Spatiotemporal Clusters of GPS Locations and Prediction of Grizzly Bear Behaviour
Abstract
Changes to grizzly bears foraging patterns caused by natural or anthropogenic alterations in their environment could have substantial consequences for both ungulate populations of prey and for bears in Alberta, where the species is considered as “Threatened”. I developed a method for identifying foraging sites of bears fitted with Geographic Positioning System collars, which allowed downloads of locations through satellites. Using spatiotemporal collar data and landscape data I created logistic regression models to predict occurrence of bedding, predation, and other foraging behaviours. I was therefore able to precisely identify sites where grizzly bears concentrated their activity and also increased the predictability of predation event locations by 2.75 times, compared to visits of random GPS-collar locations. My study also determined the natural and human factors influencing bear behavioural patterns, prominently forestry operations and human infrastructure; i.e. factors to be considered in conservation planning for grizzly bears.
Description
Keywords
Forestry and Wildlife, Ecology
Citation
Kermish-Wells, J. (2017). Spatiotemporal Clusters of GPS Locations and Prediction of Grizzly Bear Behaviour (Master's thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca. doi:10.11575/PRISM/27847