Potential Distribution Models for Predicting Human-Black Bear (Ursus americanus var. eremicus) Interactions in the Sierra de Zápaliname Natural Reserve, Saltillo, Coahuila

Objetive:This research analyzed the interactions between humans and bears in the ejidos of the Sierra de Zápaliname Natural Reserve and the urban area of Saltillo, Coahuila, Mexico. Design/Methodology/Approach: A database of approximately 481 georeferenced records of Human-Black Bear Interactions (IHO) for the year 2021 was used. Based on these records and their relationship with physiographic variables (slope, aspect, elevation, roughness) and environmental variables (normalized difference vegetation index and a humidity detection index), potential distribution models were developed using the support vector machine algorithm, which is characterized by the assumption of the Greenelian niche theory. Results: Six potential distribution maps were generated, divided into two rainy and dry periods and three types of interaction: sightings, agricultural-livestock conflicts and captures. Findings/Conclusions: In all models, a common pattern was found where the sites with the highest probability of IHO are places close to human settlements and areas where agriculture, livestock and landfill activities occur.