58 



Fishery Bulletin 100(1) 



pups to females increased from 0.75 to >1.00 

 from 1968 to 1986 (Merrick et al., 1987). The 

 South Farallon ratio is more typical of pe- 

 ripheral areas of rookeries in Alaska where 

 only 0.01 to 0.09 of females had pups com- 

 pared with main areas of rookeries where ra- 

 tios averaged 0.63 to 0.74 (Withrow, 1982). 



Similarly, high pup mortality rates observed 

 at the Farallon Islands (average of 0.49 of 

 pups born from February to August, range 

 of 0.33 to 0.90 among years; Huber et al.^) 

 are more characteristic of peripheral areas of 

 rookeries where pup mortality ranged from 

 0.30 to 1.00 compared with 6.10 to 0.12 at 

 main rookery sites (Withrow, 19821. Rooker- 

 ies had much lower pup mortality rates dur- 

 ing the first two months of life than those ob- 

 served at the Farallon Islands, including Ario 

 Nuevo Island, California (0.10, Gentry, 1970), 

 and sites in Alaska (0.03-0.14, Merrick et al., 

 1987). The frequency of premature pupping 

 (0.40 of those born; Huber et al.'') is also very 

 high compared with the frequency at rook- 

 eries in Alaska (0.09; Pitcher and Calkins, 

 1981), Oregon (0.04; Mate, 1973), at Aho Nue- 

 vo Island (0.02; Gentry, 1970). As at Aho Nue- 

 vo, most premature pups are born from F'eb- 

 ruary to May at the Farallon Islands (0.65 

 born in April with a range of February to 

 May), whereas full-term pups are born from 

 mid-May to late July (Gentry, 1970; Huber^). Causes of the 

 high rate of premature pupping at the Farallon Islands 

 are unknown but may be due to several factors known to 

 cause reproductive failure in pinnipeds, including disease 

 or exposure to pollutants (Gilmartin et al., 1976; Huber''), 

 or a prevalence of young, inexperienced, or malnourished 

 females (Pitcher et al., 1998). A high frequency of abortions 

 has been observed at haulout sites rather than at rooker- 

 ies in Alaska (Pitcher and Calkins, 1981 ). Low pup produc- 

 tion and reproductive rates, coupled with high pup mortal- 

 ity and premature pupping rates, support characterization 

 of the Farallon Islands in recent years as a haulout site or 

 peripheral rookery for this species. 



Seasonal patterns in counts 



Seasonal haulout patterns varied significantly among sexes 

 and age classes. Adult and subadult male attendance was 

 highly seasonal and males were present only during the 

 breeding season. In contrast, adult females and immature 

 individuals were present year-round and their numbers 

 peaked twice (breeding season and from late fall through 

 early winter). Many studies reported the absence of adult 

 and subadult males at California rookeries outside the 

 breeding season, including Ano Nuevo Island (Orr and 

 Poulter, 1967) and San Miguel Island ( Bartholomew, 1967), 

 and the presence of females and immature individuals at 

 rookeries year-round (Rowley, 1929; Bartholomew, 1967). 

 At Canadian rookeries, males were also generally absent 

 in the winter, but small numbers of females and young 



OReproduclive rale 



D Sex-ratio 



• Maximum pup count 



-1 — ' — 1 — ' — 1 — I — I- 



76 78 80 82 84 86 88 90 92 94 96 98 



"lear 



Figure 4 



Annual variation in reproductive rate, adult sex ratio, and ma.ximum 

 pup count during tlie breeding season (June-.Julyi at the South Farallon 

 Islands, 1977-97. Reproductive rate was defined as the ma.ximum pup 

 count divided by the maximum count of adult females per year. Adult sex 

 ratio was calculated as the maximum number of adult females divided 

 by the maximum number of adult males (bulls) counted per year Signifi- 

 cant (P<0.05) trends are shown for adult sex ratio (dashed black line) and 

 maximum pup count (bold black line). 



of the year usually remained at rookeries throughout the 

 year (Bigg, 1988). Circumstantial evidence suggests males 

 from California migrate northward or males from South- 

 east Alaska move southward in winter, or both movements 

 take place. Large numbers of males have been seen outside 

 the breeding season off northern California (Fry, 1939), 

 Oregon, Washington (Mate, 1973), and southern Vancouver 

 Island (Bigg, 1988). Total numbers of Steller sea lions are 

 also higher in the winter than in the summer off the Cana- 

 dian coast (Bigg, 1988); some winter haulouts in Canada 

 consist almost exclusively of males (Bigg, 1988). The earli- 

 est evidence for sea lion migrations was provided by the 

 recovery of north-coast native American spearheads from 

 several sea lions killed off southern California in the late 

 1800s; and in June 1870, a spearhead used by native Alas- 

 kans was found in a large male sea lion at Point Arena, 

 California (Scammon, 1874). Seasonal northward move- 

 ment has also been documented in male California sea 

 lions, which were similarly absent from southern sites out- 

 side the breeding season but which ranged up into Wash- 

 ington and British Columbia during winter (Starks, 1921; 

 Fry, 1939; reviewed by Bartholomew, 1967). 



In contrast to animals on the Farallon Islands, animals 

 of all age classes and both sexes on Aho Nuevo Island were 

 present in significant numbers only during the breeding 

 season from 1967 to 1990 (Le Boeuf and Bonnell, 1980; Le 

 Boeuf et al.^). Data from 1962 and 1963 indicated a sub- 

 stantial presence of Steller sea lions at Aho Nuevo through 

 the fall and winter (Orr and Poulter, 1965) and therefore 

 the lower numbers and, more recently, near absence of all 



