Evaluation of simultaneous selection for live body weight and against abdominal fat in broilers

Three hundred chickens from pedigree matings in a commercial female grandparent stock were slaughtered at 7 or 9 weeks of age. Data were obtained for live body weight and various carcass parts. Coefficients of variation of abdominal fat (g and g/100 g body weight) were above 30%. Coefficients of variation of the weights of muscular and skeletal parts were much lower (about 10%) and were further reduced when calculated per 100 g body weight. Heritability of abdominal fat was higher (.82) than that of live body weight (.55) and parts: breast (.55), thighs (.31), and drumsticks (.51). Phenotypic and genetic correlations between each of the heavier parts (breast, thighs, and drumsticks) and body weight were over .8; for smaller parts (legs, wings, etc.) correlations (.49 to .79) were lower. Abdominal fat showed the lowest correlation (.26 to .36) with body weight. The genetic correlations seemed free of the part-whole relationship. Simultaneous selections for live body weight and for breast weight, or against abdominal fat, were simulated using "independent culling levels". The theoretical calculations showed that adding breast weight as a selection criterion had no advantage over selection for body weight alone due to the high genetic correlation between the two traits. However, simultaneous selection for body weight and against abdominal fat, using optimum combination of culling, was economically superior to selection for body weight alone. The advantage of this two-trait selection depends on the correlation between the traits and their relative values.