Development of new tools and germplasms for improvement of wheat resistance to Fusarium head blight

Abstract

Wheat Fusarium head blight (FHB) is a devastating disease of wheat worldwide, which can significantly reduce grain yield and quality. Although the application of fungicides can reduce FHB damage, growing FHB resistant wheat is the most effective and eco-friendly approach to reduce the losses. To develop locally adapted FHB-resistant hard winter wheat germplasm, we transferred three major QTLs: Fhb1, Qfhs.ifa-5A, and Qfhb.rwg-5A.2 into two hard winter wheat cultivars, ‘Everest’ and ‘Overland’, using marker-assisted backcrossing and multiplex restriction amplicon sequencing (MRASeq). Ten ‘Overland’ background lines and nine ‘Everest’ background lines with better FHB resistance, recurrent parent similar agronomic traits were selected. They can be used as FHB resistant bridge parents for hard winter wheat breeding. To identify native FHB resistant sources, a population of 201 U.S. breeding lines and cultivars were genotyped using 90K wheat SNP arrays and phenotyped for the percentage of symptomatic spikelets (PSS), Fusarium damaged kernels (FDK) and deoxynivalenol (DON), a toxin produced by the pathogen. Genome-wide association studies (GWAS) identified significant trait associations with single nucleotide polymorphisms (SNPs) on chromosomes 1A, 1D, 2B, 3A, 3B, 4A, 5B and 5D. These marker-trait associations (MTAs) were significant for at least two of the three traits or a single trait in at least two experiments. To accelerate the evaluation of the FDK, we developed an algorithm that can separate FDK from healthy kernels with an accuracy of 90% based on color differences using image processing and unsupervised machine learning methods. Discovery and creation of the new FHB resistant germplasms and development of the fast FDK phenotyping algorithm will accelerate the improvement of U.S. hard winter wheat cultivars for FHB resistance.