BIGG: ARRIVAL OF NORTHERN FUR SEALS 



males killed was interpolated from the annual plots, 

 so as to standardize the number killed by date. The 

 mean number of males killed, and standard error 

 of the mean, were determined for each date of the 

 last-day of rounds. During 1965-72, a kill of males 

 sometimes took place twice at a haul-out site in one 

 round, but was missed at this site in the preceding 

 or following round. In these cases, one of the two 

 kills was selected randomly and transposed to the 

 other round. Occasionally, sites were visited extra 

 times without being missed in the adjacent round. 

 These data were omitted. 



The kill data used for females on hauling grounds 

 were from years in which the kills on rookeries and 

 hauling grounds were recorded separately, and in 

 which the pregnancy rates were noted. Such kills 

 took place only during 1956-59. These kills were 

 made during the 5-d rounds with the last-day dates 

 between 1 July and 20 August. The only kills on 

 rookeries for which pregnancy data could be used 

 were in 1956 and 1957. On 1-6 July 1956, a kill was 

 made on Polivina rookery. All kills made in the 

 region of this rookery on 1-21 July 1957 were in fact 

 made only on the rookery (A. Roppel 3 ). The number 

 of females killed on rookeries was set by quota, 

 rather than by all available animals being taken, as 

 on hauling grounds. No body length limits were im- 

 posed on the kill of females on rookeries in 1956 and 

 1957. 



To determine the relative number of each age that 

 returned to St. Paul Island, I reviewed the largely 

 subjective comments on abundance given in the 

 literature, and also compared the number killed 

 when arrival was believed to have been completed. 

 For the latter, the only data used were from years 

 when body length limits included at least 50% of the 

 individuals of the relevant age and when the total 

 number of living animals of a particular age did not 

 change substantially between years. The main 

 change in herd size was between 1956 and 1959, 

 when pup production on St. Paul Island decreased 

 by about 27% due to the killing of adult females dur- 

 ing the herd reduction program (York and Hartley 

 1981; Fowler 1982). Pup production changed little 

 between 1960 and 1980, although declined slightly 

 in 1981-82. The cumulative effect of harvesting a 

 cohort over several years was considered when com- 

 paring the relative number of each age killed. The 

 relative numbers of females of each age killed be- 

 tween 1956 and 1959 were biased slightly downward 

 with time by the herd reduction program during the 



3 A. Roppel, Northwest and Alaska Fisheries Center, National 

 Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., 

 Seattle, WA 98115, pers. commun. July 1983. 



intervening years. The bias was only slight because 

 of the years and ages selected for analysis, the lack 

 of time for the herd reduction program to have 

 potentially changed age distribution, and the fact 

 that most seals ages 1 and 2 remained at sea. 



RESULTS 



Effect of Body Length Limits 



The lower length limit of 107 cm for males in- 

 cluded essentially no individuals age 1, few age 2, 

 but most of those >2> yr (Table 1). The upper length 

 limit varied by year, with the smallest upper limit 

 including a few ^4 yr, and the largest, a few >6 yr. 

 I used kill data collected from the years 1969-82 to 

 describe arrival times and relative numbers of males 

 ages 1 and 2. Data from the years 1962-82 were used 

 to describe the arrival time and relative number of 

 3-yr-olds. For males >4 yr, the relative numbers 

 returning by age could not be compared with one 

 another, or with younger males, because of the 

 cumulative reduction in the size of a cohort by the 

 harvest, and the exclusion of seals by upper length 

 limits. I used data from the years 1963-72 and 

 1980-82 to describe the arrival schedule for age 4, 

 and 1964-71 for ages 5, 6, and >7. 



The lower length limit of 104 cm for females in- 

 cluded most individuals >4 yr, while the upper 

 length limit of 116-117 cm included mostly <5 yr. 

 Data collected in 1956 were used to describe the ar- 

 rival schedules for females >A yr, and 1958-59 for 

 those <5 yr. The number of females killed at age 

 3 during 1959 was not used due to an unusually low 

 pup survival in 1956 (Abegglen et al. 1959; Lander 

 1979). 



Arrival of Males on Hauling Grounds 



1 -Year-Olds 



No yearling males were taken in the kill by 5 

 August, and thus none were likely to have been on 

 hauling grounds up to this time. However, few year- 

 ling males apparently go to hauling grounds. Osgood 

 et al. (1915) and Roppel et al. (1965a) indicated that 

 yearlings of both sexes preferred rookery edges, 

 near cows and pups, and only occasionally went to 

 hauling grounds (see section on Arrival of Males on 

 Rookeries). 



2-Year-Olds 



Very few 2-yr-old males arrived by 1 July (Fig. 



385 



