Use of Real-Time Ultrasound in Pigs During the Early Finishing Phase to Predict Carcass Composition at Slaughter

© 1999 American Registry of Professional Animal Scientists. The possibility of predicting kilograms of lean pork, the percentage of lean, and lean gain per day using measurements taken during the early finishing phase was studied using real-time ultrasound measurements on 164 barrows. Initial live weight (IWT), ultrasound backfat (IUBF), and ultrasound loin muscle area (IULA) were determined at 16 to 19 wk of age (70 kg). Final measures were obtained as the average pen weight reached 110 kg (22 to 24 wk of age). Initial kilograms of lean pork (IKOL) and the percentage of lean (ILEAN) were estimated with IWT and IUBF as the independent variables. Average daily gain, backfat accretion (BFACR), and lean gain per day (LGPD) were calculated from the initial and final measures. Pearson correlation coefficients were determined among the variables at the initial and final measures. The IULA and IKOL were correlated with the amount of lean tissue at slaughter (FKOL) (r=0.305 and 0.315, respectively; P<.001). Initial estimates of backfat (IUBF) and percentage lean (ILEAN) were correlated with FLEAN (r=-0.650 and -0.661, respectively; P<0.001). The IWT and IKOL were the only measurements taken in the early finishing phase related to LGPD r= -0.210 and -0.204, respectively; P<0.05). Prediction equations for FKOL, FLEAN, and LGPD were estimated from linear regression procedures. The IWT and IUBF were predictors of FKOL (R2=0.235; P<0.001). IUBF or ILEAN were predictive of FLEAN (R2=0.409 or 0.437, respectively; P<0.001). Measurements from the early finishing phase accounted for less than 10% of the variation in LGPD, ADG, or BFACR. This study suggests that carcass fat (FUBF or FLEAN) at slaughter can be predicted using real-time ultrasound in the early finishing phase with moderate confidence, accounting for 40 to 50% of the variation. Measures of lean tissue content or muscling (FKOL or FULA) can also be predicted, but with less confidence and accounting for less of the variation (20 to 30%).