Determination and prediction of the net energy content of seven feed ingredients fed to growing pigs based on chemical composition

Two experiments were conducted to determine the net energy (NE) content of various feed ingredients (Exp. 1) and to establish prediction equations for NE for these same ingredients (Exp. 2). In Exp. 1, 36 barrows (27.5 ± 0.5 kg) were assigned to six diets (in a completely randomised design) to give six replicates per diet. The diets were a corn-soybean meal basal diet and five experimental diets containing either wheat, wheat bran, corn dried distillers grains with solubles (cDDGS), canola meal or cottonseed meal. The pigs were housed in six open-circuit chambers (six pigs at a time balanced for dietary treatment) for 16 days, the first 7 days were used for an adaptation to the diets, while for the next 5 days pigs were fed ad libitum (2400 kJ/kg BW0.6·day–1), then fed at 1 × metabolisable energy (ME) for maintenance for 3 days, followed by fasting for 24 h. The NE content of wheat, wheat bran, cDDGS, canola meal, and cottonseed meal were 11.44, 7.78, 10.21, 8.38, and 7.32 MJ/kg DM, respectively. Stepwise regression analysis performed between the chemical composition and the energy value of the ingredients, allowed for the NE to be accurately predicted from chemical characteristics. The best fit equation was: NE = 1.46 + (0.63 × gross energy) – (0.37 × acid detergent fibre), with a R2 of 0.94, a residual standard deviation (RSD) of 0.54 MJ/kg, and P < 0.01. In Exp. 2, 36 barrows (34.6 ± 1.5 kg) were randomly allotted to six diets containing various combinations of corn, soybean meal, wheat, wheat bran, cDDGS, canola meal, or cottonseed meal. The chemical composition of each diet was determined and the results were used to establish prediction equations for the NE in the six diets and 16 diets (including six diets from Exp. 1, six diets from Exp. 2 and four diets from our laboratory published data). The best fit equations were as follows: NE = –0.47 + (0.84 × ME) with an R2 of 0.94, a RSD of 0.20 MJ/kg, and P < 0.01; NE = –10.19 + (0.97 × digestible energy) + (0.08 × starch) + (0.05 × acid detergent fibre) with an R2 of 0.84, a RSD of 0.46 MJ/kg, and P < 0.01, respectively. In summary, the NE of the seven ingredients could be better measured by the respiration calorimetry estimated by substitution method. The NE prediction equation determined based on the chemical characteristics of the test ingredients was better than when determined using the chemical characteristics of the diets.