Abstract
Two experiments are reported in which the measurement of back fat in live sheep using an ultrasonic probe was evaluated. In the first experiment, 5 operators assessed 20 sheep on each of 5 days. Repeatabilities within operators across days, and within days across operators, were evaluated as intraclass correlations. These were all high, ranging from 0.89 to 0.94. In the second experiment, which involved the same 5 operators and 65 lambs, correlations between comparable live and carcass measurements ranged from 0.84 to 0.91. However, live/carcass relationships differed between operators, and in some instances deviated appreciably from the expected 1:1 ratio. After correction for body weight, correlations between live and carcass measurements were reduced but remained highly significant. Carcass fat depth measurements were slightly better as indicators of side fat percent than were ultrasonic measurements. An important outcome of the experiments was an improved appreciation of the importance of minimising fat depression when making measurements, especially for lambs.