Ranking parents by best linear unbiased prediction (BLUP) breeding values in oil palm

Abstract

Breeding values obtained from best linear unbiased prediction (BLUP) were used to rank nine pisifera (P) male parents belonging to two different but related genetic groups, utilising highly unbalanced data from three D (Dura) × P progeny-test trials of oil palm. The traits studied were fresh fruit bunch yield (FFB), total number of bunches (BNo), average bunch weight (BW), oil to bunch percentage (OB) and average annual height increment (HINC). Four P's from the AVROS genetic group had higher breeding values for FFB, BNo and OB while five P's from the Dumpy-AVROS group were superior in breeding values for BW and HINC. Selection of parents based on the average of the rankings in breeding value for BNo, OB and HINC for each parent tended to favour AVROS P's. However, selection of parents based on the aggregate genetic worth function involving relative economic values, tended to favour Dumpy-AVROS P's. The latter method is preferred to the average ranking method because of its emphasis on relative economic values and the selection index approach. It is suggested that the BLUP technique may prove to be a very useful tool in oil palm breeding.