DAWLEY and EBEL: EFFECTS OF DISSOLVED GASES ON SALMONIDS 



(American Public Health Association et al. 1971). 

 Gas concentrations at saturation (100%) were 

 taken from Weiss (1970). 



To obtain the dissolved gas levels for the various 

 tests, we adjusted the water equilibrators of each 

 tank (screens plus air bubbling boxes) until the 

 nitrogen concentration remained within ±2% of 

 the desired value. The oxygen concentration was 

 then measured and we found that the saturation 

 value was 5 to 10% lower than that of Ng+Ar for 

 each tank. This did not differ appreciably from 

 prevailing oxygen saturations in the Columbia 

 and Snake rivers which are usually 5 to 10% lower 

 than dissolved nitrogen values (Beiningen and 

 Ebel 1971; Ebel 1971). After introducing fish, 

 however, we noted that the oxygen concentration 

 dropped further (presumably because it was con- 

 sumed), resulting in values from 8 to 28% of sat- 

 uration below that of Ng + Ar, particularly in test 

 area B and the outlet area of the tank. Due to large 

 numbers of fish required for experiments on the 

 survivors of these bioassays, and the complexity of 

 changing the dissolved gas ratios of the water 

 source, we did not alter the O2 concentrations in 

 the tests but carefully documented the mid-tank 

 gas concentrations. Data affected by this drop in 

 oxygen partial pressure are discussed later in this 

 report. 



One-year-old spring chinook salmon from 

 Leavenworth National Fish Hatchery, Leaven- 

 worth, Wash., and steelhead trout from the 

 Washington Department of Game Hatchery at 

 Aberdeen, Wash., were used in the tests. Test 

 populations were acclimated to laboratory water 

 at 15°C with normal dissolved gas concentrations 

 for at least 2 wk before testing. Groups of 30 or 60 

 fish were placed simultaneously in control (100% 



atmospheric nitrogen saturation) and test tanks 

 set at 105, 110, 115, 120, and 125% of Ng + Ar sat- 

 uration and one to four replicates of each test were 

 made, depending on test level. When 60 fish 

 were being tested, 30 were in each of the two test 

 sections A and B. Fish were randomized before 

 introduction into individual test tanks. Mean sizes 

 of the fish at completion of the tests are indicated 

 in Table 1. Measurement of size at the beginning 

 of the tests was omitted to avoid placing addi- 

 tional stress on the test animals. Feeding of fish 

 during the test period began 48 h after introduc- 

 tion to test tanks; thereafter they were fed to 

 satiation once each weekday. 



Lethal exposure times to 10 and 50% mortality 

 (LEjo and LE50) were averaged for lots of test fish 

 Held in tank sections A and B during the same time 

 period, and the mid-tank gas concentrations were 

 used for analysis with the exception of the 

 steelhead groups stressed at 115% nitrogen; in 

 these tests, exposure times and gas concentrations 

 were measured separately for A and B sections of 

 the tanks. In addition, lethal exposure times to 

 100% mortality (LEjoq) for chinook and steelhead 

 at all levels of supersaturation were taken only 

 from groups held in the A section of the tanks. 



Observations of behavior, progression of exter- 

 nal signs of gas bubble disease, and mortality were 

 recorded continuously for the first 6 h then every V2 

 h for 24 h and every 3, 6, or 12 h thereaf- 

 ter-depending on test concentration-until ter- 

 mination of the bioassay at 35 days. Observations 

 of change in degree of external disease signs 

 among test fish after a recovery period in normally 

 saturated (100%) water also were made from 

 selected groups. 



Sublethal effects of supersaturation were as- 



Table l.-Comparison of mean weigrhts and lengths of surviving test and control fish held in 15°C water w^ith 

 N2 +Ar levels at 100 to 125% of saturation, February-April 1972. 



'Replicates of tests at 115-125% levels were made at various time intervals throughout the indicated test 

 period; others lasted the full indicated period. 



789 



