Table 22. — Estimates of the yearling male seal populations, year 

 classes 1961-62 and 1964-65, St. Paul Island 



4-year-old males will be available in 1968. 

 The several estimates are obviously rather 

 widely divergent, and hence we need not em- 

 phasize that the estimate of the kill of 4-year- 

 olds in 1968 is less reliable than in past years. 



FORECAST OF THE KILL OF 3-YEAR-OLD 

 MALE SEALS ON ST. PAUL ISLAND 



Three regressions used to forecast the kill 

 of 4-year-old males have also been used to 

 predict the kill of 3-year-old males. The re- 

 gressions are: 



(1) Temperature regression: K = 15.8+l .03T. 

 For the 1965 year class, T = 12, so estimated 

 kill at ages 3 and 4=28,200 with a standard 

 error of 9,800. 



(2) Pup weight regression: K=8.33W-39.56. 

 For the 1965 year class: W = 9.5 so estimated 

 kill at ages 3 and 4=39,600 with a standard 

 error of 10,600. 



(3) Count of dead pups and pup estimate 

 regression: 



K=0.9+0.21B-0.69D For the 1965 year class: 

 B=267 and D=41 so estimated kill at ages 3 and 

 4=28,700 with a standard error of at least 

 6,500. 



The estimate of the pup population in 1965 

 seems low in comparison with the estimates 

 of the number of pups born in 1964 and 1966. 



The use of the factor 0.67 to estimate the kill 

 at age 3 from estimates at ages 3 and 4 com- 

 bined yields the estimates 18,900, 26,500, and 

 19,200, respectively. 



Estimates of the number of yearlings are 

 shown in table 22. Estimates based on re- 

 coveries at age 2 have been low in all years 

 when checking has been possible. The relation 

 between the estimated number of yearling 



males based on recoveries at age 2 and the kill 

 in thousands of the same year class at ages 2 

 and 3 is: 



K, = 22.4+0. 23Y 



(2 + 3) 



where Y yearling estmate based on recoveries 

 at age 2. 



For the 1965 year class: 



Y = 38.5, so K 



(2 + 3)" 



31.3. 



Since the kill at age 2 has been 2,200, the esti- 

 mated kill at age 3 by this method for St. Paul 

 Island is 29,100. The standard error of this 

 estimate is near zero. 



COMBINED ESTIMATES OF THE KILL OF 

 3-YEAR-OLD MALE SEALS ON 

 ST. PAUL ISLAND 



The estimates and their standard errors are: 



Estimate Standard error 



Temperature regression 18,900 9,800 



Pup weight regression 26,500 10,600 



Regression on count of dead 

 pups and estimate of pup 

 population 19,200 >6,500 



Yearling estimate 29,100 (?) 



Since these estimates are reasonably simi- 

 lar and we are in doubt about the standard 

 errors of two of the estimates, it is simplest 

 and not inaccurate to use an unweighted aver- 

 age. This average is 23,000, rounded to the 

 nearest thousand. 



22 



