EBEL. DAWLEV. and MONK: THERMAL TOLERANCE 



Table 1. — Holding time and temperature of water before transfer to acclimation tanks, 

 acclimation conditions, and mean size at time of testing of three species of salmon and 

 trout. 



a Final acclimation temperature. 



tion levels were then measured for 24 hr before 

 introduction of the test fish. 



The source of supersaturation of nitrogen gas 

 in the Columbia River was entrained air. We 

 used pumps to entrain air in our test and accli- 

 mation tanks. Oxygen, therefore, was also su- 

 persaturated; the ratio of oxygen to nitrogen 

 supersaturation was similar to that recorded in 

 the river (Beiningen and Ebel, 1971). Oxygen 

 saturation varied from about 114 to 120 '?f. 



Facilities used for tolerance tests in relation 

 to supersaturation of nitrogen at surface pres- 

 sures were designed to provide a continuous 

 supply of fresh water. Sufficient cooling and 

 heating capacity was available from a water 

 heater and chiller to continuously supply 227 

 liters/min of either heated or chilled water from 

 5° to 40° C. Ten cylindrical fiber glass acclima- 

 tion tanks, about 1 m high x 214 m in diameter, 

 provided sufficient space to maintain 2000 fish 

 at each of five acclimation temperatures without 

 crowding. The test tanks were rectangular fi- 

 ber glass tanks of 113.6-liter capacity. A flow of 

 approximately 11.5 to 30.0 liters/min was main- 

 tained in acclimation tanks and 3.5 liters/min 

 in test tanks. 



Supersaturation of nitrogen and oxygen was 

 achieved in the acclimation tanks by metering 



air into the intake of two high-pressure recircu- 

 lating pumps with 42 kg/cm- back pressure on 

 the discharge side of the pump. About 0.75 liter/ 

 min air in each pump created the desired satura- 

 tion of nitrogen (115-120Sr). Supersaturation 

 of nitrogen and oxygen in the test tanks was 

 achieved by metering 0.05 liter/min into an- 

 other high-pressure recirculating pump with 

 32 kg/cm^ back pressure, which recirculated the 

 water in the coldwater supply tank to the test 

 tanks. (Source of water was from Seattle mu- 

 nicipal supply; chlorine was eliminated by char- 

 coal filters.) The final saturation value for each 

 tank was then achieved by manipulating the 

 number of equilibrating screens through which 

 the water flowed before entering the tanks. 



The test tank (9 m deep) used for our ex- 

 periments concerning water depth has been de- 

 scribed by Pugh, Groves, and Ebel (1969). 

 When tolerance tests were conducted in the deep 

 tank, the saturation levels of nitrogen and oxy- 

 gen were conti-olled in the above manner (by 

 injecting air into a recirculating pump). Be- 

 cause of the large volume of water (66,648 

 liters) , a continuous flow of fresh water was not 

 needed and the existing water was recirculated. 

 Time limitations precluded testing more than 

 two populations (hatchery coho and wild spring 



835 



