Commercial adaptation of ultrasonography to predict pork carcass composition from live animal and carcass measurements.

Live animal and carcass data were collected from market barrows and gilts (n = 120) slaughtered at a regional commercial slaughter facility to develop and test prediction equations to estimate carcass composition from live animal and carcass ultrasonic measurements. Data from 60 animals were used to develop these equations. Best results were obtained in predicting weight and percentage of boneless cuts (ham, loin, and shoulder) and less accuracy was obtained for predicting weight and ratio of trimmed, bone-in cuts. Independent variables analyzed for the live models were live weight, sex, ultrasonic fat at first rib, last rib, and last lumbar vertebra, and muscle depth at last rib. Independent variables for the carcass models included hot carcass weight, sex of carcass, and carcass ultrasonic measurements for fat at the first rib, last rib, last lumbar vertebra, and muscle depth at last rib. Equations were tested against an independent set of experimental animals (n = 60). Equations for predicting weight of lean cuts, boneless lean cuts, fat-standardized lean, and percentage of fat-standardized lean were most accurate from both live animal and carcass measurements with R2 values between .75 and .88. The results from this study, under commercial conditions, suggest that although live animal or carcass weight and sex were the greatest contributors to variation in carcass composition, ultrasonography can be a noninvasive means of differentiating value, especially for fat-standardized lean and weight of boneless cuts.