PROGRESS IN BIOLOGICAL INQUIRIES, 193 3 361 



this will not be feasible under present conditions. These aniiniils 

 are not ada})ted to transfer from one habitat to another, while their 

 small size makes them easy prey for the fish, resultin«^ in almost 

 total loss of brood stock. Shrimj) are very sensitive to differences 

 in pH and quickly die when transferred to waters showing; a material 

 difference in this respect from their original habitat. This is true 

 even though the water may have an abundant supply of dissolved 

 oxygen. 



TROUT CULITJRE 



Feeding experwients. — As in 1932, feeding experiments were car- 

 ried on at tile Pittsford (Vt.) station under the direction of R. F. 

 Lord and at the Leetown (W.Va.) station under the direction of 

 E. W. Surber. Both brook and rainbow trout were used in these 

 experiments. Several lots of Loch Leven fingerlings w^ere also car- 

 ried on experimental diets at the Leetown station. 



Since previous experiments have demonstrated conclusively that 

 better results can be obtained when certain dry products are included 

 in the diet tlian by feeding fresh meats alone, the experiments in 1933 

 were primarily designed to determine the level at which these dry 

 products can be fed most efficiently and economically. Unfortu- 

 nately, owning to the limited funds available for experimental work, 

 it was necessary to discontinue the experiments at both stations early 

 in September. 



As has been emphasized in previous reports, there is no dry prod- 

 uct available connnercially which can be fed to trout successfully for 

 any considerable length of time without the inclusion of raw meat 

 in the diet. With large fingerlings and older trout only 15 to 25 

 percent of raw meat is required to keep the fish in healthy condition. 

 Unfortunately, mixtures containing such a small percentage of meat 

 cannot be fed without considerable waste and it is consequenth' more 

 economical to include a larger amount of raw meat than is necessary 

 for the well being of the fish. The meat not only makes the ration 

 more palatable to the fish but serves as a binder to liold the fine par- 

 ticles of meal together so they can be eaten readily. When the pro- 

 portion of the meat is too small, the mixture quickly disintegrates in 

 the water and much of the dry food is lost. 



As in previous years, salmon-egg meal gave the best results of any 

 dry product used, although, with regard to growth, there was very 

 little difference between this product when used alone and a mixture 

 of equal parts salmon-egg meal and a good grade of meat meal. 

 Meat meal alone was somewhat inferior to salmon-egg meal. 



The results of the experiments show that as high as 60 percent of 

 dry meal may be economically incorporated in the diet. For in- 

 stance, one lot of yearling rainbow trout, on a diet composed of 60 

 percent salmon-egg meal and 40 percent raw pig liver, gained 153 

 percent in weight from June 7 to September 1, while a second lot, on 

 a diet composed of equal parts pig liver and salmcm eggs, showed an 

 increase of only 143 percent during the same period. The conversion 

 factor was slightly better when the larger amount of salmon-egg 

 meal M^as used, since it required 2.1 pounds of food to produce a 

 pound of trout when this product was fed at a 50 percent level, and 

 only 1.9 pounds when fed at a 60 percent level. Wlien the dry con- 



