What exactly is an urban bear? Characterizing black bear space use relative to an anthropogenic landscape gradient

Successful conservation and management of large mammals in the 21st century demands an understanding of how they interact with humans outside of protected areas across an increasingly anthropogenic landscape. A recent proliferation of studies focused on mammals operating within a range of human-altered systems highlights the need for a standardized approach in assessing human influence on the landscape. We propose a straightforward method to generate an anthropogenic-influence metric, at a 30-meter spatial resolution, composed of three readily-available inputs (land-use land-cover, human population density, and housing unit density) aggregated into an explicit, flexible rule set within a fuzzy inference framework. Coupled with Resource Selection functions, we will apply our model across northwestern New Jersey to classify 6053 spatially-explicit black bear captures and mortality recoveries as bears using space in, and exploiting resources from, Wildland, Exurban, Rural, Suburban, or Urban areas on the day of their encounter. Using location data from 35 GPS-instrumented bears during 2008-2012, we calculated average maximum linear distance between locations for each bear day with ≥ 7 locations (n = 6857 days; 107,344 locations) to determine sex-, and season-specific differences in the magnitude of average maximum daily movement. The average maximum distance between any two locations per 24-hour period significantly differed between seasons: spring (1111.7 m; as reference), summer (1667.6 m; p = 0.002), fall (1398.4 m; p = 0.044), and pre-denning (684.6 m; p = 0.016); and between sexes during spring (p = 0.035) and summer (p = 0.015). Appropriate maximum daily movement buffers will be overlaid upon all geo-referenced encounters to estimate the anthropogenic influence of the projected space accessible to bears on the day of capture or mortality. Systematic characterization of anthropogenic landscape gradients will facilitate inter-regional and/or interspecific comparisons and allow for the explicit characterization of behavioral states, necessary for behavioral state-specific demographic modeling