Comparison of real-time ultrasound measurements for body composition traits to carcass data in feedlot cattle fed zilpaterol hydrochloride

Abstract Text: The objective of this study was to compare measurements of real-time ultrasound (RTU) and carcass data to determine body composition in feedlot cattle (n = 96) fed zilpaterol hydrochloride (ZH). Black-hided cattle were weighed and scanned using RTU 32 d prior to the start of the ZH treatment period.  Resulting data were used to predict empty body fat (EBF %).  Steers (n = 48; BW = 520 ± 30.4 kg; EBF % = 26.2 ± 1.9) and heifers (n = 48; BW = 466 ± 29.5 kg; EBF % = 26.7 ± 1.7) were blocked within gender by EBF % in a complete randomized block design and randomly assigned to pen (2 pens/block; 4 hd/pen) and treatment (6 pens/treatment). Measurements of RTU were taken by a certified ultrasound technician approximately 24 h prior to slaughter using an Aloka 500-V instrument with a 17-cm 3.5 MHz transducer. Hair was clipped to less than 0.64 cm and vegetable oil was applied. The RTU measured traits consisted of 12-13^th rib backfat thickness (uBF, mean = 12.4 mm), 12-13^th LM area (uREA, mean = 93.9 cm²), and marbling score (uMARB, mean = 5.1). Intramuscular fat was converted to uMARB by using the equation: uMARB = ((769.7 + (56.69×uIMF))/100) – 5. Overall means for carcass data were 12-13^th rib backfat thickness (cBF, 13.6 mm), 12-13^th LM area (cREA, 92.0 cm²; measured using tracing paper), and marbling score (cMARB, 5.2). Marbling scores were converted to a numeric cMARB (Slight⁰⁰ = 4, Small⁰⁰ = 5, and Modest⁰⁰= 6). Data were analyzed using the PROC REG, MEANS and CORR procedures of SAS. Results show that both methods were highly correlated to each other. Overall correlations were 0.83, 0.61, 0.69 for BF, REA and MARB, respectively. The accuracy statistics for cattle (heifers and steers pooled together) fed ZH and Control were almost identical. Correlations for ZH and control cattle were 0.83, 0.60, 0.68 and 0.83, 0.59, 0.70 for BF, REA, MARB for ZH, respectively. The low correlation for REA could be a result of the method of collecting carcass REA data. Overall RTU underpredicted BF and MARB and overpredicted REA (bias = -1.32, -0.18 and 1.47, respectively). These results show that RTU can be used to predict carcass traits prior to slaughter and feeding ZH seems to have no effect in predicting carcass traits using RTU in live animals.

Keywords: ultrasound, carcass traits, zilpaterol hydrochloride