Comparison of phenotypic versus marker-assisted background selection for the SUB1 QTL during backcrossing in rice

Marker assisted backcrossing has been used effectively to transfer the submergence tolerance gene SUB1 into popular rice varieties, but the approach can be costly. The selection strategy comprising foreground marker and phenotypic selection was investigated as an alternative. The non-significant correlation coefficients between ranking of phenotypic selection and ranking of background marker selection in BC(2)F(1), BC(3)F(1) and BC(3)F(2) generations indicated inefficiency of phenotypic selection compared to marker-assisted background selection with respect to recovery of the recipient genome. In addition, the introgression size of the chromosome fragment containing SUB1 was approximately 17 Mb, showing the effects of linkage drag. The significant correlation coefficient between rankings of phenotypic selection with the percentage of recipient alleles in the BC(1)F(1) generation suggested that background selection could be avoided in this generation to minimize the genotyping cost. The phenotypically selected best plant of the BC(3)F(1) generation was selfed and backcross recombinant lines were selected in the resulting BC(3)F(4) generation. The selection strategy could be appropriate for the introgression of SUB1 QTL in countries that lack access to high-throughput genotyping facilities.