FISHERY BULLETIN: VOL. 73, NO. 4 



vaceans, barnacle nauplii, cladocerans, and other 

 small crustaceans. Chum salmon fry tended to eat 

 more larger hard-shelled organisms and 

 epibenthic organisms than did pink salmon fry. 

 The food of pink and chum salmon fry at Traitors 

 Cove in general was similar to that reported at 

 Uala and Anapka bays on the east side of the 

 Kamchatka Peninsula (Andrievskaya 1968), the 

 San Juan area of northern Washington (Annan 

 1958), the Strait of Georgia in southern British 

 Columbia (Barraclough 1967; Robinson et al. 1968), 

 Chatham Sound off the northern coast of British 

 Columbia (Manzer 1969), and Moser Bay of 

 southeastern Alaska (Chamberlain 1906). In con- 

 trast, in Puget Sound epibenthic organisms 

 (especially harpacticoid copepods) were more im- 

 portant than pelagic zooplankters to pink and 

 chum salmon fry (Gerke and Kaczynski 1972). 



Food Selection 



Salmon fry in Traitors Cove did not eat the same 

 kinds and sizes of zooplankters in the same rela- 

 tive numbers as they appeared in the samples of 

 zooplankton, i.e., the fry fed selectively. Selective 

 feeding in relation to sizes of prey and juvenile 

 chum salmon has been reported by LeBrasseur 

 (1969). The average size of the zooplankton eaten 

 bv the fish was greater than the zooplankton 

 collected by the Clarke-Bumpus sampler (Table 5). 

 A coarser net such as a no. 6 mesh (233 /u,m) would 

 probably have collected the zooplankters that were 

 usually eaten by salmon fry and would not have 

 collected so many of the small forms that are sel- 

 dom eaten such as tintinnids, rotifers, and others. 



Selective feeding by pink and chum salmon fry 

 was also demonstrated by the occurrence of cer- 

 tain food items relatively more often in the 

 stomachs of fry (Tables 3, 4) than in the plankton 

 samples (Table 8). Relatively more cladocerans, 

 decapod zoeae, and larvaceans were eaten by salm- 

 on than appeared in the plankton samples. 

 Another example of the marked disparity is the 

 barnacle nauplii which were very abundant in 

 most of the plankton samples (4 to 94% of the 

 number of plankters) but constituted only 14% of 

 the animals actually eaten by pink salmon and 

 only 3% of the number of food items eaten by chum 

 salmon. 



The high incidence of larvaceans in the stomach 

 samples, especially in the chum salmon, may be the 

 result of selective feeding on a scarce but very 

 visible plankter. Larvaceans, in particular 



Oikopleura spp., form mucous feeding nets which 

 may increase the visibility of the larvacean to the 

 salmon fry. Once learning to capture Oikopleura, 

 the fry may prefer that food item. 



Benthic and intertidal forms of mysids, 

 cumaceans, isopods, amphipods, and insects were 

 rare in the plankton samples and their presence in 

 some of the stomachs shows that pink and chum 

 salmon fry did on occasion feed in these ecological 

 niches. This type of feeding behavior could not 

 predominate at Traitors Cove because most of the 

 shoreline is rocky and precipitous and offers little 

 opportunity for benthic feeding. 



Grazing Rate 



The average number of zooplankters comsumed 

 daily by a pink salmon fry in Traitors Cove was 

 calculated from estimates of average stomach 

 contents and evacuation rates. Stomachs of pink 

 salmon collected from Traitors Cove estuary dur- 

 ing daylight contained an average of 136 

 zooplankters. Stomach evacuation required 6 and 

 16 h at temperatures of 12.8°C and 8.5°C, respec- 

 tively, although Brett and Higgs (1970) observed 

 slower stomach evacuation rates at comparable 

 temperatures in sockeye salmon fingerlings that 

 had been fed a commercial pelleted food. The fry 

 did not feed during darkness, which extended 

 from about one-half hour after sunset to one-half 

 hour before sunrise on cloudy or moonless nights. 

 The duration of feeding at Traitors Cove when fry 

 are present typically is about 16.5 h (range 15 to 18 

 h); the water temperature at 1 m ranges from 5° to 

 13°C. 



Thus, it appears that fry would consume a 

 volume of food required to fill their stomachs once 

 a day at cooler temperatures (8.5°C) and four 

 times a day at warmer temperatures (12.8°C). The 

 number of zooplankters consumed daily would, 

 therefore, range between 136 and 544 per pink 

 salmon fry for temperatures that are normal dur- 

 ing the time fry are in Traitors Cove. By the same 

 Hne of reasoning, chum salmon would consume 

 about 120 to 480 food items per fry per day in 

 Traitors Cove. 



Some insight into the availability of food for 

 salmon fry at Traitors Cove was obtained by con- 

 sidering the abundance of plankton in relation to 

 the feeding habits of the fry. For example, fry 39 

 mm long that were holding a position relative to 

 the shore while feeding in a current of 11 cm/s 

 were in effect grazing a cylindrical mass of water 



856 



