GxE and changes to the genetic architecture of flowering time in an overwintering environment

Characterizing how the expression of genetic variation changes across environments is essential for understanding evolution in complex environments, the evolution of phenotypic plasticity, and the genetic architecture of adaptation. We investigated how the phenotypic distribution, large-effect loci, and polygenetic architecture of flowering time in Arabidopsis thaliana changes when grown in an ecologically realistic environment compared to standard chamber conditions. We grew 479 A. thaliana accessions in an overwintering common garden that have previously been phenotyped in growth chambers by the 1001 Genomes Project (2016). Our findings suggest there is extensive GxE for flowering time between the overwintering common garden and the growth chamber environments. None of the large effect loci that we identified using a multivariate Genome Wide Association Study (GWAS) were shared across growth conditions. We find evidence of allelic differential sensitivity, conditional neutrality, and antagonistic pleiotropy contributing to GxE for flowering time between growth conditions. Our results also suggest that the effects of genetic variation across multiple flowering time regulatory pathways is revealed when grown in a more ecologically realistic environment. The effect sizes for the most significant 1% of SNPs are moderately correlated between the outdoor and growth chamber conditions which indicates a partially shared genetic basis. Overall, our findings highlight how making inferences about the genetic basis and evolution of complex traits may be dependent on conducting experiments in ecologically realistic environments.