Hotspots of livestock depredation by pumas and jaguars in Brazil: a biome‐scale analysis

Abstract

AbstractRetaliatory killing, motivated by depredation of domestic animals by large carnivores, has had a significant impact on wild predators’ populations. In Brazil, pumas and jaguars are the most persecuted species, so it is important to analyse the spatial patterns of depredation risk of livestock. We aimed to identify the hotspots of depredation risk by these two felids in Brazil, as well as the most important variables that determine the risk at the biome scale. We generated spatial models of livestock depredation using a database of depredation records, anthropogenic, topographic and vegetation variables, and ecological niche models. We used six algorithms to generate spatial risk models of depredation and selected those with the best performance, for inclusion in a consensus model, for each biome. Finally, we overlapped the areas of high depredation risk by both felids, so that areas considered as hotspots are those in which there is a high risk that livestock will be preyed by pumas or jaguars. Approximately 17% of the area of the biomes analysed for puma had high depredation risk, and 18% in the case of jaguar. The hotspots encompassed 6.5% of the included biomes’ area. The variables associated with the high depredation risk were different between species and biomes. Depredation risk by pumas was primarily associated with anthropogenic variables, whereas a combination of anthropogenic variables and vegetation types increased depredation risk by jaguar. Biome‐scale analysis, coupled with the availability of reliable data and the implementation of statistically robust methods, provides information specifically on the areas of highest risk and the conditions that contribute to it.