Groundnut Breeding Advancements: Efforts Towards Genomic Selection for Quicker Genetic Gains

Groundnut (Arachis hypogaea L.) is an important food crop in sub-Saharan Africa, and worldwide. Among the major causes for low yields is the susceptibility of cultivated varieties to the Groundnut Rosette Disease (GRD) and leaf spots. Genomic selection (GS), characterized by a model calibrated on phenotype and genotype information of a training population is used to predict genomic estimated breeding values (GEBVs). The essence of GS in any breeding program is to accelerate the selection progress by shortening generation interval and increase in selection intensity, thus a resource saving breeding method. Traditional breeding methods are augmented by GS that has the ability to forecast GEBVs with enough precision for selection across multiple generations that eliminates the need for extensive phenotyping and speeds up genetic gains. To support these efforts, vector-host interaction studies have been conducted, populations to support GRD markers support developed, evaluated and genotyped; an African core set genotyped, a genome-wide association analysis for loci associated key traits done, and studies on prediction models building on studies from earlier efforts, such high-density genotyping and prediction accuracy for different GS models and cross validation approaches for key traits. The valuable results on vector-host interaction forms a basis for further characterization of these genotypes using the GRD validated molecular markers to understand the physiological basis of the varied reaction to vector and disease incidence. Sequencing the genome of the aphid species on groundnut is crucial to inform the diversity of the vector and give insights on how microbial effector proteins, host targets and plant immune receptors coevolve. The validation of the GRD markers will be a breakthrough in breeding efforts through marker assisted selection for this trait, while at the same time providing genetic information to improve the prediction of GS models. The genome wide association mapping marker set was very informative, comparable to the Africa core set study. The marker set would be ideal for future development of quality check (QC) and mid-density panel markers. The prediction accuracy increased and genetic variation decreased when large-effect SNPs were fitted as fixed factors. We envisage to enhance the superiority of the GS results further through multienvironment prediction models, and more quality phenotypic details of the key traits in question. These efforts will provide an array of tools for use to achieve quick genetic gains in the groundnut breeding programs.