Genomics-enabled breeding for sorghum in West Africa

Abstract

In semi-arid regions, staple crop productivity is affected by multiple environmental, biological and socio-economic factors threatening food security. Sorghum (Sorghum bicolor) is adapted to semi-arid and sub-humid zones of West Africa (WA). This crop is cultivated over large areas, corresponding to variable and diversified local contexts. The genetic basis of local adaptation and farmer preferences and their applications in breeding need further studies. Recent genotyping methods have provided access to high-density markers and their applications in breeding. In this thesis, genomic resources of WA germplasm were developed using the genotyping-by-sequencing method (GBS) to understand the genetic diversity and to identify quantitative trait loci (QTLs) associated with yield components under pre-flowering water deficit. Evidence of local adaptation in genomic regions linked to flowering time in sub-humid zones and balancing selection grain pigmentation were found. Phenotyping of the WA sorghum association panel (WASAP) was conducted under experimental water-deficit treatments. Significant variations of yield components were observed suggesting local adaptation and drought tolerance in the WASAP. Genome-wide association studies identified novel QTLs controlling vegetative biomass and grain weight under water deficit treatment. QTLs colocalizing with known genes in various traits were also identified. Furthermore, these genomic resources were used to develop diagnostic markers for resistance to Striga hermonthica, a weed parasite of grass crops, in which resistance is known to be associated with a deletion of a few genes. Using GBS data, single nucleotide polymorphisms in linkage disequilibrium with the deletion to generate breeder-friendly markers were selected. Analyses identified eight SNPs, converted to breeder-friendly markers and tested in biparental populations and diverse germplasm using outsourced genotyping. The findings provide genetic resources to the scientific community and could facilitate genomics-enabled breeding of sorghum in sub-Saharan Africa.