Landscapes of coexistence: generating predictive risk models to mitigate human-raptor conflicts in forest socio-ecosystems

Human persecution is a worldwide threat to raptors, contributing to the decline of many species. Perceived or real predation of domestic animals is the main driver of persecution and a barrier to the success of conservation initiatives. Predictive risk models are used to identify hazards in order to target effective prevention actions, and they have been successfully applied to conflicts with top predators. The Andean temperate region of Chile is a Global Biodiversity Hotspot where diurnal raptors co-inhabit with humans in rural areas. Here, complaints from farmers on raptor attacks on poultry have steadily increased; however, there is no empirical information about the conflict. This study aims to build a predictive social-ecological risk model to identify husbandry practices and landscape attributes associated with poultry predation by diurnal raptors in Chile. We applied 100 questionnaires to local farmers about their poultry husbandry practices and raptor predation patterns. We used data from questionnaires and landscape attributes to generate risk models of poultry predation. We show that farmers maintaining an enclosure to keep the chickens, a guard dog and a high proportion of forest in their properties can reduce the risk of raptor predation. These findings can be used to prevent raptor attacks on poultry by facilitating the implementation of these management measures in southern Chile. Our results indicate that predictive models can help in identifying effective coexistence measures for human-raptor conflicts benefiting human livelihoods as well as conservation of wild predators.