Hastings and Sydeman Population status of Eumetopias jubctus at the South Farallon Islands, California 



53 



Methods 



Survey methods 



PRBO began conducting surveys of pinnipeds at the South 

 Farallon Islands in 1971. In June 1973 surveys were stan- 

 dardized and all Steller sea lions visible on or in the water 

 near the coast of the South Farallon Islands were counted 

 weekly from standard vantage points on Southeast Far- 

 allon Island: 1) atop Lighthouse Hill (110 m) with bin- 

 oculars or a 20-60x spotting scope, 2) from Cormorant 

 Blind Hill (35 m) with binoculars, and 3) from North Land- 

 ing, Sewer Gulch, and Garbage Gulch with no optical aids 

 I Fig. 1). Most surveys were conducted between 1000 and 

 1800 hours on Thursdays if visibility was adequate. Begin- 

 ning in 1977, animals were classified by age class (adult 

 male, subadult male, adult female, immature, yearling, or 

 pup) when possible, primarily by body size. Adult males 

 were distinctive as very large animals with large muscular 

 necks bearing well-developed manes of long, coarse hair on 

 the chest, shoulders, and back. Subadult males were dis- 

 tinguished from adult males by their smaller size and less 

 developed mane. Immature individuals included animals 

 of distinctly smaller size, such as young-of-the-year (after 

 November) and animals likely one to four years of age. The 

 adult female category included animals smaller than sub- 

 adult males but larger than immature individuals. Pups 

 were distinguishable from June until late November by 

 their thick, dark brown coats, which were later molted and 

 replaced with a lighter brown coat after five to six months 

 of age. Counts were conducted by numerous observers 

 over the years; several observers conducted surveys for 

 over a decade and all observers were trained in identifica- 

 tion of sea lions by age class. Counts represent minimum 

 estimates of numbers of sea lions hauled out because only 

 SS"* to 90'7f of the islands were visible from the study's 

 vantage points. 



We compared maximum counts taken during the breed- 

 ing season (June-July i in recent years (1974-97) with 

 counts from surveys conducted a single time during the 

 breeding season (once annually) and intermittently over 

 the years by CDFG from 1927 to 1970. From 1927 to 

 1938, counts of subadult or adult sea lions (i.e. excluding 

 pups) made by at least two obsei-vers from boats were 

 averaged (Bonnot, 1931, 1937: Bonnot et al., 1938). Meth- 

 ods of counting changed after 1938, such that counts af- 

 ter 1938 could only be compared cautiously with earlier 

 years. Surveys were conducted by airplane, blimp, or boat 

 in 1946 and 1947 and by airplane only from 1958 to 1970 

 (Bonnot and Ripley. 1948: Ripley et al, 1962; Carlisle and 

 Aplin, 1971). Counts from 1946 to 1970 were likely over- 

 estimates because observers assumed that all sea lions 

 north of Point Conception were Steller sea lions (many 

 sea lions may have been California sea lions, Zalophus 

 califoi-niaruis) and because pups were likely included in 

 these counts (Bonnot and Ripley, 1948; Ripley et al., 1962). 

 Counts conducted by PRBO since the 1970s targeted only 

 the South Farallon Islands, whereas CDFG counts includ- 

 ed the South and North Farallon Islands. Monitoring of 

 the North Farallon Islands since 1970. however, has been 



sparse. Although the North Farallon Islands are a known 

 haulout area for Steller sea lions, pupping rates are un- 

 known. The North Farallon Islands were surveyed during 

 the breeding season by PRBO in 1977 (when 17 adult fe- 

 males and 1 pup were counted) and in 1983 (when 92 

 adults but no pups were counted; PRBO, unpubl. data^). 

 Because of the exclusion of the North Farallon Islands in 

 recent counts, comparisons with earlier CDFG data were 

 made cautiously. 



Under the direction of D. G. Ainley and H. R. Huber, pup 

 production and pup mortality were monitored intensively 

 from 1973 to 1986, when animals were breeding in accessi- 

 ble areas. Breeding areas were checked daily for new pups, 

 and prematurely born and dead pups were noted. Breed- 

 ing areas shifted from accessible to inaccessible areas over 

 the years. From 1973 to 1975, all full-term pups were born 

 on the more accessible Saddle Rock, a small islet one- 

 quarter mile offshore (Fig. 1), and a few premature pups 

 were born on the mainland. From 1976 to 1983, females 

 pupped in the equally accessible Sea Lion Cove (Fig. 1), 

 perhaps because of reduced disturbance on Southeast Far- 

 allon Island, although one pup was obser\'ed on Saddle 

 Rock in 1981. Although photogi-aphs from the 1930s show 

 large numbers of Steller sea lions on West End (Huber^), 

 they were not obsei-ved there in recent years until 1983 

 (one female in the spring). The first pup was born on West 

 End in 1985 ( Huber"' ). Cun-ently, the majority of the popu- 

 lation is found and all pupping occurs at Indian Head and 

 Shell Beach on West End (Fig. 1); both of these areas are 

 inaccessible and difficult to monitor. 



Statistical analyses 



Statistical models have been developed that account for 

 effects of obsei-ver, and environmental and survey-related 

 covariates on counts of birds and marine mammals (Link 

 and Sauer, 1997, 1998; Calkins et al., 1999; Frost et al., 

 1999; Forney, 2000). These models can increase accuracy 

 in estimating and power in detecting population trends 

 by reducing variability in counts and correcting biases in 

 trend that result from methodological changes in sui-vey 

 design over time (such as changes in survey dates), par- 

 ticularly when few surveys are conducted during a stan- 

 dard survey window each year (Calkins et al., 1999; Frost 

 et al., 1999). However, environmental covariates could not 

 be included in the statistical models in our study when 

 the full data set was used because observers recorded the 

 times that sui-veys began and ended on only a few occa- 

 sions prior to 1983 (41 of 569 surveys, or 7.2'^^^f of surveys), 

 such that the majority of data during the first decade of 

 the time series would be excluded. To include the entire 



•» PRBO (Point Reyes Bird Observatory). 1988. Unpubl. data. 

 (Available from W. J. Sydeman, Point Reyes Bird Observatory, 

 4990 Shoreline Hwy, Stinson Beach. CA 94970.1 



' Huber, H. R. 1985. Reproduction in northern sea lions on 

 Southeast Farrallon Island, 1973-1985. Final report to the 

 Gulf of the Farallones National Marine Sanctuary, San Fran- 

 cisco. CA, 22 p. [Available from Point Reyes Bird Observatory, 

 4990 Shoreline Hwy., Stinson Beach, CA 94970.) 



