Maoer Mountain National Forest Ecological Station is located in Heilongjiang Province. By using the statistical spatial analysis method, which showed that the soil mites community and species richness of mites spatial heterogeneity. Also, which is illustrated that the spatial effect relationship with the environmental factors. The results of crossvariogram and simple Mantel test showed that positive relationships between soil mite communities, specific mite species and environmental factors were detected. Among the environmental factors, soil pH showed significant affects on spatial heterogeneity of soil mite species in both years. When considering the relationships between soil mite species, positive correlations and specific mite species were also observed. Spatial heterogeneities of soil mite communities and most of the soil mite species were regulated by structured factors, in which environmental variables might be important drivers. Thus, it is implied that abundances of soil mite community and most of the soil mite species showed aggregated spatially heterogeneity, and significant relationship between such heterogeneity and environmental factors were detected.

Metacommunity theory provides an understanding of how ecological processes regulate local community assemblies. However, few field studies have evaluated the underlying mechanisms of a metacommunity on a small scale through revealing the relative roles of spatial and environmental filtering in structuring local community composition. Based on a spatially explicit sampling design in 2012 and 2013, this study aims to evaluate the underlying processes of a soil mite metacommunity on a small spatial scale (50 m) in a temperate deciduous forest located at the Maoershan Ecosystem Research Station, Northeast China. Moran’s eigenvector maps (MEMs) were used to model independent spatial variables. The relative importance of spatial (including trend variables, i.e., geographical coordinates, and broad- and fine-scale spatial variables) and environmental factors in driving the soil mite metacommunity was determined by variation partitioning. Mantel and partial Mantel tests and a redundancy analysis (RDA) were also used to identify the relative contributions of spatial and environmental variables. The results of variation partitioning suggested that the relatively large and significant variance was a result of spatial variables (including broad- and fine-scale spatial variables and trend), indicating the importance of dispersal limitation and autocorrelation processes. The significant contribution of environmental variables was detected in 2012 based on a partial Mantel test, and soil moisture and soil organic matter were especially important for the soil mite metacommunity composition in both years. The study suggested that the soil mite metacommunity was primarily regulated by dispersal limitation due to broad-scale and neutral biotic processes at a fine-scale and that environmental filtering might be of subordinate importance. In conclusion, a combination of metacommunity perspectives between neutral and species sorting theories was suggested to be important in the observed structure of the soil mite metacommunity at the studied small scale.