Polar bears tend not to cache prey animals they have killed like grizzly bears do (Stirling 1974; DeMaster and Stirling 1981; Stirling and Derocher 1990). This may be another reason why they consume the highest reward portion of their prey first. Although they have not been observed to cache, polar bears are surplus killers. Stirling and Derocher (1990) reported seeing a polar bear kill two seals within an hour of feeding extensively on another seal. Neither of the latter two seals killed was eaten. Stirling and Øritsland (1995) also have reported surplus killing in polar bears. I once observed a young male polar bear still-hunting at a breathing hole on new autumn ice. There was a partially consumed seal nearby, and between that feeding site and where he was still-hunting were three freshly killed ringed seals stacked like cordwood. When my helicopter approached the bear to capture him, he abandoned his still-hunting site, ran to the pile of dead seals, and covered them with his body as if to protect his stash. This bear apparently had eaten his fill from the first seal but was continuing to hunt, catch, and stack seals despite a low probability that he would consume much of them.

An interesting adaptation to the carnivorous diet, and a difference between polar bears and other temperate and arctic bears, is that only the pregnant females enter dens for the entire winter. Other members of the population continue to hunt seals on the sea-ice throughout the winter. The year-around availability of seals allows denning in polar bears to be strictly a reproductive strategy (affording an acceptable environment for neonates), whereas in most bears it is largely a foraging strategy (avoiding the winter period of food unavailability).

Like other ursids, polar bears will eat human refuse (Lunn and Stirling 1985), and when trapped on land for long periods they will consume coastal marine and terrestrial plants and other terrestrial foods (Derocher et al. 1993). The significance of other foods to polar bears may be limited, however (Lunn and Stirling 1985; Derocher et al. 1993). Over most of their range, polar bears have little opportunity to take foods of shoreline or terrestrial origin. Derocher et al. (1993) found that 31% of pregnant polar bears in the Hudson Bay area fed on berries before denning in autumn. The significance of this to their productivity was not known. Ramsay and Hobson (1991) and Hobson and Stirling (1997) differed in opinions of the value of supplemental terrestrial food. In general, the significance of terrestrial foraging to polar bears is poorly understood.

Clearly the value of alternate foods for polar bears depends on their richness and digestability. Polar bears are poorly equipped to consume and digest most plant parts (Bunnell and Hamilton 1983), and it seems likely that except for fruiting bodies, plants will contribute little to their energy balance. Lunn and Stirling (1985) found that polar bears using human refuse at a dump maintained their weight or lost less weight than bears not using anthropogenic foods. Some bears using the dump even gained weight, but the supplemental food did not appear to confer a reproductive advantage (Lunn and Stirling 1985). Derocher et al. (2000) reported that some polar bears in Svalbard have become adept at catching reindeer (Rangifer tarandus). Considering the high digestibility of meat, it seems plausible that if readily available, reindeer could be an important alternate food of polar bears. Likewise, in the Beaufort Sea, dozens of polar bears each year have developed a habit of gathering at the butchering sites of bowhead whales (Balaena mysticetus) that are killed by local Native people. The value of this alternate food is apparently great, as nearly every bear seen near whale carcasses in autumn is obese.