• Version
• Download 15
• File Size 650.45 KB
• Create Date 1 January 1998
• Download

Black bears (Ursus americanus) use the landscape over relatively coarse scales compared to many other mammals. We analyzed the way these animals relate to their habitat in general terms using a statistical approach. A priori, we conceived of scale-specific and scale-free types of habitat use. The scale-specific model assumes that an individual's movements are due to dynamic responses to the individual's perceptions over a limited spatial scale that reflects the circle of perception (CP) around the animal's location. Using serial processing, the animal moves to new sites, gaining new perceptions that lead to new movements, but is constrained by site fidelity (home range, HR) tendencies. This model is of a scale-specific statistical process at the scale of CP. According to the alternative model, individuals now and then use memory maps of past experiences to take strategic, relatively directed (i.e., coarse-scaled) movements beyond the CP but within the HR using principles from parallel processing. Although the former model is more traditionally used in modeling, the latter more closely resembles an ecological understanding of animal behavior. Our model, the multi-scaled random walk (MRW), assumes (1) spatially explicit memory effects, and (2) that optimization of patch use at coarse scales is as important as optimization at finer scales over the scale range from CP to HR. Mechanistically, this property makes the habitat use process scale-free, or fractal, over this range of scales. From the 2 basic MRW assumptions, the observed part of the HR area (A) from n non-autocorrelated relocations is expected to increase proportionally with the square root of n up to n well beyond the limits of applied ecology. We verify the predictions of this model on telemetry material on black bear movements. Some implications for bear research and management are discussed.