Pacific salmon found in local waters. The 

 Chinook salmon used in the experiment 

 averaged less than 2 inches in length. These 

 fish were chinook stock from the Dungeness 

 Hatchery, Washington State Department of 

 Fisheries. The eggs were all hatched under 

 Brett's supervision and then acclimated to 

 various water temperatures, including 41° F. 

 and 50° F. " Though these fish were then sub- 

 jected to water temperatures either higher or 

 lower than temperatures at the normal time of 

 release, a 50 percent mortality occurred with- 

 in 12 hours when fish acclimated to 50° F. 

 were placed in water of 76° F. Fish acclimated 

 at 41° F. and placed in water of 72.5° F. showed 

 a 50 percent mortality within 4 hours. Brett 

 also subjected chinook salmon acclimated at 

 50° to 32° F. water. Fifty percent mortality 

 was reached more than 18 hours after placing 

 in the 32° F. water. The overall mortality is 

 not known as Brett was concerned only with 

 the length of time necessary to kill half of the 

 fish. Brett (1952, p. 283-284) found that size 

 influenced the mortality of fish subjected to a 

 sudden decrease in temperature. The dead 

 fish were smaller than the survivors, and the 

 difference between the two groups was statisti- 

 cally significant. No size influence could be 

 demonstrated in the mortality occurring with 

 an upward change in temperature. 



The temperature regime of the Columbia 

 River may range from a low of 32° F. in late 

 January or early February, to a high of over 

 70° F. in late summer. The Columbia River 

 temperature at Bonneville Dam (U.S. Army 

 Corps of Engineers, 1948-58) relative to the 

 constant 46° F. of Spring Creek Hatchery 

 water is shown for the period of January to 

 July, 1948-58 (fig. 9). These temperatures are 

 grouped into three arbitrary categories: 

 "ideal" (46° t 3° F.), "safe" (40° to 42° F., 

 50° to 52° F.) and "critical" (less than 40° F., 

 more than 52° F.). It should be noted that in 

 the li years included in the figure, the period 

 from mid-March to mid- April has always been 

 "safe" or "ideal". In most years, the first 

 week of February (time of fry release) has 

 been in the "critical" range. In the first week 

 of May, frequently the modal timeof fingerling 



" All temperatures have been converted from Centi- 

 grade to Fahrenheit degrees. 



Figure 9. — Daily water temperatures(°F.) of Columbia 

 River as measured at Bonneville Dam, January - 

 July, 1948-58, classified as "ideal", "safe", and 

 "critical" for chinook releases at Spring Creek 

 Hatchery. Period from mid- March to mid- April 

 always "safe" in these years. 



releases, temperatures are often "critically" 

 high. Actual water temperatures at the timeof 

 fry and fingerling. release into the Columbia 

 River from Spring Creek Hatchery may be 

 found in appendix table 11, which shows the 

 extreme temperature differences often existing 

 between Spring Creek Hatchery and the Colum- 

 bia River when releases are made. For 

 example, in the 1955 brood year a large fry 

 release was made into the Columbia River 

 when its temperature was 34° F. Later in the 

 same year a substantial number of fingerlings 

 were released into 56° F. water. 



The findings of Varma and Brett suggest 

 that the temperature differential between 

 hatchery water and river water may produce 

 heavy mortality following release into the 

 Columbia. This differential may be of par- 

 ticular concern with regard to releases on 

 recently hatched, unfed fry. These fish are not 

 completely developed, and the sudden shock to 

 their system may be very serious. A short 

 tempering to the necessary temperature 

 probably will meet with little success. This 

 has been indicated by Brett (1941) on brook 

 trout (Salvelinus fontinalis). No benefit could 



16 



