Olesiuk. Prey consumption of Phoca vitulina 



505 



Diet composition 



Of the 2,841 scat samples collected in the Strait of 

 Georgia, 2,765 (97.3%) contained identifiable prey. 

 Samples typically contained one to three (X=1.91) dif- 

 ferent prey species, but occasionally contained as many 

 as seven. Marine and anadromous fishes, which ac- 

 counted for 96.0% of all prey identified, were by far 

 the most prevalent prey category. The diet included at 

 least 48 species from 20 different families (Olesiuk et 

 al, 1990b). The diet was dominated by gadoids and 

 clupeids which were present in 62.0 and 59.2%, re- 

 spectively of all samples containing identifiable prey. 

 Other important families that occurred in at least 1% 

 of samples were, in decreasing order of importance, 

 salmonids, batrachoids, embiotocids, cottids, pleuro- 

 nectids, hexagrammids, scorpaenids, ammodytids and 

 osmerids. 



The second most prevalent prey category was cepha- 

 lopods, which occurred in 168 (6.1%) of all samples 

 containing identifiable prey and represented 3.5% of 

 all prey items identified. A superficial examination of 

 their beaks indicated that the vast majority were squid 

 (mainly Loligo opalacens with lesser amounts of 

 Gonatus spp. ), but at least one octopus was also con- 

 sumed. The remaining prey categories, namely crusta- 

 ceans, other molluscs, echinoderms, and birds, occurred 

 in <1% of all samples containing identifiable prey and 

 accounted for <0.5% of the total number of prey items 

 identified. 



In assessing the relative importance of each prey 

 (Eqn. 16), it was assumed that scats represented all 

 prey consumed within a 24-hour period. Since pinniped 

 gastrointestinal passage rates typically range from 5 

 to 30 hours (Pastukhov, 1975; Helm and Morejohn, 

 1979; Prime, 1979; Bigg and Fawcett, 1985; Prime and 

 Hammond, 1987; Harvey, 1989), it is likely that most 

 of the prey present in a scat sample had been con- 

 sumed within the same 24-hour period. Although cepha- 

 lopod beaks may be retained in stomachs over longer 

 periods (Bigg and Fawcett, 1985), cephalopods consti- 

 tuted only a small part of the diet and beaks were 

 usually accompanied by cephalopod eye lenses, which 

 probably pass rapidly. 



It is believed that essentially all species consumed 

 in a meal were represented in scat samples, mainly 

 because such a wide array of structures had been used 

 to identify prey. Captive studies of otolith recovery rates 

 have indicated that the fragile otoliths of small fishes, 

 such as herring, may be completely digested and hence 

 under-represented in scats (Hawes, 1983; Jobling and 

 Breiby, 1986; Jobling, 1987; da Silva and Neilson, 1985; 

 Dellinger and Trillmich, 1988; Harvey, 1989). Further- 

 more, based on a comparison between harbor seal stom- 

 ach and intestinal contents, Pitcher (1980) concluded 



that the otoliths of larger prey such as salmon would 

 also tend to be under-represented in scats because their 

 heads (i.e., otoliths) are sometimes discarded prior to 

 being consumed. However, these studies merely dem- 

 onstrate the inadequacy of relying exclusively on oto- 

 liths (see also Fig. 7), and are therefore not pertinent 

 in the present study. 



The improved resolution achieved by utilizing a wide 

 array of structures can be illustrated by examining 

 the prevalence of the above prey species in selected 

 collections in which they constituted the dominant prey. 

 For example, herring elements were identified in 86.4- 

 100% of samples (X=96.1%; 150 of 156 samples) in 5 

 selected collections in which they were the dominant 

 prey; and salmonids in 73.8-90.0% of samples 

 (X=77.7%; 73 of 94 samples) in 5 selected collections in 

 which they were the dominant prey. The slightly lower 

 prevalence of salmonids was probably due to the fact 

 they were not consumed by all seals as most of the 

 samples without salmonids contained other prey spe- 

 cies. Nevertheless, even if it were assumed that all 

 seals had consumed these prey, the prevalence of her- 

 ring in the diet would only have been underestimated 

 by a factor of 1.04, and the prevalence of salmonids by 

 a factor of 1.29. In contrast, in the same collections 

 herring otoliths occurred in only 62.7%- of the samples 

 containing herring and salmonid otoliths in only 9.6% 

 of the samples containing salmonids. Thus, the preva- 

 lence of herring would have been underestimated by a 

 factor of 1.59 and salmonids by a factor of 10.42 had 

 only otoliths been used to identify prey. 



The assertion that scat samples provided an accu- 

 rate representation of diets is further substantiated 



O 50 

 < 10- 

 8 30- 



MID HEX SFP 



SPECIES 



Figure 7 



Percentage of scat samples in which various fish prey were 

 represented by otoliths. Prey codes: GAD=gadoids; HER= 

 herring; SAL=salmonids; MID=plainfin midshipman; 

 HEX=hexagrammids; SFP=surfperches; FLF=flatfish; 

 SCP=sculpins; and ROK=rockfish. 



