Reproductive rate and body size predict road impacts on mammal abundance

It has been hypothesized that mobile species should be more negatively affected by road mortality than less-mobile species because they interact with roads more often, and that species with lower reproductive rates and longer generation times should be more susceptible to road effects because they will be less able to rebound quickly from population declines. Taken together, these hypotheses suggest that, in general, larger species should be more affected by road networks than smaller species because larger species generally have lower reproductive rates and longer generation times and are more mobile than smaller species. We tested these hypotheses by estimating relative abundances of 17 mammal species across landscapes ranging in road density within eastern Ontario, Canada. For each of the 13 species for which detectability was not related to road density, we quantified the relationship between road density and relative abundance. We then tested three cross-species predictions: that the slope of the relationship between road density and abundance should become increasingly negative with (1) decreasing annual reproductive rate; (2) increasing home range area (an indicator of movement range); and (3) increasing body size. All three predictions were supported in univariate models, with R2 values of 0.68, 0.50, and 0.52 respectively. The best overall model based on AICc contained both reproductive rate (P = 0.008) and body size (P = 0.072) and explained 77% of the variation in the slope of the relationship between road density and abundance. Our results suggest that priority should be placed on mitigating road effects on large mammals with low reproductive rates.