FISHERY BULLETIN: VOL. 82, NO. 1 



by California sea lions on the west end of San Miguel 

 Island, Calif., during spring (2-3 May 1978; 2 and 16 

 May 1979) and summer (3-4 August 1978; 30-3 1 July 

 1979). During both sample periods, scats were 

 collected from areas where mostly females and 

 juveniles of both sexes occurred and relatively few 

 (<12% of the total animals censused) adult and sub- 

 adult males were present. In order to document the 

 occurrence of prey species which were consumed at 

 or close to the time of collections, only recent scats, 

 which showed no obvious signs of desiccation, were 

 collected. Each scat was placed in a plastic bag, 

 where it was later soaked in water or a solution of 

 about 1 part liquid detergent to 100 parts water for 

 about 24 h. Each bag was shaken occasionally to 

 facilitate emulsification of the digested organic 

 material, and then rinsed with water through three 

 nested sieves with screen mesh sizes of 3.35 mm, 

 2.00 mm, and 1.00 mm from top to bottom. After 

 most of the soft digested organic material was 

 washed away, fish otoliths and cephalopod beaks 

 were removed and stored in a solution of 70% 

 ethanol. Prey totals were determined by using the 

 higher number of left or right otoliths and upper or 

 lower squid beaks. The otoliths were identified by 

 the late J. Fitch, California Department of Fish and 

 Game, Long Beach, Calif., the octopus beaks by E. 

 Hochberg, Santa Barbara Museum of Natural His- 

 tory, Santa Barbara, Calif., and the squid beaks by 

 the second author. 



The data for each of the four major prey species 

 were summarized by a three-way (2X2X2) con- 

 tingency table and tested for independence of 

 occurrence by season, year, and both season and year 

 (Fienberg 1977). 



Length measurements of these otoliths and squid 

 beaks were used to estimate the body lengths or ages 

 of the most frequently occurring prey species. 

 Although many otoliths and beaks of all sizes were 

 recovered from the scats in good condition, some 

 were not measured because they were broken or 

 showed obvious signs of damage from digestion. We 

 assumed that damage to the otoliths and squid beaks 

 collected in this study was not dependent on size. 

 Lengths of northern anchovy, Engraulis mordax, 

 were estimated from a regression equation of fish 

 lengths on otolith lengths (Spratt 1975). Length 

 information for rockfish, Sebastes spp., was obtained 

 from previously reported data (Phillips 1964) for 

 specimens (bocaccio, Sebastes paucispinis) of the 

 same age as most of the rockfish reported in this 

 study. Bocaccio was chosen as the representative 

 rockfish because it has been reported as the most 

 abundant rockfish in the waters near San Miguel 



Island (Best and Oliphant 1965). The regression 

 equation used to estimate the length of Pacific whit- 

 ing, Merluccius productus, was derived in this study 

 from specimens collected off the coast of southern 

 California by the National Marine Fisheries Service 

 (NMFS). The Pacific whiting otoliths and the corre- 

 sponding length information were provided by K. 

 Bailey of the NMFS Northwest and Alaska Fisheries 

 Center, Seattle, Wash. Market squid, Loligo opales- 

 cens, lengths were estimated from a regression equa- 

 tion of dorsal mantle length on upper hood length of 

 the beak. Upper hood measurements were chosen for 

 the estimation of squid lengths because they were 

 reported as having the highest correlation to dorsal 

 mantle length (Kashiwada et al. 1979). 



In order to detect changes in the diet which would 

 reflect apparent yearly changes in the age and size 

 composition of a specific prey-species population, we 

 compared the lengths of otoliths for 1978 and 1979 

 using the Wilcoxon rank sum test (Hollander and 

 Wolfe 1973). 



Weight estimates of the most frequently occurring 

 prey species were obtained by using the prey length 

 estimate (described above) in regression equations 

 of length and weight measurements or by obtaining 

 weight data from fish which were the same age as 

 those identified in the scats (Phillips 1964; Fields 

 1965; Dark 1975; Pacific Fishery Management 

 Council 1978). The total estimated weight for each of 

 the four major prey species was obtained by mul- 

 tiplying the weight of the average-sized prey by the 

 number of individuals represented in the scat collec- 

 tion. Differences between these estimates could 

 not be statistically analyzed because the raw data 

 for the growth curves of each species were not 

 available. 



The names of fishes follow Fitch and Lavenberg 

 (1968) and Robins (1980), and those of cephalopods 

 follow Fields (1965) and Young (1972). 



RESULTS 



We collected 224 California sea lion scats on San 

 Miguel Island during the spring and summer of 1978 

 and 1979. From 195 (87%) of those scats, we 

 recovered 2,629 otoliths and 2,061 cephalopod 

 beaks. Twenty-nine (13%) scats did not contain 

 otoliths or cephalopod beaks. The prey species iden- 

 tified in the scats are shown in Table 1 by their per- 

 centage of occurrence. The four most frequently 

 occurring prey in scats containing otoliths and/or 

 cephalopod beaks were Pacific whiting (48.7%), 

 market squid (46.7%), juvenile rockfish from 

 the Sebastes paucispinis -goodei-jordani complex 



68 



