Table 13. — Estimates of the pup population 

 based on tag recoveries, year classes 

 1951-63, Pribilof Islands, Alaska 



A preliminary estimate based on tag 

 recoveries at age 2 only. 



The number of pups born, as estimated from 

 tag recoveries, has decreased steadily from 

 778,000 in 1959 to 340,000 in 1963 (table 13). 



Estinnates of the number of yearling males 

 are available for the 1961 and 1962 year classes 

 (table 15). These animals were tagged as 

 yearlings in late September and October of 

 1962 and 1963. The errors in selection (by 

 length) of yearling males for tagging, calcu- 

 lated from age of animals taken later in the 

 kill, were 2.9 and 5.5 percent for the 1961 and 

 1962 year classes, respectively. Age was de- 

 termined for 346 tagged males recovered from 

 the 1961 year class and 199 tagged males re- 



covered from the 1962 year class. The esti- 

 mates based on recoveries at ages 2, 3, and 4 

 increased with the increase in age. The reason 

 for the increase is not known. The recoveries 

 from females are too few to provide a basis for 

 estimates. 



Estimates based on sampling live pups .-- 

 The methods used in shearing and sampling 

 pups on St. Paul Island were described by 

 Roppel, Johnson, and Chapman (1965), and 

 Roppel, Johnson, Anas, and Chapman ( 1965). 



A total of 25,868 pups were sheared on St. 

 Paul Island in 1965. As in previous years, the 

 shearing and sampling nnethod was tested for 

 consistency of results by sannpling the pups on 

 each rookery twice, and obtaining ratios of 

 sheared to unsheared seals from groups of 25 

 pups each time. The resulting estimates- - 

 208,468 and 200,82 1- -were similar (table 16). 

 An additional estimate of 247,960 was nnadeby 

 finding the ratio of sheared to unsheared pups 

 in groups of 100 on all rookeries (table 17). 



The accuracy of estimates obtained from 

 sampling live pups was checked by making total 

 counts on three small rookeries and part of a 

 large one (table 18), The estimates based on 

 sampling by groups of 25 usually were lower 

 than the total counts. The estimates based on 

 sampling by groups of 100, however, were 

 higher than the counts on the small rookeries 

 but lower than the count on the large rookery. 

 The mean of the three estimates (two based on 

 samples of 25 and one on samples of 100) was 

 in good agreennent with the counts for the 

 smaller areas, but was only 86 percent of the 

 count on the large area. The number of pups 

 born, as estimated by the two sampling 

 methods, is given by rookery in table 19. When 

 corrected by the average difference between 

 estimates and counts (estimate = 91.3 percent 

 of the count), the nnean of the three estimates 

 gives 247,500 as the number of pups alive on 



Table 14. — Estimates of the pup pcjpulation based on tag recoveries from females, year classes 



1960-62, Pribilof Islands, Alaska . 



[n=femalee killed from each year class; t=tags applied to each year class; s=tag recoveries from 

 each year class; N=estimate of year-class size at time of tagging] 



Lost-tag recoveries on St. George Island were corrected by the ratio of lost-tag to tagged 

 pups observed on St. Paul Island. 



17 



