Impacts of vegetation properties and temperature characteristics on species richness patterns in drylands: Case study from Xinjiang

Energy availability at trophic and hydrologic level dominates species richness gradients by constraining food resources, and regulating population sizes and extinction rates. Remote sensing datasets have mapped vegetation productivities as a proxy for energy availability, for example, using Dynamic Habitat Indices (DHIs). Considering the sparse vegetation across drylands, we developed indices of Land surface temperature (ILST) based on daytime land surface temperature from Terra and Aqua Moderate Resolution Imaging Spectroradiometer (MODIS), which has three components: (1) annual mean temperature (LSTmean), (2) annual maximum temperature (LSTmax), and (3) standard deviation of temperature (LSTstd). We hypothesized that the temperature variables, such as ILST, would predict species richness better than productivity proxies across drylands. Thus, our objective was to determine how well they would predict the richness of plants, mammals and birds across the Xinjiang Uygur Autonomous Region. We calculated the DHIs and ILST from the MODIS vegetation and temperature products from 2001 to 2015. We found that: (1) ILST could capture more additive information compared with DHIs in terms of the relatively high variance explanation of species richness and high variable importance, and the combination of ILST and DHIs gave better predictions than single metrics for species richness patterns. (2) Plants and birds were more sensitive to temperature than vegetation productivity, probably due to physiological tolerance and evolutionary processes. (3) LSTstd was the most important variable affecting species richness, except on mammals. High LSTstd was related to more food resources and habitats, and low LSTstd represented extreme environment and environmental stress. Combined vegetation properties and temperature variabilities are good determinants of species richness, and should be carefully considered in future research.

Automated summary

This study found that ILST performed better than DHI in predicting species richness, with a combination of ILST and DHI providing improved predictions. Plants and birds were more sensitive to temperature than vegetation productivity, with LSTstd identified as the most important variable affecting species richness.