Association Analysis for Yield-Related Traits in Soft Red Winter Wheat for the Triticeae CAP Mid-Atlantic Allele-Based Breeding Project.

Early-generation evaluations for yield and other highly quantitative traits in crop plants have traditionally been hindered by a lack of available seed and/or resources to enable robust replicated testing across environments. The extensive genotype-by-environment (GxE) interactions that are typical for quantitative traits often confound attempts to derive reliable estimates of line performance when lines are tested across a limited range of environments. Recent widespread adoption of high-throughput technologies such as genotyping-by-sequencing (GBS) for routine line genotypic assessment will in theory provide greater power to predict line performance by allowing for estimation of the phenotypic effects of individual alleles. An objective of the current research is to assess the utility of high-density, genome-wide marker information for early-generation testing in a panel of 365 lines from three regional breeding programs (Illinois, Kentucky and Virginia). Lines were evaluated in six locations for the 2013-2014 growing season: Brownstone, IL; Urbana, IL; Lexington, KY; Woodford, KY; Blacksburg, VA; and Warsaw, VA. Phenotypic traits assessed across all environments include yield, test weight, plant height, and heading date. Data on additional yield-related traits was collected in both Virginia locations, including grains m-2, heads m-2, thousand-kernel weight, harvest index, days to flag leaf senescence, days to physiological maturity, canopy-temperature depression at anthesis and during grain fill, and near-infrared grain composition analysis. Similar tests were planted at all six locations in fall 2014, and are being evaluated over the 2014-2015 growing season. A publicly-available GBS protocol utilizing a methylation-sensitive restriction enzyme was used to generate a large volume of marker data while minimizing ascertainment bias and representation of highly repetitive DNA sequences. After filtering out SNPs with a high proportion of missing data and/or a low minor-allele frequency (MAF), ~100,000 SNPs were retained for analysis. Results and conclusions from this work will be presented during the poster session.