Multistate Capture and Search Data from the Southern Beaufort Sea Polar Bear Subpopulation in Alaska, 2001-2016

The southern Beaufort Sea mark-recapture study area was partitioned into four spatial areas (states). The states were defined as large spatial regions to summarize interannual polar bear movements and the spatial distribution of capture effort, with the expectation that the combination would help explain heterogeneity in capture probabilities. The area in which most capture effort and all captures occurred during the study period 2001–2016 was delineated using longitude on the eastern and western extremes and distance from the Alaska mainland shore to the north. The eastern boundary at -140.93 degrees longitude and the western boundary at -158.49 degrees longitude were selected just beyond the most extreme longitudes at which a polar bear capture occurred. The northern boundary was defined based on search effort distance from the mainland shore. Aircraft locations while searching for bears were filtered to remove implausible locations, aircraft speeds, etc. and the distance to shore of the remaining locations was computed. The 99.5th percentile of the aircraft distances from shore equaled 104,606 m and was selected as the northern boundary as it encompassed the capture location farthest from shore and only excluded a small number of effort locations that were unusually distant from the coast. The study area was divided into eastern and western sections by the longitude -151.00 degrees, selected to lie between the operational bases of Utqiağvik, AK and Deadhorse, AK, because relatively few captures occur near that longitude. Each eastern and western section was divided into nearshore and offshore regions using a distance from shore. Three distances from shore were selected as the 50th, 67th, and 80th percentiles of the aircraft distances from shore, i.e., 31,135 m, 43,246 m, and 54,562 m respectively. Three distances were selected to allow investigation of the robustness of estimates to the value selected. The four spatial states were therefore defined by the four combinations of east/west and nearshore/offshore (i.e., nearshore-east, nearshore-west, offshore-east, and offshore-west). A fifth spatial state was defined as all areas not in States 1-4. Aircraft GPS flight-tracking data were used to compute the annual distances flown in each of the four states while searching for bears. The state in which each bear capture occurred was determined based on the capture location. Similarly, each telemetered bear was assigned to a single state each year in which location tracking data were available. Telemetered bears were assigned to the state in which a majority of their daily locations occurred during the field season. Daily locations of telemetered bears were estimated using a continuous-time correlated random walk model (R library Crawl) that input the highest-quality daily filtered tracking locations.