The use of various live animal measurements to predict carcass and meat quality in two divergent lamb breeds

Live weight, subjective scores of condition and conformation, live animal video image analysis (LVIA), ultrasound and X-ray computed tomography (CT) scanning were used to investigate the best method or combination of methods for predicting carcass and meat quality traits in live Texel and Scottish Black-face lambs. Predictors derived from CT alone accounted for a high proportion of the variance in dissected fat and muscle weight in Texel lambs (adjusted R-2 = similar to 0.8), as well as intra-muscular fat content in the loin (similar to 0.6), but lower proportions in Blackface lambs (similar to 0.7 for fat, 0.4-0.5 for muscle and intra-muscular fat), after adjusting for sire and fixed effects. Adding traits measured by other in vivo methods increased prediction accuracies (adjusted R-2) by up to 0.26, depending on trait and data set. Shear force and ultimate pH could not be accurately predicted using the traits considered here (adjusted R-2 < 0.4). Although the same methods tended to be best for predicting product quality traits between breeds, prediction accuracies differed. (c) 2008 Elsevier Ltd. All rights reserved.