County, the builders of Rocky Reach Dam, in 

 accordance with the terms of Federal Power Com- 

 mission license number 2145. Biologists of the 

 Bureau of Commercial Fisheries were designated to 

 study possible delay because of their experience with 

 a similar investigation at Rock Island Dam in 

 1953-5G (French and Wahle, 19G(i). 



The study underwent a major expansion as it 

 proceeded. Originally, the sole aim was to measure 

 delay, if any, caused to upstream-migrating adult 

 salmonids by Rocky Reach Dam. This aim was to 

 be accomplished by comparing the time required for 

 tagged sockeye salmon to migrate from the forebay 

 (reservoir) of Rock Island Dam to Zosel Dam, before 

 (1957) and after (1962 and 1963) the completion of 

 Rocky Reach Dam. If minimal influence by other 

 factors were assumed, any major change in travel 

 time could be attributed to the new structure. 



French and Wahle (1960) summarized the "pre- 

 dam" work (performed before the dam was built). 

 Because too few tagged fish were observed at Zosel 

 Dam in 1957, they estimated the travel time (10.7 

 days) from tagged sockej^e salmon tliat had been 

 released in the Rock Island Dam forebay and later 

 observed at Zosel Dam during the earlier (1953-56) 

 study. 



The tagging results in the first '"postdam" year 

 (1962) had a profound impact on our investigation. 

 Sockeye salmon migrating through the study area 

 were noticeably delayed — apparently by high water 

 temperatures of the Okanogan River. If a tem- 

 perature block were shown to exist, the assumption 

 of minimal influence by factors other than Rocky 

 Reach Dam would have been invalidated and the 

 straightforward comparison of "predam" and "post- 

 dam" travel times as a measure of delay ruled out 

 — unless the influence of the various factors could be 

 examined separately. It was necessary, therefore, 

 to confirm the existence of the temperature block, to 

 ree.xamine the estimated "predam" travel time, and 

 finally to evaluate Okanogan River temperatures as 

 well as Rocky Reach Dam as sources of delay to 

 sockeye salmon in their upstream migration. These 

 expanded objectives greatly increased the complexity 

 of the "postdam" phase of the study. 



METHODS AND MATERIALS 



The experimental procedure was as follows; (1) 

 determine, both before and after the completion of 

 Rocky Reach Dam, the time required for tagged 

 sockeye salmon to migrate from the Rock Island 



forebay to Zosel Dam; (2) determine the time re- 

 quired for tagged sockeye salmon to migrate from 

 the Rock Island forebay to Rocky Reach Dam; 

 (3) examine the variability in passage time in 

 relation to Rocky Reach Dam and the flows and 

 temperatures of the Okanogan and Columbia Rivers. 

 Tagging experiments provided the answers to 

 items (1) and (2). For item (3), these same tagging 

 data were supplemented by counts of sockeye salmon 

 made at Rock Island and Zosel Dams in years when 

 there was no tagging. 



SOCKEYE SALMON AS THE STUDY SPECIES 



Observations were confined to sockeye, the only 

 salmon that can be intercepted in significant num- 

 bers above Rocky Reach Dam while still actively 

 migrating to the spawning grounds. 



Most sockeye salmon that pass Rock Island Dam 

 are bound for spawning areas in the Wenatchee and 

 Okanogan River systems. Those that pass Rocky 

 Reach Dam are, on the other hand, mostly Okanogan- 

 bound fish, because the Wenatchee population leaves 

 the main Columbia River below Rocky Reach Dam. 

 After sockeye salmon pass Rocky Reach Dam, they 

 move to the mouth of the Okanogan River near 

 Brewster, Wash., and continue up the Okanogan 

 into Lake Osoyoos, where they remain until they 

 migrate to the spawning grounds, 10 to 15 miles 

 above the lake, in late September. We have as- 

 sumed that the Okanogan and Wenatchee popula- 

 tions pass Rock Island Dam simultaneously. The 

 close similarity in the shapes of the graphs of sockeye 

 salmon counts for Rock Island and Rocky Reach 

 Dams (examples of which are shown in figure 10, in 

 conjunction with other data) suggests that this 

 assumption is reasonable. 



TAGGING 



Sockeye salmon were tagged at Rock Island Dam 

 in 1953-57, before Rocky Reach Dam was con- 

 structed, and in 1962 and 1963, after construction. 

 The numbers of tagged sockeye salmon that were 

 later observed at Zosel Dam in 1953, 1954, 1962, and 

 1963 are presented in table 1. The effort to observe 

 tagged sockeye salmon at Zosel Dam was so variable 

 and so ineffective in 1955-57 (for reasons given 

 later) that data for these years are not included. 



The tagging procedure was the same each year. 

 Tagging was started when the daily count reached 

 about one thousand fish and continued until it 

 dropped to about one thousand near the end of the 



MIGRATION OF SOCKEYE SALMON 



133 



