slight edema over the brain. In these cases 

 we do not know the cause of death and little 

 of the nature of the injury. All of these fish 

 were sexually immature and death was pre- 

 sumably not related to the reproductive 

 cycle." 



Table 8. — Observations at time of recovery of 8 moribund 

 and recently dead fish found within 5 miles of Bonneville 

 Dam, 1951-55 



Date found 



Species 



Condition at time of recovery 



I95i: 

 September 9 Cliinook salmon. 



September 11 White sturgeon. 



1956: 

 May 3 do 



May 12 do 



Do Carp. 



Do... do... 



May 18 White sturgeon.. 



June 7 Chinook salmon. 



June 10 Chinook salmon ' 



June 17 American shad- 



June 21 White sturgeon.. 



July 9 .Sockeye salmon. 



July 12 Sockeye salmon ' 



Do do 



Do do 



July 14 -.do 



July 15 do 



July 21 Chinook salmon ^ 



.Fresh; jaw reflexes: right eye 

 turned backward into socket; 

 gills bleeding; slight bloody 

 l)ruises on pectoral flils, isthmus, 

 right opercle, and caudal pedun- 

 cle. 



.Fresh; fin reflexes; abrasions on 

 pectoral girdle and side. 



.Alive; deep gash near left ventral 

 fin. 



.Alive; peduncle severed. 



.Alive, deep dorsal gash. 



.Fresh; decapitated. 



.Severed at midl)ody. 



.Fresh; torn premaxillary, isthmus, 

 and first gill arch; clotted blood 

 in heart cavity. 



.Moribund; abrasions on head and 

 dorsal fin; blood clot between 

 first and second left gill arches. 



.Fresh; head severed; caudal pe- 

 duncle severed. 



.Fresh; body severed. 



.Fresh; abrasion on left opercle. 



.Fungused gashes on right side. 



.Fresh; no apparent injuries. 



..Do. 



..Fresh; massive wound into body 

 cavity. 



.Alive; floating. 



.Large gash on peduncle. 



' Autopsy performed. 



Since the conclusion of our study of mortality, 

 other investigators have discovered that super- 

 saturation of nitrogen in Columbia River water 

 at times of high flow may be a significant cause of 

 fish mortalities. Some of the physiological symp- 

 toms described by Wood when he autopsied the 

 fresh dead salmon are characteristic of nitrogen 

 poisoning (such as internal hemorrhaging and con- 

 gestion of spleen), which could well have been a 

 direct cause of mortality that we did not recognize. 



Ebel (1969) found that nitrogen saturation 

 levels potentially dangerous to fish always oc- 

 curred at Bonneville Dam when water was dis- 

 charged through the spillway. Nitrogen super- 

 satui'ation may occur when atmospheric nitrogen 

 is entrained and dissolved in the plunging turbu- 

 lent spillway flow. Therefore, highest nitrogen 

 saturation values coincide with peak spillway 

 flows. Nitrogen levels in tail waters below Bonne- 

 ville Dam were substantially higher than levels 

 below seven other Columbia River dams. High 

 nitrogen levels, coinciding with high flows and a 



delay of chinook salmon below Bonneville Dam 

 (Schoning and Johnson, 1956) could have been 

 the immediate cause of death of at least a portion 

 of the chinook salmon included in our estimate of 

 mortality. 



An epidemic of the bacterial disease columnaris, 

 Chondrococcus columnaris, among salmon (par- 

 ticularly sockeye) was reported in the Columbia 

 River system in late July and August 1955.^ 

 Sockeye salmon in the latter part of the run seemed 

 more heavily infected than those in earlier seg- 

 ments, probably because water temperature in- 

 creased during the summer. We do not believe 

 that columnaris was an important contributing 

 factor to death of chinook salmon in June and 

 early July 1955 because our study was completed 

 before the outbreak of the disease. Because 

 columnaris epidemics are associated with rela- 

 tively warm water, mortalities reported during 

 spring cold-temperature periods in previous years 

 were probably not caused by this disease. 



In summary, although we were able to recover 

 only a small number of freshly killed fish im- 

 mediately below Bonneville Dam, there was no 

 indication that disease contributed to death in 

 1954 and 1955. Many of the fish had severe recent 

 external injuries. Some of the dead fish without 

 massive injuries probably died as a result of nitro- 

 gen poisoning, a potential cause of death not 

 recognized by us at the time of our study. 



SUMMARY 



1. In most years since the completion of Bonne- 

 ville Dam in 1938, floating dead fish — particularly 

 chinook salmon — have been observed downstream 

 from the dam. 



2. In 1954 and 1955 the Oregon Fish Commis- 

 sion, under a contract with FWS, studied salmon 

 mortality at Bonneville Dam. 



3. The main aims of the studies were to esti- 

 mate the chinook salmon mortality at Bonneville 

 Dam and to determine the causes of death. 



4. From ratios of tagged to untagged floating 

 carcasses, we estimated that 4,412 recoverable 

 carcasses of chinook salmon that had died near 

 Bonneville Dam were in the river at the time of 

 our 1955 experiment. 



5. The chinook salmon mortality at Bonneville 



«Erling J. Ordal. 1955. Progress Report No. 13, U.S. Fish and Wildlife 

 Service, Contract No. 14-19-008-2418, 7 pp. On file Bureau of Commercial 

 Fisheries Biological Laboratory, Seattle, Wash. 98102. 



488 



U.S. FISH AND WILDLIFE SERVICE 



