Prediction of beef carcass and meat traits from rearing factors in young bulls and cull cows.

The aim of this study was to predict the beef carcass and LM (thoracis part) characteristics and the sensory properties of the LM from rearing factors applied during the fattening period. Individual data from 995 animals (688 young bulls and 307 cull cows) in 15 experiments were used to establish prediction models. The data concerned rearing factors (13 variables), carcass characteristics (5 variables), LM characteristics (2 variables), and LM sensory properties (3 variables). In this study, 8 prediction models were established: dressing percentage and the proportions of fat tissue and muscle in the carcass to characterize the beef carcass; cross-sectional area of fibers (mean fiber area) and isocitrate dehydrogenase activity to characterize the LM; and, finally, overall tenderness, juiciness, and flavor intensity scores to characterize the LM sensory properties. A random effect was considered in each model: the breed for the prediction models for the carcass and LM characteristics and the trained taste panel for the prediction of the meat sensory properties. To evaluate the quality of prediction models, 3 criteria were measured: robustness, accuracy, and precision. The model was robust when the root mean square errors of prediction of calibration and validation sub-data sets were near to one another. Except for the mean fiber area model, the obtained predicted models were robust. The prediction models were considered to have a high accuracy when the mean prediction error (MPE) was ≤0.10 and to have a high precision when the was the closest to 1. The prediction of the characteristics of the carcass from the rearing factors had a high precision ( > 0.70) and a high prediction accuracy (MPE < 0.10), except for the fat percentage model ( = 0.67, MPE = 0.16). However, the predictions of the LM characteristics and LM sensory properties from the rearing factors were not sufficiently precise ( < 0.50) and accurate (MPE > 0.10). Only the flavor intensity of the beef score could be satisfactorily predicted from the rearing factors with high precision ( = 0.72) and accuracy (MPE = 0.10). All the prediction models displayed different effects of the rearing factors according to animal categories (young bulls or cull cows). In consequence, these prediction models display the necessary adaption of rearing factors during the fattening period according to animal categories to optimize the carcass traits according to animal categories.