Independent recruitment of FRUITFULL-like transcription factors in the convergent origins of vernalization-responsive grass flowering.

Flowering in response to low temperatures (vernalization) has evolved multiple times independently across angiosperms as an adaptation to match reproductive development with the short growing season of temperate habitats. Despite the context of a generally conserved flowering time network, evidence suggests that the genes underlying vernalization responsiveness are distinct across major plant clades. Whether different or similar mechanisms underlie vernalization-induced flowering at narrower (e.g., family-level) phylogenetic scales is not well understood. To test the hypothesis that vernalization responsiveness has evolved convergently in temperate species of the grass family (Poaceae), we carried out flowering time experiments with and without vernalization in several representative species from different subfamilies. We then determined the likelihood that vernalization responsiveness evolved through parallel mechanisms by quantifying the response of Pooideae vernalization pathway FRUITFULL (FUL)-like genes to extended periods of cold. Our results demonstrate that vernalization-induced flowering has evolved multiple times independently in at least five grass subfamilies, and that different combinations of FUL-like genes have been recruited to this pathway on several occasions.

Major niche transitions in Pooideae correlate with variation in photoperiodic flowering and evolution of CCT domain genes.

The external cues that trigger timely flowering vary greatly across tropical and temperate plant taxa, the latter relying on predictable seasonal fluctuations in temperature and photoperiod. In the grass family (Poaceae) for example, species of the subfamily Pooideae have become specialists of the northern temperate hemisphere, generating the hypothesis that their progenitor evolved a flowering response to long days from a short-day or day-neutral ancestor. Sampling across the Pooideae, we found support for this hypothesis, and identified several secondary shifts to day-neutral flowering and one to short-day flowering in a tropical highland clade. To explain the proximate mechanisms for the secondary transition back to short-day-regulated flowering, we investigated the expression of CCT domain genes, some of which are known to repress flowering in cereal grasses under specific photoperiods. We found a shift in CONSTANS 1 and CONSTANS 9 expression that coincides with the derived short-day photoperiodism of our exemplar species Nassella pubiflora. This sets up the testable hypothesis that trans- or cis-regulatory elements of these CCT domain genes were the targets of selection for major niche shifts in Pooideae grasses.