Shrub encroachment enhances AMF network stability and complexity, while cropland destabilizes AMF communities in a subtropical alpine grassland

Abstract

Background

Ecosystem conversion, primarily driven by agricultural expansion, has profoundly altered ecosystem structure and function. Grasslands, characterized by deep, nutrient-rich soils that support high soil carbon content, are particularly vulnerable to conversion for agricultural purposes. This transformation significantly impacts soil microbial communities, yet the effects of such changes on the stability and complexity of arbuscular mycorrhizal fungi (AMF) networks remain poorly understood, particularly in subtropical alpine grasslands.

Results

In this study, we investigated how the conversion of natural grasslands into shrublands, artificial woodlands, and croplands affects AMF communities in a subtropical alpine region of China. Our results demonstrate that shrub encroachment increased AMF diversity by up to 25%, and enhanced network modularity and robustness by approximately 20% and 25%, respectively, compared with natural grasslands. This phenomenon may be partially attributed to deep root-mediated niche diversification and the alleviation of soil disturbance in shrubland. In contrast, conversion to cropland decreased AMF diversity by nearly 40%, destabilizing microbial networks due to increased nutrient enrichment and mechanical disturbance. Although the dominant genera Glomus and Paraglomus persisted across all systems, their relative abundance shifted (e.g., a 10–15% reduction of Glomus in croplands). Soil organic carbon, nitrogen, and phosphorus collectively explained up to 89.7% of the variation in AMF network complexity.

Conclusions

These findings address the critical knowledge gap identified in the background regarding AMF responses to land-use changes in subtropical alpine grasslands. By demonstrating that shrub encroachment enhances soil fertility and AMF network stability—contrasting sharply with cropland conversion—our results highlight the importance of preserving natural succession processes to maintain microbial-driven ecosystem functions. This aligns with global efforts to mitigate grassland degradation and supports sustainable management practices in vulnerable alpine regions.

Background

Grasslands are vital terrestrial ecosystems that contribute significantly to global carbon sequestration, nutrient cycling, and biodiversity conservation, while providing essential services such as forage for livestock and wild herbivores [1, 2]. However, these ecosystems have undergone extensive anthropogenic transformations, particularly through conversion into shrublands, artificial woodlands, and croplands [3, 4]. These land-use changes are expected to continue over the coming century, posing significant risks to the extent and integrity of natural grasslands globally [5]. In China’s southwestern subtropical alpine regions, for example, approximately 70% of newly cultivated croplands and over 20% of emerging shrublands between 2009 and 2022 have originated from natural grasslands (https://www.forestry.gov.cn/). Such large-scale transformations have been shown to significantly alter soil microbial communities, especially across different soil layers, with consequential effects on ecosystem functions and soil health [6].

Arbuscular mycorrhizal fungi (AMF) are a group of obligate symbiotic fungi that colonize plant roots and play a central role in terrestrial ecosystem functioning [7]. By facilitating nutrient exchange—particularly phosphorus and nitrogen—between soil and host plants, AMF enhance plant growth, regulate soil structure, and stabilize microbial networks [8, 9]. Their sensitivity to environmental change makes them ideal bioindicators for assessing the ecological impacts of land-use transitions. Yet, the mechanisms underlying AMF community responses to such transitions in subtropical alpine ecosystems remain insufficiently understood.

Despite extensive research on arbuscular mycorrhizal fungi in temperate [10, 11] and tropical grasslands [12, 13], the effects of land-use change on AMF diversity, network stability, and ecosystem functions remain poorly understood in subtropical alpine systems. These high-altitude ecosystems are characterized by distinct edaphic and climatic conditions, as well as unique plant–microbe interactions that may deviate from patterns observed elsewhere [14]. Previous studies have shown that shrub encroachment can increase AMF diversity and biomass by promoting plant functional heterogeneity and enhancing root-derived carbon inputs [15, 16]. Whereas agricultural intensification tends to homogenize soil microbial communities and diminish mycorrhizal associations through frequent tillage and excessive nutrient input [17,18,19]. However, such findings are often derived from lowland or semiarid regions, with limited transferability to fragile alpine systems where edaphic constraints and ecological thresholds are more pronounced [20, 21]. Moreover, existing literature has largely focused on species richness or community composition [10, 12, 22], with insufficient attention paid to AMF co-occurrence network properties—key indicators of microbial resilience and ecosystem multifunctionality. It remains unclear how different land-use trajectories, particularly shrub encroachment versus cropland conversion, influence the structural complexity and robustness of AMF networks in subtropical alpine grasslands. Addressing this gap is essential for understanding the mechanisms driving microbial responses to disturbance and for informing land management strategies that support soil health and biodiversity in vulnerable montane environments.

Furthermore, soil depth is a critical determinant of microbial community dynamics, as it governs vertical gradients in resource availability, root distribution, and microhabitat heterogeneity [23, 24]. In alpine ecosystems, shallow soils are typically nutrient-poor and susceptible to erosion, whereas deeper layers may retain moisture and organic matter but face constraints from limited root penetration and microbial activity [25]. These depth-dependent variations influence AMF symbiosis by modulating host plant access to nutrients and water [26]. For instance, topsoil soils exhibit higher microbial biomass due to concentrated root exudates and litter inputs, while subsoil layers may harbor distinct AMF taxa adapted to resource scarcity and physical isolation [26]. However, research on the vertical differentiation patterns of AMF communities in alpine ecosystems, particularly the interactions between AMF communities across different soil layers, remains limited. This gap hinders a comprehensive understanding of the mechanisms underlying the maintenance of multifunctionality in soil–plant systems mediated by AMF.

