Climate warming and high-intensity human activities threaten the stability of alpine meadow ecosystems. The stability of the soil microbial community is crucial for maintaining ecological service function. However, the effects of warming and litter removal on microbial interactions, community-building processes, and species coexistence strategies remain unclear. In this study, we used a fiberglass open-top chamber to simulate global change, and moderate grazing in winter was simulated by removing above-ground litter from all plants in the Qinghai–Tibet Plateau, China, to investigate the effects of warming, litter removal, and interactions on soil microbial communities. The treatments included (1) warming treatment (W); (2) litter removal treatment (L); (3) the combined treatment (WL); and (4) control (CK). The results show that compared with the control treatment, warming, litter removal, and the combined treatments increased bacterial Shannon diversity but reduced fungal Shannon diversity, and warming treatment significantly changed the bacterial community composition. Warming, litter removal, and the combined treatments reduced the colinear network connectivity among microorganisms but increased the modularity of the network, and the average path distance and average clustering coefficient were higher than those in the control group. Stochastic processes played a more important role in shaping the microbial community composition, and soil–available phosphorus and soil ammonium contributed more to the βNTI of the bacterial community, while total phosphorus and NAG enzyme in the soil contributed more to the βNTI of the fungal community. Notably, litter removal counteracts the effects of warming on bacterial communities. These results suggest that litter removal may enhance bacterial community stability under warming conditions, providing insights for managing alpine meadow ecosystems in the context of climate change.

Belowground multitrophic communities and their potential associations are important biological foundations for the maintenance and recovery of soil ecosystem functions. However, it remains unclear how soil micro-food web composition and network complexity are linked to soil ecosystem multifunctionality across different post-cropland restoration states. In this study, we used a space-for-time substitution approach and selected cropland and three post-cropland restoration states in the Taihang Mountains, including shrub–grassland, Populus spp. plantation, and Robinia pseudoacacia plantation, each with five plots, to examine changes in bacterial, fungal, and nematode communities, soil micro-food web network complexity, and soil ecosystem multifunctionality in topsoil and subsoil. Compared with cropland, bacterial diversity was higher in the shrub–grassland, Populus spp. and R. pseudoacacia plantations, whereas fungal diversity showed a marked decline in the Populus spp. plantation. Furthermore, the topsoil of the R. pseudoacacia plantation exhibited much higher micro-food web network complexity and soil multifunctionality than cropland, whereas differences among land-use types were relatively small in the subsoil. Structural equation modeling indicated that vegetation type, soil depth, soil moisture content, microbial diversity, nematode diversity, and soil micro-food web network complexity jointly explained 94% of the variation in soil ecosystem multifunctionality (R² = 0.94). Further random forest analysis identified soil depth as the most important predictor of soil ecosystem multifunctionality, followed by soil micro-food web network complexity; among biological variables, fungal diversity, bacterial diversity, omnivorous-predatory nematode diversity, and herbivorous nematode diversity also showed relatively high importance. Overall, these results suggest that changes in soil ecosystem multifunctionality across post-cropland restoration states were strongly soil-depth-dependent, and that soil micro-food web diversity and network complexity can serve as important biological indicators for assessing restoration outcomes.