Citation. Garshelis, D. L. 1983. The role of sampling intensity in the selection of a home range model. Fourth International Conference on Wildlife Biotelemetry. Halifax, Nova Scotia, Canada. 4:270-275.   [1298  CC]

Abstract. Numerous home range models have been proposed, ranging from polygons formed by the connection of outermost locations to probabilistic models based on the frequency distribution of locational data. Probabilistic models are gaining attention because intensity of use is considered in the perception of the home range; areas used infrequently are not regarded within the home range. One major assumption inherent in such models is that sampling effort is temporally and spatially unbiased. In my study of home range utilization of radio-transmittered sea otters (Enhydra lutris) in Alaska (1979-1982) I could not sample locations with the same intensity at night as during the day. Because sea otters frequently used different areas at night (feeding areas) than during the day (resting areas), the distribution of locational data differed from the true utilization distribution of the animals. This precluded the use of probabilistic models for examination of sea otter home ranges. Other researchers likely encounter difficulties trying to maintain a sampling intensity that is temporally and spatially homogeneous. Such difficulties may dictate the use of non-probabilistic home range models despite the advantages offered by more sophisticated models.