Abstract
There has been little study on the biogeographical patterns of microbial co-occurrence, especially in agricultural soils. Here we investigated the biogeographical patterns and major drivers of co-occurrence network topological structure, and the relative abundance of keystone taxa for soil bacterial and fungal communities using high-throughput sequencing on a set of 90 samples across a 1,092 km transect in wheat fields of the North China Plain (NCP). We found that pH was the most important environmental factor driving network topology and relative abundance of keystone taxa. For the metacommunity composed of both bacteria and fungi, and for the bacterial community alone, lower soil pH was associated with a more complex microbial network. However, the network for fungi showed no strong trend with soil pH. In addition, keystone taxa abundance was positively correlated with ecosystem function and stability, and best explained by pH. Our results present new perspectives on impacts of pH on soil microbial network structure across large scales in agricultural environments. This improved knowledge of community processes provides a step toward understanding of functioning and stability of agricultural ecosystems.