The potential for floral evolution in response to competing selection pressures following the loss of hawkmoth pollination in Ruellia humilis
John S Heywood,Joseph S Michalski,Braden K McCann,Kara J Andres,Allison R Hall,Amber D Hartman,Tessa C Middleton,Amelia Chiles,Sarah E Dewey,Cay A Miller
DOI: https://doi.org/10.1002/ajb2.16063
Abstract:Premise: In the absence of hawkmoth pollinators, chasmogamous (CH) flowers of Ruellia humilis self-pollinate by two secondary mechanisms. Other floral visitors might exert selection on CH floral traits to restore outcrossing, but at the same time preferential predation of CH seeds generates selection to increase the allocation of resources to cleistogamous (CL) flowers. Methods: To assess the potential for an evolutionary response to these competing selection pressures, we estimated additive genetic variances ( σ A 2 ${\sigma }_{{\rm{A}}}^{2}$ ) and covariances for 14 reproductive traits and three fitness components in a Missouri population lacking hawkmoth pollinators. Results: We found significant σ A 2 ${\sigma }_{{\rm{A}}}^{2}$ for all 11 floral traits and two measures of resource allocation to CL flowers, indicating the potential for a short-term response to selection on most reproductive traits. Selection generated by seed predators is predicted to increase the percentage of CL flowers by 0.24% per generation, and mean stigma-anther separation is predicted to decrease as a correlated response, increasing the fraction of plants that engage in prior selfing. However, the initial response to this selection is opposed by strong directional dominance. Conclusions: The predicted evolutionary decrease in the number of CH flowers available for potential outcrossing, combined with the apparent preclusion of potential diurnal pollinators by the pollen-harvesting activities of sweat bees, suggest that 100% cleistogamy is the likely outcome of evolution in the absence of hawkmoths. However, rare mutations with large effects, such as delaying budbreak until after sunrise, could provide pathways for the restoration of outcrossing that are not reachable by gradual quantitative-genetic evolution.
What problem does this paper attempt to address?