506 



Fishery Bulletin 91(3). 1993 



NON- ESTUARIES 



by the similarity of the prey identified in scat samples 

 and those identified in stomachs collected within and 

 adjacent to the study area. All but one of the 19 prey 

 types identified in 69 stomachs collected throughout 

 British Columbia (Fisher, 1952; Spalding, 1964) were 

 represented in the scats, and all but 2 of the 22 prey 

 types identified in 81 stomachs collected in Washing- 

 ton State (Scheffer and Sperry, 1931; Scheffer and 

 Slipp, 1944). The exceptions were sablefish and rat- 

 fish, each of which occurred in one stomach, and bur- 

 rowing crayfishes, which occurred in three stomachs. 

 Furthermore, the mean number of prey species identi- 

 fied in the scat and stomach samples was also similar 

 (X=1.91 in scats versus 2.06 in stomachs). 



Seasonal changes in diet composition outside of es- 

 tuaries are shown in Figure 8A. Salmonids composed 

 a relatively minor portion (0-5.2%; X=3.1% ) of the diet 

 in all months. The diet was domi- 

 nated by gadoids (X=45.7%; 

 94.5% of which were hake) and 

 herring (X=33.0%). There was a 

 pronounced seasonal shift be- 

 tween these two prey, with ga- 

 doids dominating during April- 

 October (54.3-73.7%) and herring 

 during December-March (58.0- 

 70.2% ). Other important prey, de- 

 fined as those constituting >1% 

 of the overall diet or >2% of the 

 diet in any month, were hex- 

 agrammids (X=3.4%; 95.8% of 

 which were lingcod and 4.2% 

 greenling), especially during 

 December-April; plainfin mid- 

 shipman (X=3.4%), especially 

 during April-June and Novem- 

 ber-December; surfperches (X= 

 2.2%; 81.3% of which were shiner 

 perch and 18.7% pile perch); 

 cephalopods (2.1%); and sand- 

 lance (0.9%). Incidental prey in- 

 cluded, in decreasing order, rock- 

 fish, flatfishes, sculpins, smelts, 

 skates, gunnels, lamprey, prickle- 

 backs, crabs, sticklebacks, cling- 

 fishes, eelpouts and sea urchins. 

 Unidentified prey accounted for 

 a mean of 2.09J of the total non- 

 estuary diet. 



Seasonal changes in diet com- 

 position within estuaries are 

 shown in Figure 8B. Salmonids 

 were consumed in all months 

 (X=10.3$ i, but were most preva- 



lent during during September-January (14.5-21.2%). 

 As was the case outside estuaries, the dominant prey 

 in estuaries were gadoids (X=42.9%; 94.2% of which 

 were hake) and, to a lesser extent, herring (X=27.3%). 

 Gadoids dominate (39.2-53.9%) in all months except 

 February-March, when herring were dominant (49.5- 

 50.6%). However, the seasonal shift in the importance 

 of these prey was not nearly as pronounced as it was 

 outside estuaries. Other important prey included 

 plainfin midshipman (X=3.6%) especially during May- 

 June (9.7-10.5% ); surfperches (X=3.6%; 91.9% of which 

 were shiner perch), flatfishes (X=2.8%), sculpins 

 (X=2.6%), all of which were most prevalent during Au- 

 gust-September (5.4-6.1%, 5.0-6.3% and 7.8-7.9% re- 

 spectively); and cephalopods (X=2.3%), especially dur- 

 ing November-March (2.9-5.5%). Other incidental prey 

 were, in decreasing order, rockfishes, sandlance, stick- 



LEGEND 



Salmonids 



Gadoids 



Pacific herring 



Sculpins 



Flatfishes 



□ 



Surfperches 



- 



Hexagrammids 



Plainfin midship: 



Sandlance 



Cephalopods 



Other 



I I Uni< 



Figure 8 



Seasonal changes in diet composition <Ai outside of estuaries; and 'Bi within estuaries. 

 Other prey include all those that constituted <3,% of the diet in all months and <l r/ r of 

 the overall diet. Note that the patterns in the legend are ordered in the sequence they 

 appear in the figures from bottom to top. 



