The relationships between species and their environment, and the role of various environmental factors (such as climatic, topographic, historical, and anthropogenic factors) in determining species distribution and composition are a central issue in ecology. Being scale dependent both spatially and temporally, these relationships are often elusive, and have been seldom quantified at multiple scales simultaneously. The goal of this study is to quantify the effects of various environmental determinants on mammal species composition in the contiguous USA, varying both the grain and extent components of spatial scale. We compiled data on the occurrence of all terrestrial mammal species (excluding bats) from GBIF, and a corresponding set of 15 environmental parameters from Worldclim and FAO. We sampled the study area at four different spatial scales (grain ranged from 100 km² to 10,000 km² and extent ranged from 90,000 km² to ~10,000,000 km²). Each scale consisted of 900 grid cells (30X30), except for the largest scale, which spanned the entire study area and consisted of 1,200 cells. We quantified the relationships between mammal species and environmental parameters using CCA analyses.

Results/Conclusions

Our results show that climate was the predominant factor affecting species composition at all scales. Its effect increased with spatial scale. The second most important factor was land use-land cover which was almost as strong as that of climate at the smallest scale, and decreased with scale. Topography and primary productivity were less important, and decreased with scale. We assume that if our analysis included smaller scales, the importance of topography would be larger. This study provides the first quantitative evidence to the high importance of climate and LULC parameters in affecting mammal species composition at large spatial scales. These results are alerting in light of increasing rates of climate and LULC changes.