Table 2. — Summary of foods eaten by juvenile sockeye salmon 

 (N = 160) in all regions of Bristol Bay, Alaska, between June 

 and September 1966 and 1967. 



Port Moller. (The 50 specimens were taken at the 

 same time and place as my samples.) Fish were 

 second in importance to copepods in terms of 

 weight of food, and over half the bulk of these fish 

 were Pacific sand lance, Ammodytes hexap- 

 terus. Larval crustaceans were the only other food 

 of major importance (by bulk) and most of these 

 were anomuran larvae eaten by juveniles in the 

 Ugashik area in July 1966. Other items eaten by 

 juvenile sockeye salmon in significant amounts 

 during their migration out of Bristol Bay were 

 euphausiids, amphipods, and insects. Insects and 

 amphipods occurred frequently in the diet but did 

 not contribute much bulk. 



I looked for differences in food selectivity be- 

 tween large and small fish among 144 juveniles 

 (6-13 cm fork length) grouped in 1-cm size 

 categories, but the results were inconclusive. 



Discussion 



The results of this study generally agree with 

 those of other investigators. The importance of 

 copepods in the diet of juvenile sockeye salmon 

 near shore in Bristol Bay is paralleled in coastal 

 waters of British Columbia (Manzer 1969) and 

 is similar to Kamchatka coasts, where copepods 

 and cladocerans were the predominant foods of 

 juvenile sockeye salmon (Synkova 1951). My 

 findings differ from those of Jaenicke and Ben- 

 nett (see footnote 2), who sampled mainly over 

 deeper waters of Bristol Bay farther offshore than 

 I did, and Dell (1963), who sampled off Port 

 Moller in Bristol Bay. Jaenicke and Bennett 

 examined the food of over 1,200 juvenile sockeye 

 salmon captured during the summers of 1969-70 

 and found that the main items (in bulk) were 

 young and larval sand lance and euphausiids. 

 Similarly, Dell reported that 45 juvenile sockeye 



salmon taken in late July 1962 contained mostly 

 larval sand lance and euphausiids. 



Nearly all of the insects I found were from 

 juvenile sockeye salmon captured in the Kvichak 

 and Egegik areas in June (Table 1). These areas 

 are contiguous to many rivers that form part of a 

 major sockeye salmon reproductive complex of 

 lakes and steams (Figure 1). According to Hart- 

 man et al. (1967), most of the migration from 

 freshwater to Bristol Bay takes place in June. 

 Most of the insects were probably ingested in 

 fresh water when the fish were migrating sea- 

 ward, suggesting that many of the juveniles 

 taken in these areas were recent immigrants 

 from fresh water. The occurrence of the only 

 empty stomachs and small average weight of food 

 per fish at Kvichak and Egegik suggest that the 

 juveniles eat very little when they first enter salt 

 water. Straty (1974) concluded that the young 

 sockeye salmon did not feed when they entered 

 Bristol Bay or that food was scarce. Reduction of 

 feeding could be caused by a number of factors 

 other than lack of food, including the physiologi- 

 cal strain of adjusting osmoregulatory ftinction 

 from a freshwater to a marine environment. 



The differences I observed in the types, relative 

 proportions, and amounts of food eaten over suc- 

 cessive months by the juvenile sockeye salmon as 

 they progressed seaward can be largely attrib- 

 uted to food availability. Near-surface waters in 

 the Kvichak area contained an average of 27 

 zooplankters per cubic meter in June, while near 

 Port Moller in July the density was 1,400-8,100 

 (see footnote 1). Straty (1974) compared zooplank- 

 ton abundance in the inner part of Bristol Bay 

 (above Port Heiden) and the outer part (below Port 

 Heiden) during 1969-71 and concluded that zoo- 

 plankton was much more abundant as one pro- 

 gressed seaward. 



Acknowledgements 



I thank Walter R. Whitworth, Professor of 

 Fisheries of the University of Connecticut, for his 

 advice and help with this manuscript, which in 

 more inclusive form was accepted as an M.S. 

 thesis (1968) at the University of Connecticut, 

 Storrs, and also thank my colleagues at the 

 Northwest Fisheries Center Auke Bay Labora- 

 tory, National Marine Fisheries Service, NOAA, 

 particularly Richard R. Straty and Herbert W. 

 Jaenicke for their assistance and advice through- 

 out the study. 



461 



