518 



Fishery Bulletin 102(3) 



2002; Tollit et al., 2003). Even after applying a DCF, 

 the estimated mean size of walleye pollock based on 

 otoliths was 10.9 cm smaller than the mean size esti- 

 mated by using all cranial structures. Because walleye 

 pollock otoliths are relatively large and have a different 

 composition than other cranial structures, the larger 

 otoliths may be regurgitated, fully digested, or crushed 

 by rocks in the stomach and not pass through in scat as 

 readily as smaller otoliths or other cranial structures, 

 thereby reducing their occurrence in scat and use in 

 generating prey-size estimates. Atka mackerel otoliths 

 are much smaller at older ages in relation to walleye 

 pollock, which may explain why the size of prey esti- 

 mated from otoliths was similar to the size estimated 

 from other cranial structures. 



The use of DCFs for all structures, including otoliths, 

 to account for erosion increased mean size estimates 

 for both walleye pollock (33.1 vs. 39.3 cm FL) and Atka 

 mackerel (30.7 vs. 32.3 cm FLi. The relatively small 

 increase in the corrected size of Atka mackerel re- 

 flects that the structures from this species were found 

 in better condition than those from pollock (Table 2), 

 as well as that correction factors were found to be 

 species-, structure-, and condition-specific (Tollit et 

 al., 2004b, this issue). Overall, our results emphasize 

 the importance of using appropriate condition-specific 

 DCFs. Other studies with captive sea lions have also 

 demonstrated that grade-specific DCFs can reduce sys- 

 tematic error and increase precision of body mass es- 

 timates (Tollit et al. 1997). For walleye pollock, there 

 was no significant difference in the degree of erosion 

 across the three size ranges for each structure within 

 each condition category (Tollit et al., 2004b, this issue). 

 We assume the DCFs can be used for fish outside of 

 this size range because the relative shape, structure, 

 and proportion of the morphological features are con- 

 sistent for both smaller and larger fish (Tollit et al., 

 2004b, this issue). Further research is necessary to 

 test whether there are differences across the size range 

 for Atka mackerel. 



Size of walleye pollock and Atka mackerel consumed by 

 Steller sea lions in the Bering Sea and Gulf of Alaska 



In general, Steller sea lions on summer rookery and 

 winter haul-out sites consumed primarily subadult and 

 adult-size walleye pollock and Atka mackerel year-round 

 in 1998-2000. Steller sea lions typically forage near 

 shore, in shallow water (<50 m) and at night (Raura- 

 Suryan et al., 2002; Loughlin et al., 2003). Likewise, 

 adult walleye pollock migrate vertically to shallower 

 depths during the night (Smith, 1981). Adult-size Atka 

 mackerel also are commonly found in nearshore coastal 

 areas during their spawning season (Zolotov, 1993). 



Juvenile walleye pollock were found in relatively high 

 numbers only in scats collected on summer haul-out 

 sites. Scats collected from summer haul-out sites likely 

 represent a larger proportion of juvenile Steller sea li- 

 ons than those collected on summer rookery or winter 

 haul-out sites. Previous studies indicate t hat juvenile 



sea lions (<4 years old) consume smaller walleye pollock 

 than adult sea lions (Pitcher, 1981; Frost and Lowry, 

 1986; Merrick and Calkins, 1996). Juvenile walleye pol- 

 lock are often found at shallow depths in bays and near 

 shore habitat (Smith, 1981). Likewise, Loughlin et al. 

 (2003) reported that juvenile Steller sea lions are typi- 

 cally shallow divers and frequently make short range 

 foraging trips (<15 km). Additional scat collections on 

 summer haul-out sites are necessary to determine more 

 conclusively prey-size selectivity for juvenile Steller sea 

 lions. 



Annual changes in the size-frequency distribution of 

 Atka mackerel consumed by Steller sea lions followed 

 changes in the size distribution of Atka mackerel re- 

 sulting from exceptionally strong year classes. Merrick 

 and Calkins (1996) also showed that the size of prey 

 consumed by Steller sea lions can reflect the size dis- 

 tribution of the fish population. From the mid-1990s on, 

 only 1999 was a strong recruitment year for walleye 

 pollock in the Gulf of Alaska (Dorn et al., 2001), but we 

 did not find a significantly greater proportion of juvenile 

 fish eaten by Steller sea lions in 2000 than in 1999 or 

 1998 perhaps because sufficient numbers of larger size 

 fish were available in regions where walleye pollock 

 were consumed. 



Historical studies of Steller sea lion prey size have 

 primarily been based on measurements of walleye pol- 

 lock otoliths found in stomach samples but often with- 

 out application of correction factors for erosion (Pitcher, 

 1981; Merrick and Calkins, 1996; Calkins. 1998). Prey- 

 size estimates based on stomach contents will likely 

 differ from estimates derived from scats because of 

 differences in digestion rates and breakage ( Jobling and 

 Breiby, 1986 i. However, results of studies examining 

 the variability in prey size with sample type are vari- 

 able. Sinclair et al. (1996) suggested that in northern 

 fur seals (Callorhinus ursinus), another otariid, small 

 otoliths tend to flush through the digestive system more 

 quickly than larger ones, resulting in a possible bias in 

 scats towards smaller otoliths. In contrast, experiments 

 with captive sea lions have shown that smaller otoliths 

 are recovered in lower relative frequencies than are 

 larger ones (Tollit et al.. 1997). Frost and Lowry (1980) 

 found no significant difference between the size of oto- 

 liths obtained from stomach and intestines of ribbon 

 seals. Overall, we believe useful comparisons of prey 

 size consumed by Steller sea lions can be made between 

 our study and earlier studies. 



Steller sea lions have been reported to consume a 

 wide size range of walleye pollock. However, in most 

 prior studies a larger proportion of juvenile fish were 

 found than what we estimated from scats. Otoliths from 

 stomach samples collected from 1975 to 1978 in the 

 Gulf of Alaska contained primarily juvenile age pollock 

 (mean FL=29.8cm; SD= 11.6; Pitcher, 1981). Undigested 

 otoliths from stomach samples collected between 1975 

 and 1981 in the Bering Sea also contained mostly juve- 

 nile fish (mean FL=29.3 cm) but had a distinct mode of 

 adult-size pollock (48 cm FL: Frost and Lowry. 1986). 

 Likewise, 43 stomach samples collected between 1985 



