Structure equation modeling (SEM) analysis showed that total soil nitrogen and total soil phosphorus had direct effects on fungal community diversity, while grazing exclusion, grassland type and SOC had indirect effects on fungal community diversity. Fraction had a significant effect on fungal community diversity by directly affecting total soil phosphorus (P < 0.05) and grassland type had a significant effect on fungal community diversity by directly affecting SOC, which in turn affected KN and TP (P < 0.05). These variables explained approximately 40% of the variation in fungal community diversity (Fig. 7a). In addition, combining direct and indirect effects revealed that soil KN was the most important predictor of change in fungal community diversity (Fig. 7b).
Fig. 7. (a) Structural equation modeling exploring the direct and indirect effects of the fraction(grazing and grazing exclusion), grassland type, plant index(coverage and Patrick index), and soil chemical properties on fungal diversity and (b) standardized effects of fraction, grassland type, plant index, and soil chemical properties on fungal diversity. Line thickness indicates the correlation strength: the thicker the line, the stronger the correlation. The red and blue arrows indicate negative and positive relationships, respectively. The numbers adjacent to the arrows represent the standardized path coefficients. The proportion of variance explained (R2) appears alongside each response variable in the model.
Discussion
4.1. Effect of grazing exclusion on soil physicochemical properties in different grassland types
Whether grazing exclusion can change the diversity and composition of soil fungal communities, soil chemistry is the dominant factor (Li et al., 2019; Qin et al., 2021). There is growing evidence that fungal communities are more sensitive than bacterial communities to carbon and nitrogen sources. (Zhang et al., 2018). Our study found that C:N did not differ significantly among the three grassland types in either the 0-5 cm or the 5-10 cm soils before grazing exclusion, probably because grazing can promote root secretions and fallen leaf decomposition through trampling, which can increase the amount of C and N entering the soil and keep the soil C and N in dynamic equilibrium (Bi et al., 2018). In the 5-10 cm soil layer, whereas the C:N was significantly higher in temperate deserts than in temperate steppes and mountain meadows after grazing exclusion, the interaction of exclusion and grassland type significantly altered the C:N, indicating that grazing exclusion reduced the soil C and N pools (Sun et al., 2020), and this result was prominent in steppes and meadows. In addition, we found that grazing significantly increased SOC in the 0-5 cm soil layer of temperate deserts and temperate steppes and in the 5-10 cm soil layer of mountain meadows; moreover grazing significantly increased SWC in the 0-5 cm soil layer of temperate steppes and in the 5-10 cm soil layer of mountain meadows, due to a higher soil fertility associated with good plant cover and greater apoplastic accumulation (Marzaioli et al., 2010; Jing et al., 2014). Aboveground biomass has a direct effect on apoplastic accumulation, which increases soil moisture. This explains the increase in SWC in both temperate steppes and mountain meadows. Furthermore, apoplastic decomposes more rapidly in moist soils, and large amounts of decomposing apoplastic can increase soil nutrients (Jing et al., 2014).
Furthermore, soil TP showed a close correlation with temperate desert, and grazing exclusion significantly increased the TP content in temperate desert, which is inconsistent with previous studies (Yang et al., 2018), This difference may be because soil nutrients still had not returned to their original levels after 9 years of grazing exclusion but were in an unstable and fluctuating state. This may be related to the change in the main vegetation types on the surface, the rate of decomposition and the amount of nutrient uptake and return by plant roots. The effects of grazing exclusion on pH, C:N, C:P and N:P in the three grasslands were not significant regardless of the soil stratum. The reasons for this phenomenon may be the removal of grazing pressure after grazing exclusion in degraded grasslands, the disappearance of conditions that maintained the spatial pattern of soil nutrients or vegetation, an increase in the spatial heterogeneity of soil nutrients, and an increase in the proportion of stochastic factors, making the research results more stochastic and the study results appear to be different (David, 1988). The grazing areas have a higher soil nutrient content due to the trampling of livestock, which leads to the crushing of apoplastic and increasing its entry into the soil, thus accelerating the circulation of organic matter; additionally, nutrients are added from the excrement of grazing livestock(Zhao, 2022). Regional differences in the soil itself may also be responsible for this difference(Yang et al., 2022).
