Multi‐Locus Mixed Model Analysis Of Stem Rust Resistance In Winter Wheat

Abstract

© Crop Science Society of America 5585 Guilford Rd., Madison, WI 53711 USA. Genome-wide association mapping is a powerful tool for dissecting the relationship between phenotypes and genetic variants in diverse populations. With the improved cost efficiency of high-throughput genotyping platforms, association mapping is a desirable method of mining populations for favorable alleles that hold value for crop improvement. Stem rust, caused by the fungus Puccinia graminis f. sp. tritici, is a devastating disease that threatens wheat (Triticum aestivum L.) production worldwide. Here, we explored the genetic basis of stem rust resistance in a global collection of 1411 hexa ploid winter wheat accessions genotyped with 5390 single nucleotide polymorphism markers. To facilitate the development of resistant varieties, we characterized marker–trait associations underlying field resistance to North American races and seedling resistance to the races TTKSK (Ug99), TRTTF, TTTTF, and BCCBC. After evaluating several commonly used linear models, a multi-locus mixed model provided the maximum statistical power and improved the identification of loci with direct breeding application. Ten high-confidence resistance loci were identified, including SNP markers linked to Sr8a, Sr9h, Sr28, and Sr31, and at least three newly discovered resistance loci that are strong candidates for introgression into modern cultivars. In the present study, we assessed the power of multi-locus association mapping while providing an in-depth analysis for its practical ability to assist breeders with the introgression of rare alleles into elite varieties.