FISHERY BULLETIN: VOL. 70. NO, 1 



frequencies needed in postencroachment tagging 

 to be do^f certain of detecting a change of one- 

 half day and 99 -^r certain of detecting a change 

 of a full day in elapsed times, if such changes 

 occurred between 1954-55 and 1964-65. If we 

 assume that the basic condition of the tagged 

 fish and the time required for tagged fish to re- 

 cover from possible effects of tagging did not 

 diflfer significantly between the 1954-55 and 

 1964-65 experiments, it follows that changes in 

 elapsed times could be attributed to the ability of 

 tagged fish to find and ascend the fishways. The 

 efficiency of the fish-passage system at Rock Is- 

 land Dam could thus be compared under pre- and 

 postencroachment conditions. Although travel 

 times were expected to be the most sensitive 

 measure of encroachment effects, it is obvious 

 that any significant drop in the percentage of 

 tagged fish passing over the dam would indicate 

 severe stress under postencroachment conditions. 



It might seem unrealistic at first to assume 

 that tagged fish recovered from the possible 

 effects of tagging equally well in the pre- and 

 postencroachment phases of the study. One 

 might expect, for example, that tagged fish were 

 released into faster moving water in 1954-55 and 

 into slower moving water in 1964-65 and that, 

 accordingly, the tagged fish required longer to 

 recover from the effects of tagging in the earlier 

 phase of the study than in the latter. If this 

 were true, we might have ended up measuring 

 diflferences in recovery time of tagged fish rather 

 than differences in the efficiency of the Rock 

 Island Dam fish ladders. 



Although water velocities were not measured 

 at the release sites, velocities measured in a 

 model of Rock Island Dam (Ward, 1965)' were 

 not uniformly diflferent under postencroachment 

 conditions than under preencroachment condi- 

 tions. In fact, velocities at the measuring point 

 nearest the right-bank release site on the simu- 

 lated model were generally higher after en- 

 croachment than before. On the other hand, 

 at the station closest to the left-bank release site, 



Ward, David A. 1965. Hydraulic model studies 

 of the Rock Island fish attraction facilities. Wash. 

 State Univ., Pullman, Div. Ind. Res., Inst. Technol 

 Res. Rep. 65/9-4.3. Vol. 1—20 p., 29 fig., Append. I-II; 

 Vol. 11—23 fig. (Processed.) 



velocities were higher at lower river flows and 

 about the same at higher flows after encroach- 

 ment. 



Observations made during the 1964-65 exper- 

 iments revealed that large numbers of tagged 

 fish tended to remain close to shore in protective 

 eddies. According to French, tagged fish be- 

 haved similarly during the 1954-55 experiments." 

 These observations tend to support the assump- 

 tion that tagged fish recovered from tagging 

 equally well in 1954-55 and 1964-65. 



The flooding of the lower portions of the fish 

 ladders at Rock Island Dam by Wanapum Reser- 

 voir was not the only factor aflfecting fish pas- 

 sage that changed between 1954-55 and 1964-65. 

 Riprap was added to the left bank of the river 

 below the dam, and the left and center fish lad- 

 ders were modified extensively. Figures 3 and 4 

 show Rock Island Dam before and after Wana- 

 pum Reservoir had been filled. New spill pat- 

 terns designed to enhance fish passage were in 

 eflfect throughout most of the 1964 and all of the 

 1965 tagging. The basic pattern was developed 

 from findings of the 1958 study (Leman and 

 Paulik, 1966) and modified slightly after ex- 

 periments with the model of Rock Island Dam 

 (Ward, 1965, see footnote 5). 



METHODS AND MATERIALS 



The basic experimental procedure was as fol- 

 lows: salmon were trapped as they passed over 

 the left ladder, transported to the release sites 

 approximately 300 m below Rock Island Dam 

 on either side of the river, then tagged and re- 

 leased. Fish counters at the dam recorded the 

 tags as the tagged fish passed the counting 

 boards after reascending the ladders. 



TAGGING 



Two diflferent traps were used to capture the 

 salmon. Sockeye were captured as they entered 

 a trap placed at the upstream edge of the count- 

 ing board in the left-bank fish ladder. Chinook 

 salmon, which would not enter this trap, had to 



" Personal communication, Robert R. French, Fi-shery 

 Biologist, Natl. Mar. Fish. Serv., Northwest Fish. Cent., 

 Seattle, Wash. 



128 