4.2. Effects of grazing exclusion on the composition and diversity of soil fungi
Our results go further than the commonly reported differences in fungal composition between grassland types (Zhang et al., 2010; Yang et al., 2022; Wang et al., 2021). In different grassland types, the response of fungal composition to grazing exclusion was basically the same. It is worth mentioning the changes in Mortierellomycota, which we found to be significantly lower after grazing exclusion in the 0-5 soil layer in mountain meadows, and highly significant higher in the 5-10 soil layer in temperate deserts. This is consistent with previous findings that grazing exclusion is more favorable to the relative abundance of Mortierellomycota(Wang et al., 2023), as a relatively unique group in the soil, it may play a role in promoting material circulation in grazing exclusion grasslands with low nutrient content(Dai et al., 2021). This answered the second scientific question of the study. Consistent with previous results (Maestre et al., 2015; Ren et al., 2017), we found that at the phylum level, Ascomycota andBasidiomycota were the dominant pre-grazing exclusion fungi in all three grassland types. After the grazing exclusion Ascomycotadecreased by 14.72% and 9.74% in temperate deserts and mountain meadows, respectively, while it increased by 4.82% in temperate steppes. In contrast to Ascomycota , the Basidiomycotaincreased by 238.8% and 47.68% in temperate deserts and mountain meadows, respectively, while it decreased by 23.12% in temperate steppes. This result may be due to the higher soil pH in temperate desert, the acidic soils of temperate steppes and mountain meadows, and the positive responses of Ascomycota and Basidiomycotafungi to changes in soil pH, with a higher abundance ofAscomycota in soils with higher pH (Tedersoo et al., 2014). The correlation analysis results also confirmed the above view, and the results showed that pH was significantly positively correlated withDothideomycetes (a class of Ascomycota subphylum), which also explained the higher relative abundance of Ascomycota in temperate deserts. Additionally, studies of soil fungi have found that the soil fungal community has a strong spatial structure along the ecological gradients, and some fungi can dominate many environments. Wind-dispersed Ascomycota taxa dominated the soil fungal communities, followed by Basidiomycota and other fungi(Bucbe et al., 2009; Egidi et al., 2019). Next, significant differences were found between among the three grassland types for the Glomeromycota andMucoromycota , and it was found that the Mortierellomycotawas not influenced by either grazing exclusion or grassland type, but the interaction between the two had a significant effect on theMortierellomycota . It has been shown that the activity ofGlomeromycota is reduced in highly acidic environments and is not efficient at taking up nitrogen or phosphorus from organic matter and converting it to inorganic components for plant use; eventually, plants gradually form new symbiotic relationships with other organic matter mycorrhizal fungi and replace the Glomeromycota (Smith et al., 2008). In this study, the organic matter content of the temperate desert soil was low and the relative abundance of Glomeromycota was lower than that in the other two types of grassland. This study validates the findings of most studies regarding the relationship between Glomeromycota and soil physicochemical properties, and supports previous studies suggesting that soil organic matter content has a suppressive effect on Glomeromycota (Pellissier et al., 2015; Smith et al., 2008). Furthermore, at the class level, the dominant fungi in the three grassland types before and after exclusion included Dothideomycetes, Archaeorhizomycetes, Sordariomycetes and Agaricomycetes, while in the temperate desert, Archaeorhizomycetes almost disappeared. This result indicates that fungi display incredible functional diversity at the class levels. Some fungi, which were just as sensitive as bacteria, disappeared (and were replaced by something else) in response to the treatment or grassland type. When some taxa disappear, other taxa are well adapted to the new niches that are created. The different responses of various fungal classes to environmental change reflect the role of environmental niche separation in selecting fungal communities (Pellissier et al., 2014).
Unlike the effect of grazing exclusion on soil fungal composition, grazing exclusion, grassland type and the interaction between the two did not significantly alter soil fungal α-diversity in either the 0-5 cm or the 5-10 cm soils, similar to previous findings (Hao et al., 2020; Yang et al., 2019; Ding et al., 2020). The response of fungal α-diversity to grazing exclusion and grassland type was not significant, essentially meaning that there were no significant differences in the taxonomic richness among the sites.(Cheng et al., 2016; Zhang et al., 2018). These results were consistent with previous findings that soil fungi were more stable than bacteria (Hamonts et al., 2017; Wang et al., 2019). Although there was no effect on fungal α-diversity, we found that the interaction of grazing exclusion and grassland type significantly altered fungal β-diversity. Changes in the β-diversity of fungi may indicate the ability of many soil microorganisms to colonize the inter-rhizosphere, where root biomass and plant C:N ratios may influence the utilization of inter-rhizosphere resources by altering root turnover or root secretions, which may in turn strongly alter the composition of the microbial community (Chase, 2010; Philippot et al., 2013). Moreover, other studies have detected significant effects of grazing exclusion on soil fungal β-diversity (Chen et al., 2020). In the current study, the strong link between soil microbial diversity and grazing exclusion and grassland type can be explained by the fact that microbial communities are often directly influenced by aboveground plant biomass and community structure (de Vries et al., 2012). For example, grazing exclusion increases aboveground plant biomass and apoplastic content, thus increasing fungal diversity (Hamonts et al., 2017). The increase in species richness induced by grazing exclusion suggests that high plant diversity results in a greater variety of organic matter ultimately flowing into the subsurface, thereby generating more ecological niches for use by different species of microorganisms and altering the diversity of microbial communities (Lodge, 1997; Brodie et al., 2003). The results showed that under the influence of different grazing methods and different grassland types, the soil fungal community composition changed, and the difference in β diversity was significant, but the change in soil fungal α-diversity was not significant, which answered the first scientific question of this study.
