Prediction of the sensory quality of longissimus beef of three cattle categories from rearing factors characterising the rearing management

Abstract

Here we used individual data on three cattle categories (young bulls, n = 154; heifers, n = 154; and cull cows, n = 341). As rearing managements are very specific to the animal category, a model was developed for each category of cattle, using binomial in logistic regression to predict meat eating quality in longissimus muscle (LM) from rearing factors applied throughout the animal’s life. The meat eating quality variable (composed of two modalities: Low vs. High) was statistically defined by a hierarchical clustering on principal components from the data of eight sensory descriptors. The heifer and cull cow models discriminated the meat quality clusters better than the young bull model, with 66.0%, 68.0%, and 59.6% success rate for external validation, respectively. Rearing factors merged as potential action levers to improve LM meat quality (odds ratio > 1), but only in the heifer model. Heifers fed concentrates during the pre-weaning period in housing and at pasture were 5.0 and 6.4 times more likely to produce higher LM meat quality, respectively. The nature of the main forage in the fattening diet was also an actionable factor for improving meat quality. Compared to heifers fed corn silage, heifers fed only pasture (1.9 times) or hay and wrapped haylage (7.8 times) or corn silage and wrapped haylage (23.4 times) were more likely to produce higher LM meat quality. Management practices at different periods of heifer life can ultimately act on sensory eating quality.

Keywords

• rearing factors
• logistic regression
• longissimus beef
• meat sensory properties
• whole life