Data from: Effects of roads and land use on frog distributions across spatial scales and regions in the eastern and central United States

Aim: Understanding the scales over which land use affects animal populations is critical for conservation planning, and it can provide information about the mechanisms that underlie correlations between species distributions and land use. We used a citizen-science database of anuran surveys to examine the relationship between road density, land use, and the distribution of frogs and toads across spatial scales and regions of the United States. Location: Eastern and Central United States Methods: We compiled data on anuran occupancy collected from 1999-2013 across 13 states in the North American Amphibian Monitoring Program, a citizen science survey of calling frogs. These data were indexed to measures of land use within buffers ranging from 300 m to 10 km. Results: The negative effects of road density and development on anuran richness were strongest at the smallest scales (300 – 1000 m), and this pattern was consistent across regions. In contrast, the relationships of anuran richness to agriculture and forest cover were similar across local scales but varied among regions. Richness had a negative relationship with agriculture/ forest loss in the Midwest but a positive relationship with agriculture in the Northeast. Anuran richness was more closely related to primary/secondary road density than to rural road density, and the negative effects of larger roads increased at smaller scales. Individual species differed in the scales over which roads and development affected their distributions, but these differences were not closely related to either body size or movement ability. Main conclusions: This study further refines our understanding of the relationship between roads and amphibian populations and highlights the need for research into the specific mechanisms by which roads affect amphibians. Additionally, we find that relationships between land use and species richness can differ substantially across regions, demonstrating that one should use caution in generalizing from one region to another, even when species composition is similar.