FISHERY BULLETIN: VOL. 73, NO. 4 



DISCUSSION 



Data from these tests indicate that the critical 

 level of supersaturation of nitrogen where 

 juvenile spring chinook and steelhead began to 

 show mortality was about 115% N2+Ar when Og 

 saturations were about 95% (111% TDG). These 

 data agree closely with the findings of Shirahata 

 (1966), who indicated that the critical level for 2- 

 mo-old rainbow trout was about 111.3%, N2+Ar 

 and 99.7% 02(109% TDG). 



Although mortality from supersaturation did 

 not occur until fish were exposed beyond 110% 

 (± 2%) N2 + Ar, swimming performance 

 measurements with juvenile chinook showed some 

 effect from stress caused by exposure to supersat- 

 uration at levels as low as 110% N2 + Ar (106% 

 TDG). We believe that one can infer from the 

 results of these tests, that something less than 

 normal survival will result when juvenile chinook 

 and steelhead are exposed for 35 days or longer at 

 or above 110% N2+ Ar (106% TDG). 



Results of our testing program indicate that 

 oxygen as well as nitrogen is responsible for caus- 

 ing gas bubble disease, even when O2 concentra- 

 tions are below saturation. The immediate 

 conclusion drawn from this observation would be 

 that total dissolved gas is the cause rather than 

 any one or combination of component atmospheric 

 gases. However, fish tolerance research by Egusa 

 (1969) and by Rucker (1975) with various ratios of 

 dissolved gas indicate that mortality from gas 

 bubble disease is not necessarily in linear correla- 

 tion with TDG. Egusa showed that oxygen sat- 

 uration values of 400 to 500% were required to 

 produce initial mortality of goldfish, Carassius 

 auratus, and an eel Anguilla japonica when ni- 

 trogen concentrations were near 100% (TDG 160- 

 180%). In earlier work with the same two species, 

 however, Egusa (1959) recorded high mortality of 

 goldfish with N2+Ar at 132% and O2 at 75% of 

 saturation (TDG 123%), and of eel with N2+ Ar at 

 124%, O2 at 66% (TDG 112). Rucker found that 

 mortality rate of juvenile salmon declines con- 

 siderably if the ratio of oxygen to nitrogen is 

 increased even though the same TDG pressure is 

 maintained. 



It is apparent from our tests and those of Egusa 

 and Rucker that the ratio of O2 and N2 must be 

 considered as well as TDG when assessing possible 

 effects from supersaturation. 



Additional information is needed to quantify 

 the effects of various gas ratios (nitrogen to 



oxygen) on tolerance limits of fish in general. It is 

 probable that most fish could tolerate higher total 

 gas pressure if the major portion of the excess gas 

 were oxygen. 



Dissolved gas measurements and resulting per- 

 centage saturations for the Columbia and Snake 

 rivers (Ebel 1969, 1971; Beiningen and Ebel 1971) 

 have been based on surface or atmospheric pres- 

 sure plus vapor pressure. Corrections for the 

 hydrostatic pressure (or depth) at which a sample 

 was taken were not made. Thus, the calculations of 

 percentage saturation were made as though the 

 samples were collected at the surface. This is con- 

 venient when limnologists or oceanographers wish 

 to compare values taken at various depths, but 

 leads to confusion when attempting to assess how 

 a given saturation measurement will affect a fish 

 at depth. 



The depth that populations of fish travel must be 

 considered when one attempts to determine the 

 effects of an exposure to supersaturated levels of 

 dissolved gases. Bubble formation in the circula- 

 tory system or tissues of fish is directly dependent 

 on the external hydrostatic pressure. For example, 

 a fish traveling at a depth of only 1 m will be 

 provided with enough hydrostatic pressure to 

 compensate for a gas pressure in excess of 10% 

 (110% saturation at surface pressures). A fish 

 traveling at 3 m can compensate for 30%, or 130% 

 saturation at surface pressures; a fish traveling at 

 10 m can compensate for an excess of 100% of 

 saturation and so on. These tests were conducted 

 in shallow tanks at essentially zero hydrostatic 

 pressure with only a few centimeters depth com- 

 pensation possible. The lethal exposure times we 

 measured could only be applied directly to fish 

 populations that could not compensate by sound- 

 ing. Much more information is needed to deter- 

 mine how a given gas level in a river affects the 

 population inhabiting the river. Information 

 regarding the behavior of fish is obviously essen- 

 tial. We believe, however, that data from our tests 

 support the 110% maximum allowable limit es- 

 tablished by the Environmental Protection 

 Agency primarily because significant mortalities 

 did not occur until concentrations exceeded 110% 

 TDG. 



Gas bubble disease signs either singly or in 

 combination with one another did not correlate 

 well with mortality. Those generated from stress 

 conditions of 120% saturation and higher seemed 

 to be nearly the same at LEjq as at LEjoo (gas 

 blisters in the fins and lateral lines of most live and 



794 



