Comparison among three approaches to evaluate winter habitat selection by white-tailed deer on Anticosti Island using occurrences from an aerial survey and forest vegetation maps

Abstract

Habitat-selection analysis involves a comparison between the proportions of different cover types that are used by the animal and the proportions that are available. Telemetry locations or animal occurrences (e.g., from aerial surveys) can provide information on habitat utilization. With telemetry data, a classical approach involves computing habitat use at the individual location sites or inside fixed circle buffers applied to the sites. We used this approach (200 m radius circles) on data from a systematic aerial survey on Anticosti Island, where 260 groups of white-tailed deer (Odocoileus virginianus) (374 animals) were counted in a 270-km2 block. We compared the selection indices obtained from site occurrences with those of two approaches that define areas of high intensity (animal concentrations): 50% fixed kernels (0.5–2 km bandwidth) and the local K function (0.5–2 km distance). The results were very consistent among the three sets of approaches, with the same cover types generally identified as those having the highest or lowest indices. White-tailed deer preferred forest stands where balsam fir (Abies balsamea) was present as high regeneration or was dominant in the tree layer (>50% basal area) and stands at the regeneration stage. In the studied landscape, there seems to be a wide range of spatial scales where the selection process can be analysed from aerial survey data.