Response of Soil Fungal Communities to the Alpine Steppe Degradation Gradient

ABSTRACT The degradation of alpine steppe on the Qinghai‐Tibet Plateau has become a critical ecological security issue. Soil fungi, as core regulators of grassland ecosystem functioning, play a pivotal role in mediating the feedback between degradation and soil ecological processes. This study targeted Stipa purpurea alpine steppe in the Qinghai Lake Basin, establishing four degradation gradients: non‐degraded, slightly degraded, moderately degraded, and severely degraded. We aimed to investigate variations in soil fungal community attributes (composition, diversity, differential taxa, co‐occurrence network complexity, keystone taxa, and functional guilds) along a degradation gradient via ITS1 region sequencing, and to evaluate their relationships with soil properties (nutrients, moisture, and pH). The results indicated that alpine steppe degradation had a limited impact on the overall composition and alpha diversity of the soil fungal community. The community was consistently dominated by the core phyla Ascomycota , Mortierellomycota , and Basidiomycota . LEfSe analysis identified 18 differential fungal taxa that varied across the degradation gradient. Saprotrophic taxa within the Sordariomycetes class were enriched in non‐degraded soils, whereas severely degraded soils were dominated by pathogenic and stress‐tolerant taxa, such as Sclerotinia and Phialophora . Changes in these differential taxa along the alpine steppe degradation gradient were significantly correlated with soil organic carbon, available phosphorus, and pH. The complexity of the soil fungal co‐occurrence network exhibited a nonlinear response, characterized by a continuous decrease from the non‐degraded to the moderately degraded stage, followed by a slight increase under severe degradation. The keystone taxa within these networks gradually shifted from carbon‐dependent saprotrophs to stress‐tolerant or pathogenic types, a change that was strongly associated with environmental filtering under conditions of soil organic carbon scarcity. Functional prediction revealed that saprotrophic fungi were the dominant functional guild across all degradation gradients, with no significant change in relative abundance. In contrast, the relative abundance of plant‐pathogenic fungi increased progressively with intensifying degradation. Random Forest modeling confirmed that the relative abundance of the plant‐pathogenic functional guild was the most sensitive indicator in response to alpine steppe degradation. This indicator showed a significant negative correlation with soil nutrients (total nitrogen, available nitrogen, soil organic carbon, etc.) and a significant positive correlation with soil pH. This study elucidates the response patterns of the soil fungal community to alpine steppe degradation and provides a theoretical foundation for the ecological restoration of degraded alpine grasslands.