Increased Prediction Accuracy Using Combined Genomic Information and Physiological Traits in A Soft Wheat Panel...

Area of Science:

• Plant Breeding
• Quantitative Genetics
• Agronomy

Background:

• Improving genetic gain for complex traits like grain yield in wheat is crucial for food security.
• Genomic selection models are powerful but can be enhanced by integrating field-based physiological data.

Purpose of the Study:

• To evaluate the predictive accuracy of multi-kernel physiological and genomic best linear unbiased prediction (BLUP) models compared to single-kernel models for soft wheat grain yield.
• To determine if integrating physiological traits and genotype-environment interactions improves prediction accuracy.

Main Methods:

• Compared multi-kernel BLUP models (physiological + genomic) against single-kernel BLUP models (physiological or genomic only) for grain yield prediction.
• Utilized field-based physiological data (canopy temperature, SPAD, membrane stability, senescence, NDVI) and SNP data (19,353 SNPs) from a soft wheat population across four environments.
• Assessed prediction accuracy and response to selection.

Main Results:

• Multi-kernel models incorporating physiological traits and/or interaction effects significantly improved grain yield prediction accuracy (35-169% increase) compared to genomic-only models when heading date was a covariate.
• Physiological and genotype-environment interaction effects captured a higher response to selection.
• Prediction accuracy improved across four environments when physiological data was integrated.

Conclusions:

• Integrating field-based physiological data with genomic data enhances prediction accuracy for soft wheat grain yield in multi-environment settings.
• This integrated approach holds significant potential for increasing genetic gain in wheat breeding programs.
• Multi-kernel BLUP models offer a promising strategy for optimizing selection for complex yield traits.