How did the animal come to cross the road? Drawing insights on animal movement from existing roadkill data and expert knowledge

Abstract

Context: Animal movement through the landscape is essential to several ecological processes, assuring genetic flow, reproductive success and population dynamics. In fragmented landscapes, species change their movement patterns according to their sensitivity to landscape cover and configuration. Increased landscape habitat loss and fragmentation affects resource distribution and habitat availability, compelling species to move more frequently through anthropized matrices. Objectives: Our aim was to provide a new prospect for what is known about animal movement through the landscape anthropogenic matrix in the Atlantic Forest biodiversity hotspot, by analyzing the potential movement distances of several mammal, bird, reptile, and amphibian species. Methods: We used recorded roadkills within the Brazilian Atlantic Forest as a source of information of species’ occurrence. We assumed the euclidean distance from the roadkill spot to the nearest forest fragment as the potential movement distance by different species through the anthropogenic matrix. Experts’ classification of species’ forest dependence degree, and animal movement information from the literature survey was used to analyze the variation in potential animal movement distance in these matrices. Results: We analyzed 90 species, the majority of them being non-forest dependent (n = 64; 71%). The potential movement distances ranged from 0 to > 1500 m, and forest-dependent species presented lower mean distances (between 772 and 978 m) than non-forest dependent species (between 673 and 1015 m). The literature survey resulted in 44 articles regarding animal movement in the matrix, referred only to 22 species. No articles were found for 20 species, mainly reptiles and amphibians. Conclusions: By comparing these values to the average distance between forest remnants in the Atlantic Forest, we expect that the distance to be traveled through the matrix to reach forest fragments may be restrictive, mainly for species with higher degree of forest dependence. Despite being descriptive, our results provide important insights on animal movement through the matrix, especially for less studied species, such as amphibians and reptiles.