The use of carcass cuts to predict beef carcass composition: A research technique

Eleven Angus, 12 Friesian and 12 Hereford steers were used to investigate the degree of accuracy and usefulness of primal cut tissues in predicting side composition. The criteria used for evaluating the cuts were: (a) standard error of estimate of the equation, (b) homogeneity of ‘b’ values among breeds, (c) appreciable bone content in cut to allow the prediction of side bone, and (d) the absence of major difficulties in the replication and dissection of cuts. Simple and multiple regression analyses showed that the most accurate predictors of carcass composition, in descending order, with standard errors of estimate of muscle, fat and bone percentages respectively, were: hindquarter plus rib cut (0.37%, 0.47%, 0.30%); hindquarter (0.73%, 0.87%, 0.49%); loin plus rib cut (0.84%, 0.88%, 0.48%); rib cut (1.13%, 1.26%, 0.59%); loin (1.24%, 1.21%, 0.72%). The most useful of four easily obtained carcass variables in improving the prediction accuracy of carcass components from multiple regression proved to be primal cut weight and fat thickness at the 12th rib, particularly the former. Both significantly reduced the standard errors of estimate of muscle, fat or bone in equations based on loin, rib cut and loin plus rib cut, but not in equations based on hindquarter plus rib cut and hindquarter. Kidney plus pelvic fat weight was of limited value, resulting only in a slight improvement in the prediction of side bone percentage using the equations based on bone percentage of the hindquarter. Carcass weight was of equal value to primal cut weight in improving the prediction accuracy of multiple regression. Five sets of part-carcass prediction equations are given, providing a choice of prediction accuracy, labour expenditure and cost for research workers whose requirements and resources may vary