Evaluation of techniques used in estimating phenotypic, genetic and environmental trends in beef cattle

Data collected between 1957 and 1975 in the University of Tennessee, Ames Plantation Angus herd was used to compare two methods of adjusting weaning weight for known sources of fixed environmental variation and four methods of estimating phenotypic, genetic and environmental trends. The method of adjusting age-constant (calculated 205-day weight) weaning weights using the coefficient of regression of weaning age on calculated 205-day weight, reported by Robertson (1974), was superior to standard procedures for calculating 205-day weight and in removing the effects of sex and age at weaning. The use of sex of calf and age of dam adjustment factors described by Robertson (1974) and readjustment of calculated 205- day weight removed 97 percent of the variation in weaning weight due to sex, age at weaning and age of dam compared to 89 percent for the Tennessee Beef Cattle Improvement Program method. Appropriate adjustment factors for non-genetic effects are important in subsequent estimates of genetic trends; therefore, only records adjusted for weaning age, sex of calf and age of dam were used in subsequent calculations. The techniques used in estimating phenotypic, genetic and environmental trends were (1) the sires-over-time technique, (2) the repeat-mating technique, (3) the dams-over-time technique and (4) the sires-and-dams-over-time technique. Three estimators involving the sires-over-time technique were described by Smith (1962) and evaluated in this study. The preferred sires-over time estimator was minus twice the pooled within-sire regression of the deviation of performance on time. Each performance record was deviated from the annual population mean. The superiority of this estimator was probably due to the fact that converting performance records to deviations from the annual population mean tended to minimize the effects of short term environmental fluctuations and more accurately express the genetic value of individuals. Estimates of the phenotypic, genetic and environ-mental trends from the sires-over-time technique, using deviations rather than records per se, were 2.89 ± .18, 2.58 ± .36 and .31 ± .40 lb, respectively. Estimates of phenotypic, genetic and environmental trends by the repeat-mating technique were 2.89 ± .18, 2.83 ± .61 and .06 ± .59 lb, respectively; 2.89 ± .18, 1.40 ± .39 and 1.49 ± .40 lb, respectively, for the dams-over-time technique; and 2.89 ± .18, 2.77 ± .36 and .12 ± .40 lb, respectively for the sires-and-dams-over-time technique. All pheno-typic and genetic trends were positive but varied between techniques. Similar magnitude of standard errors was associated with estimates of genetic and environmental trends obtained by the sires-over-time technique, when deviations of performance from the population mean were regressed on time; those obtained by the repeat-mating technique and the sires-and-dams-over-time technique indicated that these methods were equal with respect to their accuracy in estimating genetic and environmental trends.