Diet 5 contained the standard meat and viscera mixture and 10 percent 

 acetone-extracted salmon viscera meal bound with salt. This meal derived 

 from the same source as the low-teraperature-dried salmon viscera meal 

 varied only in its method of preparation. Comparison with Diets 2 and 3 

 would measure the effect of this meal on the growth rate. 



Diet 6 included 90 percent salmon viscera and 10 percent apple pomace. 

 Salmon viscera, because of its semi-liquid nature, cannot be fed without 

 the addition of some absorbing agent. Otherwise a large portion of the 

 water-soluble components of the viscera dissolve into the water and are una- 

 vailable to the fish. Apple pomace was selected as the absorbing agent 

 because previous experimental work indicated that its contribution to the 

 growth rate when added to a diet at the 10 percent level was insignificant. 

 The salmon viscera had been ground, mixed with the apple pomace, packaged 

 in small containers, and stored at -10° F. until used. During the grinding 

 and mixing operations the viscera was kept in a frozen condition. Only the 

 amount sufficient to meet the daily requiremaits of the troughs on this diet 

 was removed from cold storage on the day it was to be used. The diet was 

 allowed to soften, then was cut into small chunks and placed in the mixer 

 where the particle size was reduced to a point where it could be fed with 

 the hand ricers. This procedure was necessary in order that its method of 

 preparation be comparable to Diet 7. Comparisons with Diets 1 and 2 would 

 measure the growth potential of salmon viscera. 



Diet 7 consisted of 100 percent salmon trininings. The method of prep- 

 aration of this product differed fran normal procedures. At the cannery the 

 waste was run through a disintegrator, packaged in 50 pound boxes, and quick- 

 frozen. Its method of preparation for feeding duplicated the prooedure 

 used for Diet 6. The effect of this diet on the growth rate of blueback 

 salmon could be determined by comparison with Diets 1, 2, and 6. 



The experiment was initiated on July 9, 1947, and concluded on Septe»i- 

 ber 30, 1947. The summarized data for the experimental period will be found 

 in Table 1. 



The mortalities in all diets, with the exception of Diet 7 (100 percait 

 salmon trimmings), compared favorably with Diet 1 — the beef liver control. 

 Examinations of sample lots of fish from each diet for the presence of 

 anemia as determined by gill coloration indicated that only Diet 7 vras so 

 affected. These fish were on the verge of an acute anemia when the experiment 

 was discontined. Although the fish fed the salmon viscera diet (Diet 6) 

 showed an intense red coloration of the gills comparable to control groups 

 (Diets 1 and 2), it cannot be positively concluded that this diet contains 

 adequate amounts of the anti-anemia factor because of the short term of the 

 of the experimental period. 



Analysis of variance was used to evaluate the significance of differences 

 in gains in weight between diets and between troughs on a single diet. Us- 

 ing this procedure it was found that a highly significant difference, well 

 below the 1 percent level, existed between diets and an insignificant dif- 

 ference between the trou^s on single diets. The coefficient of variation 

 due to experimental error and biological variation between troughs on the 

 separate diets amounted to 2.6 percent. The fiducial limits of the diet means 

 were used to determine the significance of difference between diets. 



56 



