ECOLOGY AND BIOLOGY OF THE PACIFIC WALRUS 



23 



Their progress is somewhat more rapid in autumn than in the spring migration, 

 presumably due to the greater proportion of time spent in swimming toward 

 their goaL 



Adult females with dependent young are the most strongly migrator)', judging 

 from their near absence from the Bering Sea in summer. That adult males are 

 least migratory is indicated by their occurrence in large numbers at traditional 

 all-male hauling grounds on Rudder, Meechken, Arakamchechen, Big Diomede, 

 and Round islands (and formerly, on the Pribilof Islands) throughout the 

 summer. Female northern fur seals {Callorhinus ursimis) also are more strongly 

 migratory than males (Kenyon and Wilke 1953), and this is true also of many 

 species of birds. Dorst (1961) suggested that the migration of males may be 

 inhibited by gonadal recrudescence following the breeding season. Certainly, 

 recrudescent changes in the testes are well under way in most adult male 

 walruses at the time of the spring migration, whereas many of the subadults and 

 juveniles are in the peak of their rut at that time. There are indications that the 

 migratory males are predominantly of those younger age classes; most of the 

 males that remain in the Bering Sea in summer are the older adults (Fedoseev 

 and Gol'tsev 1969). 



Subadults of either sex seem most inclined to wander or to be diverted by 

 irregular ice movements. Extralimital strays and animals sighted in the 

 peripher>' of the Bering-Chukchi region are most often subadults or young 

 adults. Since 1930, the limits of wandering in the south have been Honshu, Japan 

 (Scheffer 1958), northeastern Okhotsk Sea (Moiseev 1951 in Heptner et al. 1976), 

 Atka Island in the Aleutian Chain (K. W. Kenyon, personal communication), 

 and Yakutat Bay in the Gulf of Alaska (D. G. Calkins, personal 

 communication). In the north, the limits have been the Bear Islands, East 

 Siberian Sea (Chapskii 1940), 75°52'N, 164°15'W in the central Chukchi Sea 

 (C. A. Barnes, personal communication), and Bathurst Inlet, Canada 

 (Harington 1966). 



The distribution appears to be limited by bathymetry (Fig. 17). The animals 

 ordinarily do not occur in areas where the water is more than 80 to 100 m deep, 

 presumably because they cannot feed efficiently or are not capable of diving in 

 greater depths (Vibe 1950). I have observed walruses feeding in the Bering Sea in 

 areas that ranged in depth from 25 to 79 m. 



In winter, the distribution appears to be influenced also by ice conditions, as 

 noted earlier. Comparison of the distribution of walruses in March (Fig. 6) with 

 that of the pack ice (Fig. 18) indicates that the animals tend to concentrate in 

 areas where the ice is darkest (thinnest or most dispersed) and to be scarce or 

 absent from the areas where it is whitest (heaviest and most compact). The ice 

 pattern indicated in Fig. 18 is about average for February and March. Although 

 the southward extent of the ice varies widely from year to year in the eastern 

 Bering Sea, the overall pattern of light and heavy ice prevails in most years 

 (Brower et al. 1977; J. J. Burns, L. H. Shapiro, and F. H. Fay, unpubHshed 

 data). An exception to this may have occurred in 1967, w^hen the southern edge 

 of the pack in mid- April was in the vicinity of St. Lawrence Island and Norton 

 Sound. At that time, most of the walrus population was in the St. Lawrence 

 Island to Bering Strait region (J. J. Burns, personal communication). 



The factors limiting distribution in summer are less clear, other than that the 

 100-m isobath and extent of the pack ice play major restrictive roles to the north. 



