Soil Micro-Food Web Composition and Complexity Shape Multifunctionality Across Post-Cropland Restoration States

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.