Predicting net energy in broiler chickens using the comparative slaughter technique

Two experiments were conducted to determine a prediction model of the net energy (NE) requirement from the net energy for maintenance (NEm) and the retained energy (RE) by the comparative slaughter technique and its validation, and the metabolizable energy was determined using the nitrogen-corrected total excreta collection method. A completely randomized block design was used, and in the validation, the error (%) and the analysis of the concordance correlation coefficient (CCC) were determined. In each experiment, two replicates were used (five chickens per replicate) for each treatment. The experiment included 370 Cobb 500 male chickens (one day old), distributed in two experiments (190 in the first and 180 in the second), fed with six diets, considering a basal diet (corn + soybean meal), from which was substituted on a percentage basis (weight/weight) for the ingredients frequently used in chicken feed (corn, wheat by-product, soybean meal, fish meal and soybean oil) with three feeding levels: ad libitum (AL), 85 % (AL) and 70 % (AL) in both experiments. To determine NEm, the relationship between RE and metabolizable energy intake (MEI) was established using simple linear regression. The second study estimated and validated the NE with the proposed model, where the NEm was calculated with simultaneous equations obtained from the basal diet of Study 1. The values of NEm/W^0.75, EMm/W^0.75, and the efficiency k, showed similar values among diets (p<0.05). The MEI was greater in the diet with soybean oil than others, the RE represented 45.6 % of the metabolizable energy (ME) consumed and was the greatest in the diet with soybean oil (p<0.05), the heat production on average was 104.59 kcal/day (54.35 % of the MEI), and NEm (kcal/day) was the highest in the diet with soybean oil and represented 24.90 % of the MEI, the heat increment was similar in all diets (29.45 % of the MEI). The estimated NE was similar (141.39 ± 21.13) to the observed NE (137.07 ± 22.28) (p>0.05), with an adequate linear relationship (R² = 0.996) and low error (4.01 %), the CCC was high (0.993) and a υ of 1.02. The proposed net energy model (NE(kcal/day)=79.66·W^0.75+RE), allows to estimate with precision and accuracy the NE requirement in broiler chickens.