Table 36. — Estimates of the number of pups per 

 harem bull, St. Paul Island, 1963 



Average estimates from table 35, page 28. 

 2 Pups of the year class that died on land 

 before shearing are not included. 



Table 37. --Estimates of the numbers of pups 

 based on marked to unmarked ratios obtained 

 after tagging in 1961 and 1962 and after 

 shearing in 1963, St. Paul Island, year 

 classes 1961-63 



Pups of the year class that died on land 

 before tagging or shearing are not included. 



shown that seals are affected by tagging. How- 

 ever, a means of measuring the quantitative 

 affect of tagging on the mortality rate is not 

 apparent. Mortality of tagged animals above 

 that of untagged seals would inflate pup popu- 

 lation estimates based on tag recovery data. 



In sampling pups for a marked to unmarked 

 ratio, observers may unconsciously ignore 



unmarked animals in favor of the more promi- 

 nent marked seals. If true, the pup population 

 will be underestimated. The reliability of esti- 

 mates based on sampling was partially checked 

 in 1963 by counting all live pups on Little 

 Polovina Rookery. The number of pups esti- 

 mated from sampling was only 89.9 percent 

 of the number counted. Accordingly the esti- 

 mate of the total number of pups on St. Paul 

 Island was adjusted upward. Counts of live 

 pups on several St. Paul Island rookeries 

 are planned for 1964 to provide a more 

 representative correction factor. 



Two estimates of the total pup population 

 on St. Paul Island were calculated on the 

 basis of the complete count of pups on Little 

 Polovina Rookery. These estimates, how- 

 ever, are based on two questionable assump- 

 tions: (1) That the number of pups per harem 

 bull is a constant and (2) that the mortality 

 rate of pups on land is constant for all 

 rookeries for any one year. 



Choice of one estimate as the best among 

 those available is a matter of preference, 

 depending on how much importance is placed 

 on known or possible bias inherent in each. 

 That the actual pup population in 1963 lies 

 between the estimate from tag recoveries 

 (613,000 for I960) and the estimate from 

 live pup sampling (316,000 for 1963) can safely 

 be assumed. Although the estimate from live 

 pup sampling has been corrected on the basis 

 of bias known to exist on one rookery, the 

 magnitude may be underestimated. A 20-per- 

 cent bias would increase the pup population 

 on the Pribilof Islands in 1963 to 355,000, and 

 30 percent, to 405,000. An error greater than 

 30 percent for all rookeries is believed to be 

 highly improbable. 



A maximum limit of error in an estimate 

 based on tag recoveries cannot be determined. 



The figure of 400,000 is presented as the 

 best estimate of the 1963 pup population. 



SEAL- PUP WEIGHTS 



Since 1957, seal pups have been weighed 

 annually on St. Paul Island to determine if 

 individuals from a given year class vary in 

 weight (body condition) from those of another. 

 Differences, if they exist, may be useful in 

 forecasting the survival (return) of year 

 classes through age 4. Although data from 

 year classes 1957-59 suggest a positive corre- 

 lation between mean weight of pups and return 

 of male seals (table 38), the technique cannot 

 be fully evaluated until the return from more 

 year classes is complete. 



Important secondary information has been 

 obtained from the weighing program through 

 the discovery that tagged pups weigh less than 

 untagged pups. This weight loss has been ob- 

 served each year (average 0.57 kg.) from 1957 

 through 1963 and in two additional weighings 



29 



