362 U.S. BUREAU OF FISHERIES 



stituent of the diet is still further increased, the mixture disinte- 

 grates so readily that it is very diiRcult to feed without considerable 

 waste. 



Since it is a universal practice to feed rapidly growing fish all 

 they will eat, experiments were run at both the Pittsford and Lee- 

 town stations to determine if a reduction in the amount of food 

 would result in its being utilized more efficiently. The results are 

 inconclusive, although indicating that probably a somewhat greater 

 efficiency can be obtained by feeding slightly less than the fish will 

 consume readily. In the case of brook trout fingerlings, 2.4 pounds 

 of food were required to produce a pound of fish on a diet of beef 

 liver and salmon-egg meal when the fish were fed all they would 

 eat readily. In another lot of trout on the same diet but given 25 

 percent less food than the former lot, 2.3 pounds of food were re- 

 quired for each pound of fish produced. Rainbow fingerlings made 

 a better showing, since the amount of food required to produce a 

 pound of fish was 2.8 and 2.1, respectively. In the case of brook 

 fingerlings at the Leetown station on a similar diet, 2.57 pounds of 

 food were required for each pound of fish produced, when the fish 

 were given all they would eat, and 2.67 pounds when fed 25 per- 

 cent less. The rainbow fingerlings again made a much better show- 

 ing. In this case 2.58 pounds of food were required for each pound 

 of fish produced when the fish were fed all they would eat,, while 

 only 2 pounds were required when the amount of food was reduced 

 25 percent. 



With rainbow yearlings at the Pittsford station, a reduction of 

 10 percent in the amount of food in the case of fish fed a mixture 

 of equal parts of pig liver and salmon-egg meal, resulted in 1.8 

 pounds of food to 1 pound of trout as compared with 1.9 pounds 

 in fish fed the larger amount. When fed pig liver a considerably 

 larger amount of food was required to produce a pound of fish, when 

 the amount was reduced 15 percent, 10.6 pounds being required when 

 the fish were fed the full amount and 13.7 pounds on the reduced 

 diet. 



It is scarcely necessary to point out that the growth on the re- 

 duced diets was considerabl}'' less in every case than that of fish fed 

 all they would readily consume. It should be emphasized, how- 

 ever, that in no case were the fish overfed. The fingerlings were 

 given onl}?^ as much food as they would readily eat twice a day, 

 while the yearling trout were fed onl}^ once a day. 



In an effort to clear up some of the uncertainty regarding the 

 amount of food required to support trout in nature, a number of 

 brook and rainbow trout fingerlings at the Leetown hatchery were 

 kept on natural food from May 3 to October 12. The food of these 

 fish consisted principally of the water sawbug (Asellus) with some 

 gammarus and a few snails. A supply of these organisms was kept 

 in the troughs at all times so that the fish had all they could eat. 



Within 3 weeks after the experiment was started, the color of 

 these fish was noticeably brighter and within a short time they be- 

 came the most highly colored fish at the station, in fact, the colors 

 were much more intense than those of the average wild trout. 



At the end of the experiment the average individual weight of 

 the brook trout was 33 orams and of the rainbow trout 31.5 grams. 



