C. (fig. 10). Thus, dead salmon are least likely 

 to be evident as floaters in the spring and most 

 likely to be evident in the fall. Figure 10 shows 

 that in the spring of 1955 average water tempera- 

 tures were the lowest for the 10 years shown 

 (8.9° C). Unusually cold water probably con- 

 tributed to the almost complete absence of floaters 

 during the spring run in 1955. In 1952, a period of 

 apparently high mortality (table 7, fig. 10), water 

 temperatures during the spring chinook salmon 

 migration were highest of the 10-year period. 



TURBIDITY 



Turbid water, as measured by Secchi disk 

 visibility, was investigated as a possible factor 

 contributing to chinook salmon deaths. We 

 hypothesized that fish rely partially on sight to 

 locate and negotiate the fish ladders, and reduc- 

 tion of visibility might handicap them. 



High flows in the Columbia River are charac- 

 terized by turbid water, and low flows by rela- 

 tively clear water. During low flows the water may 

 be turbid for short periods when a tributary floods 

 from heavy rains or rapid snowmelt. For example, 

 in mid-January 1953 after a flash flood on a 

 tributary, a Secchi disk visibility reading of 0.06 

 m. was recorded at Bonneville Dam when flow 

 was only 2,700 c.m.s. 



Secchi disk visibility at the dam has seldom 

 been less than 0.3 m. during periods of major 

 salmon migrations (fig. 11). From 1950 to 1955, 

 visibility was 0.3 m. or less for only short, infre- 

 quent periods from April to September, except in 

 1952, when it remained about 0.3 m. or less from 

 April 1 through May 30, throughout the spring 

 migration. The 1952 spring run suffered a heavy 

 mortality, but because the high turbidity was 

 accompanied by high flows and relatively high 

 water temperatures, it was impossible to evaluate 

 the separate effects of flow, temperature, and 

 turbidity. 



In 1955, from April 10 to June 30, Secchi disk 

 visibility varied between 0.3 and 0.8 m.; it was 

 0.6 m. at the peak of the spring chinook salmon 

 run (May 2) and 0.45 m. at the peak of the sum- 

 mer run (June 23). Few floating chinook salmon 

 were observed in May, but large numbers were 

 seen in June and July. Because turbidity differed 

 little between the two periods, we concluded that 

 it was not a major factor in mortality at the dam 

 in 1955. 



Because high turbidity and high flow usually 

 coincide, we could not evaluate the separate effect 

 of turbidity, if any. 



COMMERCIAL FISHING 



The Columbia River gill net fishery has some- 

 times been blamed for dead salmon in the river 

 because some fish escape from nets after becoming 

 enmeshed. An escaped fish usually has charac- 

 teristic net marks — encircling bands where scales 

 have been scraped off and cuts on the anterior 

 edges of fins. Hanson et al. (1950: 24) concluded 

 from observations in the fish ladders at Bonne- 

 ville Dam and at hatcheries above and below the 

 dam. that "Most injuries to the fish observed are 

 traceables (sic) to fishermen's gill nets; none of 

 the injuries were directly traceable to conditions 

 at Bonneville Dam. Most net injuries were not 

 fatal to fall-run chinook salmon at hatcheries 

 above and below Bonneville Dam in 1946." The 

 conclusion is questionable because only the fish 

 that survived after being injured were available 

 for observation — those that may have died could 

 not, of course, be sampled at the hatcheries. 



Most significant in discounting gill nets as a 

 major cause of mortality in our study is the fact 

 that 85,769 spring chinook salmon passed the 

 dam in May 1955 with little apparent mortality, 

 despite an intensive commercial gill net fishery 

 from April 30 to May 27. The spring chinook 

 salmon run during this period was the largest 

 since 1939 (the first year runs were counted at 

 Bonneville Dam). The catch was 80 percent of the 

 total run; the catch below the dam was the second 

 largest since 1939. Four stations below the dam 

 and one station near The Dalles were searched 

 intensively for floaters throughout this period, 

 but few were found and none of these bore charac- 

 teristic net marks. This is strong evidence against 

 attributing the death of floating chinook salmon 

 to the gill net fishery. 



We concluded that gill net injuries are not a 

 major cause of death of chinook salmon found 

 floating near the dam. 



FISH PASSAGE FACILITIES AT BONNEVILLE DAM 



Great effort has been made by fishery agencies 

 and especially by the U.S. Army Corps of Engi- 

 neers to discover any structure or operation at 

 the dam that might delay, injure, or kill migrat- 

 ing adult fish. Mechanical failures or routine 



CHINOOK SALMON MORTALITY IN COLUMBIA RIVER NEAR BONNEVILLE DAM 



485 



