Dissecting Quantitative Variation Introgressed into Upland Cotton Using QTL-Stacked Segregating Populations

We developed QTL-stacked populations in two elite G. hirsutum backgrounds targeting eight QTL alleles introgressed into advanced-backcross G. hirsutum lines affecting fiber elongation (three each from G. tomentosum, Gt; and G. mustelinum, Gm), fiber fineness (one from Gt), and fiber strength (one from Gm). A total of twenty-one QTL-stacked F₁ individuals (nine in GA2004230 and 12 in R01-40-08 backgrounds) were selfed to generate F₂ populations segregating at two to three QTL regions with 33 to 132 plants per population (totaling 1,509 plants), which were field tested for fiber quality traits, genotyped at target QTL regions, selectively advanced to F_2:3 generation and assessed at two different locations. Favorable shifts in average phenotypes of selected traits were partially explained by genotypes at introgressed QTL regions, with a subset consistent across backgrounds and/or generations. One-way analysis of variance in F₂ populations validated the effects of two target QTL in both backgrounds (one to three populations each), and four more in at least one background. While three of four Gm-QTL allele effects were detected, Gt-QTL effects were relatively elusive. Gm-QTL ‘qELO-1-1’ and ‘qELO-11-1’ were evidently ‘stable’ and consistently increased fiber elongation in backcross-selfed populations (prior study) and in QTL-stacked secondary segregating populations (current study). A previously undetected Gm-QTL introgressed into chromosome 11 explained a significant proportion of variation in fiber fineness in F₂ and pooled analysis of F₂/F_2:3 lines in several populations. Compared to average phenotypes of appropriate parental backgrounds, selectively advanced lines conferred fiber quality improvements as high as 23.3-24.0% for micronaire, 10.5-11.0% for strength, and 28.6-47.6% for elongation. Phenotypic differences between singly homozygous and doubly homozygous QTL-stacks were also evident in some populations. Our results attest to the promise of MAS as a molecular breeding tool for simultaneously validating QTL effects and developing QTL-stacked germplasm in different genetic backgrounds.