ECOLOGY AND BIOLOGY OF THE PACIFIC WALRUS 



21 



The southernmost record in the Bering Sea is of a single male taken at Atka, in 

 the central Aleutian Islands in 1972 (K. W. Kenyon, personal communication). 



December (Fig. 15) 



The northern part of the Bering Sea, south to about 63°N, becomes covered by 

 new ice usually in mid- to late December. Before that ice develops, large herds of 

 walruses are present in Bering Strait and in the vicinity of St. Lawrence Island. 

 Apart from one sighting near Unalaska Island in 1976 (C. A. Smith, personal 

 communication), there are no data concerning distribution or movements 

 elsewhere in or near the southern Bering Sea at that time. 



Environmental Factors 



A close correlation between the distribution of the Pacific walrus population 

 and the extent and character of the pack ice is apparent from the foregoing 

 summary. Throughout the year, these animals occur mostly in or near the 

 southern periphery of the pack or in other areas within the pack where open 

 leads and thin ice are continually present. They are not generally found where 

 heavy, snow-covered ice occupies more than 80% of the sea's surface. Where 

 much of the cover is young, gray ice, the animals' movements are almost unre- 

 stricted, for they are able to create breathing holes at will by breaking up 

 through ice up to 20 cm thick (Fig. 16). Although they can maintain holes in 

 heavy, snow-covered ice, they seem to avoid areas where this is necessary, such 

 as in the Chukchi Sea in winter, the southern Gulf of Anadyr, and areas north 

 and east of St. Lawrence Island. 



As Brooks (1954) observed, their migrations through Bering Strait in spring 

 seem to coincide with the most favorable ice movements. Southerly winds and 

 sea surface currents cause the ice to drift northward through the Strait much of 

 the time in May and June, and any walrus that rests on it is transported toward 

 the north without the slightest effort. Strong winds from the north and east also 

 are not uncommon in spring in the northern Bering Sea, and by their force the 

 ice often is moved in the "wrong" direction for several days or weeks at a time. 

 Since ice seems to be used as a resting place, irrespective of the direction in which 

 it is moving, this southerly drift often works to the disadvantage of the migrants. 

 The principal progress in migration, however, is achieved by swimming. A 

 walrus swimming at its normal speed of about 10 km/h could easily complete the 

 journey from Bristol Bay to Wrangell Island (about 1,800 km) in 7 to 8 days, if it 

 did not stop to rest. However, these animals seem not to progress with such 

 purpose but to travel at a more leisurely rate, feeding and swimming for a time, 

 then resting on the ice nearby. My impression from field observations is that they 

 swim most purposively during stormy periods and when passing through areas 

 where there is no ice on which to rest. 



Conditions during the southward migration in autumn ordinarily are much 

 different from those in the spring. Although persistent northerly winds blow the 

 ice in the "right" direction much of the time, most of that ice is too thin to 

 support resting walrus herds, and they usually swim ahead of it in the open sea. 

 On arriving in the vicinity of a traditional hauling ground, they go ashore to rest, 

 or they support themselves with their tusks along the edge of the shorefast ice. 



