FISHERY BULLETIN: VOL. 86, NO. 2 



freshwater and 2) larger individuals may have 

 migrated continuously from the estuary to the 

 sea. Conversely, growth could be overestimated if 

 predators of salmon selected small individuals 

 (Parker 1971). In addition, although the inner 

 and middle basins are semiisolated from adjacent 

 marine waters during low tide, juvenile salmon 

 from adjacent waters could easily enter the estu- 

 ary, especially the outer basin, during flood tide 

 and mix with salmon from Porcupine Creek. 



Estimates of salmon growth in estuaries and 

 nearshore marine waters are variable, but gener- 

 ally range between 1 and 2 mm/day. LeBrasseur 

 and Parker (1964) estimated pink salmon growth 

 to be 0.9 mm/day during the first 30 days at sea, 

 and Healey (1978) estimated pink salmon growth 

 during summer to be 1.0 mm/day; our estimate 

 was 1.5 mm/day. Our estimate for chum salmon 

 at 0.4 mm/day was considerably less than that of 

 Healey (1978) at 1.5 mm/day, also based on 

 change in mean length; however, our estimate for 

 coho salmon of 1.6 mm/day was similar to that of 

 Healey (1978) at 1.2 mm/day. Summer growth 

 back-calculated from scales of salmon from the 

 Sea of Okhotsk was about 1.6 mm/day for pink 

 and chum salmon (Birman 1969). 



Because of their initial small size, pink and 

 chum salmon particularly are vulnerable to 

 predators including juvenile coho salmon (Parker 

 1971). Several authors have suggested that a 

 major share of pink salmon mortality in the first 

 weeks at sea results from juvenile coho salmon 

 predation (Parker 1971; Kaczynski et al. 1973; 

 Hargreaves and LeBrasseur 1985), but such pre- 

 dation has not been found in field collections. 

 Parker (1971) demonstrated predation by juve- 

 nile coho salmon on pink salmon fry in the labora- 

 tory, and juvenile coho salmon are known preda- 

 tors of salmon fi:y in fi*eshwater (Hunter 1959; 

 Koski and Kirchhofer 1984). However, we have 

 not found any published data that show predation 

 by juvenile coho salmon on other salmon in estu- 

 aries or marine waters. Predation by juvenile 

 coho salmon on pink salmon fiy migrating ft-om 

 freshwater does occur in the tidal-influenced 

 reach of Porcupine Creek (Koski and Kirchhofer 

 1984), but such predation apparently does not ex- 

 tend into the estuarine basins. Many fishes have 

 been identified as predators of pink and chum 

 salmon in estuaries, including Pacific herring 

 (Thorsteinson 1960), sea-run cutthroat trout, 

 Salmo clarki; cod, Gadus macrocephalus ; and 

 sculpin, Leptocottus armatus, (Bax et al. 1977). 

 We speculate that predation by coho salmon on 



salmon fiy may occur only under circumstances 

 in which the coho salmon are combined with 

 small fry as they migrate from freshwater. 



The period of vulnerability of pink and chum 

 salmon fry to predation by juvenile coho salmon is 

 probably relatively short. Within the first 3 

 weeks after entering the estuary, pink salmon fiy 

 can grow larger than the prey fish of juvenile coho 

 salmon. In the laboratory, juvenile coho salmon 

 ate the smallest pink salmon available and did 

 not eat any larger than about 50 mm FL (Parker 

 1971), which coincides with the largest fish eaten 

 by coho salmon in our study. At a growth of 1 

 mm/day, pink salmon entering the estuary at 32 

 mm FL will outgrow predation by coho salmon 

 smolts in 18 days. In Porcupine Creek, most pink 

 and chum salmon migrated fi-om the stream sev- 

 eral weeks before coho salmon, which enables 

 them to grow large enough to avoid predation by 

 coho salmon in the estuary. Thus, early migration 

 and rapid growth of pink and chum salmon fiy 

 probably are important in reducing predation by 

 coho salmon. 



In the Porcupine Creek estuary, competition 

 and predation probably were slight. Competition 

 for food was minimal, as evidenced by the rapid 

 salmon growth, because of differences in prey and 

 foraging mode and because regular tidal flushing 

 probably replenished food supplies, as in Traitors 

 Cove, AK (Bailey et al. 1975). Natural stocking 

 levels in the estuary also probably were below 

 thresholds where competition for food would de- 

 press survival. Predation by coho salmon on pink 

 and chum salmon was avoided because the pink 

 and chum salmon migrated earlier than coho 

 salmon and rapidly grew too large for the 

 coho to handle. Thus, in this natural system, com- 

 petition and predation probably were unimpor- 

 tant because of moderate stocking levels, rapid 

 growth, and differences in diet and timing of mi- 

 grations. In systems with hatchery inputs, how- 

 ever, stocking levels would probably be higher 

 and salmon size and timing of migrations differ- 

 ent than in natural systems, which could increase 

 competition and predation. 



Stocking levels and timing of hatchery releases 

 of juvenile salmon in estuaries are important in 

 minimizing competition and predation (Myers 

 1980). Hatchery releases should avoid combining 

 large concentrations of pink and chum salmon fi-y 

 so as not to deplete food supplies. Conversely, re- 

 leases during low predator abundance and good 

 growing conditions — high food availability and 

 warm temperature — could increase grov^rth and 



220 



