Polar Bears In Depth
Denning
Page 4 of 8
Fidelity to Denning Locales. Although there are no historical data regarding denning fidelity, it has logically been assumed that concentrated denning areas are maintained by fidelity of individual females to those sites (Uspenski and Chernyavski 1965; LØnØ 1970; Uspenski and Kistchinski 1972; Larsen 1985). Pregnant females return, it is assumed, to areas where they have successfully denned in the past.
The greatest number of records of den-site fidelity derives from the Beaufort Sea. There, 27 polar bears were followed to more than one suspected or confirmed maternity den (Amstrup and Gardner 1994). One radio-collared polar bear was followed to four maternal denning sites, 7 were followed to three dens each, and 19 to two dens. Confirmed sequential dens were separated from their precursors by a mean of 308 km (SD = 262, n = 30), and the minimum distance was 23 km. Distances separating sequential land dens were not different from those separating sequential pack-ice dens. Bears that denned once on pack ice were more likely to den on pack ice than on land in subsequent years, and vice versa. Similarly, bears were faithful to general geographic areas. Those that denned once those that denned once in the eastern half of the Alaskan coast were more likely to den there than to the west in subsequent years. When all years were considered, denning polar bears preferred some areas, but no areas were used by collared bears in all years. Weather, ice conditions, and prey availability, all of which varied annually, probably determined where bears denned. Those annual variations and the long-distance movements of polar bears (Amstrup et al. 1986, 2000; Garner et al. 1990) make seasonal recurrence at exactly the same location unlikely.
Although Beaufort Sea polar bears were not faithful to particular denning sites, data on den distribution and fidelity of females to denning areas indicated there are both "pack-ice" and "land-denning" bears. Den substrate switching appeared to be limited. This segregation may have begun when some females were prevented from reaching land in the fall.
The only other region where data are available on fidelity to denning areas is Hudson Bay. There, pregnant females initiate their overwinter denning period in earthen dens they occupy in summer. During winter, they burrow into adjacent snow drifts (Watts and Hansen 1987). The presence of hundreds of earthen dens in the region suggests a long tradition of use. In three instances, cavities were reused by different bears, but no observations of reuse of a cavity by the same bear were reported (Ramsay and Stirling 1990). On average, bears followed to second dens chose locations 27 km (452 km) from previous attempts (Ramsay and Stirling 1990). Hence, there was greater fidelity to local areas than in the Beaufort Sea, but site-specific philopatry was not apparent.
Despite general fidelity to local areas, the overall distribution of denning along the west coast of Hudson Bay shifted markedly over a 20-year period (Ramsay and Stirling 1990). Because bears have the navigational skills to return to the same area, the reason for the shift is not clear. A similar shift appears to be occurring in the Beaufort Sea region as well, however. During the 1990s, more females appeared to choose den locations in the central and western portions of the northern Alaska shore than during the 1980s (S. C. Amstrup, unpublished data). Such shifts must reflect changing ice formation and ablation patterns, food availability, or other unidentified ecological factors. Harington (1968), Larsen (1985), and LØnØ (1970) concluded that variation in the local pattern of sea-ice movements during the preceding summer and autumn accounts for annual changes in the distribution of winter dens. Multiple-year trends in sea-ice patterns could clearly alter denning and other behavioral patterns.
The greatest number of records of den-site fidelity derives from the Beaufort Sea. There, 27 polar bears were followed to more than one suspected or confirmed maternity den (Amstrup and Gardner 1994). One radio-collared polar bear was followed to four maternal denning sites, 7 were followed to three dens each, and 19 to two dens. Confirmed sequential dens were separated from their precursors by a mean of 308 km (SD = 262, n = 30), and the minimum distance was 23 km. Distances separating sequential land dens were not different from those separating sequential pack-ice dens. Bears that denned once on pack ice were more likely to den on pack ice than on land in subsequent years, and vice versa. Similarly, bears were faithful to general geographic areas. Those that denned once those that denned once in the eastern half of the Alaskan coast were more likely to den there than to the west in subsequent years. When all years were considered, denning polar bears preferred some areas, but no areas were used by collared bears in all years. Weather, ice conditions, and prey availability, all of which varied annually, probably determined where bears denned. Those annual variations and the long-distance movements of polar bears (Amstrup et al. 1986, 2000; Garner et al. 1990) make seasonal recurrence at exactly the same location unlikely.
Although Beaufort Sea polar bears were not faithful to particular denning sites, data on den distribution and fidelity of females to denning areas indicated there are both "pack-ice" and "land-denning" bears. Den substrate switching appeared to be limited. This segregation may have begun when some females were prevented from reaching land in the fall.
The only other region where data are available on fidelity to denning areas is Hudson Bay. There, pregnant females initiate their overwinter denning period in earthen dens they occupy in summer. During winter, they burrow into adjacent snow drifts (Watts and Hansen 1987). The presence of hundreds of earthen dens in the region suggests a long tradition of use. In three instances, cavities were reused by different bears, but no observations of reuse of a cavity by the same bear were reported (Ramsay and Stirling 1990). On average, bears followed to second dens chose locations 27 km (452 km) from previous attempts (Ramsay and Stirling 1990). Hence, there was greater fidelity to local areas than in the Beaufort Sea, but site-specific philopatry was not apparent.
Despite general fidelity to local areas, the overall distribution of denning along the west coast of Hudson Bay shifted markedly over a 20-year period (Ramsay and Stirling 1990). Because bears have the navigational skills to return to the same area, the reason for the shift is not clear. A similar shift appears to be occurring in the Beaufort Sea region as well, however. During the 1990s, more females appeared to choose den locations in the central and western portions of the northern Alaska shore than during the 1980s (S. C. Amstrup, unpublished data). Such shifts must reflect changing ice formation and ablation patterns, food availability, or other unidentified ecological factors. Harington (1968), Larsen (1985), and LØnØ (1970) concluded that variation in the local pattern of sea-ice movements during the preceding summer and autumn accounts for annual changes in the distribution of winter dens. Multiple-year trends in sea-ice patterns could clearly alter denning and other behavioral patterns.