Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: Ultrasound Estimates of Loin Muscle Measures and Backfat Thickness Augment Live Animal Prediction of Weights of Subprimal Cuts in Sheep.

Authors
item Leeds, Timothy
item Mousel, Michelle
item Notter, David - VIRGINIA POLYTECHNIC
item Lewis, Gregory

Submitted to: Western Section of Animal Science Proceedings
Publication Type: Proceedings
Publication Acceptance Date: March 1, 2007
Publication Date: June 7, 2007
Citation: Leeds, T.D., Mousel, M.R., Notter, D.R., Lewis, G.S. 2007. Ultrasound Estimates of Loin Muscle Measures and Backfat Thickness Augment Live Animal Prediction of Weights of Subprimal Cuts in Sheep. Western Section of Animal Science Proceedings. 58:97-100.

Interpretive Summary: In domestic meat-producing species, the ability to predict carcass composition in the live animal is a valuable tool for producers and breeders. Noninvasive ultrasound technology to measure biological tissues was developed primarily for human medicine applications in the mid-20th century. This technology has been widely and successfully adapted by the beef and swine industries and has vastly improved the efficiency and accuracy of live animal carcass composition prediction. However, the extent and success of its use in the sheep industry has been limited to date. Widespread use of ultrasound as a predictor of carcass composition in live sheep requires that it be validated as a reliable tool. The research reported here describes the accuracy and precision of carcass fat and muscle measures obtained on the live animal using ultrasound. Additionally, this research suggests that these ultrasound measures improve our ability to predict weights of high-value meat products in the live animal.

Technical Abstract: The efficacy of live animal, real-time, B-mode ultrasound (US) estimates of carcass traits as (partial) predictors of carcass composition warrants investigation in sheep of varying genetic and environmental backgrounds. Our objectives were to 1) evaluate US estimates of corresponding carcass measures using correlations (r) and statistics established for beef and swine (prediction SE [SEP]; repeatability SE [SER]; and bias [TB]); and 2) estimate variation in weights of subprimal cuts (roast-ready rack, trimmed loin, and boneless leg), after accounting for live BW, explained with US loin muscle and backfat (BF) measures. Wethers (n = 172) from four sire breeds were reared in an extensive system, finished on a concentrate diet, and harvested at a mean weight of 62.9 (SD = 9.5) kg. Before harvest, 12th/13th rib transverse US images were captured using an ALOKA SSD-500V US device with a 3.5-MHz, 14.5-cm linear array transducer and standoff. Images were interpreted using ImageJ software (v1.36b). After a 24-h chill, carcasses were ribbed, measured for loin muscle area (LMA) and BF, and fabricated. Weights of subprimal cuts were described using linear models with BW and US loin muscle and BF measures as predictors. Prediction models that maximized R2 and included only significant terms (P < 0.05) were identified. For Objective 1, SEP, SER, and TB for BF were 0.14, 0.08, and 0.07 cm, respectively, and r was 0.81. The SEP, SER, and TB for LMA were 1.55, 1.31, and -0.004 cm2, respectively, and r was 0.75. For Objective 2, the best prediction models for trimmed loin and boneless leg weights included BW and US LMA and BF as predictors. The best prediction model for roast-ready rack weight included BW and US LMA. The BW explained 70.3, 69.9, and 72.9% of variation in trimmed loin, boneless leg, and roast-ready rack weights, respectively; US estimates explained an additional 4.1, 5.6, and 2.3% of variation in these weights. Based on these data, US estimates of carcass measures obtained by a trained technician can be reliable and can augment our ability to predict weights of subprimal cuts in live sheep.

Last Modified: 10/21/2014
Footer Content Back to Top of Page