Nematode trophic networks drive stoichiometry-dependent plant-soil feedback across alpine grassland restoration chronosequences

• Nematodes shift from nutrient competitors to facilitators as grassland soils transition from high to low C:N ratios. • Fungivorous nematodes enhance nitrate-N and labile P in low C:N soils. • Nematodes reshape microbial networks from P to N cycling as restoration progresses. • Nematode grazing selectively reduces fungal species richness without affecting bacterial diversity. The reassembly of soil trophic networks is critical for grassland restoration on the Qinghai-Tibetan Plateau, where alpine grassland ecosystems face severe degradation. Yet the mechanisms by which nematode-microbe interactions regulate nutrient cycling across successional stages remain unclear. Using controlled microcosm experiments with native communities from 2-year (early-stage; high C:N) and 12-year (late-stage; low C:N) restored grasslands, we demonstrate that nematode functional guilds drive plant-soil feedback via stoichiometric controls on microbial turnover. In early-stage soils, the combined bacterivorous ( Acrobeloides spp.) and fungivorous ( Aphelenchoides spp.) nematodes reduced plant biomass (39.9% aboveground, 23.8% belowground) while increasing labile phosphorus (83.6%), indicating nitrogen limitation of primary production. Conversely, in late-stage soils, nematodes enhanced plant growth (84.0% aboveground, 98.3% belowground) and nitrogen accumulation (65.1%), with fungivore-alone treatment increasing nitrate-N (92.9%) and labile P (35.9%). Microbial analyses revealed nematode-induced restructuring of co-occurrence networks, with early-stage modules linked to phosphorus mobilization and late-stage modules to nitrogen cycling. Nematodes selectively reduced fungal richness while maintaining bacterial diversity, consistent with preferential hyphal grazing. PLFA profiling showed a 16.3% decline in fungi:bacteria ratios under combined treatment in early-stage soils. Our findings highlight that nematodes drive restoration outcomes through stoichiometric thresholds, shifting from nutrient competition (high C:N) to facilitation (low C/N) via fungal grazing and microbial network restructuring. These results provide a predictive framework for managing soil fauna to optimize ecosystem recovery.