Genomic Selection for Ascochyta Blight Resistance in Pea

Area of Science:

• Plant breeding and genetics
• Agricultural science
• Quantitative genetics

Background:

• Ascochyta blight resistance in pea (Pisum sativum L.) is crucial but challenging to assess due to environmental influences and pathogen variability.
• Genomic selection (GS) offers a promising approach for predicting complex traits like disease resistance in plants.

Purpose of the Study:

• To evaluate the efficacy of genomic selection (GS) for predicting ascochyta blight resistance in pea.
• To compare different GS models, data quality thresholds, and methods for integrating multi-environment trial data.

Main Methods:

• Genotyping-by-sequencing was used to acquire nucleotide polymorphism data.
• Cross-validation was employed to assess prediction accuracy across various GS models, including genomic best linear unbiased prediction (GBLUP) and Bayesian Reproducing Kernel Hilbert Spaces regression (RKHS).
• Marker × environment interactions and different data integration strategies (bivariate, spatial, single-stage) were evaluated.

Main Results:

• The highest prediction accuracy for ascochyta blight disease score (ASC) was 0.56, achieved with GBLUP using mean ASC values and a 70% data quality threshold.
• GBLUP and Bayesian RKHS models showed slightly superior performance compared to other models.
• Missing data thresholds had minimal impact on prediction accuracy, while marker × environment interactions improved cross-environment predictions.

Conclusions:

• Genomic selection is a valuable tool for pea breeding programs aiming to enhance ascochyta blight resistance.
• GS can accurately predict breeding values for un-phenotyped lines and improve estimates for phenotyped lines, facilitating the development of resistant pea cultivars.