Targeted breeding for drought-prone environments

Key components of targeted breeding for drought-prone environments are discussed. The "Breeder's Equation" provides a framework to operationalise technologies that improve predictive breeding and accelerate genetic gain. Multi-Environment Trials (METs) are used to evaluate genotypes and construct training data sets for genomic prediction. However, to ensure the success of targeted breeding, they must include the high frequency agricultural drought environments that occur within the on-farm Target Population of Environments (TPEs). Enviromics and Phenomics technologies have advanced and can be applied to characterise the key agricultural drought environment-types, including distinguishing between vegetative, flowering and post-flowering water-deficits that expose contributions of traits to crop productivity. These design considerations collectively are referred to as MET-TPE alignment. Targeted breeding strategies are discussed using forms of the Breeder's Equation that include a term to quantify the trait genetic correlation between METs and the TPE as a measure of MET-TPE alignment. This extended Breeder's equation separates within-MET prediction accuracy from the MET→TPE transfer of prediction and selection outcomes. With improved MET-TPE alignment improved transfer of genetic gain to the on-farm TPE can be achieved. Conversely, the transfer diminishes with poor MET-TPE alignment and can fail for many scenarios even when within-MET cross-validation for genomic prediction looks good.