Reproductive isolating barriers among co-occuring species of Collinsia

Background/Question/Methods

When closely related species co-occur in sympatry, prezygotic isolating barriers are predicted to minimize the opportunity for heterospecific gene flow between species.  In Collinsia, many species are known to co-occur in sympatry, to co-flower, and to share pollinators.  However, few hybrid offspring are observed in natural populations.  Thus, I examined the degree to which prezygotic isolating barriers minimize the potential for gene flow between two sympatric sister species of Collinsia; C. linearis and C. rattanii.  I examined geographic, phenological, pollinator, and mating-system isolation between C. linearis and C. rattanii.  I quantified the geographic and elevational range overlap across the entire range of both species using herbarium collection data.  In four sympatric sites in southern Oregon, I measured the overlap in flowering phenology.  Pollinator observations were conducted in four sympatric site and two allopatric - C. rattanii sites to look at the extent of pollinator movement between species in sympatry, and to compare the visitation rates of C. rattanii in sympatry and allopatry.  Finally, across 12 sites, I collected seeds from 17 allopatric and sympatric populations of C. rattanii and C. linearis to determine if the timing of autonomous selfing differed between sympatric and allopatric populations of each species.
Results/Conclusions

I found that species overlapped over large portions of their range, with the majority of the geographic and elevational range of C. linearis shared with C. rattanii.  There was almost complete overlap in flowering duration and substantial overlap in peak flowering time between C. linearis and C. rattanii in sympatric sites.  Pollinators moved between species 5% of the time when movements were scored on a per flower basis, and 8% of the time when the movements were scored on a per plant basis.   In general, pollinators visited the larger flowered C. linearis more than the smaller flowered C. rattanii.  However, in allopatry, a greater proportion of available flowers of C. rattanii received pollinator visits, suggesting competition for pollinators in sympatry.  When C. rattanii was sympatric with C. linearis it self-pollinated at a significantly earlier stage, whereas, C. linearis self-pollinated later.  These results suggest that there may be reinforcement of reproductive isolation in sympatry via shifts in the timing of selfing.