minimized, which is already being accomplished 

 by the hatchery practice of holding salmon 

 until they are at a migratory stage. 



SUMMARY 



Chinook salmon fry reared in a hatchery 

 and released into the Elokomin River in May 

 1962 and 1963 were available to predators for 

 about 2 days. Fourteen species of fish were 

 collected, but the predator species were lim- 

 ited to four sculpins, two trout, and one salmon. 



Sculpins were significant predators; the 

 extent of salmon loss, however, varied within 

 and between the species of sculpin. The tend- 

 ency of sculpins to prey on salmon increases 

 with body size and mouth width. Because of 

 the large size it attains the prickly sculpin 

 is extremely predatory and inflicted the great- 

 est estimated mortality on chinook salmon fry. 

 Torrent and reticulate sculpins were of less 

 importance as predators. Despite having the 

 largest population, the coastrange sculpin was 

 an insignificant predator. Estimates of chi- 

 nook salmon fry eaten by sculpins were 58,868 

 out of a release of 1.5 million in 1962 and 

 27,969 out of 2.2 million and 4,561 out of 

 127,000 in 1963; respective percentage losses 

 were 3.9, 1.3, and 3.6. The mean length of 

 the chinook salmon released was about 60 mm., 

 but the mean length of those found in sculpins 

 was about 43 mm. 



Salmon released from Oxbow Hatchery 

 were little affected by predation from sculpins 

 in the 0.2 km. of stream to the Columbia River, 

 because numbers of predators and intensity 

 of predation were low and the duration of pre- 

 dation was short. 



In Herman Creek the prickly sculpin was 

 generally more predaceous, larger, and more 

 abundant than reticulate sculpin. The data 

 support the theory that improvement of hatch- 

 ery techniques tends to reduce predation more 

 effectively than control of predators. Chinook 

 salmon fed the Oregon moist pellet in 1963 

 appeared to be of higher quality at release and 

 had lower mortality following release in the 



Elokomin River than those fed a wet diet in 

 1962. Liberating chinook salmon en masse 

 at night apparently reduces percentage pre- 

 dation. The proximity of migrating salmon 

 fry to predators and the high rate of predation 

 in streams suggest that the transport of hatch- 

 ery fish around concentrations of predators 

 may be of value. 



ACKNOWLEDGMENTS 



This study was carried out under the 

 Bureau of Commercial Fisheries Columbia 

 River Operation Studies Program. R. J. Gerke, 

 now with the State of Washington Department 

 of Fisheries, helped throughout the study; S. 

 Murai of the Bureau of Commercial Fisheries 

 Biological Laboratory helped with the statis- 

 tical computations; and T. J. Wogoman, 

 student, helped collect predators in the Elok- 

 omin River in 1962. E. R. Jeffries, Fish Com- 

 mission of Oregon, and R. E. Noble and E. 

 Miller, State of Washington Department of 

 Fisheries, supplied me with data on hatchery 

 releases and information on the hatchery fish. 



LITERATURE CITED 



BURROWS, ROGER E. 



1964. Effect of fingerling stamina on adult 

 survival. Proc. Northwest Fish Cult. 

 Conf., Oreg. St. Univ., Corvallis, Oreg., 

 Dec. 2-3, 1964: 54-55. [Processed.] 



DAIBER, FRANKLIN C. 



1956. A comparative analysis of the win- 

 ter feeding habits of two benthic stream 

 fishes. Copeia 1956: 141-151. 



ELLIS, C. H., and R. E. NOBLE. 



1960. Barging and hauling experiments 

 with fall chinook salmon on the Klickitat 

 River to test effects on survival. Wash. 

 State Dep. Fish., 70th Annu. Rep.: 

 57-65. 



13 