To fill these knowledge gaps, this study investigates how different land-use trajectories—namely natural grassland, shrub encroachment, artificial afforestation, and cropland conversion—affect the diversity, composition, and co-occurrence network complexity of AMF communities in a subtropical alpine grassland of Southwest China. By integrating high-throughput sequencing, soil physicochemical profiling, and molecular ecological network analysis across multiple soil layers, we aim to unravel the ecological mechanisms that drive AMF responses to land-use change. Specifically, we ask: (1) how do AMF community diversity and network properties vary across land-use types and soil depths; (2) what are the dominant environmental drivers underlying these patterns; and (3) to what extent do natural succession and anthropogenic disturbance differentially shape microbial network stability and resilience. These insights will not only enhance our understanding of AMF-mediated ecosystem processes in fragile alpine environments but also provide scientific evidence to support sustainable land-use planning and ecological restoration in subtropical grassland systems.

Results

Soil properties

Conversion of natural grassland to shrubland (SL), artificial woodland (AL), and cropland (CL) significantly altered soil physicochemical properties (Fig. 1 and Additional File 1: Table S1). Compared to natural grassland, soil water content (SWC), total phosphorus (TP), total organic carbon (TOC), dissolved organic carbon (DOC), total nitrogen (TN), ammonium nitrogen (NH4+-N), and available potassium (AK) were significantly higher in shrubland (all P < 0.05). In contrast, SWC, TOC, DOC, and TN were significantly lower in cropland (all P < 0.05). Soil pH, available phosphorus (AP), and nitrate nitrogen (NO3−) were markedly higher in artificial woodland soils than in natural grassland (all P < 0.05), and exceeded those in shrubland and cropland (all P < 0.05). Soil nutrient levels consistently declined with increasing soil depth, except for total carbon (TC) (Additional File 1: Table S1). Strong correlations were observed among soil moisture, pH, carbon, nitrogen, and phosphorus contents (Additional File 1: Table S2).

[IMAGE OMITTED: SEE PDF]

AMF community composition

Alpha diversity analysis of AMF communities, rarefied to a total of 3,127,834 sequences, captured most of the observed amplicon sequence variant (ASV) richness. Compared to natural grassland, the Simpson, Chao1, and ACE indices indicated higher AMF species diversity in shrubland and artificial woodland (P < 0.05) and lower diversity in cropland (P < 0.05; Fig. 2A). AMF species diversity significantly declined with increasing soil depth (P < 0.05; Fig. 2A). All alpha diversity indices aligned in showing SL > AL > GL > CL (P < 0.05). Simpson and Chao1 were emphasized to represent dominant taxon evenness and total richness, respectively.

[IMAGE OMITTED: SEE PDF]

Ecosystem conversion had a pronounced impact on AMF community structure, as indicated by principal coordinates analysis (PCoA). The four ecosystems were clearly separated, with significantly greater dispersion between natural grassland and shrubland or cropland systems, but smaller dispersion between natural grassland and artificial woodland (Fig. 2B). Generally, smaller differences in AMF community structure were observed across soil depths (Fig. 2B).

As hypothesized, ecosystem conversion significantly shaped AMF community structure (Fig. 3). Dominant genera Glomus and Paraglomus showed significant abundance shifts (P < 0.05), with Glomus declining by 10–15% in croplands and Paraglomus increasing 8–12% in shrublands from surface soil to deep soil (0–30 cm, Fig. 3A, Additional File 2: Table S3). Compared to natural grassland, the relative abundance of the dominant genus Glomus decreased in shrubland and cropland, while Paraglomus increased (Fig. 3A). The relative abundance of dominant AMF species increased with soil depth in shrubland, whereas it decreased in cropland (Fig. 3A). AMF taxa were classified at the genus and species levels, and biomarker AMF taxa were ranked by their importance to model accuracy (Additional File 1: Fig. S1). Venn diagrams illustrated unique ASVs for each ecosystem type and soil depth (Fig. 3B). Unique ASVs were more abundant in shrubland (685), artificial woodland (968), and cropland (1064) than in grassland (615). At the 0–10 cm soil depth, unique ASVs (1383) outnumbered those at 10–20 cm (926) and 20–30 cm (735).

[IMAGE OMITTED: SEE PDF]

Across the four ecosystem types and three soil depths, 23 and 16 biomarkers were identified (log10 LDA score > 4.0), representing significant differences in species (Fig. 3C). Random forest analysis ranked AMF taxa by importance to model accuracy, with Paraglomus, Glomus, Acaulospora, and Paraglomus being the most important genera in natural grassland, shrubland, artificial woodland, and cropland, respectively. Relative abundances of Paraglomus and Glomus were significantly higher than other taxa in all ecosystems (Additional File 1: Fig. S2, Additional File 2: Table S3).

AMF network community structure

Molecular ecological networks (MENs) were constructed for each ecosystem type and soil depth, yielding seven distinct networks (Fig. 4). Overall topological properties revealed significant differences in microbial co-occurrence patterns across ecosystems and soil depths (Additional File 1: Table S4 and Additional File 2: Table S5). Compared to natural grassland, modularity values, connectivity, and average K were significantly higher in shrubland (P < 0.001) and significantly lower in cropland (P < 0.05). No significant difference was observed between natural grassland and artificial woodland for these indicators (P > 0.05; Additional File 2: Table S5). Based on within-module connectivity (Zi) and among-module connectivity (Pi), 10 module hubs, 11 network hubs, and 72 connector hubs were identified across ecosystems and soil depths (Fig. 4B and D, Additional File 2: Table S6). Shrubland had significantly more “hub” nodes than natural grassland. Key species details are listed in Table S5 (Additional File 2: Table S5). All key species in these networks were dominant genera, including Glomus, Paraglomus, and Claroideoglomus (Additional File 2: Table S6).

[IMAGE OMITTED: SEE PDF]

Network robustness analyses showed significantly greater robustness in shrubland than in grassland (P < 0.05), and significantly lower in cropland (P < 0.05) when 50% of module hubs were removed (Fig. 5A). Robustness significantly decreased with soil depth (Fig. 5C). Conversely, network vulnerability was lower in shrubland than in grassland but greater than in cropland. Network stability (robustness and vulnerability) and complexity (AvgK, AvgCC, Links, and Con) exhibited significant positive correlations (Fig. 5E). Robustness was positively correlated with AvgK, AvgCC, Links, and Con, but negatively correlated with HD, GD, and RM. In contrast, vulnerability was negatively correlated with AvgK but positively correlated with HD and GD (Fig. 5E and Additional File 2: Table S6).

