Genomic selection to accelerate breeding for complex stem rust resistance

Genomic selection (GS) is a relatively new breeding technology which aims to predict breeding values of un-observed individuals using prediction models trained with existing phenotypic observations and genome-wide markers. GS is particularly suited for quantitative trait breeding, where it can lead to more rapid population improvement over cycles of selection by enabling selection to occur before phenotypic evaluations are conducted. We are currently implementing recurrent selection for complex stem rust resistance to race lineage Ug99 in two elite spring wheat populations: one selected using GS, another selected using phenotypic selection (PS). Although selection accuracy using GS may be less than selection accuracy using PS, we can complete 3 cycles of GS vs. 2 cycles of PS over 2 years. Thus, given our GS selection accuracy is at least 2/3 the selection accuracy of PS, we can achieve equal or greater gain from selection per unit of time using GS. With the recent emphasis on complex stem rust resistance as a longer-term solution, we hope our work provides a basis for the implementation of GS methodology as a way to gain the upper hand in the plant-pathogen arms race.