Factors Influencing Ocean Catches of Salmon, Oncorhynchus spp., 

 off Washington and Vancouver Island 1 



R. A. LOW, JR. 2 and S. B. MATHEWS 3 



ABSTRACT 



The relative influence of various factors on ocean fishing success was evaluated for pink, Oncorhynchus 

 gorbuscha, chinook, O. tshawytscha, and coho, O. kisutch, salmon off Washington and Vancouver Island. In 

 addition, an evaluation was made of the practicality of predictive models for ocean catch. For each species, predic- 

 tive regression equations were developed and their reliability evaluated in terms of the average percentage error of 

 predicted catches from actual catches. 



Pink salmon catches were significantly correlated with indices of brood year abundance and the average in- 

 dividual weight of fish caught in terminal areas during the brood year. Average error of predicted catches ranged up- 

 ward nt ± 25" . Success for chinook salmon in year /was highly associated with Columbia River hatchery' releases of 

 fall brood year groups i— 3 and i - 4, Canadian purse seine catches of immature chinook salmon in Canadian area 20 

 during August of year r — 1, and troll catch per effort during the fall of year /" - 1 . Washington troll and sport catches 

 of chinook salmon were also significantly correlated with the amount of nominal fishing effort. Coho salmon catches 

 were significantly associated with level of Fishing effort, indices of brood year abundance of Columbia River wild 

 coho salmon, and Columbia River jack returns the preceding year. The average error of predicted annual troll coho 

 salmon catches off the central Washington coast was ± 15% for 1966-75. 



INTRODUCTION 



Pink, Oncorhynchus gorbuscha, chinook, O. tshawytscha, 

 and coho, O. kisutch, salmon returning to Washington waters 

 are subjected to intensive ocean fisheries off the west coast of 

 Vancouver Island and Washington. In inside waters, the level of 

 exploitation by commercial netters and recreational anglers is 

 based on the remaining harvestable portion of the runs. One of 

 the salmon management objectives of the Pacific Fishery 

 Management Council is to provide all ocean and inside fisheries 

 the continuing opportunity to harvest salmon. In order to 

 achieve equitable allocation of fishing opportunities in addition 

 to conservation of salmon resources, management agencies 

 must understand the relative effects of various factors on ocean 

 fishing success. Because terminal area allocations for several 

 species are based in part on preseason estimates of ocean 

 interceptions, reliable predictive models for ocean catch are also 

 desirable. Previous studies of effects of diverse conditions on 

 abundance of and/or fishing success for the species mentioned 

 have been limited to analyses of one or a few categories of data 

 for a particular stock in a limited area. Very few of these studies 

 have addressed the ocean fisheries. 



Trends in the Washington troll fishery through 1975 were 

 described by Wright (1970, 1976). Haw et al. (1967) described 

 the development of the Washington salmon sport fishery prior 

 to 1965, while Phinney and Miller (1977) analyzed trends in the 

 ocean sport fishery since 1970. United States Department of 

 Commerce et al. (1977) provided a detailed account of the 

 development and status of the ocean salmon fishery off Cali- 

 fornia, Oregon, and Washington. Milne and Godfrey (1964) and 



'Contribution Number 533 of the College of Fisheries, University of Washing- 

 ton, Seattle, WA 98195. 



'College of Fisheries, University of Washington, Seattle, WA 98195; present ad- 

 dress: Marine Resources Research Institute, South Carolina Department of 

 Wildlife and Marine Resources, Charleston, SC 29412. 



'College of Fisheries, University of Washington, Seattle, WA 98195. 



Godfrey (1969, 1970) discussed aspects of the Canadian troll 

 fishery off the west coast of Vancouver Island. 



Although fisheries managers have traditionally assumed that 

 abundance of pink salmon is associated with brood year abun- 

 dance in a Ricker spawner-recruit curve, other associations have 

 been reported. One factor that may influence the abundance of 

 offspring is the average individual weight of spawning females, 

 since higher fry survival rates appear to be associated with larger 

 brood females (Skud 1973). Survival of out-migrant fry in 

 estuarine assembly areas may also reflect a prey-predator rela- 

 tionship with yearling coho salmon (Hunter 1959). During their 

 marine residence, pink salmon from the Fraser River system and 

 Puget Sound tributaries appear to follow a similar migratory 

 route (Neave et al. 1967; Royce et al. 1968). The oceanographic 

 condition most likely to be associated with survival of juveniles 

 and catchability of maturing fish is surface sea temperature 

 (Vernon 1958; Favorite 1961; Birman 1964; International Pacific 

 Salmon Fisheries Commission 1974). The average individual 

 weight of maturing pink salmon varies inversely with run 

 strength in inside waters (International Pacific Salmon Fisheries 

 Commission 1974). Ocean catches could be related to all of 

 these elements. 



A Ricker-type spawner-recruit relationship has also been 

 assumed for major chinook salmon stocks (Van Hyning 1973), 

 but again other factors are involved. Contributing to production 

 are hatchery releases, of massive proportions in recent years 

 (Wahle et al. 1975). During out-migration, juvenile fall chinook 

 salmon, particularly Columbia River fish, may be detrimentally 

 influenced by warmer-than-normal surface coastal sea 

 temperature (Vernon 1958; Van Hyning 1968) and lower-than- 

 normal upwelling rates (Van Hyning 1968). The location and 

 extent of upwelling may also affect catchability of maturing 

 chinook salmon because of the influence on abundance and 

 distribution of forage. Tully (1954) found that the only variables 

 significantly associated with troll catch rates off the west coast 

 of Vancouver Island were wind velocity and direction. Because 



