FISHERY BULLETIN: VOL. 78, NO. 2 



wire tag was considered a fish that had lost its tag 

 after marking. Steelhead and chinook salmon 

 were kept in separate compartments in the tank 

 truck whenever both species were hauled simul- 

 taneously. All fish were transported in a truck of 

 18,900 1 (5,000 gal) capacity (Smith and Ebel 1973) 

 that was equipped with aeration, refrigeration, 

 and filtration systems. Load densities were gov- 

 erned by the size of the daily catch and were usu- 

 ally <0.12 kg/1 (1 lb/gal) except during periods 

 when unmarked fish were hauled. Maximum load 

 density was kept <0.18 kg/1 (1.5 lbs/gal) of water. 

 Water chemistry measurements were taken 

 from the truck at the time of release for every load 

 transported. Concentrations of ammonia, nitro- 

 gen, dissolved oxygen, carbon dioxide, pH, and 

 total alkalinity were recorded for possible correla- 

 tions with delayed mortality information. All re- 

 leases were made at dusk. Records of mortality 

 were kept during marking and at time of release; a 

 sample of 50-100 fish was taken from each trans- 

 ported load and held for 48 h at Bonneville Dam to 

 provide an indication of delayed mortality. This 

 procedure was repeated during downstream mi- 

 grations in 1971, 1972, and 1973. 



Evaluation of Returning Adults 



The effect of transportation on the survival and 

 homing of adult fish was evaluated by comparing 

 returns of transported and nontransported fish to 

 the sport, commercial, and Indian fisheries in the 

 lower Columbia River; to Little Goose Dam on the 

 lower Snake River; to Rapid River Hatchery, 

 Pahsimeroi Hatchery near Salmon, Idaho, and 

 Dworshak Hatchery; and to the spawning grounds 

 throughout the Snake River drainage. 



All adult fish migrating upstream at Little 

 Goose Dam must ascend one ladder located on the 

 south side of the dam. An adult tag detection and 

 fish separating device that intercepted tagged 

 salmon and steelhead and diverted them into a 

 holding pen was installed in this ladder in 1972 

 (Ebel 1974). Tagged fish from our study were read- 

 ily identified by the missing adipose fin. All fish 

 were anesthetized and further examined for 

 brands. If the brand was recognizable, the origin of 

 the fish could be determined without having to 

 extract the magnetic tag from the snout. 



Fish with recognizable brands were then 

 weighed and measured, dart-tagged or jaw-tagged 

 (Slatick 1976), and released to provide further in- 

 formation in the event of recapture upstream and 



to identify fish that fell back over the dam and 

 ascended the ladder a second time. If a fish was 

 known to be tagged but had a brand that was 

 indistinguishable, it was held until maturity in 

 holding tanks at the dam and artificially spawned. 

 The tag was then extracted after spawning, and 

 the test or control group was determined from the 

 color code. Data obtained from these fish were 

 combined with those obtained from reading 

 brands. 



The Columbia River gillnet fishery below Bon- 

 neville Dam, the Indian fishery above the dam, 

 and the sport fishery (primarily below the dam) 

 provided samples of chinook salmon throughout 

 the spring run. The samples yielded information 

 concerning the returns to the lower river of 

 marked fish originating primarily in Idaho. Clo- 

 sure of the summer fishery on chinook salmon 

 during all 3 yr and the spring fishery in 1974 and 

 1975 prevented sampling of this segment of the 

 run in the lower river. The sport and commercial 

 fisheries of the lower Columbia River and the sport 

 fishery above Little Goose Dam provided samples 

 of steelhead. 



Surveys of spawning grounds were conducted 

 with the cooperation of the Washington Depart- 

 ment of Fisheries, Fish Commission of Oregon, 

 and the Idaho Department of Fish and Game. Most 

 of the surveys were in the Snake River drainage of 

 Idaho, but hatcheries and spawning grounds of 

 spring and summer chinook salmon in the upper 

 Columbia River were also checked for strays. 



The G statistic. Student's ^-test, and analysis of 

 variance were used for analysis of most return 

 data. 



RESULTS 



Factors Influencing Assessment of Data 



Tag loss, tag detector efficiency, transport mor- 

 tality, and delayed mortality were factors that 

 influenced the assessment of the experimental 

 data. Comparisons of tests and control releases 

 could be biased if a differential effect among any of 

 these factors occurred between test and control 

 releases and was not considered in the analysis. 

 For example, if tag loss was greater in control 

 releases than in test releases, percentage return 

 would be biased in favor of the test release if the 

 data were not adjusted for this loss. 



During the 3 yr of this study, average annual 

 initial tag loss ranged from 0.45% in 1973 to 10.4% 



494 



