Anthropogenic and environmental influences on mammalian alpha and beta diversity in a hardwood forest landscape

Species richness within a site (i.e., alpha diversity) and between sites (i.e., beta diversity) are important measurements of biodiversity utilized in wildlife research and management. However, novel occupancy modeling approaches, especially at large spatial scales and in conjunction with other analyses of biodiversity, remain scarce in studies of mammalian biodiversity across forested landscapes. We assessed alpha and beta diversity of mammals across a 16,058-km2 region of southern Illinois, USA, between January and April of the years 2008–2010. We deployed camera traps, measured habitat variables, applied hierarchical occupancy modeling and meta-analysis techniques to investigate alpha diversity, and used partial redundancy analyses and partial Mantel tests to investigate beta diversity. We collected 86,486 photographic captures of mammals. Alpha diversity values ranged from 3 to 10; models incorporating Simpson's diversity index of patch types, distance to major road, and agriculture clumpiness index were most supported. Forest proximity and percentage forest cover explained minor variation in beta diversity, and we found weak, positive correlations between beta diversity and percentage forest cover and forest proximity index. Mammalian biodiversity was strongly influenced by generalist species and highest with (1) moderate levels of habitat heterogeneity, (2) low to moderate levels of anthropogenic influence, and (3) nearby forest cover. Anthropogenic influences and the presence of agriculture appeared to increase species richness by providing novel food resources and additional habitat that generalist species could exploit while also being tolerant of any accompanying landscape fragmentation or disturbance. While dominated by generalist species, our local mammalian community had an affinity for forest cover, reinforcing the importance of forested habitat to these species in hardwood forest systems. Our study demonstrates the usefulness of hierarchical modeling approaches and the importance of examining biodiversity through varying measurements.