246 



NORTH AMERICAN FAUNA 74 



Table 40. Size of the Pacific walrus population, estimated from aerial 

 surveys between 1958 and 1975. 



Estimated population 

 (thousands) 



Date 



Range 



Median 



Source 



August-September 1958 



— 



40 



P. G. Nikulin (in Fedoseev 1962) 



February-March 1960 



78-113 



95 



Kenyon (1960a) 



April 1960 



70-100 



85 



Kenyon (1960a) 



September-October 1960 





50 



Fedoseev (1962) 



March 1961 



70-100 



85 



K. W. Kenyon (unpublished data) 



September-October 1964 



47-71 



59 



Gol'tsev (1968) 



April 1968 



73-110 



92 



K. W. Kenyon (unpublished data) 



September-October 1970 





101 



Gol'tsev (1972) 



April 1972 



85-162 



123 



Kenyon (1972) 



September-October 1975 



140-200 



170 



J. A. Estes and V. N. Gol'tsev 









(unpublished data) 



American surveys conducted about the same time. The reasons for this are not 

 yet clear but seem to be related to procedural differences. Whereas the American 

 estimates were based on extrapolation from counts along randomly placed strip 

 transects, the Soviet estimates were based on photographic counts of animals on 

 the hauling grounds plus conservative guesses as to the numbers at sea (not 

 sampled). Extrapolation from the American strip transects to the total area 

 occupied could have led to overestimates or underestimates, due to the nonran- 

 dom distribution of the animals (Estes and Gilbert 1978). The Soviet counts only 

 of animals on the haulouts probably were low, due to underestimation of the 

 proportion of the population at sea. In both the Soviet and the American surveys, 

 the population estimates were based mainly on animals that were sighted out of 

 the water, on land or ice, for there was no known method to account for those 

 under water and not visible to the observers. This probably contributed sub- 

 stantially to underestimation in all instances. 



Because walruses are large and their coloration contrasts well with the 

 background, especially when they are lying on the ice, they are easily sighted 

 from low-flying aircraft. Walruses seem to spend much more time out of the 

 water in all seasons than do most other pinnipeds (Fay and Ray 1968), but they 

 are more widely distributed than is optimal for short-term aerial surveys, and 

 their patchy distribution can lead to large errors in extrapolation to areas not 

 covered by the surveys. Their tendency to congregate requires aerial photog- 

 raphy for optimal counts of large herds. Insufficient knowledge of their activity 

 rhythm has made estimation of numbers not sighted (in the water) difficult at 

 best and generally untenable. 



Whereas the Soviet and American estimates shown in Table 40 were derived by 

 different methods and gave different results, each set suggests a trend of rapid 

 increase of the population over the past 20 years. Some of that trend could have 

 been due to improvement in survey methods, greater knowledge of distribution, 

 and increased effort. The more recent surveys have been more extensive than 

 earlier ones, and have made greater use of aerial photography and more sophis- 



