In Vitro Estimation of Apparent Metabolizable Energy

A number of experiments were carried out to determine if a two stage in vitro method developed for monogastrics could, using in vitro digestible energy (IVDE), estimate the nitrogen-corrected apparent metabolizable energy (AMEn) of poultry diets. For a total of 11 diets assayed with adult roosters the average apparent metabolizable energy was 2.937 kcal/g, while the in vitro determined digestible energy was 2.969 kcal/g of diet. Regression analysis showed a significant (P<.05) correlation (r = .93) between IVDE and AMEn. The prediction equation generated from regression analysis was: Y = 1.0380X − .143 [r = .93, residual standard deviation (rSD) = ± .145 kcal/g], where Y = apparent metabolizable energy in kilocalories per gram and X = in vitro digestible energy in kilocalories per gram. When percent ether extract (EE) was included in the regression model, the accuracy of predicting AMEn was increased, and the prediction equation became: Y = .9169X1 + .0518X2 − .0213, (r = .98, rSD = ± .073 kcal/g), where Y = apparent metabolizable energy in kilocalories per gram, X1= in vitro digestible energy in kilocalories per gram, and X2 = % EE of the diet. Including other parameters of proximate analysis did not significantly (P>.05) increase the correlation nor improve the accuracy of prediction. When actual IVDE values were compared to predicted values generated from regression equations as estimates of AMEn, the IVDE values were more accurate (root-mean-squared-error = ± .119 kcal/g) than predicted values (rSD = ± .145 kcal/g) for the 11 diets assayed.