Association Mapping for Drought Tolerance Traits in Spring Wheat Grown Under Rainfed and Irrigated Conditions.

Drought tolerance is an integrative trait which involves expression of many genes or QTLs. Understanding and identifying genes responsible for drought tolerance may accelerate the breeding process for the trait. The International Maize and Wheat Improvement Center (CIMMYT) has developed a spring wheat association mapping panel (n=294) specifically for the identification of genes and markers associated with high temperature and drought tolerance. In 2010 and 2011, we evaluated these lines under rainfed conditions and full irrigation in Greeley, CO and Melkassa, Ethiopia (total of five environments) for yield and yield components, flag leaf characters, and canopy reflectance indices. A total of 287 lines were genotyped with Diversity Array Technology (DArT) markers to identify markers associated with drought related traits under different water regimes. Significant differences among lines were observed for most traits at each environment and across environments. Best Linear Unbiased Predictors of each line were used to calculate associations between 1863 markers and each trait employing a mixed linear model with population structure and kinship-matrix in the model as covariates. A total of 21 markers/regions from 13 chromosomes were associated (P<0.05) with yield in at least three of a total of six data sets (five environments plus combined data). A strong marker-yield association (P<0.001) was obtained for marker wpt3457 on chromosome 5B (73 cM) and this marker also showed weak (P<0.05) to moderate (P<0.01) associations with major yield component traits such as grain number, harvest index, test weight and kernel weight. Similarly, marker wpt6531 on chromosome 2D (67.3 cM) was associated with yield and major yield components in 2-4 environments for each trait. In addition, both grain number and yield were significantly associated with markers on chromosome 1A (87.9 cM), 3D (51.6 cM) and 7B (69.6 cM), which is consistent with the strong phenotypic correlation (r=0.86) between these two traits. Most marker-trait associations for biomass were environment-specific but markers on chromosome 5B (108.5 cM), 6B (57.6 cM) and 7B (210.9 cM) showed significant associations (P<0.05) in three or more environments. However, more marker-trait associations were obtained for normalized vegetation index (NDVI) and green area index (GA) on a total of 13 and 7 chromosomes, respectively. Markers on chromosome 1A, 1B, 2D and 6B were associated (P<0.05) with leaf senescence. Although several marker-trait associations were obtained at each rainfed condition in this study, reliable identification of genes or QTLs for drought tolerance requires additional testing across multiple locations and years.