The fasting heat production of broilers is a function of their body composition

Abstract

The fasting heat production (FHP; net energy expenditure under maintenance) is a key value to determine the net energy value of the feed and is calculated from measurements in respiratory chambers or predicted as a function of the metabolic BW (MBW). This study aimed to determine the influence of body composition on the FHP of broilers through a modeling approach. Six experiments (Exp 1–6) were conducted to develop (Exp 1–4) and externally validate (Exp 5 and 6) predictive models. In Exp 1–4, broilers of a single genetic line in floor pens were subjected to three dietary treatments to induce differences in body composition. The FHP (calorimetry chambers) and the body protein-to-fat ratio (PFR; Dual-Energy X-ray Absorptiometry) were recorded at nine time points between 7 and 55 days. Linear models were adjusted (considering the pen as the experimental unit), starting with a model containing the MBW as the sole predictor. Further modifications were tested by including the PFR and the age of the birds. A random split with a refitting approach was applied. Four models were internally validated and refitted to the whole data. They considered the FHP a function of the MBW (or BW) and the PFR, including or not the age of the birds, and showed high precision (R² > 0.98) and accuracy (error < 2.5%). All models demonstrated that birds at the same BW and age have higher FHP the leaner they are. Body protein and fat are positively and negatively correlated to FHP, respectively. In Exp 5, broilers of two genetic lines fed three dietary treatments were tested for FHP and body composition at 22 and 41 days. In Exp 6, broilers fed continuous dietary treatments in a dose–response design were tested for FHP and body composition at 17, 20, and 41 days. The models were tested on the data from Exp 5 and 6 considering the treatments as the validation unit. The four selected models showed high prediction precision (validation R² > 0.92) and accuracy (model prediction error < 3%) and were validated externally. In conclusion, at the same age and BW, the higher the PFR, the higher the FHP. Including the PFR improved the strength of the models.