The characteristics of soil microbial co-occurrence networks across a high-latitude forested wetland ecotone in China.

To understand the effect of seasonal variations on soil microbial communities in a forested wetland ecotone, here, we investigated the dynamics of the diversities and functions of both soil bacterial and fungal communities inhabiting three wetland types (forested wetland, shrub wetland and herbaceous vegetation wetland) from forest-wetland ecotone of northern Xiaoxing'an Mountains spanning different seasons. β-diversity of soil microbial communities varied significantly among different vegetation types (-, , , and wetlands). We totally detected 34 fungal and 14 bacterial indicator taxa among distinctive groups by using Linear discriminant analysis effect size (LEfSe) analysis, and identified nine network hubs as the most important nodes detected in whole fungi, bacteria, and fungi-bacteria networks. At the vegetation type-level, bacterial and fungal microbiome living in wetland soil possessed fewer positive interactions and lower modularity than those in other types of wetland soil. Furthermore, we also discovered that ectomycorrhizal fungi were dominant in the fungal microbiota existing in forested and shrub wetland soils, whereas arbuscular mycorrhizal fungi were predominated in those residing in herbaceous vegetation wetland soil. The distribution of the predicted bacterial functional enzymes also obviously varied among different vegetation-types. In addition, the correlation analysis further revealed that the key fungal network modules were significantly affected by the contents of total N and soil water-soluble K, whereas most of the bacterial network modules were remarkably positively driven by the contents of total N, soil water-soluble K, Mg and Na. Our study suggested that vegetation type are substantive factors controlling the diversity, composition and functional group of soil microbiomes from forest-wetland ecotone of northern Xiaoxing'an Mountains.