Aim:Collisions between wildlife and vehicles are recognized as one of the major causes of mortality for many species. Empirical estimates of road mortality show that some species are more likely to be killed than others, but to what extent this variation can be explained and predicted using intrinsic species characteristics remains poorly understood. This study aims to identify general macroecological patterns associated with road mortality and generate spatial and species-level predictions of risks. Location: Brazil.Time period: 2001-2014.Major taxa: Birds and mammals. Methods:We fitted trait-based random forest regression models (controlling for survey characteristics) to explain 783 empirical road mortality rates from Brazil, representing 170 bird and 73 mammalian species. Fitted models were then used to make spatial and species-level predictions of road mortality risk in Brazil, considering 1,775 birds and 623 mammals that occur within the continental boundaries of the country.Results: Survey frequency and geographical location were key predictors of observed rates, but mortality was also explained by the body size, reproductive speed and ecological specialization of the species. Spatial predictions revealed a high potential standardized (per kilometre of road) mortality risk in Amazonia for birds and mammals and, additionally, a high risk in Southern Brazil for mammals. Given the existing road network, these predictions mean that >8 million birds and >2 million mammals could be killed per year on Brazilian roads. Furthermore, predicted rates for all Brazilian endotherms uncovered potential vulnerability to road mortality of several understudied species that are currently listed as threatened by the International Union for Conservation of Nature. Conclusion:With a rapidly expanding global road network, there is an urgent need to develop improved approaches to assess and predict road-related impacts. This study illustrates the potential of trait-based models as assessment tools to gain a better understanding of the correlates of vulnerability to road mortality across species, and as predictive tools for difficult-to-sample or understudied species and areas. K E Y W O R D Sbird, Brazil, life history, machine-learning models, mammal, random forest, road-associated mortality, species traits

Aim:The road network is increasing globally but the consequences of roadkill on the viability of wildlife populations are largely unknown. We provide a framework that allows us to estimate how risk of extinction of local populations increases due to roadkill and to generate a global assessment that identifies which mammalian species are most vulnerable to roadkill and the areas where they occur.

No abstract

No abstract