Revegetation approaches shifted soil bacterial co‐occurrence in a desertified Tibetan alpine grassland

Abstract Introduction There is much uncertainty about how various revegetation approaches influence soil microbial co‐occurrence patterns in the severely degraded grasslands in the cold regions. Objectives To evaluate the restoration effects of different revegetation measures on soil multifunctionality and microbial co‐occurrence networks in the desertified alpine grassland, and reveal the key drivers influencing soil bacterial communities and their co‐occurrence relationships. Methods We assessed the shifts in soil multifunctionality and bacterial co‐occurrence pattern of 22 year‐old artificially revegetated grasslands in a desertified Tibetan grassland, taking the undegraded natural and desertified grasslands (DGs) as controls. Results Soil multifunctionality of revegetated grasslands was significantly higher than that of the DG but lower than that of the natural grassland (NG). The revegetated grasslands significantly enhanced the network stability compared to the natural and DGs. Notably, the herb revegetated grassland decreased the vulnerability (by 70.9–77.9%) and conversely increased the complexity of the soil bacterial network compared to that of NG and DGs, whereas the shrub revegetated grassland showed less reduction in vulnerability and no significant difference in complexity compared to the NG. The interactions of vegetation and soil attributes explained much more (32%) of the variation in soil bacterial network topological characters than their individual explanations (5–9%). Conclusions Although shrub revegetation was more effective in improving soil multifunctionality, herb revegetation was preferable for advancing complexity and stability of the soil bacterial co‐occurrence network when restoring the desertified alpine grassland.