of the fish products which were available in quantity on the Pacific 

 Coast, 



The offal from salmon canneries offers a relatively untapped 

 source of fish food. Although the hatcheries, located close to can- 

 nery operations in the United States, utilize the bulk of the local 

 product, Jones et al. (I9I48) estimate that 100,000,000 pounds of sal- 

 mon offal annually goes to waste in Alaska. 



Two products are prepared from salmon cannery waste for fish food,, 

 The entire offal consisting of the heads, collars, tails, fins and 

 viscera is one preparation, and the salmon viscera, alone, which con- 

 tains the gastro-intestinal tract, the gonads, and a portion of the 

 liver is the other. 



Salmon Viscera 



Experiments conducted by Donaldson and Foster (1939) and Phillips 

 and ^ewitt (19^5) indicated that high levels of salmon viscera in the 

 diet were capable of maintaining chinook salmon Cingerling for con- 

 siderable periods. Phillips and Hewitt were of the opinion that beef 

 liver was a necessary supplement to salmon viscera if anemia was to 

 be prevented. Donaldson and Foster found that they could rear, finger- 

 ling for a 20-week period on a diet composed of 100 per cent salmon 

 viscera. 



Salmon viscera proved to be one of the most promising products 

 tested in the Leavenworth fpeding trials. To overcome the feeding 

 difficulties encountered by Phillips and Hewitt (19U5), salmon vis- 

 cera was fed in conjunction with dry meals to serve as absorbing a- 

 gents or in diets which included a binding agent, to prevent the loss 

 of the water-soluble components. In no trial was salmon viscera fed 

 at the 100 per cent level. When fed at water temperatures above $0 

 degrees F., high levels of salmon viscera (90 per cent or more in the 

 diet) exhibited an excellent growth potential, in every instance 

 superior to the beef liver control (Table 2, Diets 6 and 7 5 Table 3, 

 Diet 2 j and Table i|. Diet 11). 



The high mortalities encountered in lots fed salmon viscera diets 

 during I9I1I4 and 19U8 stemmed from two entirely unrelated sources. In 

 19Wi, mechanical factors and partially spoiled viscera were the two 

 primary causes of mortality. The extremely small" fish used in the ex- 

 periment (2j631 per pound) did not adjust readily to the abrupt change 

 in diet consistency, color, and particle size encountered in the high 

 viscera diets. The high mortality encountered in the first six weeks 

 of the 19UU experiments consisted of emaciated fish which showed every 

 evidence of starvation. During the last two weeks of this experiment 

 the mortality showed another upswing. In this instance the presence of 



lU 



