temperature' did affect the efficiency with which the dry meals were 

 utilized by the fish and, also, that the mortality was responsible 

 for a depression in the total gain. This latter effect is par- 

 ticularly apparent in the case of the Cortland diet where the gain 

 in total weight was comparable to the beef liver control but the 

 average weight of the fish closely paralleled that" of those fed the 

 meat and viscera mixture. In this instance, however, the retard- 

 ation effect of the cold water period is clearly demonstrated by 

 the fact that the average weight of the fish on the Cortland diet 

 did not approach that of those fed the meat- viscera-meal mixture. 

 In the 1 Q U7 trials, at warm-water temperatures these two diets were 

 comparable. 



Unlike the mortalities which occurred in the high-viscera diets 

 during the cold-water period, no symptoms of a vitamin deficiency 

 could be identified upon examination of moribund fish from Diets 2 

 or I4.0 The factor responsible for these high losses has not been 

 indentified. 



The exact wd.ter temperature at which dry meals make a measurable 

 contribution to the growth rate of blueback salmon cannot be determined 

 accurately from either Figure 2 or 3, A comparsion of the curves in 

 the two graphs shows that the point of divergence of the meat-viscera- 

 meal mixture from the meat and viscera control differs. In Figure 2, 

 the total weight of the fish on the meal diet (Diet 2) is depressed 

 by the mortality which continued after a growth response is dis- 

 cernable in the average weight of the fish as shown in Figure 3» It 

 is evident, however, that dry meals are not utilized effectively and 

 appear to be detrimental to blueback salmon when fed at water tem- 

 peratures approximating U5 degrees F„ 



Potential Production Diets 



The ultimate aim of feeding experiments is the development of 

 practical diets which may be used in fish-cultural operations. Several 

 potential diets have been developed from the leavenworth feeding trials 

 some of which have been tested in actual production operations. 



As a starting diet for first-feeding fingerling, the beef liver- 

 hog liver combination (Table 2, Diet 3) has proved superior to either 

 product fed separately or in combination with hog spleen, 'This diet 

 has the additional advantage of adaptability in that the feeding con- 

 sistency may be altered to a form compatible with the size of the 

 fish with a minimum of effort. Varying consistencies may be pro- 

 duced by the degree to which the diet is beaten just prior to being 

 fed. The beef liver-hog liver ration has been used on the three 

 hatcheries of the Grand Coulee Project since 19 U5 as a starting diet 

 for blueback, chinook, and silver salmon with excellent results. 



29 



