Projections of polar bear (Ursus maritimus) sea ice habitat distribution in the polar basin during the 21st century were developed to understand the consequences of anticipated sea ice reductions on polar bear populations. We used location data from satellite‐collared polar bears and environmental data (e.g., bathymetry, distance to coastlines, and sea ice) collected from 1985 to 1995 to build resource selection functions (RSFs). RSFs described habitats that polar bears preferred in summer, autumn, winter, and spring. When applied to independent data from 1996 to 2006, the RSFs consistently identified habitats most frequently used by polar bears. We applied the RSFs to monthly maps of 21st‐century sea ice concentration projected by 10 general circulation models (GCMs) used in the Intergovernmental Panel of Climate Change Fourth Assessment Report, under the A1B greenhouse gas forcing scenario. Despite variation in their projections, all GCMs indicated habitat losses in the polar basin during the 21st century. Losses in the highest‐valued RSF habitat (optimal habitat) were greatest in the southern seas of the polar basin, especially the Chukchi and Barents seas, and least along the Arctic Ocean shores of Banks Island to northern Greenland. Mean loss of optimal polar bear habitat was greatest during summer; from an observed 1.0 million km2 in 1985–1995 (baseline) to a projected multi‐model mean of 0.32 million km2 in 2090–2099 (−68% change). Projected winter losses of polar bear habitat were less: from 1.7 million km2 in 1985–1995 to 1.4 million km2 in 2090–2099 (−17% change). Habitat losses based on GCM multi‐model means may be conservative; simulated rates of habitat loss during 1985–2006 from many GCMs were less than the actual observed rates of loss. Although a reduction in the total amount of optimal habitat will likely reduce polar bear populations, exact relationships between habitat losses and population demographics remain unknown. Density and energetic effects may become important as polar bears make long‐distance annual migrations from traditional winter ranges to remnant high‐latitude summer sea ice. These impacts will likely affect specific sex and age groups differently and may ultimately preclude bears from seasonally returning to their traditional ranges.

N. 2003. Functional responses in polar bear habitat selection. -Oikos 100: 112-124.Habitat selection may occur in situations in which animals experience a trade-off, e.g. between the use of habitats with abundant forage and the use of safer retreat habitats with little forage. Such trade-offs may yield relative habitat use conditional on the relative availability of the different habitat types, as proportional use of foraging habitat may exceed proportional availability when foraging habitat is scarce, but be less than availability when foraging habitat is abundant. Hence, trade-offs in habitat use may result in functional responses in habitat use (i.e. change in relative use with changing availability). We used logistic and log-linear models to model functional responses in female polar bear habitat use based on satellite telemetry data from two contiguous populations; one near shore inhabiting sea ice within fjords, and one inhabiting pelagic drift ice. Open ice, near the ice edge, is a highly dynamic habitat hypothesised to be important polar bear habitat due to high prey availability. In open ice-polar bears may experience a high energetic cost of movements and risk drifting away from the main ice field (i.e. trade off between feeding and energy saving or safety). If polar bears were constrained by ice dynamics we therefore predicted use of retreat habitats with greater ice coverage relative to habitats used for hunting. The polar bears demonstrated season and population specific functional responses in habitat use, likely reflecting seasonal and regional variation in use of retreat and foraging habitats. We suggest that in seasons with functional responses in habitat use, polar bear space use and population distribution may not be a mere reflection of prey availability but rather reflect the alternate allocation of time in hunting and retreat habitats.M. Mauritzen, Norwegian Polar Inst., Zool. Mus.,

In environments with high spatiotemporal variability in resources, animals may exhibit nomadic movements for resource tracking as opposed to long-term area fidelity. Polar bears (Ursus maritimus) inhabit the dynamic sea ice, preying on seals, and demonstrate considerable intraspecific variation in space use. We studied patterns of fidelity and annual range size for 74 adult female polar bears captured in the Norwegian Arctic that were tracked for up to 5 years using satellite telemetry data. We used the autocorrelation structure of movements and distance between observations at a 1-year interval as measures of fidelity. The female polar bears had a circannual migration pattern. Annual range size varied with reproductive state and geographic location of the range. Females entering maternity dens had smaller annual ranges than females not entering dens. Nearshore females had smaller annual ranges than pelagic females, demonstrating different space-use strategies. Repeatability of movement patterns indicated strategy specialization. We suggest that the different space-use strategies result from variation in habitat and prey selection and in sea-ice dynamics. Factors affecting population and predator-prey dynamics may interact differently with the different space-use strategies and yield strategy-dependent outcomes, therefore a knowledge of movement strategies may be important for understanding polar bear population dynamics.Résumé : Dans les milieux où la variabilité spatiotemporelle des ressources est élevée, les animaux ont parfois des habitudes nomades pour chercher leurs ressources et ils ne montrent pas de fidélité à long terme à une région en particulier. Les Ours blancs (Ursus maritimus) habitent les glaces maritimes en mouvement, chassant les phoques, et on observe une variation intraspécifique considérable dans leur utilisation de l'espace. Nous avons étudié les patterns de fidélité et la taille annuelle des domaines de 74 femelles adultes de l'Ours blanc capturées dans l'Arctique norvégien où elles ont été suivies par télémétrie par satellite, certaines jusqu'à 5 ans. Comme mesures de la fidélité, nous avons utilisé la structure des autocorrélations des déplacements et des distances entre les points d'observation à intervalles de 1 an. Les Ours blancs femelles ont un pattern de migration circannuel. La taille du domaine annuel d'un ours varie en fonction de son état reproducteur et en fonction de la position géographique de son domaine. Les femelles qui gagnent les terriers de mise bas ont des domaines annuels plus restreints que ceux des femelles qui n'entrent pas en terrier. Les femelles près des côtes ont des domaines annuels plus petits que les femelles pélagiques, ce qui reflète l'existence de diverses stratégies d'utilisation de l'espace. La reproductibilité des patterns de déplacement indique l'existence d'une spécialisation de ces stratégies. Nous croyons que les diverses stratégies d'utilisation de l'espace sont dues à des variations dans l'habitat et la sélection des proies et à des variations da...

Summary1. Animal populations, defined by geographical areas within a species' distribution where population dynamics are largely regulated by births and deaths rather than by migration from surrounding areas, may be the correct unit for wildlife management. However, in heterogeneous landscapes varying habitat quality may yield subpopulations with distinct patterns in resource use and demography significant to the dynamics of populations. 2. To define the spatial population structure of polar bears Ursus maritimus in the Norwegian and western Russian Arctic, and to assess the existence of a shared population between the two countries, we analysed satellite telemetry data obtained from 105 female polar bears over 12 years. 3. Using both cluster analyses and home-range estimation methods, we identified five population units inhabiting areas with different sea-ice characteristics and prey availability. 4. The continuous distribution of polar bear positions indicated that the different subpopulations formed one continuous polar bear population in the Norwegian and western Russian Arctic. Hence, Norway and Russia have a shared management responsibility. 5. The spatial population structure identified will provide a guide for evaluating geographical patterns in polar bear ecology, the dynamics of polar bear-seal relationships and the effects of habitat alteration due to climate change. The work illustrates the importance of defining population borders and subpopulation structure in understanding the dynamics and management of larger animals.

No abstract