas those in fungal networks increased. Moreover, microbial community diversity exhibited a pattern of initial decline followed by recovery after land use conversion. Collectively, our results suggest that microbial network characteristics serve as the most critical predictors of the potential denitrification-derived N2O emissions during rice to vegetable conversion. This finding highlights that microbial communities are not merely passive responders to environmental change but actively mediate biogeochemical processes through their interaction networks. Given the ongoing global shifts in land use patterns, integrating microbial network features into predictive models of soil greenhouse gas emissions will be essential for improving the accuracy of N2O emission assessments under future land use change scenarios.
Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.
Author contributions
CL: Funding acquisition, Writing – original draft. ZZ: Conceptualization, Writing – original draft. XC: Investigation, Writing – original draft. QT: Methodology, Writing – review & editing. QZ: Supervision, Writing – review & editing. JT: Writing – review & editing, Supervision.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This study was financially supported by the Southwest Medical University Technology Program (2024ZKY021).
Acknowledgments
We are also grateful for the guidance and support provided by the guidance and support provided by Dr. Jun Shan (Institute of Soil Science, Chinese Academy of Sciences) and the Planning Program of the 2025 Luzhou Social Science Application Research Quality Improvement Project (LZSK25GH046).
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The reviewer XRY declared a shared affiliation with the author CL to the handling editor at the time of review.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmicb.2026.1750894/full#supplementary-material
Footnotes
References
Allison, S. D., and Martiny, J. B. H. (2008). Resistance, resilience, and redundancy in microbial communities. Proc. Natl. Acad. Sci. U. S. A. 105, 11512–11519. doi: 10.1073/pnas.0801925105,
Babur, E., Dindaroglu, T., Uslu, Ö., Gozukara, G., and Ozlu, E. (2025). Long-term effects of land use conversion on soil microbial biomass and stoichiometric indices in eastern Mediterranean karst ecosystems (1981-2018). Land Degrad. Dev. 36, 5666–5680. doi: 10.1002/ldr.70029
Baker, N. R., Zhalnina, K., Yuan, M. T., Herman, D., Ceja-Navarro, J. A., Sasse, J., et al. (2024). Nutrient and moisture limitations reveal keystone metabolites linking rhizosphere metabolomes and microbiomes. Proc. Natl. Acad. Sci. USA 121:e2303439121. doi: 10.1073/pnas.2303439121,
Bao, Y., Dolfing, J., Li, X., Chen, R., Cui, X., Li, Z., et al. (2024). Functional redundancy across space and time in litter-degrading fungal communities. J. Sustain. Agric. Environ. 3:e12086. doi: 10.1002/sae2.12086
Booth, J. M., Fusi, M., Marasco, R., Michoud, G., Fodelianakis, S., Merlino, G., et al. (2019). The role of fungi in heterogeneous sediment microbial networks. Sci. Rep. 9:7537. doi: 10.1038/s41598-019-43980-3,
Camenzind, T., Aguilar-Trigueros, C. A., Hempel, S., Lehmann, A., Bielcik, M., Andrade-Linares, D. R., et al. (2024). Towards establishing a fungal economics spectrum in soil saprobic fungi. Nat. Commun. 15:3321. doi: 10.1038/s41467-024-47705-7,
Carr, A. V., Otwell, A. E., Hunt, K. A., Chen, Y., Wilson, J., Faria, J. P., et al. (2025). Emergence and disruption of cooperativity in a denitrifying microbial community. ISME J. 19:wraf093. doi: 10.1093/ismejo/wraf093,
