One such main-stem works is Rock Is- 

 land Dam located 452 miles from the mouth of 

 the Columbia River. Its 22 to 33 -foot height 

 was increased 20 feet in 1952. Yearly average 

 counts of 12,000 chinooks were made through its 

 fishways between 1934 to 1954. Further up- 

 stream from Rock Island Dam is Grand Coulee 

 Dam which is 595 miles from the sea. It became 

 a total block to salmon in 1939, but Rock Island 

 counts averaging 8,500 chinooks for 6 years 

 previous to its construction indicated that the 

 area above Grand Coulee was no longer recipient 

 of large runs. Further downstream is McNary 

 Dam located 292 miles from the mouth of the 

 Columbia River. Fish ladders were placed in 

 operation at this dam in 1951 and power turbines 

 were installed somewhat later so that returns 

 from downstream migrants would not influence 

 the period of comparisons in this report. Bonne- 

 ville Dam, constructed in 1938, is a lower river 

 dam which is located just above the boundary of 

 tidal influence . Approximately 70 percent of the 

 Columbia River chinooks must pass the fish- 

 passage facilities placed there for them and the 

 young must return the same route without spe- 

 cial protection. 



inantly 4 and 5 years . In this regard, Mitchell 

 G. Hanavan's Grand Coulee marking experiments 

 reveal 62 percent four -year olds and 36 percent 

 five-year olds caught in May of 151 marks col- 

 lected in the fishery. June and July recoveries 

 from the same experiments were 20 percent 3's, 

 42 percent ^s, 33 percent 5's, and 5 percent 6's 

 from 1,153 marks collected in the fishery. The 

 same type of experiment conducted by Harlan B. 

 Holmes showed about two -thirds of the fall re- 

 turns were 4 -year olds from a collection of 

 2,677 marks. From these marking experiments 

 a predominant 4 or 5 -year cycle was determined 

 for spring, summer, and fall from hatchery raised 

 stock. Spring season ratio of return to escape- 

 ment for the period before Bonneville Dam was 

 2.3:1 compared to 2.6:1 after the construction of 

 Bonneville Dam. In the summer a return -escape- 

 ment ratio of 4.4:1 existed before Bonneville 

 construction and a ratio of 2.7:1 after. The fall 

 return-escapement ratio before Bonneville Dam 

 was 4.3:1 and 3.2:1 after. Combining all seasons 

 the ratio of return to escapement was 3.8:1 before 

 and 3.0:1 after construction of Bonneville Dam. 

 These ratios were derived from data summarized 

 in table 12. 



Indication of the effect of dams on sur- 

 vival of downstream migrants may be gained by 

 comparing size of returns before and after the 

 existence of Bonneville Dam . Since the effect 

 on the smolts would be reflected in the adult 

 migration after three years, 1941 was chosen as 

 the division between returns instead of 1938. 

 The estimate of spring abundance for the period 

 between 1928 and 1941 averaged 3.5 million 

 pounds compared with 4 million pounds between 

 1942 and 1954. The summer average for the 

 period 1928 to 1941 was 5 million compared with 

 1.5 million pounds from 1942 to 1954. The fall 

 return, 1928 to 1941, averaged 14 million pounds 

 while the 1942 to 1954 return was also 14 million 

 pounds. On an annual basis, returns averaged 

 22.5 million pounds before 1941 and averaged 

 19.5 million pounds after 1941. These data are 

 presented in table 10. 



Perhaps more conclusive information can 

 be gained by comparing the returns with escape- 

 ments, table 11. A lapse of four years is used 

 between return and escapement on the basis of 

 marking experiments which reveal a variety of 

 maturing ages of chinooks but which are predom- 



DISCUSSION AND CONCLUSIONS 



In an evaluation of the effect of water -use 

 construction on chinook salmon in the Columbia 

 River, Bonneville Dam was selected because of 

 its strategic location in the main channel of the 

 river and because it provides a greater number 

 of years of successive fish-passage data for this 

 study. In such an evaluation, it is pertinent to 

 consider the ratios of return to escapement since 

 these ratios are an indirect measure of mortality 

 of young salmon passing over the dam. Examina- 

 tion of the data (table 12) shows that pre -Bonneville 

 survival was better than post -Bonneville survival 

 for the sum of the seasonal runs. 



The before and after disparity was great- 

 est for the summer runs, in which all fish spawn 

 above Bonneville Dam, thus making their young 

 vulnerable to injury in passing over the dam. The 

 fall run, with 80 percent of the fish spawning 

 above the dam, was intermediate in degree of 

 change in survival rate . For the spring run, with 

 only 67 percent of the fish spawning above the dam, 

 there was an actual improvement in the survival 

 rate. There is some indication of post -dam 



18 



