Table 4. — Estimates of total single recoveries in the Okanogan system by age group and release area, Rock Island Dam sockeye 



lagging, 1954-56 



We observe from the estimated confidence limits 

 that the lower limits bracket zero in most instances 

 and range to about 5 percent. The upper limits 

 range from 20 to about 46 percent. 



Another procedure was to estimate mortality 

 rates based on recoveries of 4-year-old sockeye at 

 Tumwater Dam (table 8). Compared with the 

 proportions of tagged fish recovered from area 

 AL (left bank above the dam) releases, tagged fish 

 from left bank-below releases suffered mortalities 

 of about 21, 0, and 26 percent, respectively, for 

 1954, 1955, and 1956. Tagged fish released on 

 the right bank below the dam suffered mortalities 

 of about 42, 23, and 29 percent, respectively, for 

 the same 3 years. 



Compared with the proportions of tagged fish 

 recovered from the area AR releases (right bank 

 above the dam), the tagged fish released at the 

 right and left banks below the dam suffered no 

 mortalities. 



Combining the data for the 3 years, we observe 

 on the basis of area AL releases that tagged fish 

 released below the left fishway suffered a mortality 



of about 22 percent and that tagged fish released 

 below the right bank fishway suffered a mortality 

 of about 39 percent. No mortality was indicated 

 on the basis of releases in area AR ; however, com- 

 paratively few fish were released there. 



Point estimates of mortality rates, obtained 

 from comparisons of proportions of tagged fish 

 recovered from total releases below the dam with 

 all releases above the dam are about 29 percent in 

 1954, percent in 1955, 25 percent in 1956, and 

 21 percent for all years. 



Considering the confidence limits of these 

 mortality rates based on Tumwater Dam re- 

 coveries, we observe the lower limits to bracket 

 zero in two instances and range to 28 percent. The 

 upper limits range from about 32 to 66 percent. 



It is readily apparent that the analyses of the 

 sockeye tagging data gave conflicting results of 

 mortalities caused by the dam. On the one hand, 

 many experiments indicated substantial mortal- 

 ities, and mortalities apparently were greater for 

 fish released below the right bank fishway. On 



LOSS AND DELAY OF SALMON PASSING ROCK ISLAND DAM 



349 



