Dissecting the contribution of environmental influences, plant phenology, and disease resistance to improving genomic predictions for fusarium head blight resistance in wheat

Abstract

Environmental factors like temperature and humidity are presumed to greatly influence Fusarium head blight FHB infections in wheat. Anther retention AR, on the other hand, is a morphologically neutral trait that shares a common genetic basis with FHB resistance. In this study, our aims were to: (i) Evaluate two types of corrections of FHB severity scores, namely method-1 via linear regression on flowering time (FT), and method-2 via a best-subset multiple linear regression analysis comprising FT plus accumulated thermal time variables; and (ii) assess the performance of multi-trait genomic selection (MT.GS) models for FHB severity assisted by AR. The forward prediction scenarios where GS models were trained with data from the previous years revealed average prediction accuracies (PA) of 0.28, 0.33, and 0.36 for FHB severity scores that were uncorrected or corrected by method-1 and method-2, respectively. FHB severity scores free from the influences of both environment and phenology seemed to be the most efficient trait to be predicted across different seasons. Average PA increments up to 1.9-fold were furthermore obtained for the MT.GS models, evidencing the feasibility of using AR as an assisting trait to improve the genomic selection of FHB resistance breeding lines.