with chinook salmon indicated no mortalities 

 caused by the dam. 



Residts show that mortalities caused by the 

 dam are neither substantial nor consistent over 

 the years, or that our tagging methods could not 

 detect mortalities accurately. Confidence limits 

 for the point estimates of mortalities ranged from 

 zero (in many instances) to about 28 percent for 

 the lower limits; upper limits ranged from 20 to 

 66 percent. 



Comparatively few tagged fish used the right 

 ladder (also observed for untagged runs), even 

 when they were released near it. Apparently, 

 the majority of these salmon ascended other 

 ladders. An increased percentage of the salmon 

 was attracted to the right ladder in 1957 when 

 spillway gates on the right side of the dam were 

 opened during the fish migration season. Pre- 

 viously these were closed to reduce water turbu- 

 lence near the right ladder entrance. It may be 

 that the added flow attracts fish to this bank and 

 into the fishway. 



While altering the right bank fishway might 

 induce more fish to use the right ladder, we 

 cannot say it would result in markedly improved 

 facilities with respect to mortalities. 



DELAY 



Delay of migrating adult salmon runs is very 

 important to the survival of the runs. As 

 Thompson (1945) and Talbot (1950) pointed out, 

 delays to sockeye salmon runs in the Fraser River 

 caused mortalities and a serious decline in the runs. 

 Schoning and Johnson (1956), in a study at Bonne- 

 ville Dam on the lower Columbia River, reported 

 that migrating chinook salmon were delayed 2.6 to 

 3 "days. The 2- to 4-day delay at Rock Island 

 Dam may not be important in itself, but if the 

 delay is multiplied by a series of dams, serious 

 losses of anadromous fish populations may result. 



We do not know if upper river salmon runs 

 would survive the series of delays in migration 

 that may result when all Columbia River dams are 

 completed. It may be that delays, if short-termed 

 would be compensated for by easier and faster 

 travel through the reservoirs created by the dams. 

 These salmon do not spawn immediately upon 

 reaching the spawning areas. There is an interval 

 before spawning, of from 1% to 2K months after 

 the salmon pass Rock Island Dam. This ripening 

 period has been noted on other river systems. 



Howard (1948) found that the period for sockeye 

 in Cultus Lake, British Columbia was about 1 

 month. Schaefer (1951) noted the time for sock- 

 eye in the Harrison River system of British Colum- 

 bia, was a month or less. The ripening period for 

 sockeye in Lakelse Lake, British Columbia, aver- 

 aged 54 days (Fisheries Research Board of Cana- 

 da, 1954). In the Bristol Bay area of Alaska, 

 some sockeye remain in the lakes up to 3 months 

 before spawning. The ripening period evidently 

 is a necessary adjunct to the migration time. 



In this study at Rock Island Dam, we dis- 

 covered a 2- to 4-day delay to migrating salmon. 

 Upstream tag observations at Zosel Dam indicated 

 that fish tagged and released below the right ladder 

 were delayed longer than fish tagged and released 

 below the left ladder. This was confirmed by 

 comparing returns of left and right bank tag 

 releases at Rock Island for 1954. The longer 

 delay was absent in 1955 and 1956. Thus, while 

 we do not have positive evidence of a greater delay 

 for fish released below the right ladder, such a 

 delay is indicated. 



Altering the right-bank fishway may lessen 

 delay at the dam because of the possible increased 

 chances of salmon finding a ladder sooner. How- 

 ever, the same delay pattern was noted at Bonne- 

 ville Dam (app.), where ladders at both sides of the 

 dam were designed to be equally attractive. From 

 this study, we cannot conclude that altering the 

 right bank fishway would decrease materially the 

 overall delay at Rock Island Dam. 



SUMMARY AND CONCLUSIONS 



We conducted a tagging program at Rock Island 

 Dam during 1953-56 to determine whether the 

 dam caused loss or delay of salmon passing the 

 dam, and whether such loss or delay was associ- 

 ated with the failure of fish to find and use the 

 dam's right bank fishway. In this study, tagged 

 salmon were released below the left and right bank 

 fishways and above the dam. Fish counters at 

 the dam identified different tag groups by noting 

 the different colors and kinds of tags on salmon 

 crossing the counting boards. Tags also were 

 identified at upstream points on migration routes 

 and during spawning surveys. 



A comparison of upstream tag recoveries from 

 the different release areas gave conflicting results 

 regarding losses of fish. Point estimates of losses, 

 for some combinations of years and release areas 



364 



U.S. FISH AND WrLDLIFE SERVICE 



