Table 5 contains a summary of releases of tagged 

 sockeye by age group and release area and recov- 

 eries at Tumwater Dam, in the Okanogan River, 

 and in other areas combined. The estimated total 

 recoveries of 3- and 4-year-old tagged sockeye in 

 the Okanogan system also are shown. 



DETERMINATION OF MORTALITIES 

 AT ROCK ISLAND 



If there were a substantial loss of fish below Rock 

 Island Dam, whether caused by a failure of fish 

 to find a fish ladder or for other reasons, there 

 should be a greater upstream recovery of fish 

 tagged and released above the dam than of fish 

 similarly tagged and released below the dam. 

 Mortalities also should be reflected by the number 

 of tagged fish passing the dam compared to the 

 number released. We found, as shown later, that 

 the fish counters' records of tagged fish passing 

 Rock Island Dam from releases below the dam did 

 not give completely reliable data on fish mor- 

 talities (assuming unaccounted-for tags as mortal- 

 ities caused by the dam) . Results of the upstream 

 tag recovery comparisons are presented first, 

 followed by the results obtained at Rock Island 

 Dam. 



RETURNS FROM, UPSTREAM 



Sockeye 



Analysis of the tagging data to estimate mortal- 

 ities is complicated by the presence of different 

 races of sockeye in the tagging groups, by different 

 age groups, and by the unequal effort expended 



recovering tags from different races and age 

 groups. The sockeye run separates a short 

 distance above Rock Island Dam, with some fish 

 going into the Wenatchee River system and some 

 fish up the Columbia River to the Okanogan River 

 system. A few sockeye show up at the Entiat 

 and Winthrop fish hatcheries and in Icicle Creek 

 below the Leavenworth fish hatchery. As stated 

 previously, 3-year-old sockeye are restricted to the 

 Okanogan system and 4-year-old fish are found in 

 all areas. We could effectively observe or recover 

 all tagged fish passing Tumwater Dam on the 

 Wenatchee River while on the way to spawning 

 areas. In the Okanogan system, however, tag 

 recoveries were made during stream surveys of 

 the spawning areas and sampling at Zosel Dam on 

 the Okanogan River, Thus, the recovery effort 

 for tags on fish in the Wenatchee and Okanogan 

 systems was not equal. Adding to these compli- 

 cations is the very probable loss of some tagged 

 fish due to a tagging mortality or to straying. 



Taking a simple approach, we have estimated 

 the mortality rates from the ratios of the recov- 

 ery proportions of tagged fish recovered from 

 above- and below-dam releases. We made the 

 following general assumptions: (1) The chance of 

 recovering tagged fish is the same regardless of 

 tagging date and tagging site, (2) racial propor- 

 tions in the tagging lots are the same for the 

 different areas of release, and (3) the chance of a 

 tag recovery or observation is independent of tag 

 type or color. 



Table 3. — Estimated numbers of 3- and 4-year-old tagged sockeye salmon in recovery areas 2 and 3, by year and release area 



LOSS AND DELAY OF SALMON PASSING ROCK ISLAND DAM 

 774-711 O— '66 4 



345 



