Evaluation of Overnight Shrinkage in Reduction of Variability of Weights and Gains of Beef Cattle1

Abstract

Overnight shrinkage at three different ages — weaning, yearling, and long-yearling — was evaluated in Angus (AA), Brahman (BB), and reciprocal cross calves (AB, BA) over a 4-yr period to determine changes in precision of estimates of weights and gains. Approximately 454 steers and heifers were evaluated at 205 d, 431 were evaluated at 365 d, and 231 heifers were evaluated at 540 d. Calves were spring-born and managed on either common bermudagrass [Cynodon dactylon (L.) Pers.] or endophyte-infected (Acremonium coenophialum Morgan-Jones and Gams) tall fescue (Festuea arundinacea Schreb.). Breed group differences in shrinkage, as a percentage of full body weight, were small but statistically significant. Percentage shrink at weaning in AA exceeded AB, BA, and BB (P<.05), and BA exceeded AB calves (P<.10). Similarly, AA exceeded other breed groups in percentage shrink at 365 d (P<.01) and 540 d (P<.10). Percentage shrink was larger on tall fescue than on bermudagrass at 205 d (P<.12) and 540 d (P<.01) but was larger on bermudagrass at 365 d (P<.01). Although most variances in actual shrunk weights and gain were slightly less than variances in full weights, differences were not statistically significant from variances of full weights and gains. Additionally, reductions in variances associated with shrinkage were similar to reductions expected when a random variable is multiplied by a positive constant between zero and one. Coefficients of variation for shrunk weights and gains were generally larger than coefficients of variation for full weights and gains, suggesting that shrinkage does little for reduction in relative variation. These data suggest that overnight shrinkage is not beneficial in increasing the precision of comparisons of group means of weights or gains of beef cattle. Consequently, in the interests of enhancing the well-being of experimental animals and avoiding unnecessary costs, shrunk weights for purposes of comparing group means should be avoided.