Ranking the value of germplasm: new oil palm (Elaeis guineensis) breeding stocks as a case study

Abstract

Plant breeders constantly need to introduce desirable new alleles to refresh breeding stocks. This first requires an assessment of potential new sources of material and then identification of genotypes most able to augment existing stocks. Genetic distance analysis is widely used for both purposes, although it measures both haplotype diversity and novel allele abundance. Here, we present a more tailored approach to address these problems. Using oil palm as an exemplar, simple metrics of allelic and genetic richness, graphical genotyping and multivariate analysis were deployed to determine the overall value of Ghanaian germplasm to supplement Sumatra Bioscience (SumBio) breeding material. We next compared three methods to rank individuals. The first was based on multivariate genetic distance. However, we also developed two new systems: Global Allelic Divergence (GAD), based on novel allele abundance, and Genome Scan Allelic (GSA) divergence, which additionally considers genome context. Ghanaian material exhibited increased allelic richness, higher heterozygosity and a higher proportion of private alleles than extant SumBio breeding stocks. Graphical genotyping revealed Ghanaian material as allele-rich in genomic regions that were allele-poor in SumBio breeding stocks. Multivariate analysis showed a collective distinctness and increased variability of Ghanaian plants. Ranks of individuals varied between GSA, GAD and genetic distance. The GAD and GSA ranks correlated strongly with each other but only poorly with the genetic distance-based ranks. We conclude that GSA and GAD are superior ranking systems to identify individuals most likely to introduce valuable new alleles, whilst genetic distance analysis identifies individuals likely to require least backcrossing.