Abstract
Data from 8 central beef bull performance tests, including a total of 288 bulls, were analyzed. Three models were used to analyze sources of variation in start weight and final weight (at .apprx. 8 and 18 mo.). All models accounted for the effects of test-year and the linear regression of weight on age, along with effects for herds, sires within herds and bulls within sires (the error term). Models 1 and 2 are accounted for herd in terms of the test bulls'' herds of origin, and Model 3 accounted for it in terms of sires'' herds of origin. Models 1 and 2 differed in that herd effects were classified across test-years in Model 1 and within test-years in Model 2. Overall linear regressions of age were significant for start weight but not final weight. Correlations between the 2 weights were high in spite of 270 kg gain being achieved on test. The regression of final weight on start weight was 0.79 kg .cntdot. kg-1. The sizes of overall and residual regressions were very similar, and it was shown that adjusting final weight for start weight may not remove herd genetic variance. Effects of bulls'' herd of origin were significant and persistent up to final weight on test. Bulls were able to compensate on test for differences in age, but there was little compensation of final weight for start weight. Effects of sires'' herd of origin were significant only for start weight. Other central test research results are discussed and compared with these.