The study was undertaken to assess farmer and consumer perceptions on grain quality preferences, characterize (physicochemically and with gene markers) rice germplasm from Africa and the U.S.A for their quality, map QTLs for grain quality and agronomic traits in a BC1F2 population derived from a cross between Digang and Jasmine 85 using genotyping-by-sequencing (GBS), and determine the efficiency of SNP markers derived from GBS for selection of grain quality loci in the BC1F2 population.
Good grain quality and high yields were very important to famers in all the four communities. Most consumers interviewed preferred imported rice due to the poor grain appearance and undesirable cooking and eating qualities of locally produced rice. The farmers’ regarded non availability of adequate numbers of improved varieties with good grain qualities as one of the top four challenges they faced. Farmers’ grain quality preferences were consistent with those of consumers. Both groups preferred long grain aromatic rice that cook soft and fluffy. Most farmers cultivated Jasmine 85 because of its aroma and good taste. Other important traits that were of interest to the farmers included disease resistance, plant height (short or tall depending on site), ease of threshing and tolerance to weeds.
A wide diversity was observed in all the grain traits studied. Variation in these traits was significantly associated with molecular markers for the aroma, alk, and waxy genes. The phenotypic variances explained by the markers were higher in the U.S. germplasm than in theAfrican germplasm.
Some African accessions exhibited heterozygosity at marker loci and will have to be purified to improve the efficiency of marker-assisted-selection (MAS) for grain quality.Forty-two QTLs were identified for 19 grain quality and agronomic traits in the BC1F2 population. The major QTLs for apparent amylose content (AAC) and paste viscosity parameters including peak viscosity (PV), hot paste viscosity (HPV), cool paste viscosity (CPV), breakdown (BD), setback (SB), and consistency viscosity (CS) were found in the vicinity of the waxy loci. The waxy region explained 20.4, 48.1, 56.3 and 62.3% of the phenotypic variation in AAC, PV, CPV and HPV, respectively. Alkali spreading value (ASV), peak time (PeT), set back ratio (SB ratio) and stability (ST) were mainly controlled by the alk locus on chromosome 6. QTLs for setback ratio (qSBR-6, qSBR-11) and stability (qST-6) are being reported for the first time. The major QTLs detected for plant height/culm length, grain length/shape and aroma were coincident with the SD1, GS3 and BADH2 genes, respectively. In general, alleles with increasing effects came from the higher value parent although this was not the case in all situations.
Genotyping-by-sequencing provided markers that were tightly linked to genes of interest. GBS markers were used to select the background of Digang and foreground selection for aroma, AAC and ASV. Eight individuals were selected based on percent genotypic background of the recurrent parent (Digang) (?70%), presence of aroma, AAC (15-22%) and ASV (3-7). Given the quantitative nature of ASV and AAC, the use of single markers for these traits was not efficient in predicting their respective phenotypes. The aromatic phenotype was consistently predicted by the peak marker for aroma (S8_20185312), making it very useful for marker-assisted breeding.