each third year after approximately age 10. Brooks (1954) and Fay (1955) 
report, respectively, crude annual birth rates (calves per adult female) 
of 33 percent and 39 percent. These rates tend to support the triennial 
breeding. 
At birth,the young are approximately 4 feet long and are capable 
of swimming immediately. A calf depends on the female for an exclusively 
milk diet for at least 18 months, and it is not at all unusual to find 
cows accompanied by 2-year-old calves. There is an extremely close tie 
between mother and calf, and if one is killed the other remains in the 
vicinity as long as possible. Calves deprived of their mothers before 
being weaned are assumed by natives to be adopted by another female. No 
evidence of adoption has been presented, and such adoption is so unlikely 
as to be of no importance to the population; that is, virtually all 
calves deprived of their mothers die. 
Mortality 
Mortality among walruses is probably almost entirely caused by 
human depredations. Undoubtedly some walruses, especially calves, are 
killed by killer whales, occasionally some are killed by polar bears, 
and walruses of all ages occasionally may be crushed by ice. There is 
no indication of food shortage and consequent malnutrition. None of 
these agencies is considered of importance at the present time. The 
details of human harvest will be presented in a later section. At this 
point, however, the fact of mortality and not the agency of mortality is 
of primary concern. i 
Fay (1955) has constructed a "catch curve" of male walruses 
taken near Gambell and Savoonga on St. Lawrence Island in the spring of 
1953. His data consist of 155 bulls older than calves, and his ages are 
based on cementum layers in the teeth. The approximate survivorship 
curve that he interpolates has a slope of 0.12 and can be interpreted 
as an average annual mortality rate of 12 percent. 
The data on which the curve is based were from walruses taken 
by Eskimo hunters. Because bull walruses are hunted primarily for their 
ivory rather than for meat, there is a rather constant selection toward 
the older age classes which have the largest ivory tusks. Thus, the 
slope of the survival curve probably tends to be underestimated, and the 
mortality rate estimated is thus too low. 
The "catch curves" for females are based on even fewer animals 
than those for the males. In addition, the method of aging is based en- 
tirely on tusk length, and thus probably is less accurate than is the 
case with the males. However, for ages 4 to 15 inclusive, where the 
aging method seems the most satisfactory, the slopes of the survivorship 
curves give estimates of 15 percent mortality in 1953 and 16 percent in 
1958. Data for 1953 are from Brooks (1954) and for 1958 from Kenyon; 
both sets are from Little Diomede Island. Combining the data for both 
years yields an estimate of 15 percent, which is greater than that for 
10 
