Assessment Of Dairy Cow Evaluation Methods

Abstract

A total of 956 Friesian cows with lactation records covering different parities in the period 1966 to 1987, and sired by 39 bulls, were genetically evaluated for 305-day milk yield by computing for each cow the indices: Adjusted Least Squares Mean (ALSM) , Expected Real Producing Ability (LRPA), Expected Breeding Value (EBV) and Predicted Breeding Value (PBV) . They were then ranked according to their four indices, respectively. A rank correlation test was carried out with the view to determining whether the indices ranked cows differently. The rank correlation between ALSM and ERPA or EBV was found to be 0.99, while ERPA and EBV had, as expected, a rank correlation of 1.00 for the parameter combinations of h2=0.20 and r=0.43 and h?=0.25 and r=0.45. On the other hand, the rank correlation between PBV and ALSM, ERPA or EBV was 0.74 when using the parameter estimates of h2=0.20 and r=0.43 obtained in this study and 0.80 when using the average parameter estimates of h2=0.25 and r=0.45. The results showed that Predicted Breeding Value is the most appropriate index of dairy cow evaluation for selection.