Efficacy of GPS cluster analysis for predicting carnivory sites of a wide‐ranging omnivore: the American black bear

摘要: The capacity to describe and quantify predation by large carnivores expanded considerably with the advent of GPS technology. Analyzing clusters locations formed facilitates detection events identifying characteristics which distinguish sites. We present a performance assessment cluster analysis as applied scavenging an omnivore, American black bear (Ursus americanus), on ungulate prey carrion. Through field investigations 6854 from 24 individual bears, we identified 54 sites where bears while predating or elk (Cervus elaphus), mule deer (Odocoileus hemionus), cattle (Bos spp.). developed models for three data sets predict whether was at carnivory site vs. non-carnivory (e.g., bed non-ungulate foraging sites). Two full-season contained logged either 3-h 30-min intervals April November, third set interval through July corresponding calving period elk. Longer fix resulted in fewer Clusters were more likely be if they occurred open edge habitats, early season, mean distance between all pairs within less, endured longer time. less initiated morning night compared day. top each performed well successfully predicted 71–96% field-verified events, 55–75% non–carnivory 58–76% overall. Refinement this method will benefit further application across species ecological systems.