Data from: The timing of molecular and morphological changes underlying reproductive transitions in wild tomatoes (Solanum sect. Lycopersicon)
Vosters, Stacey L. et al. (2014), Data from: The timing of molecular and morphological changes underlying reproductive transitions in wild tomatoes (Solanum sect. Lycopersicon), Dryad, Dataset, https://doi.org/10.5061/dryad.9540r
Molecular mechanisms underlying the transition from genetic self-incompatibility to self-compatibility are well documented, but the evolution of other reproductive trait changes that accompany shifts in reproductive strategy (mating system) remain comparatively poorly understood. A notable exception is the transition from exserted styles to styles with recessed positions relative to the anthers in wild tomatoes (Solanum Section Lycopersicon). This phenotypic change has been previously attributed to specific mutation in the promoter of a gene that influences style length (style2.1); however, whether this specific regulatory mutation arose concurrently with the transition from long- to short-styles, and whether it is causally responsible for this phenotypic transition, has been poorly investigated across this group. To address this gap, we assessed 74 accessions (populations) from 13 species for quantitative genetic variation in floral and reproductive traits as well as the presence/absence of deletions at two different locations (StyleD1, StyleD2) within the regulatory region upstream of style2.1. We confirmed that the putatively causal deletion variant (a 450bp deletion at StyleD1) arose within self-compatible lineages. However the variation and history of both StyleD1 and StyleD2 was more complex than previously inferred. In particular, although StyleD1 was statistically associated with differences in style length and stigma exsertion across all species, we found no evidence for this association within two species polymorphic for the StyleD1 mutation. We conclude that the previous association detected between phenotypic and molecular differences is most likely due to a phylogenetic association rather than a causal mechanistic relationship. Phenotypic variation in style length must therefore be due to other unexamined linked variants in the style2.1 regulatory region.