FISHERY BULLETIN: VOL. 80, NO. 4 



(1981) reported that juvenile chum from near- 

 shore pelagic habitats of Puget Sound fed on 

 euphausiids, crab larvae, and gammarid amphi- 

 pods on a weight basis; coho fed largely on larvae 

 of decapod crustaceans; and chinook fed on eu- 

 phausiids. 



The qualitative range in variability of diet 

 present during our 2-wk sampling period was 

 similar to that found by the above authors from 

 various months, years, and geographical loca- 

 tions. At some times and locations, euphausiids 

 were dominant prey; at others, amphipods and 

 fishes. This variability suggests that juvenile 

 salmon are opportunistic, feeding on abundant 

 prey available at a particular time and place. 



The main prey items of our juvenile salmon 

 comprise three general size groups: 1) Fishes 

 having an average length of 29 mm, 2) euphausi- 

 ids and Cancer magister megalopae, ranging in 

 length from 7 to 10 mm, and 3) hyperiid amphi- 

 pods between 4 and 6 mm.The fact that juvenile 

 salmonids ate large numbers of euphausiids 

 agrees with what is known about the abundances 

 of various-sized planktonic prey sampled in 

 coastal waters of Oregon during period years 

 (Table 9). Over the range of 7-10 mm, euphausi- 

 ids were the most abundant prey item. Shrimp 

 larvae and C. magister megalopae were abun- 

 dant only during limited periods, usually only 

 June. 



The predominant euphausiid eaten was T. 

 spinifera, a neritic species. Euphausia pacifica, 

 although a more abundant species of euphausiid 

 in the North Pacific, is more oceanic and is not 



Table 9.— Abundance of salmonid prey aver- 

 aged from plankton samples collected during 

 June and July at stations located 1, 3, 5, and 10 

 mi off Newport, Oreg. Zooplankton are aver- 

 aged over the years 1969-72 (Peterson and Miller 

 1976); crab larvae over the years 1969-71 (Lough 

 1975, 1976); and larval fish from 1971 only 

 (Richardson and Pearcy 1977). Plankton tows 

 are step-oblique through the entire water col- 

 umn, during daytime, using a 0.2 m diameter 

 bongo net (0.24 mm mesh) for zooplankton and a 

 0.7 m bongo net (0.5 mm mesh) for fish larvae. 



common in shallow shelf waters (Hebard 1966; 

 Peterson and Miller 1976; Youngbluth 1976) and 

 was found in low numbers in our salmonid stom- 

 achs. Juvenile salmonids collected off Oregon fed 

 predominantly on subadult individuals, possibly 

 because adult euphausiids migrate into deeper 

 waters during the day (Alton and Blackburn 

 1972) when salmon presumably feed. Subadult 

 euphausiids are abundant in the upper 20 m of 

 the water column during both day and night 

 (Peterson 5 ). 



The large numbers of hyperiid amphipods and 

 the paucity of copepods in the diet of juvenile 

 salmon were surprising. Hyperiids are neither 

 abundant in Oregon coastal waters nor are they 

 particularly large compared with other more 

 common planktonic taxa (Table 9). The average 

 length of the amphipods (4.5 mm) is not much 

 greater than Calanus marshallae (stage 5 cope- 

 podites and females, 3.0-4.0 mm TL (total 

 length)). The ratio of amphipods to C5 C. mar- 

 shallae abundance was 1:14 in plankton samples 

 (Table 9) but 4 : 1 in the stomachs of juvenile coho. 

 Frequency of occurrence in juvenile coho stom- 

 achs was 44% for amphipods compared with only 

 6% for Calanus. Similarly, the copepod C. cris- 

 tatus (8 mm TL), with an average abundance 

 about the same as hyperiids, was seldom eaten. 

 Length alone may not be adequate for assessing 

 size-selective predation in juvenile salmon. 

 Okada and Taniguchi (1971) found that the 

 upper size limit of prey may be determined by 

 prey width. This may be relevant because hyperi- 

 ids are generally much broader at their widest 

 dimension than copepods of the same length. 



One hypothesis to explain the high selectivity 

 of amphipods by juvenile coho salmon concerns 

 their peculiar swimming behavior and pigmen- 

 tation. In the laboratory, hyperiids caught in 

 coastal waters were extremely active swimmers 

 (Peterson 6 ). Most species have a large, heavily 

 pigmented (black) compound eye, which could 

 increase their detection by a visual predator, as 

 shown for freshwater fish (Zaret and Kerfoot 

 1975). Copepods, on the other hand, lack the 

 visual contrast of amphipods and are less active 

 swimmers, generally swimming upwards and 



5 W. T. Peterson, Marine Sciences Research Center, State 

 University of New York-Stony Brook, Stony Brook, NY 11794, 

 unpubl. data, 1977. 



6 W. T. Peterson, Marine Sciences Research Center, State 

 University of New York-Stony Brook, Stony Brook, NY 11794, 

 pers. obs. 1978. 



848 



