ECOLOGY AND BIOLOGY OF THE PACIFIC WALRUS 



243 



survey in an acceptably short period of time (Chapman et aL 1977). The walrus 

 is no exception, and indirect methods for estimating population size, such as 

 mark-recapture and interpretation of catch statistics, have not been developed 

 sufficiently to supplant direct counts. 



Mark-Recapture 



Some efforts have been made to develop methods for marking walruses but 

 principally for the sake of tracing their migrations, rather than estimating 

 population size. Mansfield (1958a) marked 115 animals in the Canadian Arctic 

 with a metal disk-tag having a harpoon-like point that penetrated the skin. 

 Krylov (1965Z?) used a similar tag to mark 500 walruses in the Bering and 

 Chukchi seas. Brooks (1954) marked 11 calves in Bering Strait with a metal 

 strap-tag attached to the hind flipper, and J. W. Brooks, K. W. Kenyon,and I 

 (unpublished data) applied 12 more of those tags to males at Round Island in 

 1958. None of the Canadian, Soviet, or American tags has been recovered, 

 probably because too few animals were marked. The metal tags also may have 

 caused inflammation of the tissues and damage from freezing, both of which 

 could result in sloughing of the tag, not long after application. The technical 

 difficulties in marking a large enough proportion of the population for 

 estimation of its size by the mark-recapture method appear insurmountable at 

 present, but the possibility of marking large numbers on the coastal hauling 

 grounds in autumn should be explored. Some lesser amount of marking would be 

 appropriate also for tracing migrational patterns, as noted earlier, and this 

 probably can be accomplished in the near future by satellite tracking. 



Catch Statistics 



Interpretation of walrus catch statistics appears to have little value for 

 estimating actual population size, although it may be useful as an indicator of 

 minimum size and of major changes in status. Fay (1957) estimated that the 

 population before 1860 must have contained at least 200,000 animals, based on 

 the reported size of the late 19th century harvests and on assumptions as to the 

 composition and gross productivity of the population. That the v/alrus popula- 

 tion declined drastically by 1880 as a result of those harvests was widely recog- 

 nized. Marked decline was indicated also by the poor hunting success and 

 consequent mass mortality of Eskimos at St. Lawrence Island in the winter of 

 1878-79 (Rosse 1883; Allen 1895; Petrof 1900). Until then, the islanders' annual 

 catch of walruses had been sufficient to support about 1,500 people, about half 

 to two-thirds of whom died in the famine that winter (Muir 1917). The scarcity 

 of walruses there at that time apparentiy was partly due to adverse ice and 

 weather conditions which excluded the animals from some areas and prevented 

 the hunters from ranging far from home (J. Otiohuk, personal communication). 

 Had the walrus population been larger, however, adequate numbers of animals 

 probably would have been available, for that kind of reliability would have been 

 a prerequisite for the founding of the walrus hunting culture on St. Lawrence 

 Island more than 2,000 years earlier. 



The irregular and occasionally insufficient numbers of walruses taken at St. 

 Lawrence Island in the 1940's to 1960's (Table 39) also suggested that the walrus 

 population at that time was smaller than was required for reliable subsistence 