Chen, S. T., Chee-Sanford, J. C., Yang, W. H., Sanford, R. A., Chen, J. Q., Yan, X. Y., et al. (2019). Effects of triclosan and triclocarban on denitrification and N2O emissions in paddy soil. Sci. Total Environ. 695:8. doi: 10.1016/j.scitotenv.2019.133782,
Chen, Z. X., Fu, S., Sun, D., and Yang, M. (2025). How land use change affects greenhouse gas emission in China. J. Soils Sediments 25, 3473–3484. doi: 10.1007/s11368-025-04158-9
Chen, Q., Long, C. Y., Chen, J. W., and Cheng, X. L. (2021). Differential response of soil CO2, CH4, and N2O emissions to edaphic properties and microbial attributes following afforestation in Central China. Glob. Change Biol. 27, 5657–5669. doi: 10.1111/gcb.15826,
Chen, Z., Tu, X., Meng, H., Chen, C., Chen, Y., Elrys, A. S., et al. (2021). Microbial process-oriented understanding of stimulation of soil N2O emission following the input of organic materials. Environ. Pollut. 284:117176. doi: 10.1016/j.envpol.2021.117176,
Chu, Z. H., Tang, X., Li, Y. C., Li, J., Xiong, W., Yin, Y. H., et al. (2025). Simultaneous removal of organic matter and inorganic nitrogen in baijiu wastewater by methanotrophic denitrification. Bioresour. Technol. 418:131956. doi: 10.1016/j.biortech.2024.131956,
Cornell, C. R., Zhang, Y., Ning, D. L., Xiao, N. J., Wagle, P., Xiao, X. M., et al. (2023). Land use conversion increases network complexity and stability of soil microbial communities in a temperate grassland. ISME J. 17, 2210–2220. doi: 10.1038/s41396-023-01521-x,
Fan, K. K., Delgado-Baquerizo, M., Guo, X. S., Wang, D. Z., Wu, Y. Y., Zhu, M., et al. (2019). Suppressed N fixation and diazotrophs after four decades of fertilization. Microbiome 7:143. doi: 10.1186/s40168-019-0757-8,
Faust, K., and Raes, J. (2012). Microbial interactions: from networks to models. Nat. Rev. Microbiol. 10, 538–550. doi: 10.1038/nrmicro2832,
Felipe-Lucia, M. R., Soliveres, S., Penone, C., Fischer, M., Ammer, C., Boch, S., et al. (2020). Land-use intensity alters networks between biodiversity, ecosystem functions, and services. Proc. Natl. Acad. Sci. U. S. A. 117, 28140–28149. doi: 10.1073/pnas.2016210117,
Hou, J. Q., Yu, C. Z., Ye, M. Y., Guo, Z. Y., Xin, Y. J., Meng, F. H., et al. (2025). Short-term hydrothermal fermentation amendments enhance labile organic carbon in deep soil: synergistic effects of organic carbon, enzymes, and microbes. Environ. Technol. Innov. 40:104348. doi: 10.1016/j.eti.2025.104348
Huang, C., He, Y., Zhou, L., Liu, R., Chen, H., Du, Z., et al. (2024). Opposite effects of soil pH on bacteria and fungi β diversity in forests at a continental scale. J. Environ. Manag. 370:122428. doi: 10.1016/j.jenvman.2024.122428
Huang, T. Y., Ju, H. J., Huang, M. Y., Kuo, Q. M., and Su, W. T. (2025). Optimal nitrite degradation by isolated Bacillus subtilis sp. N4 and applied for intensive aquaculture water quality management with immobilized strains. J. Environ. Manag. 374:123896. doi: 10.1016/j.jenvman.2024.123896,
Jameson, B. D., Murdock, S. A., Ji, Q. X., Stevens, C. J., Grundle, D. S., and Juniper, S. K. (2023). Network analysis of 16S rRNA sequences suggests microbial keystone taxa contribute to marine N2O cycling. Commun. Biol. 6:212. doi: 10.1038/s42003-023-04597-5,
Ji, M., Tian, H., Wu, X., Li, J., Zhu, Y., Wu, G., et al. (2021). Enhanced N2O emission rate in field soil undergoing conventional intensive fertilization is attributed to the shifts of denitrifying guilds. Pedosphere 31, 145–156. doi: 10.1016/S1002-0160(20)60050-9
