Improving genomic prediction for plant disease using environmental covariates

Area of Science:

• Agricultural Science
• Plant Pathology
• Genetics

Background:

• Fusarium Head Blight (FHB) significantly impacts wheat and barley yield and quality.
• Disease resistance is complex, influenced by genetics, environment, and genotype-by-environment interactions (GxE).
• Predicting FHB resistance across diverse environments is challenging for breeding programs.

Purpose of the Study:

• Investigate GxE in spring wheat to improve FHB resistance prediction.
• Evaluate methods for predicting genotype performance in untested environments.
• Identify genetic markers associated with FHB resistance and environmental sensitivity.

Main Methods:

• Compared Finlay-Wilkinson regression (FW), joint-genomic regression analysis (JGRA), and mixed models.
• Incorporated environmental covariates to compute an environment index and relationship matrix.
• Benchmarked against a baseline genomic selection (GS) model without environmental covariates.

Main Results:

• JGRA demonstrated higher accuracy than GS for within- and across-environment predictions.
• Mixed models performed comparably to JGRA for within-environment predictions.
• JGRA identified significant markers for FHB resistance and environmental sensitivity, enabling location-specific breeding value predictions.

Conclusions:

• Incorporating environmental covariates enhances predictive ability for selecting resistant genotypes.
• Leveraging GxE interactions improves disease management strategies in wheat breeding.
• This approach offers a cost-effective method for breeders to exploit GxE for improved FHB resistance.