RESUMEN
Roads are the second largest anthropogenic cause of mortality for most vertebrates. Previous research has analyzed the factors influencing roadkill either by species or by group of species based on some species-specific characteristics. However, to gain a comprehensive understanding on the consistency of findings within and between taxa, it is necessary to conduct an analysis that encompasses both individual species and group of species. This study aims to assess the role of taxonomic level in determining the spatial, climatic and temporal drivers using roadkill data for 70 species. We used generalized linear models to examine the association between roadkill and land cover and climate across 26 individual species and groups of reptiles, birds and mammals. Temporal patterns were evaluated using circular statistics. Our study revealed variations in spatial, climatic and temporal factors among taxa. For reptiles, spatial patterns differed between the class/order and species level, while climatic and temporal patterns remained consistent among these taxonomic levels. Spatial and climatic patterns were not consistent between class, order and associated species for birds and mammals. Temporal patterns, on the other hand, were consistent between each order and its associated species. In general, we found that roadkill incidence was positively associated with silviculture and agriculture cover as well as proximity of rivers. Accumulated insolation and mean minimum temperature were positively associated with roadkill, while mean temperature exhibited a negative association. Our study emphasizes the importance of analyzing roadkill variables on a species-specific basis. In the case of endangered species with low roadkill rates, it is essential to consider their assessment alongside other species exhibiting similar behavior and ecological needs.