Microorganisms play a key role in maintaining the stability of vegetation-soil-microbial systems and terrestrial geochemical processes in alpine meadows. To investigate the effects of different management practices on the structure and function of microbial communities, the present study used metagenomic sequencing to investigate the structure and function of soil microbial communities in the southern Qilian Mountains of the Qinghai-Xizang Plateau in response to the management practices of fenestration (FE), winter grazing (WG), transition zone between natural and artificial grasslands (TZ), and artificial unicast oats (AU). The management measures significantly changed vegetation diversity and soil physicochemical properties. The prokaryotic community structure was considerably similar in FE and WG, as well as in TZ and AU. Near-natural (FE) and artificial establishment (AU) disturbances changed the fungal community structure. Enzymes related to carbon metabolism did not respond significantly to the management measures, whereas those related to nitrogen metabolism did not respond significantly in TZ and AU. The relative abundance of enzymes involved in nitrogen metabolism was higher under TZ and AU than under FE and WG. We concluded that grassland management measures altered the structure of aboveground graminoid and leguminous vegetation communities and belowground biomass allocation, resulting in changes in K uptake, causing significant changes in the structure of fungal communities and nitrogen-metabolizing enzymes; moderate disturbance (WG) was beneficial for maintaining the stability of microbial communities in alpine grasslands.

The impacts of human-driven environmental changes on the stability of natural grasslands have been assessed by comparing differences between manipulative warming and grazing plots and reference plots. However, little is known about whether or how ambient climate regulates the effects of manipulative treatments. A 36-year observational dataset shows that there is a nonlinear response of community stability to ambient climate. Manipulative warming and grazing decrease community stability with experiment duration through an increase in legume coverage and/or decrease in species asynchrony, due to exceeding the threshold of background annual mean air temperature with decreasing background annual mean air temperature through time during the 10-year experiment period. Moreover, the temperature sensitivity of community stability is more sensitive under the ambient treatment than under the manipulative treatments. Therefore, our study emphasizes the importance of the context dependency of the response of community stability to human-driven environmental changes.