4.3. Potential drivers of soil fungal community diversity
SEM analysis showed that grazing exclusion had a significant effect on fungal community diversity by directly affecting total soil phosphorus, and that grassland type had a significant effect on fungal community diversity by directly affecting SOC, which in turn affected KN and TP. In line with our findings, other studies have also shown that soil TP and KN are closely related to soil fungal diversity (Lindahl et al., 2015; Lauber et al., 2008). In contrast, a recent experiment showed that soil fungal abundance and diversity differed between grassland habitats, but that this difference was not correlated with geographical distance, and that certain environmental factors, including climate, soil pH, nitrogen and phosphorus, were the main influences affecting the distribution, abundance and diversity of soil fungal communities (Pellissier et al., 2014). In our study, combining direct and indirect effects revealed that soil KN was the most important predictor of changes in fungal community diversity, and the strong association between soil KN and fungal diversity was justified because KN provides the main energy source for soil microorganisms (Lindahl et al., 2015). For example, nitrogen in the soil provides sufficient elemental nitrogen for fungal symbiosis, and plants no longer import too much carbon into the subsurface for nitrogen acquisition, which can have an impact on fungal composition and biomass (Norris et al., 2013). However, it has been shown that fungal diversity does not correlate with soil KN and TP, suggesting that environmental factors may not be limiting factors in the structural diversity of soil fungal communities at the larger scales studied (Li., 2015). In the present study, SOC had an indirect effect on fungal community diversity, similar to the results of previous studies (Li., 2015). SOC provides nutrients and energy for plant growth and soil microbial life (Li et al., 2006), and its content reflects the effectiveness of limiting resources in the soil and is the ‘source’ of nutrients for soil fungal community growth, the level of which affects soil fungal community structure and diversity. In addition, the natural environment in different regions can also affect soil fungal diversity, and harsh environments such as drought can reduce fungal diversity and abundance (Maestre et al., 2015). In summary, the results answered a third scientific question: soil nutrient content was a major factor affecting fungal diversity, especially soil KN, TP and SOC. However, this paper still has some limitations, and our results do not account for seasonal and annual differences in soil fungal communities, which still need to be analyzed further in depth. Future experimental studies could explore the potential effects of climate change on soil fungal community diversity and microbial assembly processes in different grazing exclusion years or seasons to provide a more theoretical basis for clearing the blind spots in this field.
Conclusions
Grazing exclusion is one of the most commonly used measures to rehabilitate degraded grasslands globally. Therefore, assessing its ecological impact is important for grassland conservation and biodiversity management. As revealed by the data from our study, the composition of soil fungal communities differed between grassland types More importantly, the dominant fungal taxa such as Glomeromycetes, Orbiliomycetes, Tremellomycetes, Ustilaginomycetes, Eurotiomycetes, Orbiliomycetes, Dothideomycetes, and Archaeorhizomycetes shed new light on the involvement of soil microbes in different grassland types. The response of fungal composition to grazing exclusion was basically the same in the different grassland types, except for theMortierellomycota . Under the influence of both grazing exclusion and grassland type, there was no significant change in soil fungal alpha diversity, but there were significant differences in fungal beta diversity. The changes in soil KN and TP caused by grazing exclusion were closely related to soil fungal diversity. These results suggest that soil nutrient content was the main factor influencing fungal diversity, especially soil KN, TP and SOC. Our research provides a long-term perspective for better understanding and managing different grasslands, as well as a better scientific basis for future research on grass-soil-microbe interactions.