Abstract/Summary

Multiple environmental factors are known to shape species distributions at the global scale, including climate and topography, but understanding current extents of occurrence and biodiversity patterns requires considering anthropogenic factors as well. Numerous studies have explored the relationship between contemporary human activities and different biodiversity metrics, but the influence of past activities, such as land-use, remains poorly understood despite being one of the oldest human impacts. Here we evaluate the role of past land-use modifications in the current distribution and conservation status of mammals worldwide using spatial data characterizing human land use from c.B.C.6000 to c.A.D.2000. First, we applied a clustering method that revealed three generalized past human land-use trajectories that represent low-, recently- and steadily-used areas widely represented across the globe. Second, we fitted boosted regression trees to predict total and threatened mammalian richness, globally and within trajectory-clusters, testing the role of environmental factors and multiple human land-use metrics reflecting: total used area at different time spans, rates of land-use change, and the occurrence of remarkable land-use shifts. Environmental factors were identified as the main correlates of current mammalian richness, but several proposed metrics of past land-use were also relevant predictors. Overall, these results highlight the likely existence of a land-use legacy in some regions of the world that has influenced the distribution of extant mammals, particularly of those currently classified as threatened. Even if we cannot change that legacy, our results show that we need to account for past human impacts to understand present biodiversity patterns and, arguably, to guide future actions.