Jiang, Z. W., Yang, S. H., Pang, Q. Q., Xu, Y., Chen, X., Sun, X., et al. (2021). Biochar improved soil health and mitigated greenhouse gas emission from controlled irrigation paddy field: insights into microbial diversity. J. Clean. Prod. 318:128595. doi: 10.1016/j.jclepro.2021.128595
Jones, C. M., Putz, M., Tiemann, M., and Hallin, S. (2022). Reactive nitrogen restructures and weakens microbial controls of soil N2O emissions. Commun. Biol. 5:273. doi: 10.1038/s42003-022-03211-4,
Li, W. H., Gao, S. L., Qu, W. Y., Li, J. F., Javed, T., Ho, S. H., et al. (2025). Long-term mulched drip irrigation enhances cotton yield by improving soil quality, microbial community structure, and N conversion. Agric. Water Manag. 314:109527. doi: 10.1016/j.agwat.2025.109527
Li, T. T., Liu, X. Y., Tian, J. H., Yuan, W. P., Wang, X. H., Yang, X. Q., et al. (2025). Methane and nitrous oxide budget for Chinese natural terrestrial ecosystems. Natl. Sci. Rev. 12:nwaf094. doi: 10.1093/nsr/nwaf094,
Li, C., Wei, Z., Wang, X., Ma, X., Tang, Q., Zhao, B., et al. (2024). Biochar mitigates the stimulatory effects of straw incorporation on N2O emission and N2O/(N2O+N2) ratio in upland soil. J. Environ. Manag. 369:122318. doi: 10.1016/j.jenvman.2024.122318
Li, C. L., Wei, Z. J., Yang, P. P., Shan, J., and Yan, X. Y. (2022). Conversion from rice fields to vegetable fields alters product stoichiometry of denitrification and increases N2O emission. Environ. Res. 215:114279. doi: 10.1016/j.envres.2022.114279
Li, L. Z., Zhao, C. S., Chen, Q. F., Liu, T., Li, L., Liu, X. Z., et al. (2022). Study on microbial community structure and soil nitrogen accumulation in greenhouse vegetable fields with different planting years. Agronomy 12:1911. doi: 10.3390/agronomy12081911
Li, X., Zhao, R. T., Li, D. D., Wang, G. Z., Bei, S. K., Ju, X. T., et al. (2023). Mycorrhiza-mediated recruitment of complete denitrifying Pseudomonas reduces N2O emissions from soil. Microbiome 11:45. doi: 10.1186/s40168-023-01466-5,
Lin, X. B., Zhu, B. Q., Peng, Y. Y., Lin, G. M., Sun, D. Y., and Ye, W. W. (2025). Responses of microbial communities and greenhouse gas production to land use change in mangrove wetland sediments. Limnol. Oceanogr. 70, 1210–1225. doi: 10.1002/lno.70028
Liu, J. D., Fu, X. R., Yuan, R. F., Chen, Z. B., Zhou, B. H., and Chen, H. L. (2022). Carbon sources derived from corncobs enhanced nitrogen removal in SBBR treating low C/N domestic sewage. Process. Saf. Environ. Prot. 166, 628–637. doi: 10.1016/j.psep.2022.08.063
Liu, C., Li, C. A., Jiang, Y. Q., Zeng, R. J., Yao, M. J., and Li, X. Z. (2023). A guide for comparing microbial co-occurrence networks. iMeta 2:e71. doi: 10.1002/imt2.71,
Liu, X., Wang, M., Liu, B., Chen, X., An, L., Nie, Y., et al. (2025). Keystone taxa mediate the trade-off between microbial community stability and performance in activated sludges. Nat. Water 3, 723–733. doi: 10.1038/s44221-025-00451-6
Ma, J. Y., Arneth, A., Smith, B., Anthoni, P., Xu, R., Eliasson, P., et al. (2025). Soil nitrous oxide emissions from global land ecosystems and their drivers within the LPJ-GUESS model (v4.1). Geosci. Model Dev. 18, 3131–3155. doi: 10.5194/gmd-18-3131-2025
Ma, B., Wang, Y. L., Ye, S. D., Liu, S., Stirling, E., Gilbert, J. A., et al. (2020). Earth microbial co-occurrence network reveals interconnection pattern across microbiomes. Microbiome 8:82. doi: 10.1186/s40168-020-00857-2,
