Abstract. - Feeding habits of 

 juvenile coho Oncorhynchiis kisutch, 

 Chinook 0. tshawiftscha. chiim 0. keta, 

 and sockeye 0. nerka salmon were 

 examined from collections taken off 

 the Oregon- Washington coast dur- 

 ing the summers of 1980-85. The 

 major prey of both coho and chinook 

 salmon juveniles were larval and 

 juvenile fishes, although a substan- 

 tial proportion of the diet of coho 

 salmon consisted of invertebrates 

 such as euphausiids, decapod larvae, 

 and hyperiid amphipods. Juvenile 

 chum and sockeye salmon had a 

 more varied diet consisting general- 

 ly of smaller prey, such as juvenile 

 euphausiids, copepods, amphipods, 

 and chaetognaths. Diet overlap was 

 highest between coho and chinook 

 salmon. Both dietary overlap and 

 diversity varied substantially among 

 cruises and individual collections. 



Pronounced seasonal and interan- 

 nual variations occurred in the util- 

 ization of the major prey taxa by 

 coho and chinook salmon which may 

 have been related to highly variable 

 oceanographic conditions prevailing 

 during this period. Areal (latitudinal 

 and cross-shelf) variations were of 

 lesser importance in the diets of 

 juvenile coho and chinook salmon. 



Trophic Relations of Juvenile 

 Pacific Salmon off the Oregon 

 and Washington Coast 



Richard D. Brodeur 



Fisheries Research Institute WH-IO, University of Washington 

 Seattle, Washington 98195 



and 

 Alaska Fisheries Science Center, National Marine Fisheries Service, NOAA 

 7600 Sand Point Way NE, Seattle, Washington 981 15-0070 

 Present address: Pacific Biological Station, Department of Fisheries and Oceans 



Nanaimo, BC, Canada V9R 5K6 



William G. Pearcy 



College of Oceanography, Oregon State University 

 Corvallis, Oregon 97331 



Manuscript accepted 29 May 1990. 

 Fishery Bulletin, U.S. 88:617-636. 



Until recently, much of the inter- 

 annual variability in the survival of 

 Pacific salmon Oncorhynchus spp. 

 was believed to be a result of condi- 

 tions occurring in freshwater. It was 

 surmised that by releasing high and 

 relatively constant numbers of hatch- 

 ery smolts in freshwater, much of the 

 interannual variation in salmon pro- 

 duction could be circumvented. Con- 

 trary to this assumption, a number of 

 recent studies have found that a sub- 

 stantial portion of the total natural 

 mortality for most salmon species 

 may occur during estuarine and early 

 marine residence (Parker 1968, 1971; 

 Mathews and Buckley 1976; Bax 

 1983; Furnell and Brett 1986; Fisher 

 and Pearcy 1988). For coho salmon 

 0. kisutch originating from Oregon 

 and Washington rivers, peak mortal- 

 ity appears to happen very early in 

 the ocean, perhaps within the first 

 month, and year-class size appears to 

 be well established by the end of the 

 first summer in the ocean (Pearcy 

 1988). 



We still do not know the exact 

 cause of this early ocean mortality. 

 Annual variability in ocean survival 

 of Oregon-Washington salmon 

 stocks appears to be quite large. Both 

 depensatory environmental factors, 

 such as upwelling (Gunsolus 1978, 



Nickeison 1986), and compensatory 

 biotic interactions such as compe- 

 tition or predation (Peterman 1982, 

 Peterman and Routledge 1983, 

 McGie 1984) have been implicated in 

 the early ocean mortality of juvenile 

 salmon. In either situation, early 

 ocean feeding may be important, and 

 perhaps critical, to growth and sur- 

 vival of salmonids. 



Knowledge of the feeding habits 

 and diet variability of juvenile salmon 

 in the coastal ecosystem off Washing- 

 ton and Oregon is incomplete. Peter- 

 son et al. (1982) described the diets 

 of juvenile coho, chinook 0. tshatvy- 

 tscha, and chum 0. keta salmon for 

 one month (June 1979) and three dif- 

 ferent geographical areas off south- 

 ern Washington and Oregon. Em- 

 mett et al. (1986) examined seasonal 

 variations in juvenile coho and 

 chinook salmon feeding habits from 

 three time-periods during 1980, but 

 did not examine geographic varia- 

 tions in any detail. There were sub- 

 stantial within and between-study 

 differences in the food composition of 

 these salmonids, although the rela- 

 tive importance of the geographical 

 and temporal components of the vari- 

 ability could not be assessed because 

 of the different sampling designs of 

 the two studies. 



617 



