95th ESA Annual Meeting (August 1 -- 6, 2010)

COS 102-2 - Mating-system evolution can change the ability of a population to respond to selection

Thursday, August 5, 2010: 1:50 PM
411, David L Lawrence Convention Center
David E. Carr, Blandy Experimental Farm, University of Virginia, Boyce, VA, Terry A. Junek, Department of Entomology, Texas A&M University, College Station, TX and Micky D. Eubanks, Entomology, Texas A&M University, College Station, TX
Background/Question/Methods

The transition to self-pollination from cross-pollinating ancestors is one of the most widespread evolutionary phenomena in angiosperms. The evolution of selfing mating-systems will have important effects on the genetics of the population in transition, partitioning relatively more of the genetic variation among individuals, fixing genetic variation within individuals, and thereby allowing non-additive genetic variation (dominance and epistasis) to play a bigger role in affecting phenotypic variation.  The response of populations undergoing mating-system evolution toward selfing may be constrained or enhanced relative to the response to selection in a randomly breeding population.  We examined greenhouse populations of Mimulus guttatus that were under artificial selection for either high rates of selfing or high rates of outcrossing and tested for differences in their responses.  Stigma-anther separation was under direct selection, with the more selfing populations being selected for decreased separation and more outcrossing populations being selected for increased separation.  Simultaneously we imposed selection on an important anti-herbivore trait – trichome density.  Half of the populations evolving toward a high rate of selfing experienced selection for increased trichome density, and the other half experienced selection for reduced density.  Similarly, populations evolving toward high outcrossing experienced selection for either increased or decreased trichome density.

Results/Conclusions

We examined three base populations of Mimulus guttatus, one (SV) from Snell Valley, CA, (RH) from the Red Hills Recreation Area and the other (DP) from the Don Pedro Reservoir in Tuolumne County, CA.  After three generations of selection, traits in all populations differed significantly from unselected controls.  The response of stigma-anther was largely symmetrical, with populations selected for increased or decreased stigma-anther separation showing approximately equal divergence from control lines.  The response of populations to selection on trichome density, however, showed a strong dependence on whether populations were evolving toward selfing or outcrossing.  In population SV, the increase in trichome density of selfing populations was significantly reduced relative to the response seen in outcrossing populations.  In populations RH and DP, the decrease in trichome density was significantly enhanced in selfing populations relative to the response seen in outcrossing populations.  These results suggest that mating-system evolution can potentially constrain or enhance the response to selection but that these effects are either population-specific or are specific to the starting conditions in populations.