The evolution of specialized floral phenotypes in a heterogeneous pollination environment: Fitness trade-offs in a Dudleya (Crassulaceae) species complex

The assumption that flowers readily evolve specializations for pollination by particular animals has been central to a standard view of pollinator-mediated adaptive divergence in angiosperms. Stebbins' Most Effective Pollinator Principle formalized this assumption in proposing that a plant should evolve specializations to its most effective pollinator. I develop an optimality model which predicts that plants may evolve adaptations to relatively ineffective pollinators when doing so entails little loss in pollination service from more effective ones. In this model, the sole criterion for evolving specialization is that the fitness gain exceeds losses from reduced service by other pollinators. The potential for the evolution of floral divergence depends on the form of pollinator-mediated fitness trade-offs, and not, as has been argued, on differences in the mean effectiveness of different pollinators.; I investigated trade-offs in pollination success in a complex of Dudleya species that is generalized for pollination. To test whether trade-offs can lead to adaptive floral divergence among geographically isolated populations, I used arrays of inflorescences to measure pollination success as a function of floral traits at four sites in Southern California. Pollinator assemblages varied among sites and between years, but flowers that were longer, produced more nectar, and had more inserted reproductive parts received more pollen at all sites. Pooling arrays with functionally similar pollinators revealed that the marginal performance gain of longer flowers increased as pollinators shifted from small bees and flies, to large bees, to hummingbirds, but trade-offs were absent because longer flowers performed best in all environments.; To test the prediction that hummingbird pollination should favor narrower flowers than bumblebee pollination and to measure trade-offs more precisely I manipulated flower width and pollination environment of D. greenei in a controlled setting. When hummingbirds were the sole pollinator, narrow flowers performed best by most measures, but pollination success did not depend strongly on flower width in a pure “bumblebee environment.” Together, the two studies indicate that hummingbirds are strong agents of natural selection on several aspects of floral phenotype, but that adaptation to hummingbirds entails little loss in pollination service from insects within these species of Dudleya.