of recoverable carcasses, we can, if we assume 

 that all carcasses released in the powerhouse 

 channel were recoverable, make a crude estimate 

 of the fraction of recoverable carcasses from the 

 spillway releases. 



If none of the carcasses released at the spillway 

 had been rendered unrecoverable, the proportion 

 of recoveries from the spillway releases should 

 have equaled the proportion of recoveries from 

 the powerhouse releases. On this assumption, the 

 expected number of recoveries from spillway re- 

 (635) (6) 



leases is 



130 



= 29.3. Because only two car- 



casses were actually recovered, we can estimate 

 that the proportion of recoverable carcasses for 

 fish dying by being swept over the spillway is only 



2 

 — — = 0.0683 and that the proportion of unre- 



coverable carcasses is 0.9317. Because this esti- 

 mate is based on small numbers of recoveries, 

 however, sampling error could be large. 



1955 EXPERIMENT 



In 1955, 1,169 carcasses were released. They 

 were all male fall chinook salmon collected in 

 1954 from Bonneville, OixBow, and Spring Creek 

 Hatcheries. They were killed by a blow on the 

 head and placed in frozen storage a few hours 

 later; as in 1953, the carcasses were ice glazed to 

 retard dehydration and oxidation during over- 

 winter storage. 



When the carcasses were removed from storage 

 for the 1955 experiment, they were handled in 

 the same manner as the carcasses for the 1954 

 experiment except that they were tagged with 

 sequentially numbered cellulose-acetate Petersen 

 disks fastened by nickel pins through the caudal 

 peduncle. 



• Figure 6 demonstrates the rationale on which 

 our entire 1955 study was based: the number of 

 chinook salmon floaters is an indication of the 

 total number of dead chinook salmon in the river. 

 Few floaters were recovered until late June, when 



-1 — I — r^ 

 5 15 25 



APRIL 



I I I — T^ — ^r — I I I I I I I I r 

 5 15 25 5 15 25 5 15 25 



MAY JUNE JULY 



Figure 6. — Daily counts of chinook salmon and total riverflow at Bonneville Dam and number of untagged 

 chinook salmon floaters recovered per hour, Bonneville Dam to St. Helens, 1955. 



CHINOOK SALMON MORTALITY IN COLUMBIA RIVER NEAR BONNEVILLE DAM 



473 



