Genomic prediction for grain yield and biotic stress resistance in field pea (Pisum sativum L.)

Field pea ( Pisum sativum L.) is a nutritionally important pulse crop that contributes to food security, sustainable cropping systems, and the growing demand for plant-based proteins. However, genetic gain for complex traits such as grain yield and disease resistance remains limited under conventional breeding, particularly in the face of climate change and evolving biotic stresses. Genomic selection (GS) is a promising approach to accelerate genetic improvement, yet its large-scale evaluation in large-scale breeding programs has been limited. Here, we present the first comprehensive assessment of GS in the Australian National Field Pea Breeding Program, using a decade (2013–2022) of multi-environment data from 3,199 advanced lines and cultivars. Six key traits were analyzed, including grain yield (GY) and resistance to major diseases, including ascochyta blight, bacterial blight, downy mildew, pea seed-borne mosaic virus (PSbMV), and bean leaf roll virus (BLRV). Lines were genotyped using a multispecies Pulse 30K SNP array. Genomic prediction was evaluated using GBLUP models fitted with and without genotype × environment (G × E) interactions, as well as bivariate models exploiting genetic correlations between traits. Across traits and models, prediction accuracy ranged from 0.21 to 0.72. Including G × E interactions increased GY prediction accuracy by 3.03%, while bivariate models provided moderate additional gains by leveraging correlations with disease resistance traits. Overall, our results demonstrate that GS can be integrated effectively into a field pea breeding program to enhance disease resistance and stabilize yields across environments.