Abstract/Summary

Social and economic development are often accompanied by expansion in infrastructure including roads. The presence of roads is prevalent globally with general plans for expansion particularly in megadiverse areas such as Latin America. Given the growing road network and the extraordinary biodiversity in this region, many species could be impacted by mortality due to vehicle collisions. Nevertheless, deficient road ecology research has resulted in limited knowledge of these impacts, which also hinders the development of conservation and mitigation strategies. For this thesis, I bring together different approaches at different spatial scales to evaluate roadkill impacts on Latin American and Caribbean wildlife. In the first chapter of this thesis, I present the close relation that humans have had with roads since ancient times, and how the increasing mortality of species on roads has called the attention of researchers since early 1920’s until current times, leading to the development of sophisticated methods for data analysis and new mitigation measures. For the second chapter, I developed predictive traits-based models for Latin American and Caribbean birds and mammals, which revealed a higher mortality for larger birds and medium-sized mammals with early maturity ages, more clutches/litters per year, higher population densities, and diet and habitat generalist habits. Spatial predictions revealed that roads across Central America, northern Andean regions, eastern Brazil, Uruguay, central-eastern Argentina, and southern Chile harbor higher mortality. For the third chapter, using predicted roadkill rates for birds and mammals from chapter 2, information on road network, species conservation status, and availability of road ecology studies, I identified priority areas and species for research and conservation in road ecology in Latin America and the Caribbean. For the fourth chapter, I analysed the spatial distribution of 1,125 roadkill records (that I collected across 240 km of roads in the Ecuadorian Amazon) in relation to landscape and road-related features. Results showed that wildlife mortality is more likely in straight sections of roads near rivers, and that for birds and mammals, proximity to forests and herbaceous-shrubby vegetation can also increase mortality. I also identified taxa-specific roadkill hotspots across the study area. For the fifth chapter, I conducted an assessment of wildlife road mortality at country level in Ecuador presenting a citizen science project I founded. For the assessment I compiled 5,010 roadkill records from systematic and non-systematic surveys, and identified the species and areas for which, further research and conservation actions are necessary. The final chapter presents a general discussion. Overall, this thesis constitutes a comprehensive evaluation of the impacts of roadkill on wildlife in Latin America, which can guide research and conservation actions for species and areas. As the applied approaches are widely replicable, this document can be also conceived as a reference for future works in other regions of the world.