Predicting nutrient digestibility and energy value for broilers.

Digestibility coefficients of nutrients, metabolizable energy (ME), net energy (NE) and the ratio of NE to ME (NE/ME) of 20 diets were measured in broiler chickens (1 to 21 d). Dietary nutrients were formulated to keep similar ME/nutrient ratios, except for dietary protein, fat, and fiber using corn, soybean meal, animal protein blend, barley, poultry oil and an enzyme mixture of xylanase, glucanase, and phytase. Digestibility coefficients of nutrients and ME were measured in battery cages under free-access of feed, while NE was measured in floor pens feeding 75% of recommended ME intake each day. NE for maintenance was calculated on basis of mean metabolic weight using a coefficient from a previous study and NE for gain was calculated by body protein and fat gains using dual-energy x-ray absorptiometry. Digestibility coefficients of protein and neutral detergent fiber (NDF) were curvilinearly related to dietary protein and NDF, respectively, while digestibility coefficients of fat and starch were linearly correlated to dietary fat and starch, respectively. The inclusion of enzymes increased the digestion coefficient of NDF to predict the digestibility of protein, NDF, fat, and starch. MEn/gross energy ratio averaged 72.5% and was correlated to protein, fat, NDF, and starch. ME values were accurately predicted from chemical characteristics, where best equations were obtained from digestible nutrients. Energetic efficiencies of ME were 72% (NE/MEn) and 68% (NE/ME) and varied by about 20 and 18%, respectively. Ratios of energetic efficiency were 68% for digestible carbohydrates; 86% for digestible fat; and 76% (NE/MEn) and 59% (NE/ME) for digestible protein. According to the lowest residual standard deviation the best nutrient components to predict energy were digestible nutrients for predicting ME values (41 kcal/kg); digestible protein intake, fecal organic matter, and body fat and protein for predicting heat increment values (111 kcal/kg); and combination of ME and crude nutrient for predicting NE values (140 kcal/kg).