[IMAGE OMITTED: SEE PDF]

Drivers of soil AMF community stability and complexity

Multiple regression and random forest analyses revealed interdependencies between soil environmental factors, core AMF species abundance and diversity, and community stability and complexity (Fig. 6). Soil physicochemical properties had a relatively lower impact on core AMF communities in natural grassland but a higher impact in cropland (Fig. 6A). Soil variables (DOC, TOC, TN, AK, NH₄⁺, AP, TC, MBC, TP, SWC) were positively correlated with robustness, Nodes, RM, HD, and GD, but negatively correlated with AvgK, Links, AvgCC, and Con (Fig. 6B). Complexity and stability indices were strongly interrelated (Additional File 1: Table S7). Soil properties explained 78.5% and 89.7% of the variation in AMF network complexity and stability, respectively (Fig. 6C). AP emerged as the most important predictor for AMF network complexity, while pH was the key predictor for stability. Redundancy analysis (RDA) showed that the first two axes explained 27.68% and 19.72% of the variance in soil AMF community composition (Fig. 6D).

[IMAGE OMITTED: SEE PDF]

Discussion

Effects of ecosystem conversion on soil physicochemical properties

The conversion of natural grasslands to shrublands, artificial woodlands, and croplands significantly alters soil physicochemical properties, reflecting the profound impacts of land-use changes on soil health and ecosystem function. Our study shows that shrub encroachment notably increases soil water content (SWC), total phosphorus (TP), total organic carbon (TOC), dissolved organic carbon (DOC), total nitrogen (TN), ammonium nitrogen (NH4+-N), and available potassium (AK) compared to natural grasslands (Fig. 1 and Additional File 1: Table S1). Conversely, the transformation of grasslands into croplands results in marked reductions in SWC, TOC, DOC, and TN. Artificial woodlands show intermediate changes, with increases in soil pH, available phosphorus (AP), and nitrate nitrogen (NO3−-N) compared to natural grasslands, shrublands, and croplands.

In this study region, shrubs generally possess deeper and more extensive root systems than herbaceous plants, improve water retention, and reduce soil erosion [27]. This deeper rooting enhances SWC and facilitates nutrient uptake and redistribution across the soil profile. On the other hand, croplands, due to intensive tillage and crop harvesting, lose soil water and suffer structural degradation, leading to decreased SWC [28]. Higher TOC in shrub-encroached soils likely stems from rapid nutrient cycling and the high recalcitrance of shrub litter, promoting long-term carbon sequestration [29, 30]. In contrast, cropland conversion results in elevated nitrate (NO3−-N) and available phosphorus (AP) levels, largely due to fertilization. The sharp increase in soil pH in croplands reflects nitrogen accumulation, which can alter soil chemistry and microbial dynamics [31].

These differences highlight the contrasting dynamics in soil properties, underscore the differential impacts of natural successional processes versus anthropogenic disturbances. Shrub encroachment promotes soil resilience by increasing organic carbon and nutrient retention, while cropland conversion risks nutrient depletion and soil degradation, disrupting C:N:P stoichiometry. This shift in stoichiometry can reduce microbial carbon-use efficiency and accelerate soil organic matter mineralization [31], reshaping microbial communities and affecting soil nutrient status and ecosystem functioning [30, 32].

Responses of AMF community structure to natural grassland conversion

Arbuscular mycorrhizal fungi (AMF) play an essential role in maintaining ecosystem sustainability [9]. Our findings show that AMF diversity is higher in shrub-encroached areas compared to natural grasslands, while croplands exhibit a significant reduction in AMF diversity (Fig. 2A). The increase in AMF diversity under shrub encroachment can be attributed to shrub’s different plant functional traits, such as root architecture and carbon allocation, which create diverse microhabitats for AMF [33, 34]. Enhanced litter input and organic matter in shrublands provide a richer substrate for AMF colonization and growth, fostering a more diverse and robust AMF community [8]. Additionally, nitrogen-fixing shrubs enrich soil nitrogen levels in subsoil layers (10–20 cm and 20–30 cm, Additional File 1: Table S1) and soil total nitrogen (Fig. 1C), which can alleviate nitrogen limitation, promoting a more diverse AMF community [35]. In contrast, the conversion of grasslands to croplands leads to a significant reduction in AMF diversity. Intensive agricultural practices, including fertilization and frequent tillage, disrupt AMF hyphal networks and diminish the mutualistic relationships between AMF and host plants [36, 37]. High levels of available nutrients, especially nitrogen and phosphorus, can suppress AMF colonization by reducing the plant’s dependence on mycorrhizal associations [17]. Furthermore, repeated mechanical disturbances fragment AMF networks, limiting their ability to form extensive and resilient mycelial networks necessary for sustaining diverse AMF communities.

In this study region, characterized by shallow and rocky soils, nutrient availability is naturally limited, especially at greater soil depths [32]. This limitation further reduces AMF diversity in subsurface layers, as observed in our study [38]. Despite these challenges, the dominant genus Glomus remained prevalent across all ecosystem types, followed by Paraglomus, consistent with prior reports from various grassland ecosystems [39, 40]. The persistence of these core AMF taxa highlights their functional importance and resilience to land-use changes. Linear discriminant analysis effect size (LEfSe) identified Glomus and Paraglomus as key biomarkers across ecosystems, underscoring their pivotal roles in maintaining AMF community structure and function [8, 9]. These findings suggest that while land-use changes significantly affect AMF diversity and community structure, certain core AMF taxa demonstrate resilience, likely due to their broad ecological niches and functional versatility [41]. This resilience offers opportunities for ecosystem restoration and AMF community recovery through targeted conservation efforts. By reducing agricultural intensification or promoting natural succession processes, AMF assemblages can be restored, contributing to the overall health and resilience of grassland ecosystems [42, 43].

