Integrating envirotyping and genetic modeling to dissect crossover G×E interaction and stability in maize
Marcos Antônio de Godoy Filho, Michael Keith Butterfield, Maurício dos Santos Araújo, José Tiago Barroso Chagas and José Baldin Pinheiro
Abstract: Genotype × environment (G×E) interaction is a primary challenge in plant breeding. This study integrates envirotyping with genetic modeling to dissect G×E for maize yield using the Genomes to Fields dataset. We utilized PaCMAP dimensionality reduction to classify 272 environments into four target populations of environments, which segregated into two mega-environments (Southern and Northern U.S.). A strong crossover interaction was detected between these mega-environments, resulting in low performance stability (quantified by the S index), primarily driven by an inverse relationship between growing degree days and silking, as well as yield: longer-cycle hybrids were advantageous in the South, while shorter-cycle hybrids excelled in the North. Our findings provide quantitative evidence that this phenological response, linked to photoperiod sensitivity, is a key mechanism underlying the G×E interaction. This framework offers a robust tool for defining breeding targets and accelerating genetic gain.