Table 15. --Estimates of the number of yearling male seals, year classes 

 1964-67, from recoveries of marked male seals, Pribilof 

 Islands, Alaska 



Years Number- 



1964 

 1R 

 IB 

 1R 

 1R 

 1R 



1965 

 IS 

 IS 

 IS 

 IS 



1966 

 IT 

 IT 

 IT 



1967 

 1U 



728 



1966 

 1967 

 1968 

 1969 

 Pooled 



1967 



1968 



1969 



Pooled 



1968 



1969 



Pooled 



1969 



2 3, 533 



3 34,613 



4 16,912 



5 2, 149 

 57, 207 



2 2, 940 



3 22, 978 



4 12, 769 



2 2, 040 



3 20,47 1 

 22, 511 



40 

 222 



64 



7 



333 



88 

 263 



91 

 442 



63 

 198 

 261 



1, 129 



529 



64, 301 

 113, 506 

 192, 374 

 22 '>, 496 

 125, 065 



37,719 



98, 639 



158, 420 



98, 818 



23, 185 

 74,026 



61, 754 



46, 113 



1/ Number recovered includes those of unknown age. 



2/ Number of marks applied adjusted to compensate for double 

 tag loss and according to the error in determining age at time of tagging. 

 See example in section "Number of male seals ages 1 to 2 years. " 



To determine if there is an indication of a 

 change in the basic relationship between fur 

 seals and their environment or other factors, 

 we look first at the survival of males since 1961, 

 based on comparable data (table 17). There 

 has been no particular trend in the number 

 of pups born each year since that year but, 

 nevertheless, the percentage that survived 

 through the first year of life has fluctuated 

 widely. The percentage of the yearling group 

 represented by the male harvest, however, has 

 also been quite variable, ranging in the four 

 available data points from 41.4 to 58.7. In 

 addition, survival from birth to age 1 and from 

 age 1 to ages included in the male harvest (2 

 through 5) are strongly negatively correlated 

 (r = — 0.873). This negative correlation 

 could have a biological basis, but it is probably 

 influenced largely by random errors in the 

 yearling population estimate. This estimate 

 appears in both survival percentages, as the 

 numerator in the first and the denominator in 

 the second. Moreover, the yearling estimate 



is subject to the usual sampling errors and also 

 the possibility of errors made during determi- 

 nation of ages. In view of these circumstances, 

 we should look at the relationship between the 

 male harvest and the number of pups surviving 

 their first summer. The percent survival 

 (table 18) is the fraction of the pup popula- 

 tion censused in August of their birth year that 

 is represented by the harvest of males. In- 

 cluded in table 18 are two additional variables 

 that have been used in forecasting — the mean 

 annual air temperature for the 12-month per- 

 iod ending 30 June of the year of birth and 

 the counts of dead pups. 



These series are, of course, very short for 

 use in empirical forecasting; the linear rela- 

 tion between air temperature and percent sur- 

 vival is rather poor (r 2 = 0.175) but that 

 between the count of dead pups and percent 

 survival is good (r 2 = 0.773). If the latter 

 relation is calculated, one obtains 



S = 37.9-0.2D 



19 