Effects of ecosystem conversion on AMF community complexity and stability

Our multiple regression and random forest analyses reveal complex interdependencies between soil environmental factors, core AMF species abundance and diversity, and community stability and complexity (Fig. 6). Variables such as DOC, TOC, TN, AK, NH₄⁺, AP, microbial biomass carbon (MBC), and TP are positively correlated with AMF network robustness, while high nutrient levels and soil moisture may disrupt network structures. Excessive nutrient enrichment and moisture can decrease AMF network complexity, reducing functional efficiency and increasing vulnerability [44, 45]. The strong interrelationship between complexity and stability indices underscores the interconnected nature of these network properties, where enhanced complexity typically promotes greater stability and vice versa [46]. Previous studies have shown that soil nitrogen (N) and phosphorus (P) are the most significant predictors of AMF community and network structure in temperate grasslands, arid and semi-arid grasslands, and alpine meadows [10, 47,48,49]. Elevated nutrient levels, typically resulting from fertilization, may influence AMF community composition by favoring specific species, which in turn affects network interactions and complexity [50]. However, our results indicate that in subtropical alpine grasslands, soil pH, in addition to N and P, is a key predictor of AMF community network stability.

Soil pH, as a fundamental determinant of soil chemistry, governs nutrient availability and microbial activity [51] and may, therefore, play a crucial role in shaping the overall stability of AMF communities. In our study area, soils are generally weakly acidic, but the unique karst geology and rock weathering processes in the region make soil pH highly susceptible to change [38]. This variability likely contributes to the observed differences in the drivers of AMF community and network structure between subtropical alpine grasslands and other ecosystems, including temperate grasslands, arid and semi-arid grasslands, and alpine meadows [11, 52, 53]. Elevated nutrient levels, particularly those resulting from fertilization, may further interact with pH fluctuations, influencing AMF community composition by selectively favoring certain species (Figs. 1 and 3). Specifically, changes in pH can alter the bioavailability of nutrients like phosphorus and nitrogen, which are essential for AMF growth and symbiotic interactions with plants [54]. For instance, lower pH often enhances the availability of these nutrients but may also suppress certain AMF species that are less tolerant of acidic conditions, thereby shifting community structure [13, 55]. Conversely, neutral or slightly alkaline pH levels may favor a different subset of AMF species adapted to those conditions [54]. In summary, these findings emphasize the critical role of soil pH in shaping AMF community structure in subtropical alpine grasslands, underscoring the distinctiveness of subtropical ecosystems. Future integrated soil management strategies should, therefore, account for the complex interactions between soil nutrients, pH, and microbial dynamics to ensure the long-term health and resilience of these ecosystems.

Broader implications and recommendations

Our findings emphasize the importance of shrub encroachment in supporting diverse and stable AMF communities within subtropical alpine grasslands. Given the degradation of subtropical alpine grasslands in southwestern China due to agricultural intensification and climate change, promoting targeted shrub establishment may offer an effective strategy for ecological restoration and soil improvement [30, 56]. Shrub-dominated landscapes enhance soil carbon stocks, improve soil structure, and strengthen microbial resilience, contributing to climate change adaptation and mitigation efforts [57, 58]. However, shrub expansion must be carefully managed to prevent biodiversity loss and maintain ecosystem balance.

Recent studies have shown continued encroachment results in substantial loss of herbaceous diversity at medium and high extents, with a loss of richness that is not replaced [59]. For AM-plant symbionts, changes in plant diversity are likely to result in shifts in AMF colonization and communities [8]. Nevertheless, shrub encroachment has increased the diversity and network complexity of AMF in our study area, which may be attributed to alterations in soil structure and nutrient dynamics caused by the invasion. Deep-rooted shrubs are likely to alter soil structure, providing conditions for soil microorganisms to inhabit deeper layers, thereby inducing changes in microbial structure and function [60]. Similar studies have shown that shrub invasion directly leads to increased soil nutrients, enzyme activity, and microbial biomass in deeper soil layers [61], which is consistent with our observation that nutrient levels at 10–20 cm depth exceed those at 0–10 cm (Additional File 1: Table S1). However, further research is needed to understand how shrub invasion influences AMF through changes in plant communities, a key avenue for future exploration of AMF community structure and functional responses to ecosystem type transitions.

From a global perspective, enhancing AMF diversity and network complexity holds considerable potential for improving soil health, increasing nutrient retention, and sequestering carbon in degraded ecosystems [44]. AMF play a pivotal role in carbon sequestration in ecosystems, particularly in subtropical alpine grasslands where land-use conversion can significantly alter soil properties and microbial communities [62]. As the study suggests, shrub encroachment not only enhances soil nutrient status and AMF diversity but also potentially contributes to increased carbon storage in the soil. AMF facilitate carbon cycling by forming symbiotic relationships with plant roots, enhancing plant growth and root biomass [63]. This interaction leads to greater carbon allocation into the soil through root exudates and dead plant material, thereby improving soil organic carbon stocks.

Long-term research integrating soil microbial ecology with management interventions is essential to refine restoration approaches across diverse environmental conditions. Such research should focus on understanding the mechanisms underlying AMF modularity and resilience in shrub-encroached systems, as well as the long-term impacts of various land-use practices on AMF community dynamics. Additionally, the development of land-use policies promoting sustainable agro-pastoral management practices can aid in the restoration and conservation of AMF communities, thereby safeguarding soil ecosystem functions and enhancing the resilience of grassland ecosystems.

Conclusions

This study highlights the significant effects of land-use conversion on soil properties and AMF communities in subtropical alpine grasslands. Shrub encroachment enhances soil nutrient status, AMF diversity, and network complexity, while cropland conversion leads to declines in soil quality and AMF diversity. Soil properties, particularly available phosphorus and pH, are key drivers of AMF community structure. Promoting shrub encroachment could improve soil nutrient status and microbial resilience, supporting the sustainability of the subtropical alpine ecosystem. However, managing agricultural intensification is crucial to preserve soil microbial integrity. These insights inform conservation strategies and sustainable land-use policies for ecosystem restoration in the face of ongoing environmental change.

