Factors Affecting the Observed Densities of Ringed Seals, <i>Phoca hispida</i>, in the Alaskan Beaufort Sea, 1996-99

Abstract

Aerial surveys were conducted during late May and early June 1996-99 in the central Beaufort Sea of Alaska, using strip-transect methods. The purpose of these surveys was to quantify and model the effects of environmental covariates on ringed seal counts and to provide density estimates that would be useful for evaluating trends in seal abundance. Total survey effort included 40-88 transect lines per year covering 1198-2701 km². Observed densities ranged from 0.81 seals/km² in 1996 to 1.17 seals/km² in 1999. We examined the effects of habitat, weather, and time of day on observed seal densities, using univariate chi-square goodness-of-fit tests. We also used a multivariate generalized linear model to estimate the relationship between seal counts and covariates. Three habitat-related variables - water depth, location relative to the fast ice edge, and ice deformation - had substantial and consistent effects. The highest densities occurred at depths between 5 and 35 m. Densities were also highest in relatively flat ice and near the fast ice edge, declining both shoreward and seaward of that edge. Univariate analysis suggested that observed densities were generally highest at about 1200 h Alaska daylight time, but time was not a significant variable in the generalized linear models. Analyses of the effects of weather factors on seal counts were inconclusive. This was likely at least partially because temperature and wind speed were measured at survey altitude, rather than on the ice surface, and surveys were conducted only in weather considered suitable for hauling out. The final multivariate model did not account for a substantial proportion of the variation in seal counts. We think this result was largely due to date-related variation in the proportion of seals hauling out, an issue our surveys were not suited to address.