Ma, X., Wang, J., Zhang, K., Yang, Z., Gao, Y., Wu, A., et al. (2025). Enhanced multifunctionality of sandy loam soil with co-application of biochar and organic manure is driven by microbial network complexity rather than community diversity. Agric. Ecosyst. Environ. 393:109812. doi: 10.1016/j.agee.2025.109812
Mastny, J., Bárta, J., Kastovská, E., and Picek, T. (2021). Decomposition of peatland DOC affected by root exudates is driven by specific r and K strategic bacterial taxa. Sci. Rep. 11:18677. doi: 10.1038/s41598-021-97698-2,
Moreira, G., Dos Reis, J. B., Pires, E. C., Barreto, C. C., and do Vale, H. M. (2025). Fungal guilds reveal ecological redundancy in a post-mining environment. Mining 5:28. doi: 10.3390/mining5020028
Muhammad, I., Lv, J. Z., Wang, J., Ahmad, S., Farooq, S., Ali, S., et al. (2022). Regulation of soil microbial community structure and biomass to mitigate soil greenhouse gas emission. Front. Microbiol. 13:868862. doi: 10.3389/fmicb.2022.868862,
Piontelli, E., Santamaria, M. A. T., and Caretta, G. (1981). Coprophilous fungi of the horse. Mycopathologia 74, 89–105. doi: 10.1007/bf01259464,
Qiu, L. J., Gou, X., Kong, Y. M., Tu, F. Y., Peng, X., Xu, L., et al. (2023). Nitrogen addition stimulates N2O emissions via changes in denitrification community composition in a subtropical nitrogen-rich forest. J. Environ. Manag. 348:119274. doi: 10.1016/j.jenvman.2023.119274
Ren, Y. L., Ge, W., Dong, C. B., Wang, H. Y., Zhao, S., Li, C. L., et al. (2023). Specialist species of fungi and bacteria are more important than the intermediate and generalist species in near-urban agricultural soils. Appl. Soil Ecol. 188:104894. doi: 10.1016/j.apsoil.2023.104894
Shakoor, A., Pendall, E., and Macdonald, C. A. (2025). Microbial mechanisms of interactive climate-driven changes in soil N2O and CH4 fluxes: a global meta-analysis. J. Environ. Manag. 376:124380. doi: 10.1016/j.jenvman.2025.124380,
Shen, J., Tao, Q., Dong, Q., Luo, Y. L., Luo, J. P., He, Y. T., et al. (2021). Long-term conversion from rice-wheat to rice-vegetable rotations drives variation in soil microbial communities and shifts in nitrogen-cycling through soil profiles. Geoderma 404:299. doi: 10.1016/j.geoderma.2021.115299
Su, R., Zhao, D. Y., Zhang, X. M., Guo, Z. X., Cheng, J. X., Yu, Z. B., et al. (2024). Seasonal wet-dry transition and denitrifying communities contributed to nitrous oxide emissions in the water-level fluctuating zone of the largest freshwater lake in China. J. Hydrol. 644:132077. doi: 10.1016/j.jhydrol.2024.132077
Sun, S. J., Qiao, Z. Y., Tikhonenkov, D. V., Gong, Y. C., Li, H., Li, R. H., et al. (2025). Temporal dynamics and adaptive mechanisms of microbial communities: divergent responses and network interactions. Microb. Ecol. 88:94. doi: 10.1007/s00248-025-02596-z,
Uwiragiye, Y., Wang, J., Huang, Y. Y., Wu, L. P., Zhou, J. K., Zhang, Y. H., et al. (2024). Global ecosystem nitrogen cycling reciprocates between land-use conversion and its reversal. Glob. Change Biol. 30:e70385. doi: 10.1111/gcb.17537,
Van Lent, J., Hergoualc’h, K., and Verchot, L. V. (2015). Reviews and syntheses: soil N2O and NO emissions from land use and land-use change in the tropics and subtropics: a meta-analysis. Biogeosciences 12, 7299–7313. doi: 10.5194/bg-12-7299-2015
Wagg, C., Schlaeppi, K., Banerjee, S., Kuramae, E. E., and Van der Heijden, M. G. A. (2019). Fungal-bacterial diversity and microbiome complexity predict ecosystem functioning. Nat. Commun. 10:4841. doi: 10.1038/s41467-019-12798-y,
Wang, J., Uwiragiye, Y., Cao, M. M., Chen, M. Q., Fallah, N., Huang, Y., et al. (2025). Global land use change impacts on soil nitrogen availability and environmental losses. Environ. Sci. Technol. 59, 17595–17605. doi: 10.1021/acs.est.5c03285,
