Study for the Use of Animal Models in Genetic Evaluation of Japanese Black using Akaike's Bayesian Information Criterion

The use of animal models in genetic evaluation was studied using the carcass records of 947 Japanese Black steers collected by the field testing program in S prefecture. The analyzed traits were carcass growth rate (kg/day), marbling score, rib eye area (cm2), yield estimates and carcass price per weight (yen/kg). The fixed effects considerd were year and month at slaughter, age at slaughter, fatted prefecture, slaughter house and fatted prefecture-slaughter house. Using these effects, nine analytical models were assumed in each of animal, maternal grandsire and sire models, and then the best fit one was selected among them for each trait using Akaike's Bayesian information criterion (ABIC). The sire model predicted 88 sires and the maternal grandsire model predicted 179 sires. The animal model predicted 179 sires, 937 dams and 974 steers. As a result of the analysis, animal models showed the best fit for carcass growth rate and carcass price per weight. However, for the other traits, any animal models was not selected as the best fit one. Especially, the values of ABIC for the animal models including three fixed effects were much larger than those of the other models. These results suggested that attention should be paied to use animal models in genetic evaluation of Japanese Black using the restricted pedigree information.