Helianthus argophyllus is annual plant and an obligate out-crosser. Throughout its native range in Texas the species breeds synchronously, but in a small region along the Gulf of Mexico, flowering can be bimodal. Here inhabit two flowering syndromes that correlate with habitat, forming two ecotypes: one short and early flowering -found mostly in islands- and one tall and late flowering -found throughout the species distribution. Their main genetic difference is a presence/absence polymorphism of a FLOWERING LOCUS T homolog, HaFT1, found within a region of low recombination. The mainland ecotype (late flowering) lacks HaFT1.
Common gardens suggest that the flowering time of H. argophyllus is highly heritable. But how does it affect phenology in natural populations? What are the effects of HaFT1 on flowering phenology (and reproductive isolation) across time in different environments?
To know HaFT1 effects in phenology, we sampled populations across the distribution of HaFT1, in three time points of the reproductive season for two years. Each time, we classified the reproductive stage of each individual. To measure its effects in reproductive isolation, we genotyped 762 plants and their seeds for HaFT1. We compared all results between ecotypes, habitats and time points in the reproductive season.
Results/Conclusions
HaFT1 has a highly restricted geographic distribution within H. argophyllus, only present within 15 km from the Gulf of Mexico, being more frequent in the island habitat than the mainland. Beginning of flowering had a logaritmic relationship with distance to the sea.
Phenology was the result of an interaction between ecotype and habitat. The same genotype flowers differently in each habitat. For example, in all habitats, the island ecotype started and reached its peak of flowering before the mainland ecotype, but in the mainland it continued to flower beyond what we considered the normal reproductive season.
Differences in ecotype abundance result in asymmetrical reproductive isolation (RI). Due to differences in phenology, RI changes between ecotypes, habitats and time. The island ecotype has more conspecific crosses early in the season in the island habitat, and late in the season in the mainland. The mainland ecotype shows the exact opposite pattern, though, its numeric advantage results in always higher RI, less hetero-specific crossing for this ecotype.
We think this is the first time such a small-scale mosaic of phenology and reproductive isolation is documented. And while gene flow could imperil the survival of the island ecotype, plasticity could help maintain it.