Methods

Study site and sampling

The study was conducted in a natural subalpine grassland located in Longli County on the Guizhou Plateau, Southwest China (106°51′−106°52′ E, 26°19′−26°20′ N). Over the past 30 years, this region has experienced increasing anthropogenic disturbances, resulting in a mosaic of land-use types, including agricultural cultivation and afforestation. The area has a subtropical monsoon climate characterized by hot, wet summers and concentrated rainfall during June–July and September–October. The mean daily maximum and minimum temperatures are 22.68 ± 2.4 °C and 6.05 ± 1.02 °C, respectively, with an annual precipitation of 1100 ± 10.8 mm over a 50-year period (1970–2020) [4].

In August 2023, four representative land-use types were selected based on vegetation characteristics and land-use history: natural grassland (GL), shrubland (SL), artificial woodland (AL), and cropland (CL). GL: The grassland in this region has remained open. The plant community is dominated by AMF-host herbaceous species, mainly Poaceae and Asteraceae, with over 95% vegetation cover. SL: developed through natural succession of abandoned grassland after 2005, following a decline in local grazing pressure. This allowed woody species such as Rosaceae and Leguminosae to gradually establish and dominate. The current canopy closure is about 60%, and the vegetation retains an abundant reservoir of AMF-host plant species. AL: was established in 2010 as part of a national reforestation policy (“Grain for Green” program), converting grassland into coniferous plantations. Dominant tree species include Pinus massoniana, planted at 3 × 3 m spacing and subject to periodic thinning (last conducted in 2020). The understory’s herbaceous vegetation supports AMF symbiosis. CL: has been under continuous cultivation since 1998, following mechanical plowing of grassland to a depth of ~ 20 cm. A maize–wheat rotation system has been practiced over the past 10 years, with average fertilization rates of 19–22 g m−2 years−1 for nitrogen, 6–10 g m−2 years−1 for phosphorus, and 9–12 g m−2 years−1 for potassium (https://stjj.guizhou.gov.cn/). Both maize and wheat are AMF-host crops, although repeated tillage and fertilization may constrain AMF colonization.

For each land-use type, four 10 × 10 m plots were established with a minimum spacing of 2 m. Soil samples were collected at 0–10 cm, 10–20 cm, and 20–30 cm depths using a 2.5-cm-diameter soil corer. Each sample was a composite of five subsamples collected in a Z-pattern and sieved through a 2-mm mesh to remove roots, stones, and debris. A total of 48 composite soil samples (4 land-use types × 3 depths × 4 replicates) were obtained. Each sample was divided into two portions: one stored at 4 °C for DNA extraction and the other air-dried at room temperature for physicochemical analyses.

Soil properties

Soil properties were analyzed for all replicates, with three technical replicates per sample. Total carbon (TC), total organic carbon (TOC), and total nitrogen (TN) were quantified using an elemental analyzer (Vario EL cube, Elementar, Langenselbold, Germany). To determine TOC, samples were first acidified with 2 M HCl to remove inorganic carbon prior to analysis. DOC was measured in K2SO4 extracts (12.5 g of lyophilized soil was extracted with 50 ml of 0.5 M K2SO4) using an organic carbon analyzer (vario TOC, Elementar, Langenselbold, Germany). Total phosphorus (TP) was measured using the molybdate colorimetric method after digestion with perchloric acid and reduction with ascorbic acid. Soil water content (SWC) was determined by oven-drying for 48 h at 65 °C or until the weight no longer changed. Soil pH was measured in a 1:5 soil:water (w/v) suspension using a pH meter. Ammonium (NH4+) and nitrate (NO3−) concentrations were assessed using a continuous flow analyzer (Alliance Flow Analyzer; Futura, Frépillon, France) after extraction with 2 M KCl. Microbial biomass carbon (MBC) was determined via the chloroform fumigation method (Makarov et al., 2015). Available phosphorus (AP) and potassium (AK) were quantified colorimetrically and by flame photometry, respectively, after appropriate extractions.

DNA extraction and sequencing data processing for AMF

DNA was extracted from 0.5 g of soil using a bead mill and SDS lysis method combined with the MoBio PowerSoil DNA Isolation Kit (MoBio Laboratories, a QIAGEN company, Carlsbad, CA, USA). AMF community profiling involved amplifying the 18S rRNA gene using the primer pair AMV4.5NF (5′-AAGCTCGTAGTTGAATTTCG-3′) and AMDGR (5′-CCCAACTATCCCTATTAATCAT-3′). A two-step PCR protocol was employed to minimize amplification bias, conducted as follows: initial denaturation at 94 °C for 3 min, followed by 30 cycles of 94 °C for 45 s, 45 °C for 60 s, 72 °C for 60 s, and a final extension at 72 °C for 10 min [64]. PCR products were pooled, quantified using PicoGreen, and sequenced on an Illumina MiSeq platform by Beijing Personal Biotechnology Co., Ltd. (Beijing, China). Bioinformatic processing was performed using QIIME 2 2019.4, following modifications provided in the official tutorials (https://docs.qiime2.org/2019.4/tutorials/, accessed on 10 October 2022).

Data analysis

Soil physicochemical properties were analyzed using ANOVA followed by Tukey’s HSD test. Estimates of α-diversity indices (Shannon, Simpson, ACE, and Chao1) were calculated and compared across ecosystems and depths via Kruskal–Wallis tests using R (version 4.3.2). Differential abundance of dominant genera was assessed using likelihood ratio tests (FDR-adjusted P < 0.05). Principal coordinate analysis (PCoA) was employed to assess the structural variation of AMF communities across different soil depths, utilizing Bray–Curtis distance metrics via the “vegan” R package. Unique and shared amplicon sequencing variants (ASVs) across each study system and soil layer were illustrated through Venn and UpSet diagrams using the “VennDiagram” and “UpSetR” packages, respectively. Additionally, differential ASVs abundance between ecosystems (utilizing the same thresholded ASVs tables) was analyzed using likelihood ratio tests (LRT) with the “edgeR” package.

