The use of ultrasound can predict adjusted fat thickness (FAT) and marbling (MAR) with reasonable accuracy in tropically adapted cattle, Texas researchers found. And, further refinement of the process, based on breed, management and sex, may improve such projections.

Heifers produced from a three-breed diallele mating system (Angus, Brahman and Romosinuano) were mated to Mashona (a Sanga-type breed originating in eastern Zimbabwe) bulls. The resulting calves were weaned at an average of 74 days in central Florida, backgrounded for 40 days, and trucked 750 miles for growing and finishing in central Texas.

The calves, stratified by sex and breed type, were assigned to confinement (FY) or pasture (PS) growing treatments for an average of 141 days. They were then finished on a common diet and harvested at about 300 (Group 1) or 345 (Group 2) days on feed.

Equations derived from published relationships between carcass traits and growth were used to predict FAT and MAR at slaughter from one ultrasound measurement.

Ultrasound FAT and MAR and body weight measurements were taken 74 or 126 days prior to slaughter for groups 1 and 2. Final BW was estimated based on average daily gain for the 28 days after ultrasound measures. Carcass weight change was estimated as 0.8 (body weight change to slaughter) and used to project final FAT and MAR.

The projections were compared to actual values obtained after a 48-hour chill. Percentage differences were analyzed as response variables indicative of projection accuracy with harvest group, breed, treatment, sex and sex X treatment as effects in the model.

Harvest group did not influence carcass measures (P > 0.11) or projection accuracy (P > 0.14). Breed affected MAR (P < 0.01) and projected MAR accuracy (P=0.03). Treatment influenced MAR (P < 0.01) and MAR projection accuracy (P=0.02).

PS calves had greater MAR (481 vs. 410 ± 17.6); MAR was over-predicted by 13.5% in FY calves. Heifers had greater FAT and MAR than steers (P < 0.01) and lower carcass weight (P < 0.01).

Sex differences in carcass traits resulted in differences in prediction accuracies for FAT (P < 0.01; heifers -9.5%, steers 17.5%, SE = 4.1%) and MAR (P < 0.01; heifers -2.1%, steers 17.5%, SE = 3.7%).

— JE Sawyer, AD Herring, DK Lunt, DT Dean, RD Rhoades and DG Riley, Texas A&M University-College Station; McGregor Ag Research Center, McGregor, TX; and USDA-ARS Subtropical Ag Research Station, Brooksville, FL.