26 



NORTH AMERICAN FAUNA 74 



predator, and slow to reproduce. By using the moving ice, walruses are 

 continually transported to new feeding grounds while they rest. They can find 

 suitable hauling-out sites near almost any food source that they locate on the 

 Bering-Chukchi shelf. By staying with the ice, they are able to exploit the 

 benthic resources of nearly the entire shelf, rather than being limited to areas 

 within foraging range of coastal hauling grounds. 



A major advantage of ice also is its distance from terrestrial predators, es- 

 pecially man. This advantage was clearly demonstrated during the 18th and 

 19th centuries, when most of the herds that regularly frequented coastal and 

 insular sites in the Bering and Chukchi seas were extirpated (Fay 1957). Herds 

 that stayed with the ice sur\'ived in some numbers, probably because they were 

 less accessible and their location less predictable. In general, the same can be said 

 of walrus populations in other parts of the Arctic; some that resided almost 

 entirely on coastal hauling grounds in summer are now extinct or close to 

 extinction (Bel'kovich and Khuzin 1960; Reeves 1978). 



Concentrations and Migrations 



The use of ice as a means to exploit the Bering-Chukchi shelf requires most of 

 the walruses to migrate long distances during the annual ad\ ance and retreat of 

 the pack. In his description of those migrations, Belopol'skii (1939) theorized that 

 the population was not unified but was made up of three separate segments, 

 which he called the Kresta, Wrangell, and American groups. He supposed that 

 the Kresta group separated from the main population in April-May and that it 

 moved westw^ard along the southern coast of Chukotka, arriving at Kresta Bay in 

 June or July. The Kresta group was believed to retrace its route late in the 

 summer and pass northward toward Bering Strait, where it rejoined the rest of 

 the population. The main population was thought to divide into Wrangell and 

 American groups somewhere along the spring migration route, one travelling to 

 the vicinity of Wrangell and Herald islands and the other to Point Barrow and 

 eastward to Mackenzie Bay. Both were believed to arrive at their destinations in 

 July or August. On the return trip, in autumn, the Wrangell and American 

 groups were believed to reunite near Bering Strait and to be joined there by the 

 Kresta group. This concept of a tripartite summering population was upheld by- 

 Brooks (1954), who noticed two waves of migrants reaching Bering Strait in May 

 and June and presumed that these were the Wrangell and American groups. 



The data summarized in Figs. 4-15 indicate the existence usually of at least 

 two major concentrations during the winter in the Bering Sea, one near St. 

 Lawrence Island and the other in Bristol Bay. These concentrations have been 

 detected repeatedly in aerial and shipboard surveys conducted in February to 

 April 1960-61, 1968, 1972, and 1976. In each instance they were found to 

 comprise mainly adult females with their young, together with a few adult males 

 (Table 1). Another, smaller concentration repeatedly has been found near the 

 Pribilof Islands. I believe that the smaller group is part of the Bristol Bay 

 concentration, because it was not detected in February, and it seemed to move 

 westward during March and April (Figs. 5, 6, 7). It was composed mainly of 

 immature and subadult males (Table 1). A fourth concentration, near Cape 

 Navarin, was discovered in April 1972 by Kenyon (1972); apparently it was not 



