Optimized bio-fertilization drives the synergistic recovery of vegetation productivity and soil multifunctionality by reshaping fungal interaction patterns in degraded alpine mines

Restoring degraded alpine mining ecosystems is critically constrained by soil infertility and the functional decoupling of plant–soil–microbe interactions. While optimized bio-fertilization represents a promising restoration strategy, the deterministic mechanisms linking vegetation recovery, fungal community assembly, and soil multifunctionality (SMF) remain poorly understood. This study investigated the effects of a three-year restoration experiment at the Muli coal mine on the Qinghai–Tibet Plateau, evaluating the integrated response of plant communities and soil functional networks to various fertilization regimes. Our results demonstrated that the optimized regime (W3J1, comprising 375 kg·hm⁻² of forage-specific fertilizer and 350 kg·hm⁻² of microbial inoculant) elicited the most robust ecological recovery, increasing plant height to 35.87 cm and aboveground biomass to 352.67 g·m⁻², which represents a significant increase of 80.23% compared to the control (195.67 g·m⁻²). Concurrently, the SMF index in W3J1 reached a peak of 0.95, effectively reversing the functional impairment observed in the degraded control (SMF = -2.32). Regarding fungal α-diversity, the W3J1 treatment significantly enhanced community complexity, recording a peak Shannon index of 3.50 and Pielou’s evenness of 0.55, compared to 3.01 and 0.48 in the control, respectively. Furthermore, the W1J3 treatment achieved the highest observed OTU richness (693) and Chao1 index (822), representing significant increases of approximately 33%-37% relative to the control. Notably, excessive nutrient inputs in the W3J3 treatment failed to yield additional benefits, as SMF and fungal diversity indices stabilized or declined due to resource imbalances. Random Forest modeling identified vegetation density as the paramount predictor of fungal diversity, significantly outperforming soil physicochemical variables.Structural Equation Modeling (SEM) further elucidated a significant cascading pathway (R² = 0.67) where fertilization directly promoted fungal diversity ( β  = 0.76) and initiated vegetation establishment, which subsequently facilitated the deterministic restructuring of the fungal community. In conclusion, the optimized co-application of microbial agents and fertilizers facilitates ecosystem reconstruction by orchestrating the coupling between vegetation density and fungal stability, providing a theoretical basis for the sustainable restoration of severely degraded alpine mines.