Causes of natural mortalities among polar bears are largely unknown. Because polar bears spend most of their time on drifting seaice, dead animals are likely to go undiscovered and cause of death for animals that are discovered is seldom discernible. Therefore, we are forced to extrapolate from a very few observations to understand natural mortality patterns and causes. Accidents involving unskilled young must be a common cause of natural death in the harsh arctic environment (Derocher and Stirling 1996). Starvation of independent young as well as very old animals must account for much of the natural mortality among polar bears. Age-specific differences in hunting success rates have been reported by Stirling and Latour (1978) for polar bears of the central Canadian High Arctic. Cubs of any age spent little time hunting, and were not effective at taking seals in the spring of the year. During summer, the success rate of 2-year-olds was similar to that of adults, although they spent much less time hunting. Young of the year and yearlings were less successful than adults. Cubs abandoned prior to the normal weaning age of 2.5 years likely have poor survival (Stirling and LaTour 1978). That conclusion is corroborated by the dearth of observations of independent bears <2 years old in all populations except Hudson Bay. Also, age structure data show that subadults aged 2­5 years survive at lower rates than adults (Amstrup 1995), probably because they are still learning hunting and survival skills. I once observed a 3-year-old subadult that weighed only 70 kg in November. This was near the end of the autumn period in which Beaufort Sea bears reach their peak weights (Durner and Amstrup 1996), and his cohorts at that time weighed in excess of 200 kg. This young animal apparently had not learned the skills needed to survive and was starving to death. As they age, polar bears that avoid serious injury must simply get too old and feeble to catch food, and thus literally die of old age. Local and widespread climatic phenomena that make seals less abundant or less available also can significantly affect polar bear populations (Stirling et al. 1976; Kingsley 1979; DeMaster et al. 1980; Amstrup et al. 1986).

Injuries sustained in fights over mates or in predation attempts also may lead to natural mortalities of polar bears. Some injuries are immediately fatal. I have seen three instances where a bear has killed another and consumed it. Broken teeth and even broken jaws may frequently result from fighting and failed predation attempts. In brown and black bears, such injuries commonly are not life-threatening. L. Aumiller (Alaska Department of Fish and Game, pers. commun.) has observed several brown bears at Alaska's McNeil River Sanctuary with jaws that had broken and healed in a variety of distorted conformations. D. Garshelis (Minnesota Department of Natural Resources, pers. commun.) captured a 2-year-old black bear with a missing lower jaw. The jaw and all lower teeth were destroyed by gunshot wounds that had largely healed when Garshelis examined the bear in its winter den. The bear was radio-tracked through the following spring and summer and killed by a hunter the following autumn as a normal-size 3-year-old. Brown bears and black bears often survive on a diet including plant parts, fish, insects, small animals, and carrion. A videotape made by the hunter revealed how ingenious the young Minnesota black bear was in feeding without a lower jaw. These and other observations of injured brown and black bears (D. Moody, Wyoming Game and Fish Department, Pers. Commun.; M. Haroldson, USGS Interagency Grizzly Bear Study Team, pers. commun.) suggest they regularly survive with severely damaged mouth-parts, perhaps because of their great adaptability and the small particle size of most of their foods.