Improving Marker Assisted Selection in Soft Winter Wheat for Fusarium Head Blight Resistance with QTL Validation, Genome-Wide Association, and Genomic Selection

Fusarium head blight (FHB) is a devastating disease of wheat with worldwide distribution. An increase in the frequency and intensity of outbreaks of the disease revealed that current North American cultivars lack sufficient levels of resistance. The development and release of resistant varieties has been complicated by the quantitative nature of resistance and high levels of genotype by environment interaction. In order to increase the resistance levels in elite U.S. cultivars, regional uniform scab screening nurseries have been established to evaluate adapted cultivars and breeding lines. In addition, mapping in bi-parental populations has been conducted to identify resistance QTL that may be deployed by marker assisted selection. The first objective of our research was to evaluate the effect of QTL introgressed from exotic sources into an elite soft red winter wheat cultivar. The resistance genes/QTL Fhb1 and Qfhs.umc-2DLwere introgressed from the Chinese Spring wheat germplasm W14 by accelerated backcrossing to NC-Neuse that has moderate FHB resistance. Selected BC2F2:3 lines having Fhb1 and/or Qfhs.umc-2DL had lower severity of FHB than was observed on backcross-derived lines without either resistance gene. Although lower levels of FHB were observed on backcross lines with the QTL, they were not significantly more resistant than the recurrent parent NC-Neuse. In addition, lines having both QTL were not significantly more resistant than lines having Fhb1, indicating that the effects of the two resistance genes may not be additive.;Experiments using backcrossing to evaluate QTL in different backgrounds are useful for validation of QTL effects but can be expensive and time-consuming to conduct. The second objective of our research was to use genome wide association analysis to validate QTL effects in a population of diverse lines and at the same time identify new QTL. A set of a set of 258 lines were selected from entries in the regional scab screening nurseries and genotyped with DArT markers distributed throughout the genome, as well as SSR and STS markers previously reported as linked to QTL for FHB resistance. Since no major sub-groups were identified in these germplasm using the program STRUCTURE, a combination of the kinship matrix and principal components was used to correct for population structure when performing association mapping. Data collected on FHB incidence, FHB severity, and Fusarium Damaged Kernels (FDK) from evaluation of the 2008, 2009 and 2010 regional scab screening nurseries were standardized and used as phenotypes in the analysis. Markers located in 23 regions on 14 chromosomes were significantly associated with at least one scab resistance trait. Multiple QTL were identified on chromosomes 1A, 1B, 2D, 3B, and 6A. The effect of the Fhb1 resistance gene that was known to segregate in these germplasm was confirmed. In addition, 16 of the QTL co-incided with previously mapped QTL for FHB resistance.;The higher accuracy of genomic selection over conventional marker assisted selection suggests genomic selection would be an efficient method to breed for scab resistance. The genomic selection model ridge regression was found to produce higher accuracies than Bayes Cpi.