An integrated satellite technique to evaluate grizzly bear habitat use

I present a method that combines 2 previously described remote-sensing techniques: Landsat-derived vegetation types (Craighead et al. 1986, 1988) and National Oceanic and Atmospheric Administration (NOAA) Tires satellite-derived locations of grizzly bears (Ursus arctos horribilis). This research was completed on a 5,931-km(2) study area north of the Squirrel River, a tributary of the Kobuk River, in northwestern Alaska. Six satellite radiocollared grizzly bears were located a total of 1,624 times from 1986 to 1988. Habitat use was quantified and statistically evaluated by superimposing bear locations and home ranges on a map of vegetation cover types. I acknowledge the variability of the remote measurements and describe a technique to estimate the central tendency of a sample set of vegetation complexes about bear occurrences. The inference of selection or avoidance was made from the juxtaposition of bear and habitat. The analyses showed that individual bears clearly selected for specific habitat types, but as a group the bears were quite diverse in habitat use. This indicates that habitat needs of the studied grizzly bears were very broad and that their area requirements were expansive.