No abstract

Climate change can influence interspecific interactions by differentially affecting species-specific phenology. In seasonal ice environments, there is evidence that polar bear predation of Arctic bird eggs is increasing because of earlier sea ice breakup, which forces polar bears into nearshore terrestrial environments where Arctic birds are nesting. Because polar bears can consume a large number of nests before becoming satiated, and because they can swim between island colonies, they could have dramatic influences on seabird and sea duck reproductive success. However, it is unclear whether nest foraging can provide an energetic benefit to polar bear populations, especially given the capacity of bird populations to redistribute in response to increasing predation pressure. In this study, we develop a spatially explicit agent-based model of the predator-prey relationship between polar bears and common eiders, a common and culturally important bird species for northern peoples. Our model is composed of two types of agents (polar bear agents and common eider hen agents) whose movements and decision heuristics are based on species-specific bioenergetic and behavioral ecological principles, and are influenced by historical and extrapolated sea ice conditions. Our model reproduces empirical findings that polar bear predation of bird nests is increasing and predicts an accelerating relationship between advancing ice breakup dates and the number of nests depredated. Despite increases in nest predation, our model predicts that polar bear body condition during the ice-free period will continue to decline. Finally, our model predicts that common eider nests will become more dispersed and will move closer to the mainland in response to increasing predation, possibly increasing their exposure to land-based predators and influencing the livelihood of local people that collect eider eggs and down. These results show that predator-prey interactions can have nonlinear responses to changes in climate and provides important predictions of ecological change in Arctic ecosystems.

Grizzly bears have recently become more common on the Arctic Islands in the Inuvialuit Settlement Region, concurrently with a period of environmental change. Over the last decade, grizzly bear – polar bear hybrids have been confirmed within this region, triggering extensive discussion and speculation regarding the impact of hybridization on the parent species. Through harvests, sightings, and captures, we document an increase in the presence of grizzly bears and combine field observations of hybrids with genetic analysis and parentage analysis to identify four first-generation (F1) hybrids and four offspring of F1 hybrids and grizzly bears (backcross-to-grizzly-bear individuals). We trace these eight hybrid individuals to a single female polar bear who mated with two grizzly bears. We sampled one of her mates on the sea ice in the High Arctic and deduced the genotype of the other from his five offspring. The two male grizzly bears are sires of both the F1 generation and the backcross-to-grizzly-bear generation. So what initially appeared to be a sudden spate of hybridization in the western Canadian Arctic originated with the unusual mating between three non-hybrid parents. The breakdown of species barriers may start with atypical mating preferences of select individuals; however, the story we present can be traced to a single female polar bear who, along with three of her known F1 offspring, has been killed.

We examined habitat characteristics of 101 polar bear (Ursus maritimus Phipps, 1774) den sites and 83 adjacent unoccupied sites in western Hudson Bay, Canada, between mid-August and early October 2001 and 2002. Bears denned almost exclusively in peat banks (n = 100) along the edges of creeks, rivers, and lakes adjacent to open lichen tundra sites. Den sites differed from unoccupied sites by having greater tree cover (P = 0.002), less moss cover (P < 0.001), and less herbaceous cover (P = 0.005). The presence of tree roots improved substrate stability, providing support to den structures. Den entrance azimuths were weighted toward a southeasterly aspect (P < 0.005), away from the prevailing northwest winds. To identify habitats with the greatest relative probability of having a den, a resource selection function (RSF) model was developed using remote sensing imagery and 1245 known den locations. High normalized difference vegetation index and brightness values derived from Landsat imagery, which were in close proximity to water, corresponded well with polar bear den sites. Identification of critical denning areas through the use of RSF will provide resource managers with a valuable tool for ensuring the protection of denning habitat, and consequently female bears and their young.