Empirical evaluation of marker-assisted selection for improving quantitative traits in sweet corn.

摘要

This thesis consists of two related experiments concerning the application of marker-assisted selection (MAS) in plant breeding programs. The first was designed to compare the efficiency of MAS and phenotypic selection (PS) in improving quantitative traits in sweet corn. MAS and PS were applied to three F_2:3 base populations (C₀) with either of the su1, se1 or sh2 endosperm mutations. Twenty-percent of the families were selected to constitute MAS and PS C ₁ composites of single and multiple traits including seedling emergence, kernel sucrose and tenderness, and overall liking. Selection efficiencies were evaluated based on gains and costs, where a total of 52 paired comparisons were made between MAS and PS C₁ composites. In 38% of the comparisons, MAS resulted in significantly higher gain than PS, while PS gain was significantly higher than MAS in 4% of the cases. The average MAS gain across all three composite populations was nearly twice that of PS. MAS was found most appropriate when traits are difficult and costly to measure.; In the second study, three QTL alleles associated with enhanced seedling emergence in an F_2:3 shrunken2 population were introgressed into three sweet corn elite inbreds using marker-assisted backcrossing. Three BC₂ populations were developed and one family in each population that contained all three QTL was crossed to the recurrent parents resulting in three BC₂F₁ populations. Across the three populations, BC₂F₁ families heterozygous for the beneficial QTL alleles linked to umc139, bnl9.08, or php200689 markers displayed an average of 48.6%, 35.6%, and 36.2% increase in seedling emergence, respectively, over the unmodified F₁s. The introgressed QTL alleles exerted similar effects in the BC₂F₁ generation as was observed in the original F _2:3 mapping population.; Results generated from this study suggest that incorporating DNA markers to traditional breeding programs can expedite selection progress and reduce costs. Also, data suggests that beneficial QTL identified in one population can exert similar effects in other genetic backgrounds.