Progressively lower levels of salmon viscera vdth a corresponding 

 increase in the amount of equal parts beef liver and hog liver proved to be 

 decidedly beneficial. The highest level of beef and hog liver, 66.6 percent 

 in Diet 8, produced gains that were significantly greater than either the 30 

 percent level or the l5 percent of beef and hog liver (Tab3.e 1, Diets 8, 17, 

 and 18). The 30 percent level of meats, in turn, although producing no sig- 

 nificant difference at the end of the cold-water period, yielded a signif- 

 icantly greater mean lot weight at the end of the experiment than did the l5 

 percent level of meats. The mortalities also seemed to indicate that the 

 higher the level of meats the less the mortality rate. There were no sig- 

 nificant differences to substantiate this statement, but there was a trend 

 toward higher mortalities with the higher levels of salmon viscera. 



As practical diets during cold-water, the relative value of these 

 high-level salmon viscera rations was demonstrated by a comparison of these 

 diets with the beef liver control. Diet 1. At the end of the first 12 weeks, 

 all of the high-level salmon viscera fed lots made gains that were as good or 

 better than the lot on straight beef liver, and there were no statistically 

 significant differences in mortalities (Table 1, Diets 1, lU, l6, and 17). 



The production diets used at this station during the cold-water per- 

 iod, Diets 2 and 8, if compared with the high-level salmon viscera rations, 

 produced superior weight gains thab were significant in most cases at the end 

 of 12 weeks and significant in all cases at the end of 2U weeks. 



During warm-water periods, the feeding of high-level salmon viscera 

 diets without a salmon waste meal resulted in comparatively low growth rates. 

 liVhat these high-level salmon viscera diets could have produced in the way of 

 growth if meal had been added is a matter of conjecture. A reasonable esti- 

 mate, based on comparison of Diet lU with Diet 2, which had the same meat- 

 viscera base as Diet 3 but no added salmon waste meal, would be that the final 

 mean lot weight of Diet lU with meal \d uld have been less than the final mean 

 weight of the meat-viscera-meal mixture, (Diet 3) • 



SUGGESTED PRODUCTION DIETS DERIVED FROM THE FIRST EXPERIMENT 



The following suggestions for practical hatchery diets were designed, 

 not only for immediate use, but also to stimulate interest in products such as 

 salmon eggs and vacuum-dried salmon viscera meal that are not commercially avail- 

 able but could be if s\ifficient demand were created. Primary emphasis was given 

 to rapid growth and high survival. 



The dollar value of the various diets was left to the reader to dec- 

 ide on the basis of current market prices, conversions, and on the rate of 

 growth and survival desired. 



The diets for cold water rearing had low, comparable mortalities. 



