Structural equation modeling reveals the dominant role of rhizosphere microbial α diversity relative to network complexity in explaining soil multifunctionality during alpine grassland restoration

Aims Alpine grasslands on the Tibetan Plateau are highly sensitive to human disturbances and climate change; however, the understanding of how long-term ecological restoration influences ecosystem functioning through belowground biological processes remains limited. Methods Natural alpine grasslands, artificially restored grasslands with different restoration ages (2000S, 2006S, 2013S, and 2018S), and severely degraded black-soil-patch grasslands (BOBG) were investigated. Plant biomass allocation, soil microbial diversity, microbial co-occurrence network structure, soil multifunctionality indices, and structural equation modeling were used to examine how soil microbial α diversity and network complexity vary along the restoration gradient and how they are associated with changes in soil multifunctionality. Results With increasing restoration age, plant resource allocation gradually shifted from aboveground tissues to root systems. Soil bacterial and fungal α-diversity, microbial network complexity, and soil multifunctionality increased significantly along the restoration gradient. Long-term restored grasslands were characterized by highly connected and structurally complex microbial networks, whereas networks in severely degraded grasslands were comparatively simplified. Soil multifunctionality was significantly positively associated with bacterial α diversity. Conclusions Compared with fungal diversity, bacterial α diversity showed a higher relative importance in explaining variations in soil multifunctionality. These results highlight the importance of soil bacterial α diversity for soil multifunctionality in artificially restored black-soil-patch grassland ecosystems.