Correlation networks were constructed for each land-use type and soil depth, totaling six networks per site, through the Molecular Ecological Network Analysis Pipeline (MENAP) hosted by the Institute for Environmental Genomics, University of Oklahoma (http://ieg4.rccc.ou.edu/MENA/). The random matrix theory (RMT)-based approach distinguished system-specific, non-random associations from random ones, thereby producing networks robust against random noise. Each network was independently constructed, with nodes representing ASVs and edges denoting tentative associations based on correlations among the abundance profiles of connected nodes. These networks were visualized using Gephi 0.9.2. Robustness was measured by randomly removing 50% of taxa from each of the empirical molecular ecological networks (MENs) and by removing 50% of module hubs from each of the empirical MENs.

Distance-based redundancy analysis (db-RDA) and Mantel tests were utilized to explore correlations within microbial communities in relation to soil properties, analyzing relative abundances of ASVs through the “vegan” and “LinkET” packages, respectively. Pearson’s correlation coefficient measured the relationships between microbial community diversity and soil properties. Subsequently, a random forest analysis was conducted to ascertain the importance of each soil parameter on AMF community composition, employing mean squared error assessments via the “randomForest” package.

Data availability

Raw sequence data are available at the National Center for Biotechnology Information, project number: PRJNA1266346.

Abbreviations

AMF:

Arbuscular mycorrhizal fungi

GL:

Natural grassland

SL:

Shrubland

AL:

Artificial woodland

CL:

Cropland

SWC:

Soil water content

TC:

Total carbon

TOC:

Total organic carbon

DOC:

Dissolved organic carbon

MBC:

Microbial biomass carbon

TN:

Total nitrogen

NH4 +-N:

Ammonium nitrogen

NO3 −-N:

Nitrate nitrogen

TP:

Total phosphorus

AK:

Available potassium

AP:

Available phosphorus

PCR:

Polymerase chain reaction

ASVs:

Amplicon sequence variant

ANOVA:

Analysis of Variance

db-RDA :

Distance-based redundancy analysis

PCoA:

Principal coordinate analysis

MENs:

Molecular ecological networks

References

Bai Y, Cotrufo MF. Grassland soil carbon sequestration: Current understanding, challenges, and solutions. Science. 2022;377(6606):1–8.

Borer ET, Harpole WS, Adler PB, Arnillas CA, Bugalho MN, Cadotte MW, Caldeira MC, Campana S, Dickman CR, Dickson TL, et al. Nutrients cause grassland biomass to outpace herbivory. Nat Commun. 2020;11(1):6036.

Peng Z, Qian X, Liu Y, Li X, Gao H, An Y, Qi J, Jiang L, Zhang Y, Chen S, et al. Land conversion to agriculture induces taxonomic homogenization of soil microbial communities globally. Nat Commun. 2024;15(1):3624.

Ding L, Shang Y, Zhang W, Zhang Y, Li S, Wei X, Zhang Y, Song X, Chen X, Liu J et al: Disentangling the effects of driving forces on soil bacterial and fungal communities under shrub encroachment on the Guizhou Plateau of China. Sci Total Environ 2020, 709.

Soliveres S, Maestre FT, Eldridge DJ, Delgado-Baquerizo M, Quero JL, Bowker MA, Gallardo A. Plant diversity and ecosystem multifunctionality peak at intermediate levels of woody cover in global drylands. Global Ecol Biogeogr. 2014;23(12):1408–16.

Louisson Z, Hermans SM, Buckley HL, Case BS, Taylor M, Curran-Cournane F, Lear G. Land use modification causes slow, but predictable, change in soil microbial community composition and functional potential. Environ Microbiome. 2023;18(1):30.

van der Heijden MGA, Bardgett RD, Van Straalen NM. The unseen majority: Soil microbes as drivers of plant diversity and productivity in terrestrial ecosystems. Ecol Lett. 2008;11(3):296–310.

Lanfranco L, Fiorilli V, Gutjahr C. Partner communication and role of nutrients in the arbuscular mycorrhizal symbiosis. New Phytol. 2018;220(4):1031–46.

Powell JR, Rillig MC. Biodiversity of arbuscular mycorrhizal fungi and ecosystem function. New Phytol. 2018;220(4):1059–75.

Zhang C, Xiang X, Yang T, Liu X, Ma Y, Zhang K, Liu X, Chu H. Nitrogen fertilization reduces plant diversity by changing the diversity and stability of arbuscular mycorrhizal fungal community in a temperate steppe. Sci Total Environ. 2024;918: 170775.

Kohout P, Doubková P, Bahram M, Suda J, Tedersoo L, Voříšková J, Sudová R. Niche partitioning in arbuscular mycorrhizal communities in temperate grasslands: a lesson from adjacent serpentine and nonserpentine habitats. Mol Ecol. 2015;24(8):1831–43.

Rodríguez-Echeverría S, Teixeira H, Correia M, Timóteo S, Heleno R, Öpik M, Moora M. Arbuscular mycorrhizal fungi communities from tropical Africa reveal strong ecological structure. New Phytol. 2017;213(1):380–90.

Li X, Qi Z, Yu X, Xu M, Liu Z, Du G, Yang Y. Soil pH drives the phylogenetic clustering of the arbuscular mycorrhizal fungal community across subtropical and tropical pepper fields of China. Appl Soil Ecol. 2021;165: 103978.

Hu A, Wang J, Sun H, Niu B, Si G, Wang J, Yeh C-F, Zhu X, Lu X, Zhou J, et al. Mountain biodiversity and ecosystem functions: interplay between geology and contemporary environments. ISME J. 2020;14(4):931–44.

Ambrosino ML, Torres YA, Garayalde AF, Armando LV, Lorda GS, Velázquez MS. Shrub effect in semiarid Monte rangelands: Variation of arbuscular mycorrhizal fungi communities associated with different functional groups of plants. Appl Soil Ecol. 2025;209: 106000.

Zhao Y-D, Hu X, Pan P-Y. Positive feedback relationship between shrub encroachment and arbuscular mycorrhizal fungi in the Inner Mongolia grassland of northern China. Appl Soil Ecol. 2022;177: 104525.

