Evolvability and constraint in the evolution of three‐dimensional flower morphology

Øystein H. Opedal,Laura S. Hildesheim,W. Scott Armbruster
DOI: https://doi.org/10.1002/ajb2.16092
IF: 3.325
2022-11-14
American Journal of Botany
Abstract:Premise Flower phenotypes evolve to attract pollinators and to ensure efficient pollen transfer to and from the bodies of pollinators or, in self‐compatible bisexual flowers, between anthers and stigmas. If functionally interacting traits are genetically correlated, response to selection may be subject to genetic constraints. Genetic constraints can be assessed by quantifying standing genetic variation in (multivariate) phenotypic traits and by asking how much the available variation is reduced under specific assumptions about phenotypic selection on functionally interacting and genetically correlated traits. Methods We evaluated multivariate evolvability and potential genetic constraints underlying the evolution of the three‐dimensional structure of Dalechampia blossoms. First, we used data from a greenhouse crossing design to estimate the G matrix for traits representing the relative positions of male and female sexual organs (anthers and stigmas) and used the G matrix to ask how genetic variation is distributed in multivariate space. To assess the evolutionary importance of genetic constraints, we related standing genetic variation across phenotypic space to evolutionary divergence of population and species in the same phenotypic directions. Results Evolvabilities varied substantially across phenotype space, suggesting that certain traits or trait combinations may be subject to strong genetic constraint. Traits involved functionally in flower‐pollinator fit and autonomous selfing exhibited considerable independent evolutionary potential, but population and species divergence tended to occur in phenotypic directions associated with greater‐than‐average evolvability. Conclusions These results are consistent with the hypothesis that genetic constraints can hamper joint trait evolution towards optimum flower‐pollinator fit and optimum self‐pollination rates.
What problem does this paper attempt to address?