signs, thus did not exist on mortalities from the 

 higher test levels because of shorter time duration. 

 Therefore, incidence rate was lower but is entirely 

 dependent on the time under stress. Exophthalmia 

 appeared frequently in the 3rd and 4th months, 

 similar in incidence to blisters on the head. Blisters 

 at the mid-line of the vertical surface occurred fre- 

 quently on chinook mortalities taken from the deep 

 and shallow tanks at the highest levels. This fre- 

 quency decreased, however, as testing progressed 

 and by the 4th month there was no evidence of this 

 sign; it appeared to be related to recent yolk absorp- 

 tion. Blisters between the fin rays occurred at a very 

 low incidence (40%) at the highest test levels com- 

 pared to what had been previously observed by other 

 investigations in other tests with larger salmonids 

 of other races and species. 



Comparison of condition factors between live 

 and dead fish seems to indicate that the larger fish 

 were more susceptible to gas bubble disease. This 

 agrees with earlier research by Shirahata (1966) and 

 Meekin and Turner (1974). The species tested by 

 these researchers (rainbow trout, salmon, and steel- 

 head, respectively), became less tolerant with age 

 and growth, starting as button up fry. 



Although a portion of the test groups in the 

 deep tanks remained at sufficient depth to increase 

 their resistance time, the depth did not provide 

 sufficient compensation to prevent mortality, par- 

 ticularly at levels above 120% 



The mortality rates and G.B.D. signs of both 

 species also indicated that less hydrostatic compen- 

 sation was derived due to depth disposition than 

 expected when the mean depth of the fish groups 

 is considered. Thus, individual fish must move sub- 

 stantially from the observed mean depth of the test 

 lot. If this did not occur, the effect of hydrostatic 

 compensation would have resulted in a calculated 

 reduction in effective supersaturation of 12-16% for 

 chinook and 17-20% for steelhead. Since the actual 

 mortality rates indicated that only a 10% and 10-15% 

 (chinook and steelhead, respectively) reduction 

 occurred, we can assume that the fish were moving 

 randomly about within the tank. 



Application to the River Environment 



Certain observations made during these bio- 

 assays have important implications relative to the 

 experience of naturally migrating populations of 

 juvenile salmonids. The Columbia River system is 

 of major concern in our research efforts, thus the 

 following discussion is centered on the implica- 

 tions to fish in the Columbia. 



Most areas where salmon and steelhead 

 incubate and develop in the Columbia River system 

 are located above dams and, therefore, would be 

 little affected by supersaturation. However, spring 



chinook and steelhead on tributaries of the Willam- 

 ette River, a major tributary of the Columbia, 

 make heavy use of areas below dams. Also, hatch- 

 ery water sources in some instances are either 

 below dams that may produce supersaturation 

 during the rearing period or are taken from wells 

 yielding water with high dissolved gas content. In 

 these instances the early stages of life such as the 

 period of incubation become quite important. 



Data from our bioassays and others — Shirahata 

 (1969) and Meekin and Turner (1974) — indicate that: 

 1) yolk sac fry sustain injuries at low levels of dis- 

 solved gas which become fatal as the yolk is nearly 

 absorbed; and 2) that after fry have "buttoned up" 

 tolerance to supersaturation becomes quite high 

 but decreases gradually thereafter until time of 

 seaward migration. Nebeker (1973) indicated that 

 tolerance of adult salmonids to supersaturation is 

 slightly less than that of juvenile migrants. Sig- 

 nificant changes in tolerance at various life 

 stages obviously occur and the effect varies de- 

 pending on the life stage. Equilibration of hatchery 

 water sources is thus extremely important at cer- 

 tain stages of fish development. Also, spillway 

 discharges at certain times will have more effect 

 on survival of downstream juvenile migrants than 

 at others, thus management policies should consider 

 the changing effects of these discharges. 



The spring freshet on the lower Columbia and 

 Snake Rivers coincides with juvenile salmonid 

 outmigrations as well as some adult upstream 

 migrations. Freshet conditions are variable from 

 year to year, but heavy spillway discharges usually 

 occur every year for some duration creating super- 

 saturation from 120% to 140%. During years of high 

 flow these levels occur throughout long stretches 

 of the river (650 km and more), resulting in 

 long-term exposure of some stocks. Rates of juve- 

 nile migration indicate that at least 28 days is 

 required for travel from Little Goose Dam to the 

 Columbia River estuary during the highest flows. 

 Thus, even if fish are compensating for super- 

 saturation by sounding a significant portion of the 

 population is subjected to levels of dissolved gas 

 supersaturation exceeding 120%. 



Data on depth distribution of migrating juve- 

 nile fish within the Snake River near Lower Monu- 

 mental Dam (Smith 1974) indicate that 58% of the 

 chinook and 36% of the steelhead were in the upper 

 3.7 m of the water column. Mean depths for these 

 portions of the migrating stocks were 1.30 m and 

 1.33 m, respectively. This would compensate for 

 14.5-14.8% effective saturation, which means 

 that at higher levels of supersaturation (135% or 

 greater) both stocks of fish would be exposed to 

 levels of gas concentration above 120% for at least 

 28 days during periods of high flow. 



Long-Term Exposure of Salmonids 9 



