| dc.description.abstract |
Five ultrasonic machines (Meritronics, Sonalyser, Renco, new Medata and old Medata) for predicting carcass Introscope P2 backfat depth from measurement on the live pig were compared. The effect on predictive accuracy of mislocating P2 on the live pig was also investigated. The Meritronics was the most accurate machine (r2, 0.83; r.s.d., 1.32; n, 262) and was used as the comparison standard for the other units. The order of precision of the other units compared with the Meritronics was n r2 r.s.d. Sonalyser 150 0.81 v. 0.85 1.34 v. 1.19 Renco 108 0.75 0.85 1.55 1.20 New Medata 100 0.47 0.71 1.66 1.23 Old Medata 35 0.37 0.70 2.40 1.64 Apart from the Meritronics, all units underestimated carcass P2. This was attributed to an inability to detect the third fat layer. Depth of backfat influenced the accuracy of some machines. No difference in accuracy was found between 2 operators using the Sonalyser. Mislocation of the P2 site on the live pig anterior or posterior to carcass P2 reduced the accuracy of the Meritronics ( P<0.05) and Sonalyser (P<0.10). Dorsal or ventral deviation from P2 had no significant effect. The results indicate the Meritronics, the Sonalyser and the Renco satisfactorily predict backfat thickness of pigs before slaughter and that predictive accuracy is maximised by taking the live pig measurement within 20 mm axially and 15 mm laterally of the carcass P2 site. |
|