56 



Fishery Bulletin 100(1) 



Table 1 



Linear rates of change in counts of Steller sea lions on the South P^arallon Islands by season, sex. and age class. Rate of change per 

 year was calculated by 1) removing the effect of date on counts (i.e. by using residuals from regi'ession of date on counts), and 2) 

 log-transforming (log^ ) the sum of the residuals added to the mean count,  mdicates significant trends (P<0.05) from regi'essions. 

 ;! = sample size. 



Age class 



change 



/^> 



SEi/3,,.J 



All animals 

 All months 



Adult females 

 All months 



Breeding season (May-Jul) 

 Late fall through early winter 



Males (breeding season! 

 All males 

 Bulls 

 Subadult males 



Immature sea lions 

 All months 

 Breeding season 

 Late fall through early winter 



-0.44 



-0.0044 



0.0027 



0.03* 



1134 



Annual abundance trends 



Although vi.sual inspection showed that residuals of log- 

 transformed counts were not normally distributed, a 

 square-root transform of counts normalized residuals. 

 We report trend estimates from log-transformed counts 

 because results of statistical tests for trends with log- and 

 square-root-transformed data were identical. 



After removing seasonal effects (i.e. residuals from re- 

 gression of date on count were used), all sexes and age 

 classes showed significant trends but in different di- 

 rections (Table 1). When data from all sexes and age class- 

 es were pooled, the trend was complex (variables year- 

 through year^: all P<0.05; ;! = 1134; Fig. 3A). Counts were 

 highest in the late 1970s to early 1980s and declined only 

 slightly during the late 1980s to early 1990s. A slight 

 (-0.4'% per year) linear decrease in total counts (pooling 

 over months, sexes, and age classes) with year was signifi- 

 cant (P<0.05; Table 1 ). Adult female counts declined signif- 

 icantly although the rate of decline slowed in recent years 

 (variables yea;- and yea?-; P<0.01; h=866; Fig. 3B). Counts 

 of adult females declined significantly during the breeding 

 season (variables year and year-: P<0.01; 11=217; Fig. 3B). 

 whereas no trend during late fall and early winter was 

 evident (P=0.80; Table 1 1. Linear rate of decline for adult 

 females was -3.2'7f per year for all months combined and 

 -5.9% per year during the breeding season (both P<0.001; 

 Table 1). 



In contrast, numbers of males during the breeding sea- 

 son increased linearly with year at a rate of l.V/r per year 

 (P<0.05; ;i=217; Table 1). This increase was due to an in- 

 creased number of subadult males (1.9% per year; P<0.05; 

 Table 1; Fig. 3C), whereas the numbers of adult males 

 were stable (P>0.60; Table 1). Counts of immature indi- 



viduals also increased slightly (0.6% per year; Table 1) 

 but significantly when counts from all months were pooled 

 (variables yea/- and year-: P<0.01; ;?=866; Fig. 3D). The in- 

 crease was due to the greater numbers of immature in- 

 dividuals from late fall through early winter in recent 

 years (linear trend=5.0%' per year, Table l;yea7- and year^: 

 P<0.01; ;;=280; Fig. 4D). However, numbers of immature 

 individuals present during the breeding season declined 

 at a rate of -4.5% per year (Table l;year: P<0.01; ;;=217; 

 Fig. 3D). 



Trends in pup production, reproductive rate, and 

 adult sex ratio during the breeding season 



Maximum pup count from surveys declined significantly 

 from the mid-1970s to the mid-1980s from 15 to 2-4 pups 

 and has remained low in recent years (year and year-: 

 P<0.003; Fig. 4). After adjusting for pups not seen during 

 surveys, reproductive rates of adult females ranged from 

 2.0% to 21.2'7( among years, with an average rate of 10.7% 

 (Fig. 4). Although reproductive rate appeared to decline 

 in the 1980s and recovered to 1970s levels in the 1990s, 

 no trend was discernible (year and higher-order terms: 

 P>0.20; Fig. 4). The ratio of adult females to adult males 

 during the breeding season ranged among years from 

 10.3:1 to 1.8:1, with an average of 5.2:1 (Fig. 4). The ratio 

 of adult females to adult males declined significantly and 

 linearly with vear (P<0.001). 



Discussion 



Although the Farallon Islands are an important haulout 

 area for Steller sea lions in California, numbers of ani- 



