Estimation of carcass composition by ultrasound measurements in 4 anatomical locations of 3 commercial categories of lamb.

The objectives were to study the relationship between in vivo ultrasound measurements and cold carcass measurements at 4 anatomical points along the backbone of lambs and to determine appropriate regression equations to estimate carcass composition by using ultrasonic measurements at each anatomical point. The lambs (n = 114) used were suckling lambs (BW = 11.09 kg), light lambs (BW = 22.43 kg), and wethers (BW = 32.03 kg), representing a wide range of BW. Measurements of subcutaneous fat and skin thickness and of muscle depth and muscle width were taken over the 10th to 11th and 12th to 13th thoracic vertebrae and over the first to second and third to fourth lumbar vertebrae. These measurements were taken at one-third of the musculus longissimus thoracis et lumborum (LMT) width with the probe perpendicular to the backbone. The left sides of the carcasses were dissected into muscle, fat, and bone. The weight of lean tissue increased (P < 0.001) at a rate of approximately 500 g for each kilogram of carcass weight increase. Pelvic fat weight increased (P < 0.001) slightly with increasing carcass weight (11.8 g), whereas kidney fat and subcutaneous fat showed great gains (P < 0.001; 40.3 and 134.4 g, respectively). Ultrasound LMT width of light lambs remained constant along the backbone, whereas LMT width of suckling lambs and wethers increased (P < 0.001) from the cranial to the caudal direction. Ultrasound LMT depth and fat thickness between the 10th and 11th thoracic vertebrae were greater (P < 0.01) than measurements taken at other backbone locations. The greatest difference between ultrasound and carcass measurements was in LMT width, with differences between ultrasound and carcass measurements always being greater in LMT depth than in fat thickness. Carcass LMT width was more closely correlated with carcass lean than with other tissues, especially at both thoracic locations (r = 0.80 and r = 0.71). In general, skin thickness measured by ultrasound was poorly correlated (from r = 0.19 to r = 0.33) with all carcass tissues because of slight variations in skin thickness. Ultrasound LMT depth was more closely correlated with carcass tissues than was carcass LMT depth (from r = 0.53 for bone to r = 0.71 for lean), whereas ultrasound fat depth and carcass fat depth presented similar correlations (from r = 0.49 for bone to r = 0.72 for intermuscular fat). Regression coefficients for predicting lean were 0.95 to 0.96, for predicting subcutaneous fat were 0.67 to 0.75, for predicting intermuscular fat were 0.81 to 0.84, and for predicting bone were 0.78 to 0.88. This study was not conclusive regarding predicting carcass composition in relation to an optimal anatomical position, given that all the anatomical locations studied allowed accurate regression equations. Body weight was the most important predictor of carcass composition, with a slight improvement in regression equations when using ultrasound. However, ultrasound muscle and fat depths were correlated with carcass muscle and fat depths and with the tissular composition of carcass.