Genomic Prediction of Quantitative Adult Plant Resistance to Stripe Rust in a Winter Wheat Breeding Program
Primary author: Lance Merrick
Co-author(s): Arron Carter; Xianming Chen; Brian Ward
Faculty sponsor: Arron Carter
Primary college/unit: Agricultural, Human and Natural Resource Sciences
Stripe rust (Puccinia striiformis f. sp. tritici) is one of the most damaging diseases of wheat and has resulted in massive reduction in yield and economic losses globally. Quantitative adult plant resistance (APR) is detected in mature plants, associated with non-specific resistance, and considered to be a durable form of resistance. Quantitative APR is controlled by varying numbers of additive resistance alleles and thus is a good candidate for genomic prediction. The goal of this research was to create a genomic prediction model of quantitative stripe rust resistance for advancing early-generation lines to advanced yield trials in a winter wheat breeding program. We created prediction models using breeding lines from four years (2016-2019) and two breeding populations consisting of doubled-haploid and F5 derived lines. The prediction models used genotype-by-sequencing single-nucleotide polymorphism markers for random effects and KASP markers for known resistance genes for fixed effect covariates. Prediction models used were optimized for training population size, marker density, and statistical model to find the most efficient and accurate method to integrate genomic prediction into a breeding program. Genomic prediction will aid the breeding program for the evaluation and selection of stripe rust resistance in years and environments with limited disease incidence and reduce the need for replicated phenotyping. In doing this, genomic prediction will increase the genetic gain for quantitative stripe rust resistance within wheat breeding populations.