Breeding Values and Genetic Parameters Estimation for Crude Palm Oil Yield Traits in Two Elaeis guineensis Jacq. Heterotic Groups Based on a Large‐Scale Multi‐Trial Genetic Design Planted in North Sumatra

Oil palm ( Elaeis guineensis ) accounts for the largest part of global edible oil production, and its genetic improvement is one of the main levers to increase yields while minimizing the social and environmental impacts. The oil palm breeding scheme implemented by PalmElit and its partners relies on the exploitation of the general combining abilities (GCAs) of parent palms from two heterotic groups. The GCAs are estimated for complex yield‐related traits through incomplete factorial mating designs planted in large‐scale trials. A total of 140 parents from the heterotic group A (GA) and 150 parents from the heterotic group B (GB) were evaluated through 450 test crosses assessed over a 9‐year period for 12 yield‐related traits. Breeding values (BVs) were estimated as pedigree‐based best linear unbiased predictions (P‐BLUP) using linear mixed models integrating pedigree information. Parental BVs were used to generate the best possible outcomes for A × B commercial hybrid production. The 45 best predicted combinations from the 10% best parents gave an adult yield ranging from 8.9 to 9.6 t crude palm oil (CPO) ha −1 yr. −1 which represents a genetic gain of 21.3% compared to the mean of all trialled crosses. In comparison, genetic progress obtained from the 10% best trialled crosses only reached 10.7% demonstrating the advantage of a breeding strategy based on accurate parental BV estimation and cross prediction. Overall genetic progress on oil yield in the best predicted crosses was found to be mostly attributable to the improvement of the oil extraction rate (OER), which averaged 29.2% due to the stringent pre‐selection on bunch quality traits applied to the parents. In addition, the genetic parameters calculated for all measured traits were used to characterize the breeding populations of restricted origins (BPROs) within both heterotic groups and define an optimal recombination strategy.