THE RELATION OF METABOLIZABLE ENERGY VALUES TO CHEMICAL COMPOSITION OF WHEAT AND BARLEY FOR CHICKS, TURKEYS AND ROOSTERS

Abstract

Six separate regression equations have been calculated to relate the metabolizable energy (ME) value of 16 samples of wheat and 16 samples of barley, determined with young chicks, growing turkeys and adult roosters, to their chemical composition. The simplest equation accounted for 83% of the variation in the mean ME values of the barley samples determined with chicks merely by using the crude fiber level in the barley. The prediction equation for the ME values of the barleys for the turkeys included crude fiber, crude protein and ash. In the case of the wheat, the prediction equations used crude fiber, crude protein, ash, fat, nitrogen-free extract and selenium, although selenium could be eliminated without great loss of precision for wheat samples with less than 1.32 ppm selenium. Nitrogen-free extract as a measure of available carbohydrate gave higher coefficients of determination to the regression equations than more specific assays of starch and sugar. The ME values estimated in the present study were compared with those estimated by applying prediction equations published earlier by a number of other workers. These equations fell into two groups: those designed to predict the ME values of individual feeds, and those designed to predict the ME values of a whole range of feeds and mixed diets. There was better agreement between the predicted values from the first group of equations and the observed values for the wheats and barleys than for the ME values predicted by the more general equations. However, even the equations specifically developed for wheats underestimated the ME values of the wheat samples determined with roosters, and it is suggested that there may be merit in developing prediction equations for each feed for use with each class of poultry.