Brodeur and Pearcy Trophic relations of juvenile salmon off Oregon and Washington 



627 



to sea in large numbers during favorable meteorolog- 

 ical conditions, can also contribute substantially to the 

 diets of juvenile coho salmon and other salmonids 

 (Brodeur 1989). 



Fishes were much less important in the diets of juve- 

 nile chum and sockeye salmon; however, the mean and 

 maximum size of the individuals of these species ex- 

 amined in this study were smaller than those of coho 

 and chinook salmon. Juvenile euphausiids were a major 

 component in the diet of chum and sockeye salmon in 

 this study. Peterson et al. (1982) found juvenile eu- 

 phausiids were a major prey of chum juveniles off 

 Oregon, whereas copepods, larvaceans, and hyperiid 

 amphipods were most important off British Columbia 

 (Manzer 1969; Healey 1978, 1980). Both chum and 

 sockeye consumed a greater number of chaetognaths 

 than coho and chinook, and the importance of these and 

 other soft-bodied prey may be greatly underestimated 

 in most studies because they are probably digested very 

 rapidly in salmon stomachs (Black and Low 1983). 



Patterns in dietary variability 



The pronounced interannual differences in the diets of 

 most species were expected, even when the collections 

 from same months and areas were examined. Ocean- 

 ographic conditions varied greatly among the years, 

 with both relatively strong (1982 and 1985), weak (1983 

 and 1984), and highly variable (1981) upwelling occur- 

 ring during the study period (Fisher and Pearcy 1988). 

 In addition, a strong El Nifio event dominated ocean 

 conditions in coastal waters of the northeast Pacific 

 during the summer months of 1983 and 1984, greatly 

 affecting primary and secondary productivity and fish 

 production (Mysak 1986, Pearcy and Schoener 1987). 

 Fish prey, as a proportion of the total diet by weight, 

 was generally invariant for the same months among 

 the different years. However, the species composition 

 alternated between coastal and offshore taxa (as iden- 

 tified by Richardson and Pearcy [1977]) depending on 

 the prevalent hydrogi'apliic regime in the various years. 

 Several coastal taxa (Ammodytes hexapterus, Clupea 

 harengus pallasi, Hemilepidotus spinosus, and Osmeri- 

 dae) were more prevalent during strong upwelling 

 years, whereas offshore taxa (Engraulis mordax, 

 Sebastes spp., Ronquilusjordani, and pleuronectid lar- 

 vae) were eaten more frequently during poor upwell- 

 ing years. Although ichthyoplankton collections were 

 not made during every year of the stomach sampling, 

 one limited study lends support to our diet observa- 

 tions. In a series of plankton tows along one transect 

 off the central Oregon coast in 1983, Brodeur et al. 

 (1985) found high abundances of offshore fish taxa at 

 inshore stations compared with past studies. The lar- 

 vae of the northern anchovy were unusually abundant 



that year, and juvenile anchovy made up the majority 

 of the diet-by-weight of juvenile coho and chinook 

 salmon during September 1983. Osmerid larvae, gen- 

 erally the dominant larvae inshore off Oregon (Richard- 

 son and Pearcy 1977), were found in low abundance 

 in both the plankton collections and fish stomachs dur- 

 ing 1983. 



Several invertebrate taxa showed substantial inter- 

 annual variation. The pteropod L. helicina was one of 

 the most important prey consumed in 1981, but was 

 relatively unimportant in other years. The dominant 

 inshore euphausiid T. spimfera was extremely abun- 

 dant in the stomachs following periods of active upwell- 

 ing, but was rarely consumed during the El Niiio of 



1983 and early 1984 (Brodeur 1986). Several species 

 of decapod larvae were present in greater numbers in 

 the diets of all salmon species during 1983 and early 

 1984. Many other El Nino-related anomalies that were 

 observed in the diet of coho salmon during 1983 were 

 described by Pearcy et al. (1985). The diets of coho 

 salmon during 1984 showed above-average abundances 

 of terrestrial insects, which presumably were blown 

 offshore by anomalous winds during the summer of 



1984 (Brodeur 1989). 



Strong seasonal variations in feeding habits of juve- 

 nile coho and chinook salmon were evident in the years 

 that had multiple cruises over the 5-month sampling 

 period. Some of the variability may have been due to 

 the seasonal increase in the mean size of the salmon, 

 which allowed a greater size range of prey to be con- 

 sumed later in the summer (Brodeur In press). How- 

 ever, much of the diet variation may have been due to 

 seasonal variations in the abundance of meroplanktonic 

 prey (e.g., decapod and fish larvae). The timing and 

 duration of spawning, larval development, and settle- 

 ment to benthic juvenile habitat are relatively fixed for 

 most meroplanktonic species (Lough 1975, Richardson 

 and Pearcy 1977, Parrish et al. 1981) such that their 

 seasonal occurrence in the plankton and the diets of 

 juvenile salmon are relatively predictable. Exceptions 

 may occur during anomalous years such as during an 

 El Nino (Bailey and Incze 1985, Brodeur et al. 1985). 

 Several of the major holoplanktonic taxa (e.g. , euphau- 

 siids, hyperiids, and pteropods) also showed a consis- 

 tent seasonal succession of developmental stages, but 

 many species were present throughout the summer 

 period (Brodeur 1990). 



It was not surprising that geographic (north-south) 

 variations in the diet composition of most salmon 

 species were not as substantial as temporal variations. 

 because most of the prey species are distributed 

 throughout the range of latitudes we sampled. Similar 

 results were found for adult salmonids and nonsal- 

 monid species by Brodeur et al. (1987a). However, 

 oceanographic regimes may be quite variable between 



