Affiliations: 1) Institute for Plant Sciences, University of Cologne; 2) Cluster of Excellence on Plant Sciences, University of Cologne
Keywords: Speciation & hybridization
Colors fulfill important functions that allow populations to adapt to their environment. In addition to their obvious functions, e.g., camouflage, pollinator attraction and UV protection, colorants can alter physiological traits in plants. Hence, when observing color changes as result of adaptation, the actual target trait of selection remains often unclear. The domestication of plants is a well-suited model to study plant adaptation and to identify the genetic basis and regulatory networks of adaptation. While most wild plants were domesticated only once, others were recurrently selected in different geographic locations. These species represent ideal models to study how repeatable evolution is and what determines its outcome. We study the ancient grain crop amaranth which has been domesticated three times in Central and South America from one wild ancestor. All three grain species display a distinct seed color compared to wild Amarnathus species. While all wild amaranths have dark seeds, cultivated amaranths have pale seeds. Despite strong gene flow between crop species and wild relatives, the pale seed color has likely been selected independently, on different genetic backgrounds but altering the same genomic regions. We were able to map the genetic control of the seed color adaptation to two genomic regions and identify a MYB transcription factor gene as potential regulator for the seed color change. A long-read transcriptome assembly and differential gene expression analysis identified variation in flavonoid-pathway genes between individuals with pale and dark seeds. Our results link the genomic change in a transcription factor gene to altered expression in color pathway genes and metabolites with multiple physiological functions. We show that white seeds have reduced seed dormancy, which likely increased their fitness in agricultural environment. While the color itself might have been under selection due to human preference, we speculate that the actual trait under selection was seed dormancy. This shows that trait changes can be the result of pleiotropic effects of metabolic networks rather than selection on the observed trait.