Balami S, Vasutova M, Godbold D, Kotas P, Cudlin P: Soil fungal communities across land use types. iForest 2020, 13:548–558.

Ciccolini V, Ercoli L, Davison J, Vasar M, Opik M, Pellegrino E: Land-use intensity and host plant simultaneously shape the composition of arbuscular mycorrhizal fungal communities in a Mediterranean drained peatland. FEMS Microbiol Ecol 2016, 92(12).

Sepp S-K, Jairus T, Vasar M, Zobel M, Opik M. Effects of land use on arbuscular mycorrhizal fungal communities in Estonia. Mycorrhiza. 2018;28(3):259–68

Xu M, Li X, Cai X, Li X, Christie P, Zhang J. Land use alters arbuscular mycorrhizal fungal communities and their potential role in carbon sequestration on the Tibetan Plateau. Sci Rep. 2017;7(1):3067.

Chen K, Zhang J, Muneer MA, Xue K, Niu H, Ji B. Plant community and soil available nutrients drive arbuscular mycorrhizal fungal community shifts during alpine meadow degradation. Fungal Ecol. 2023;62: 101211.

Vogelsang KM, Reynolds HL, Bever JD. Mycorrhizal fungal identity and richness determine the diversity and productivity of a tallgrass prairie system. New Phytol. 2006;172(3):554–62.

He H, Xu M, Li W, Chen L, Chen Y, Moorhead DL, Brangarí AC, Liu J, Cui Y, Zeng Y, et al. Linking soil depth to aridity effects on soil microbial community composition, diversity and resource limitation. CATENA. 2023;232: 107393.

Philipp L, Sünnemann M, Schädler M, Blagodatskaya E, Tarkka M, Eisenhauer N, Reitz T. Soil depth shapes the microbial response to land use and climate change in agroecosystems. Appl Soil Ecol. 2025;209: 106025.

Donhauser J, Frey B: Alpine soil microbial ecology in a changing world. FEMS Microbiol Ecol 2018, 94(9).

Wang Z, Zhao J, Xiao D, Chen M, He X. Higher colonization but lower diversity of root-associated arbuscular mycorrhizal fungi in the topsoil than in deep soil. Appl Soil Ecol. 2024;194: 105195.

Liu M, Li D, Hu J, Liu D, Ma Z, Cheng X, Zhao C, Liu Q: Altitudinal pattern of shrub biomass allocation in Southwest China. PLoS One 2020, 15(10).

Li K, Yang J, Wang J, Wang Z, Zeng Y, Borrelli P, Hubacek K, Hu Y, Xu B, Fang N, et al. Human-altered soil loss dominates nearly half of water erosion in China but surges in agriculture-intensive areas. One Earth. 2024;7(11):2008–18.

Montané F, Romanyà J, Rovira P, Casals P. Aboveground litter quality changes may drive soil organic carbon increase after shrub encroachment into mountain grasslands. Plant Soil. 2010;337(1):151–65.

Du Z, Zheng H, Penuelas J, Sardans J, Deng D, Cai X, Gao D, Nie S, He Y, Lü X, et al. Shrub encroachment leads to accumulation of C, N, and P in grassland soils and alters C:N: P stoichiometry: A meta-analysis. Sci Total Environ. 2024;951: 175534.

Alberti G, Vicca S, Inglima I, Belelli-Marchesini L, Genesio L, Miglietta F, Marjanovic H, Martinez C, Matteucci G, D'Andrea E et al: Soil C:N stoichiometry controls carbon sink partitioning between above-ground tree biomass and soil organic matter in high fertility forests. iForest 2015, 8:195–206.

Ding L, Wang P, Zhang W, Zhang Y, Li S, Wei X, Chen X, Zhang Y, Yang F: Shrub encroachment shapes soil nutrient concentration, stoichiometry and carbon storage in an abandoned subalpine grassland. In: Sustainability 2019 11.

Williams A, Langridge H, Straathof AL, Muhamadali H, Hollywood KA, Goodacre R, de Vries FT. Root functional traits explain root exudation rate and composition across a range of grassland species. J Ecol. 2022;110(1):21–33.

Li JH, Zhang J, Li WJ, Xu DH, Knops JMH, Du GZ. Plant functional groups, grasses versus forbs, differ in their impact on soil carbon dynamics with nitrogen fertilization. Eur J Soil Biol. 2016;75:79–87.

Gao XL, Li XG, Zhao L, Kuzyakov Y. Regulation of soil phosphorus cycling in grasslands by shrubs. Soil Biol Biochem. 2019;133:1–11.

Verbruggen E, Xiang D, Chen B, Xu T, Rillig MC. Mycorrhizal fungi associated with high soil N: P ratios are more likely to be lost upon conversion from grasslands to arable agriculture. Soil Biol Biochem. 2015;86:1–4.

Xiang D, Verbruggen E, Hu Y, Veresoglou SD, Rillig MC, Zhou W, Xu T, Li H, Hao Z, Chen Y, et al. Land use influences arbuscular mycorrhizal fungal communities in the farming-pastoral ecotone of northern China. New Phytol. 2014;204(4):968–78.

Li S, Liu C, Chen J, Wang S. Karst ecosystem and environment: Characteristics, evolution processes, and sustainable development. Agric Ecosyst Environ. 2021;2021(306): 107173.

Li J, Meng B, Chai H, Yang X, Song W, Li S, Lu A, Zhang T, Sun W. Arbuscular mycorrhizal fungi alleviate drought stress in C3 (Leymus chinensis) and C4 (Hemarthria altissima) grasses via altering antioxidant enzyme activities and photosynthesis. Front Plant Sci. 2019;10(4):499.

Ma X, Li J, Ding F, Zheng Y, Chao L, Liu H, Liu X, Qu H, Bao Y: Changes of arbuscular mycorrhizal fungal community and Glomalin in the rhizosphere along the distribution gradient of Zonal Stipa populations across the arid and semiarid steppe. Microbiol Spectrum 2022, 10(5).

