Comparison of animal model with other conventional methods of sire evaluation for milk production in Karan Fries cattle

Abstract

First lactation milk production records of (1122) Karan Fries cattle sired by 112 bulls were analyzed to estimate breeding values of sires using animal model (DFREML), best linear unbiased prediction (BLUP) and least squares methods (LSM) of sire evaluation. The average breeding value of sires was almost similar and ranged fronl 2942.67 kg (BLUP) to 2947.79 kg (DFREML). The lowest error variance and highest coefficient of determination in DFREML method revealed that this method of sire evaluation was most efficient and accurate. The coefficient of variation in DFREML method was closest to the CV (%) of unadjusted data, indicating highest stability of DFREML over other two methods of sire evaluation. The heritability estimate of milk yield by DFREML was higher than the estimates of BLUP and LSM. The rank and simple correlations among all three methods of sire evaluation were high and statistically highly significant, indicating that there was higher degree of similarity (about 90-97%) in ranking of sires by different methods.