rg= FRACTION OF CARCASSES OF FISH DYING AFTER COUNTING THAT ARE RECOVERABLE 



Figure 8. — Values of M, the fraction of the chinook salmon run dying near Bonneville Dam 

 at the time of the 1955 experiments, corresponding to pairs of values for f„ and r,,. 



mum point estimate of the level of chinook salmon 

 mortality at Bonneville Dam. As f^ increases, M 

 also increases; the rate at which M increases 

 depends strongly on the value of r.,. For large 

 values of ra, corresponding to an assumption that 

 only a small percentage of the carcasses of fish 

 that die by being swept over the spillway are 

 rendered unrecoverable, values of M increase only 

 slightly as fa increases. On the other hand, if a 

 large percentage of these carcasses are rendered 

 unrecoverable, ra is small and M increases rapidly 

 as fa increases. (Note in figure 8 that points in 

 the region above and to the right of the line 

 for I& = 1.00 correspond to values of fa and r„ 

 that are incompatible with data collected in our 

 1955 experiment.) 



FACTORS ASSOCIATED WITH MORTALITY 

 AND THEIR SIGNIFICANCE 



We have established in preceding sections that 

 many chinook salmon died near Bonneville Dam 

 at the time of our 1955 experiment. The next 

 logical step is to examine factors associated with 

 these deaths to determine the actual cause or 

 causes. 



In addition to the 1955 period, we also examined 

 available information for other periods between 

 1943 and 1956 when high or low mortality was 

 apparent (table 7). The seven periods of appar- 

 ently high mortality were September 1943, Sep- 

 tember 1950. May 1952. July 1954, September 

 1954, June-July 1955, and June-July 1956. Low 

 mortality periods were September 1946 and 

 April-May 1955. 



CHINOOK SALMON MORTALITY IN COLUMBIA RIVER NEAR BONNEVILLE DAM 



481 



