Comparison of sire evaluation methods in Frieswal cattle

Abstract

Breeding values of 72 Frieswal sires with first lactation 300 days or less milk yield records (range 200 to 300 days) on 5 or more progeny maintained at 6 Military Farms were estimated and ranked by 4 different sire evaluation methods, viz. least squares method (LSM), simple regressed least squares (SRLS), best linear unbiased prediction (BLUP) and derivative-free restricted maximum likelihood (DFREML). DFREML method was the most efficient method due to its lowest error variance. The rank correlation of DFREML with LSM, SRLS and BLUP were 0.907, 0.909 and 0.956, which were significant (P&lt0.01). It can be concluded that Frieswal sires could be evaluated by DFREML method for estimating their breeding values with comparatively more accuracy and reliability. However, significant and high magnitude of rank correlation between DFREML and BLUP revealed that BLUP method could be used as next best method for evaluating Frieswal sires.