Wang, C., Zhang, Z., Li, J. Y., Shi, R. Y., Xu, R. K., and Shen, R. F. (2024). Rare microbes are associated with crop biomass in an acid soil with different amendments. Plant Soil 504, 79–94. doi: 10.1007/s11104-024-06811-7
Wei, Z. J., Shan, J., Well, R., Yan, X., and Senbayram, M. (2022). Land use conversion and soil moisture affect the magnitude and pattern of soil-borne N2, NO, and N2O emissions. Geoderma 407:115568. doi: 10.1016/j.geoderma.2021.115568
Wen, T., Xie, P., Yang, S., Niu, G., Liu, X., Ding, Z., et al. (2022). ggClusterNet: an R package for microbiome network analysis and modularity-based multiple network layouts. iMeta 1:e32. doi: 10.1002/imt2.32,
Wu, D. M., Ren, C. Q., Ren, D., Tian, Y. J., Li, Y. P., Wu, C. Y., et al. (2023). New insights into carbon mineralization in tropical paddy soil under land use conversion: coupled roles of soil microbial community, metabolism, and dissolved organic matter chemodiversity. Geoderma 432:116393. doi: 10.1016/j.geoderma.2023.116393
Xiao, Z. G., Yang, S. T., Wang, X., Meng, L., Zhu, T. B., Müller, C., et al. (2025). Forest conversion effect on soil nitrogen transformation and nitrous oxide emissions worldwide. Geoderma 459:117373. doi: 10.1016/j.geoderma.2025.117373
Xiong, R., Gao, N., Huang, W., Zhang, X., and Shen, W. (2025). Recent progress in microbial production and consumption of nitrous oxide in agricultural soils. World J. Microbiol. Biotechnol. 41:235. doi: 10.1007/s11274-025-04464-x,
Xiong, R. N., Qian, D., Qiu, Z. J., Hou, Y. X., Li, Q., and Shen, W. S. (2024). Land-use intensification exerts a greater influence on soil microbial communities than seasonal variations in the Taihu Lake region, China. Sci. Total Environ. 943:173630. doi: 10.1016/j.scitotenv.2024.173630
Yang, H. X., Li, Y., Li, J. S., Zhao, X. Y., Zhu, K. J., Zhang, W. R., et al. (2025). Land-use changes alter the soil microbial community structure and key C, N, and S metabolic functional potentials in the Yuncheng salt Lake wetland. Catena 260:109479. doi: 10.1016/j.catena.2025.109479
Yang, T., Tedersoo, L., Liu, X., Gao, G. F., Dong, K., Adams, J. M., et al. (2022). Fungi stabilize multi-kingdom community in a high elevation timberline ecosystem. iMeta 1:e49. doi: 10.1002/imt2.49,
Yessentayeva, K., Reinhard, A., Berzhanova, R., Mukasheva, T., Urich, T., and Mikolasch, A. (2024). Bacterial crude oil and polyaromatic hydrocarbon degraders from Kazakh oil fields as barley growth support. Appl. Microbiol. Biotechnol. 108:189. doi: 10.1007/s00253-024-13010-y,
Yonatan, Y., Amit, G., Friedman, J., and Bashan, A. (2022). Complexity-stability trade-off in empirical microbial ecosystems. Nat. Ecol. Evol. 6, 693–700. doi: 10.1038/s41559-022-01745-8,
Yu, S. T., Chen, X. G., Wang, T. S., Yao, S. H., and Peng, X. H. (2024). Linking main ecological clusters of soil bacterial-fungal networks and nitrogen cycling genes to crop yields under diverse cropping systems in the North China plain. Chem. Biol. Technol. Agric. 11:96. doi: 10.1186/s40538-024-00617-6
Yuan, Y., Dai, X., Wang, H., Xu, M., Fu, X., and Yang, F. (2016). Effects of land-use conversion from double rice cropping to vegetables on methane and nitrous oxide fluxes in southern China. PLoS One 11:e0155926. doi: 10.1371/journal.pone.0155926,
Zeng, Q. C., Lebreton, A., Auer, L., Man, X. W., Jia, L. K., Wang, G. S., et al. (2023). Stable functional structure despite high taxonomic variability across fungal communities in soils of old-growth montane forests. Microbiome 11:217. doi: 10.1186/s40168-023-01650-7,
