The common bean is a tropical facultative short-day legume that is now grown in tropical and temperate zones. This observation underscores how domestication and modern breeding can change the adaptive phenology of a species. A key adaptive trait is the optimal timing of the transition from the vegetative to the reproductive stage. This trait is responsive to genetically controlled signal transduction pathways and local climatic cues. A comprehensive characterization of this trait can be started by assessing the quantitative contribution of the genetic and environmental factors, and their interactions. This study aimed to locate significant QTL (G) and environmental (E) factors controlling time-to-flower in the common bean, and to identify and measure G × E interactions. Phenotypic data were collected from a biparental [Andean × Mesoamerican] recombinant inbred population (F11:14, 188 genotypes) grown at five environmentally distinct sites. QTL analysis using a dense linkage map revealed 12 QTL, five of which showed significant interactions with the environment. Dissection of G × E interactions using a linear mixed-effect model revealed that temperature, solar radiation, and photoperiod play major roles in controlling common bean flowering time directly, and indirectly by modifying the effect of certain QTL. The model predicts flowering time across five sites with an adjusted r-square of 0.89 and root-mean square error of 2.52 d. The model provides the means to disentangle the environmental dependencies of complex traits, and presents an opportunity to identify in silico QTL allele combinations that could yield desired phenotypes under different climatic conditions.

Predicting the phenotype from the genotype is one of the major contemporary challenges in biology. This challenge is greater in plants because their development occurs mostly post-embryonically under diurnal and seasonal environmental fluctuations. Current phenotype prediction models do not adequately capture all of these fluctuations or effectively use genotype information. Instead, we have developed a dynamic modular approach that captures the genotype, environment, and Genotype-by-Environment effects to express the time-to-flowering phenotype in real time in Phaseolus vulgaris. The module we describe can be applied to different plant processes and can gradually replace processes in existing crop models. Our model can enable accelerated progress in diverse breeding programs, particularly with the prospects of climate change. Finally, a gene-based simulation model can assist policy decision makers in matters pertaining to prediction of food supplies.