2.5 Spatial scaling pattern analyses
Two typical spatial scaling patterns were analyzed, including TAR and DDR. The community richness (alpha-diversity) and Bray-Curtis community similarity (beta-diversity) were respectively used as diversity metrics for TAR and DDR analyses. The slope coefficients between log-transformed diversity indices and log-transformed geographic area/distance were calculated for TAR and DDR. Both the diversity indices of TAR and DDR were extended to different diversity orders using Hill numbers, which are a parametric family of diversity indices differentiated by the parameter q (Chao, Chiu, & Jost, 2014). The extended TAR and DDR were respectively termed as DARq and DDRq. The R package “hillR”(Chiu & Chao, 2014) (https://github.com/dajiang/hillR) was used for both alpha- and beta-diversity indices calculation under different diversity order q .
In addition, we also correlated the diversity metrics with environmental heterogeneity to investigate how microbial spatial scaling patterns were affected by environmental conditions. Here, the Euclidean distance based on normalized environmental variables was calculated to represent the environmental heterogeneity between different samples. All 19 environmental variables were included for environmental heterogeneity calculation. The analysis, as well as TAR and DDR, were carried out for the whole community, the abundant and the rare subcommunities.