Zhang, S. Q., Cao, Z. J., Liu, S. Y., Hao, Z. P., Zhang, X., Sun, G. X., et al. (2025). Soil fungal diversity, community structure, and network stability in the southwestern Tibetan plateau. J. Fungi 11:389. doi: 10.3390/jof11050389,
Zhang, H. K., Fang, Y. Y., Chen, Y. C., Li, Y., Lin, Y. X., Wu, J. S., et al. (2022). Enhanced soil potential N2O emissions by land-use change are linked to AOB-amoA and nirK gene abundances and denitrifying enzyme activity in subtropics. Sci. Total Environ. 850:158032. doi: 10.1016/j.scitotenv.2022.158032,
Zhang, X. C., Myrold, D. D., Shi, L. L., Kuzyakov, Y., Dai, H. C., Hoang, D. T. T., et al. (2021). Resistance of microbial community and its functional sensitivity in the rhizosphere hotspots to drought. Soil Biol. Biochem. 161:108360. doi: 10.1016/j.soilbio.2021.108360
Zheng, H. J., Li, Y., Liu, D. Y., Yuan, J. J., Chen, Z. M., Li, J. J., et al. (2025). Wetland restoration following agricultural abandonment supports diversity and reduces stochasticity in soil fungal communities due to anaerobic-induced interspecific associations. Soil Biol. Biochem. 209:109909. doi: 10.1016/j.soilbio.2025.109909
Zheng, N. G., Yu, Y. X., Wang, J., Chapman, S. J., Yao, H. Y., and Zhang, Y. Y. (2020). The conversion of subtropical forest to tea plantation changes the fungal community and the contribution of fungi to N2O production. Environ. Pollut. 265:115106. doi: 10.1016/j.envpol.2020.115106,
Zhou, X., Sun, H., Heinonsalo, J., Pumpanen, J., and Berninger, F. (2022). Microbial biodiversity contributes to soil carbon release: a case study on fire disturbed boreal forests. FEMS Microbiol. Ecol. 98:fiac074. doi: 10.1093/femsec/fiac074,
Zhou, X. G., and Wu, F. Z. (2021). Land-use conversion from open field to greenhouse cultivation differently affected the diversities and assembly processes of soil abundant and rare fungal communities. Sci. Total Environ. 788:147751. doi: 10.1016/j.scitotenv.2021.147751,
Zhu, M. H., Fanin, N., Wang, Q. K., Xu, Z. C., Liang, S., Ye, J., et al. (2024). High functional breadth of microbial communities decreases home-field advantage of litter decomposition. Soil Biol. Biochem. 188:109232. doi: 10.1016/j.soilbio.2023.109232
Zhu, Q., Liu, L., Wang, C., Wan, Y., Yang, R., Mou, J., et al. (2023). Carbon and nitrogen fractions control soil N2O emissions and related functional genes under land-use change in the tropics. Environ. Pollut. 335:122370. doi: 10.1016/j.envpol.2023.122370,
Zhu, K., Lu, Y. C., Ran, H. Y., Zavattaro, L., Liu, Y., Wang, G., et al. (2025). Land use intensity modulates soil C, N and stoichiometric ratios across profiles: the mediating role of pH buffering capacity. Land Degrad. Dev. 1:218. doi: 10.1002/ldr.70218
Keywords: bacterial community, fungal community, land conversion, network complexity, nitrous oxide
Citation: Li C, Zhou Z, Chen X, Tang Q, Zhang Q and Tang J (2026) Shifted microbial network characteristics govern soil N2O emission following paddy-to-vegetable land conversion. Front. Microbiol. 17:1750894. doi: 10.3389/fmicb.2026.1750894
Edited by:
Yanfu Bai, Sichuan Agricultural University, ChinaReviewed by:
Xiao Ru Yang, Chinese Academy of Sciences (CAS), ChinaYalong Kang, Yunnan Agricultural University, China
Copyright © 2026 Li, Zhou, Chen, Tang, Zhang and Tang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Qingbi Zhang, c3dtdXRneXhAMTYzLmNvbQ==; Jieshi Tang, amllc2hpX3RhbmdAMTI2LmNvbQ==