Gene-based model to predict heading date in wheat based on allelic characterization and environmental drivers

Abstract

While numerous wheat phenology prediction models are available, most of them are constrained to using variety-dependent coefficients. The overarching objective of this study was to calibrate a gene-based model to predict wheat heading date that allows breeders to select specific gene combinations that would head within the optimal window for a given environment independently of varietal genetic background. A dataset with a total of 49 Argentine wheat cultivars and two recombinant inbred lines was chosen to cover a wide range of allelic combinations for major vernalization, photoperiod, and earliness per se genes. The model was validated using independent data from an Argentine wheat trial network that includes sites from a wide latitudinal range. Ultimately, using this gene-based model, simulations were made to identify optimal gene combinations (ideotypes) × site combinations in contrasting locations. The selected model accurately predicted heading date with an overall median error of 4.6 d. This gene-based crop model for wheat phenology allowed the identification of groups of gene combinations predicted to produce heads within a low-risk window and can be adapted to predict other phenological stages based on accessible climatic information and publicly available molecular markers, facilitating its adoption in wheat-growing regions worldwide.