Predicting the metabolizable energy and metabolizability of gross energy of conventional feedstuffs for Muscovy duck using in vitro digestion method

Abstract In total, two experiments were conducted to evaluate the effectiveness of an in vitro digestion method for predicting the metabolizable energy (ME) and metabolizability of gross energy (ME/GE) values using in vitro digestible energy (IVDE) and the digestibility of gross energy (IVDE/GE) content, respectively, of conventional feedstuffs for Muscovy ducks. In experiment 1, the apparent metabolizable energy (AME), true metabolizable energy (TME), AME/GE, and TME/GE of eight-grain feedstuff samples (two corn samples, three sorghum samples, and three barley samples) and eight protein feedstuff samples (two soybean meal samples, three cottonseed meal samples, and three rapeseed meal samples) were determined by the tube-feeding method with six different ducks for each sample. In experiment 2, a computer-controlled simulated digestion system (CCSDS) contain simulated digestive fluid was used to determine the enzymatic hydrolysis energy value of feedstuffs, which was defined as IVDE in our study. The simulated gastric fluid containing pepsin and simulated small intestinal fluid containing amylase, trypsin, and chymotrypsin for the in vitro gastric and intestinal digestion, respectively. The IVDE and in vitro digestibility of GE (IVDE/GE) of 16 feedstuff samples were determined using the CCSDS with five replicates per sample. The results showed that the IVDE and IVDE/GE were positively correlated with ME and ME/GE of feedstuffs, respectively. The coefficient of determination of eight regression models in predicting ME (grain feedstuffs: AME = 1.050 × IVDE− 0.9293, TME = 1.032 × IVDE + 0.6478; protein feedstuffs: AME = 1.331 × IVDE− 6.685, TME = 1.269 × IVDE−3.490) and ME/GE (grain feedstuffs: AME/GE = 1.069 × IVDE/GE− 6.516, TME/GE = 1.068 × IVDE/GE + 0.7764; protein feedstuffs: AME/GE = 1.093 × IVDE/GE −19.21, TME/GE = 1.196 × IVDE/GE − 13.25) of feedstuffs for Muscovy ducks ranged from 0.8610 to 0.9921. The accuracy of the regression model was acceptable as the difference between measured and predicted ME and ME/GE values was less than 0.45 MJ/kg (100 kcal/kg) and 2.62% for 14 of the 16 feed samples, respectively. In conclusion, the in vitro digestion method can be used to predict the ME and ME/GE of conventional feedstuffs for Muscovy ducks with acceptable accuracy.