It is also known that decompression to levels 

 present in the turbines can be harmful to juvenile 

 Pacific salmon. In a study of the effects of neg- 

 ative pressure on fingerling chinook salmon, 

 Holmes and Donaldson' found that test fish suf- 

 fered an average of 3.5% mortality when de- 

 compressed from 18 psi to —2 psi. Muir (1959) 

 reported that fingerling coho salmon exposed to 

 negative pressures of 12.7-73.7 cm/Hg for 0.1- 

 2.3 sec appeared to be stunned momentarily but 

 concluded that decompression to partial vacuum 

 was not likely to kill the fish unless accompanied 

 by cavitation. Harvey (1963) concluded from 

 his studies and comprehensive review of the lit- 

 erature that the effects of decompression de- 

 pended on the condition of the fish and physical 

 factors in the environment as well as on the mag- 

 nitude of the vacuum. 



Immobilized fingerling salmon and trout were 

 first observed at Ice Harbor Dam in 1966. 

 Marked juvenile coho salmon and rainbow trout 

 were released in a turbine intake and recovered 

 in a funnel net attached to the draft tube exit. 

 Some of the fish recovered swam normally; 

 others were dead or in distress. Either the 

 turbine or net could conceivably have produced 

 the effects observed. Therefore, laboratory ex- 

 periments were conducted to isolate and examine 

 the effects of pressure change alone. The present 

 paper reports the results of this research and 

 subsequent observations of immobilized fish at 

 Ice Harbor Dam. 



METHODS AND MATERIALS 



Laboratory Study 



In this study chinook salmon fingerlings were 

 substituted for rainbow trout as the latter spe- 

 cies was not readily available. The chinook 

 salmon fingerlings were downstream migrants 

 caught in the gatewells of McNary Dam, Uma- 

 tilla, Oreg. The coho salmon were nonmigrants 

 taken from the White Salmon Federal Hatchery, 

 White Salmon, Wash. All fish were transported 

 to the test site by tank truck and held in a fish- 

 holding facility (described below) for at least 

 2 days before they were used in decompression 



' Holmes, H. B., and I. J. Donaldson. A study of the 

 effect of pressure changes upon salmon fingerlings as 

 applied to passage through spillway at Mayfield Dam, 

 Cowlitz River, Wash. (Unpubl. manuscr., 18 p.) 



tests. Fork lengths of the two lots of fish were 

 approximately the same; the chinook salmon 

 ranged from 119 to 180 mm and the coho salmon 

 from 135 to 190 mm. The slight size difference 

 is believed to have had little or no influence on 

 the comparative response of the two species to 

 negative pressure. 



The experimental apparatus consisted of two 

 main components: (1) the fish-holding-recir- 

 culating system and (2) the pressure chamber- 

 vacuum system. The basic arrangement (except 

 for the vacuum pump and filter) is illustrated 

 in Figure 1. A round wooden tank served as 

 the holding facility. Water from the water main 

 of Pasco, Wash., treated with sodium thiosulfate 

 to neutralize the chlorinity, was continuously re- 

 circulated through a sand and charcoal filter and 

 aerated by a diaphragm compressor. The water 

 temperature was maintained at 9.0° C ± 0.5° C 

 by a portable refrigeration system. 



The pressure chamber was built from a 

 129.4-cm length of heavy-duty iron pipe, 30.5 cm 

 in diameter. End plates of clear plastic, 2.5 cm 

 thick, permitted observation of the test fish. 

 Access to the chamber was gained through a 

 port made from a 10.2-cm pipe sleeve fitted with 

 a flange and cover plate. Additional sleeves pro- 

 vided plumbing attachments for water intake 

 and discharge lines, air exhaust port, pressure 

 gauge, and vacuum line. A 3,785-liter (1,000- 

 gallon) tank served as a vacuum reservoir and 

 was connected to the pressure chamber by a rub- 

 ber hose 1.9 cm thick. The vacuum was applied 

 by activating a hand-operated, 2.5-cm ball valve 

 on the chamber end of the hose. 



The amount of negative pressure applied to 

 the chamber was controlled by evacuating the 

 reservoir to the desired level as indicated on a 

 gauge. A second gauge plumbed into the cham- 

 ber monitored the pressure change. The vac- 

 uum was relieved by closing the vacuum line 

 valve and venting the chamber by opening a 

 second .6-cm ball valve. 



About 1 sec was required to obtain 61 cm of 

 mercury vacuum in the chamber when it was 

 filled with water; 21/2 sec were required to reach 

 71 cm of mercury vacuum because of the large 

 amount of dissolved gasses liberated by the low 

 negative pressure. At all negative pressures, 

 it took one-half second to relieve the vacuum by 

 opening the relief valve and closing the vacuum 

 line valve. 



