Environmental variables
There is unified opinion among ecologists that patterns of biodiversity across space (i.e., latitudinal or elevational gradient) is as a result of local adaptation to abiotic and biotic variables but not spatial gradient per se . With the aim to detect which variables have lead community assembly, we extracted six climatic variables (annual mean temperature, AMT; temperature seasonality, TS; annual precipitation, AP; precipitation seasonality, PS; net primary productivity, NPP and potential evapotranspiration, PET), which have been frequently reported relating to small mammals biodiversity (Wen et al. 2016b; Stevens & Gavilanez 2015; Wu et al. 2013).
    Lacking in detailed coordinate information for each sampling site and elevational bands, data extraction for each environmental variable was accomplished with local and regional digital elevation model (DEM). We constructed polygon layer for each local or regional elevation band by constraining vertical and horizontal scope. The mean values of environmental variable within sampling site or band polygons were extracted for each assemblage (details to see Du et al. 2017). AMT, TS, AP and PS were all downloaded from Worldclim (http://www.worldclim.org/) (last accessed in July, 2016) (Fick & Hijmans 2017) at 30 seconds resolution. NPP, and PET were obtained from MODIS products (MOD17A3 and MOD16A3) (Mu et al. 2011; Mu et al. 2007) accessed from LP DAAC (Land Processes Distributed Active Archive Center; https://lpdaac.usgs.gov/) (last accessed in July, 2016). We performed layer mosaic and projection transformation with ENVI (ver. 4.7) (VIS 2011) and ArcMap (ver. 10.0) (ESRI 2010).