| Summary: | Habitat selection models are commonly used to inform species conservation and management decisions; however, such models are context dependent, and results may vary depending on how close a population is to the carrying capacity. Despite acknowledgment that habitat selection is density dependent, relatively few researchers account in their analyses for changes in population density over time. Using a long-term dataset (11 years) for GPS-tracked Plains Bison (Bison bison) at Grasslands National Park, Saskatchewan, Canada (n = 22), I examined seasonal habitat selection during a period of natural population growth following reintroduction and a period of population size manipulations. I used resource selection function (RSF) and latent selection difference function (LSD) analyses to model interactions between selection for vegetation productivity and distance from roads and population density. Bison showed decreased avoidance of roads as density increased and increased avoidance of roads following reductions in population density at most spatio-temporal scales examined. The relationship between selection for vegetation productivity and density was highly seasonally variable: bison selected for abundant forage when density was low and became less selective at high density during and immediately after calving. Consistent with predictions of density-dependent habitat selection, bison were free to select for abundant forage in areas far from human activity when density was low and were required to become less selective as density increased during seasons when the herd is most vulnerable and nutritional requirements are high. My study highlights the importance of considering changes in population density when using habitat selection models to inform decisions on wildlife population management.
|
|---|