Genetic and phenotypic relationships among carcass measurements in beef cattle

Carcass data from three large beef-cattle breeding experiments in New Zeala nd were analysed. Regressions of saleable meat weight on carcass and livewe ights were estimated as parameters for a modelling exercise where pasture w as converted to carcass and liveweight, finishing cattle at different end p oints. The data were also used in order to estimate heritabilities for nine carcass traits and the relationships among them. Slaughter data were from 1962 cattle, sired by 199 different bulls from 13 sire breeds. Relationship s were first estimated between hot carcass weight and either carcass compon ents or pre-slaughter weight, and then an "animal" model was used to estima te genetic parameters. On a log-log basis, the overall regression of carcas s weight on pre-slaughter weight was 1.089 +/- 0.008. The corresponding log -log regression of saleable meat weight on carcass weight was 1.002 +/- 0.0 07, bone weight on carcass weight 0.779 +/- 0.015, and trimmed fat weight o n carcass weight 1.265 +/- 0.041. The regression slopes indicate the percen tage change in one trait as a result of a 1% increase in another trait. The values for saleable meat weight or bone weight were reasonably consistent across trials, whereas the regression for trimmed fat weight varied with tr ial. Multivariate heritability estimates were 0.48 +/- 0.07 for hot carcass weight, 0.31 +/- 0.06 for dressing percent, 0.42 +/- 0.10 for eye muscle a rea, 0.38 +/- 0.12 for carcass fat depth, 0.49 +/- 0.12 for weight of high- priced cuts, 0.48 +/- 0.07 for saleable meat weight, 0.51 +/- 0.07 for bone weight, and 0.30 +/- 0.05 for trimmed fat weight. Genetic correlations sho wed a strong negative relationship between meat percentage and trimmed fat percentage (-0.94), whilst genetic correlations of each with bone percentag e were not significantly different from zero. The relationship between hot carcass weight and fat percentage was small and positive in sign (significa nt (0.05 +/- 0.02) for the phenotypic correlation but not significant (0.06 +/- 0.12) for the genetic correlation).