| Abstract: | Two trials were conducted with mature cows to develop prediction equations for estimating carcass composition from live animal measurements. Trial 1 involved 82 animals that were used to develop these equations. Subjective criteria (frame and condition scores) also were utilized along with live weight in development of prediction equations. Equations using subjective independent variables predicting energy and carcass tissue weights apparently were slightly more variable in precision (R2 = .69 to .91) than those developed with objective variables (R2 = .74 to .90). Equations predicting percent compositions from subjective variables had lower coefficients of determination (R2 = .49 to .76) than those developed with objective variables (R2 = .63 to .83). Forty-one animals similar to those used in Trial 1 served as a test group in Trial 2. Measured carcass composition was regressed on composition predicted by equations developed in Trial 1. These equations generally had slopes and intercepts with confidence intervals that included 0 and 1, respectively, indicating that most of the prediction equations were unbiased. This work indicates that carcass composition can be predicted from both objective and subjective measurements, with some accuracy and precision. |
