of the drift, it is apparent why the diet composi- 

 tion, in synchrony with the drift, varies over 

 a 24-h period. 



Mason (1966) demonstrated nocturnal feeding 

 activity of coho salmon fry. He attributed the 

 night feeding habits of coho salmon to high retinal 

 cone sensitivity, believing that it aided them in 

 utilizing the increased drift of aquatic inverte- 

 brates during periods of low light intensity. 

 Jenkins^ observed that rainbow trout fed most 

 actively at midday and that the diet at this time 

 was composed mainly of terrestrial invertebrates 

 which predominated in the drift. However, these 

 fish were overyearlings, so a direct comparison 

 with this study (fry) cannot be made since as 

 salmonids grow older (and larger), not only do 

 they occupy different habitats (Saunders and 

 Smith 1962) but they also utilize different prey 

 species (Kallenberg 1958), in addition to eating 

 larger prey when available (Allen 1969). Tippets 

 and Moyle (1978), however, did demonstrate that 

 during the summer, rainbow trout fry fed predom- 

 inantly on drifting invertebrates during the day in 

 the McCloud River, Calif. In Orwell Brook in 1977, 

 the diurnal diet of steelhead fry during the sum- 

 mer was found to be more closely associated with 

 the composition of the bottom fauna rather than 

 with the composition of the drift (Johnson and 

 Ringler 1980). These differences in the feeding of 

 steelhead may be due to social interaction be- 

 tween coho salmon and steelhead fry in Orwell 

 Brook. In sympatry, during the summer, coho 

 salmon occupy pools while steelhead utilize riffles 

 (Hartman 1965; Allee 1974). In allopatry, al- 

 though steelhead fry occupy both habitats, they 

 prefer pools (Hartman 1965). If riffle versus pool 

 occupancy influences the prey selection of coho 

 salmon and steelhead as suggested by Johnson 

 and Ringler (1980), the diets of steelhead fry in 

 allopatric (predominately pools) and sympatric 

 (predominantly riffles) situations with coho 

 salmon may be expected to differ. 



Although subyearling coho salmon actually 

 consumed more food (10.6 mg daily meal) than 

 subyearling steelhead (7.8 mg), because coho 

 salmon were much heavier (Table 1), their daily 

 ration was substantially less than steelhead (1.7 

 and 2.8% daily rations). Elliott and Persson (1978) 

 suggested that, when using their formula to esti- 



•''Jenkins, T. M., Jr. 1970. Behavior-ecology. In Progress in 

 sport fishery research, p. 138-141. U.S. Fish Wildl. Serv., Bur. 

 Sport Fish Wild). 



mate daily food consumption, sampling intervals 

 should be 3 h or less; larger intervals may result in 

 inaccurate estimates. Since our sampling interval 

 was 4 h and rates of gastric evacuation were 

 determined from the literature (i.e., rainbow trout 

 fry fed oligochaetes, Windell et al. 1976) using 

 mean water temperatures (16°-21° C) for each 4-h 

 interval, our results are, at best, rough estimates 

 of the daily meal and daily ration of subyearling 

 coho salmon and steelhead. However, our esti- 

 mates are well within the range of juvenile sock- 

 eye salmon, Oncorhynchus nerka, in a lacustrine 

 environment in Washington (Doble and Eggers 

 1978). 



Johnson and Ringler (1980) found that the 

 diurnal diet (0900-1500 h) of coho salmon and 

 steelhead fry did not overlap substantially (i.e., 

 CaS=0.60) from June through September in Or- 

 well Brook during 1977. Interestingly, the only 

 time that the diet of these two species did not 

 overlap significantly in this study was during 

 approximately the same period, 0800-1600 h 

 (Table 1). The highest degree of overlap in the 

 diets of coho salmon and steelhead occurred from 

 2000 to 0400 h when both species were feeding 

 mainly on aquatic invertebrates. Inspection of 

 diet overlap during twilight feeding periods (0400- 

 0800 and 1600-2000 h) indicates significant over- 

 lap (x = 0.67), which is distinctly intermediate 

 between diurnal (x = 0.48) and nocturnal 

 ix = 0.87) periods. The large diel fluctuations in 

 diet similarity between these two species indicate 

 that diel studies are necessary when examining 

 aspects of their trophic ecology, at least when the 

 species occur sympatrically. Daytime food studies 

 could lead to the erroneous speculations that the 

 diet of sympatric juvenile coho salmon and steel- 

 head is not similar and that terrestrial inverte- 

 brates (at least during certain periods of the year) 

 are the major component in the diet of coho 

 salmon. Conversely, examination of the stomachs 

 contents during nocturnal and crepuscular 

 periods would indicate a great deal of similarity in 

 diet and would not reflect the importance of 

 allochthonous material in the daytime diet of coho 

 salmon. 



Acknowledgments 



We thank N. Ringler and R. Sloan for their 

 helpful comments on the manuscript. We also 

 thank A. Crawford who typed the manuscript and 

 G. Furman who prepared the figure. 



375 



