Towards an understanding of the drivers of broad-scale patterns of rarity-weighted richness for vertebrates

To combat biodiversity loss, conservation planners prioritize sites with high complementarity (ability to represent all or most species in a few sites), but little work has been done to understand the traits that influence site complementarity. Here we focus on the drivers of rarity-weighted richness (RWR), a metric of site endemism that has previously been demonstrated to be a reliable metric of site complementarity. Our aims are to assess how environmental variables individually contribute to explaining global patterns of RWR. After quantifying RWR for 18,020 terrestrial grid cells for amphibians, reptiles, birds and mammals, we used random forest models to identify associations between RWR and predictors reflecting current environment (topography, soils, climate), evolutionary history, and human footprint. Slope, long known to affect plant diversity, had strongest positive association with RWR values for amphibians, birds and mammals; human footprint was the primary driver of RWR for reptiles. RWR increased with slope, levels of human impact, diurnal temperature oscillation, land cover diversity, actual evapotranspiration, and cold season precipitation, Surprisingly, RWR increased with human footprint, perhaps because human activities cause species to have small ranges or because human activities and small-ranged species tend to occur under the same environmental conditions. Our study provides evidence that climate variables, including both temperature and precipitation—well known to drive patterns of species richness—also generate and maintain gradients of RWR at a global scale. As climate changes in the coming decades, regions of high RWR might also change, depending on the extent to which the spatial patterns of climate also change. Elucidating the patterns of RWR may improve the way in which sites are prioritized, so that all or most species can be conserved in affordable areas.