Lahrach Z, Legeay J, Ahmed B, Hijri M. The composition of the arbuscular mycorrhizal fungal bacteriome is species dependent. Environ Microbiome. 2024;19(1):77.

Verbruggen E, van der Heijden MGA, Rillig MC, Kiers ET. Mycorrhizal fungal establishment in agricultural soils: factors determining inoculation success. New Phytol. 2013;197(4):1104–9.

Gao C, Montoya L, Xu L, Madera M, Hollingsworth J, Purdom E, Hutmacher RB, Dahlberg JA, Coleman-Derr D, Lemaux PG, et al. Strong succession in arbuscular mycorrhizal fungal communities. ISME J. 2019;13(1):214–26.

Ma X, Geng Q, Zhang H, Bian C, Chen HYH, Jiang D, Xu X. Global negative effects of nutrient enrichment on arbuscular mycorrhizal fungi, plant diversity and ecosystem multifunctionality. New Phytol. 2021;229(5):2957–69.

Qiu L, Zhang Q, Zhu H, Reich PB, Banerjee S, van der Heijden MGA, Sadowsky MJ, Ishii S, Jia X, Shao M, et al. Erosion reduces soil microbial diversity, network complexity and multifunctionality. ISME J. 2021;15(8):2474–89.

Zhou J, Ning D. Stochastic community assembly: does It matter in microbial ecology? Microbiol Mol Biol Rev. 2017;81(4):e00002-00017.

Egerton-Warburton LM, Johnson NC, Allen EB. Mycorrhizal community dynamics following nitrogen fertilization: A cross-site test in five grasslands. Ecol Monogr. 2007;77(4):527–44.

Fan DD, Ji MK, Wu JS, Chen H, Jia HZ, Zhang XZ, Zhuang XL, Kong WD: Grazing does not influence soil arbuscular mycorrhizal fungal diversity, but increases their interaction complexity with plants in dry grasslands on the Tibetan Plateau. Ecol Indic 2023, 148.

Murray TR, Frank DA, Gehring CA. Ungulate and topographic control of arbuscular mycorrhizal fungal spore community composition in a temperate grassland. Ecology. 2010;91(3):815–27.

Lekberg Y, Arnillas CA, Borer ET, Bullington LS, Fierer N, Kennedy PG, Leff JW, Luis AD, Seabloom EW, Henning JA. Nitrogen and phosphorus fertilization consistently favor pathogenic over mutualistic fungi in grassland soils. Nat Commun. 2021;12(1):3484.

Rousk J, Bååth E, Brookes PC, Lauber CL, Lozupone C, Caporaso JG, Knight R, Fierer N. Soil bacterial and fungal communities across a pH gradient in an arable soil. ISME J. 2010;4(10):1340–51.

Haleem Khan AA: Chapter 19 - Arbuscular mycorrhizal community analysis from a grassland ecosystem. In: Microbial Symbionts. Edited by Dharumadurai D: Academic Press; 2023: 343–369.

Li Q, Xiang X, Du Y, Li Y, Lin L, Zhang F, Guo X, Cao G. Arbuscular mycorrhizal fungal community structure following different grazing intensities in an alpine grassland. Soil Sci Soc Am J. 2021;85(5):1620–33.

Liu M, Shen Y, Li Q, Xiao W, Song X. Arbuscular mycorrhizal fungal colonization and soil pH induced by nitrogen and phosphorus additions affects leaf C:N: P stoichiometry in Chinese fir (Cunninghamia lanceolata) forests. Plant Soil. 2021;461(1):421–40.

Zhou Y, Chen K, Muneer MA, Li C, Shi H, Tang Y, Zhang J, Ji B. Soil moisture and pH differentially drive arbuscular mycorrhizal fungal composition in the riparian zone along an alpine river of Nam Co watershed. Front Microbiol. 2022;13: 994918.

Eldridge DJ, Bowker MA, Maestre FT, Roger E, Reynolds JF, Whitford WG. Impacts of shrub encroachment on ecosystem structure and functioning: towards a global synthesis. Ecol Lett. 2011;14(7):709–22.

Zheng W, Guo X, Zhou P, Tang L, Lai J, Dai Y, Yan W, Wu J. Vegetation restoration enhancing soil carbon sequestration in karst rocky desertification ecosystems: A meta-analysis. J Environ Manage. 2024;370: 122530.

Zhao J, Yang W, Ji-Shi A, Ma Y, Tian L, Li R, Huang Z, Liu Y-F, Leite PAM, Ding L, et al. Shrub encroachment increases soil carbon and nitrogen stocks in alpine grassland ecosystems of the central Tibetan Plateau. Geoderma. 2023;433: 116468.

Wieczorkowski JD, Lehmann CER. Encroachment diminishes herbaceous plant diversity in grassy ecosystems worldwide. Global Change Biol. 2022;28(18):5532–46.

Chandregowda MH, Murthy K, Bagchi S. Woody shrubs increase soil microbial functions and multifunctionality in a tropical semi-arid grazing ecosystem. J Arid Environ. 2018;155:65–72.

Yu S, Liu X, Chen X, Sun M, Cao Y, Hu J, Yang L, Hu J: Effects of shrub encroachment on grassland community and soil nutrients among three typical shrubby grasslands in the alpine subhumid region of the Qinghai-Tibet Plateau, China. Front Ecol Evol 2022, 10.

Zhang B, Li S, Chen S, Ren T, Yang Z, Zhao H, Liang Y, Han X. Arbuscular mycorrhizal fungi regulate soil respiration and its response to precipitation change in a semiarid steppe. Sci Rep. 2016;6(1):19990.

Zhu Y-G, Michael Miller R. Carbon cycling by arbuscular mycorrhizal fungi in soil–plant systems. Trends Plant Sci. 2003;8(9):407–9.

Xiao D, Chen M, He X, Nie Y, Jiang N, Zhang W, Hu P, Wang K. Soil nutrients and vegetation along a karst slope gradient affect arbuscular mycorrhizal fungi colonization of roots rather than bulk soil AMF diversity. Plant Soil. 2023;489(1–2):139–54.