Phylogenetic relationships of North American Asclepias L. and the role of pollinators in the evolution of the milkweed inflorescence

Abstract

I utilize a variety of methods, from field studies of selection to phylogenetic systematics, to study the historical processes involved in the evolution of reproductive structures of milkweeds (Asclepias L. (Asclepiadaceae)). In an experimental study of a natural population of Asclepias tuberosa L., I explore the mechanisms by which pollinators select on the arrangement of flowers on a plant. I present an evolutionary model for the manner in which pollinators select on inflorescence design (the number of flowers in inflorescence units), considering reproductive success gained through both male and female function. In this model, the maximum fitness is achieved when proportional changes in male and female reproductive success with changes in inflorescence-unit size are balanced. The model is used to predict optimal inflorescence-unit size in Asclepias tuberosa, which roughly match inflorescence-unit sizes observed in nature. In studies of selection on plant reproductive characters in natural populations, the pollinators responsible for selection remain unidentified or are inferred from floral morphology. In the same experimental population of Asclepias tuberosa in which I studied selection on inflorescence design, I measured the effectiveness of all common pollinators. Although A. tuberosa possesses characteristics of butterfly-pollinated plants, and many butterfly species are found at its flowers, the most effective pollinators are bees. Furthermore, the most effective pollinators in the study population varied between years. These results suggest that inferences about the identity of effective pollinators should be based on careful measurements of effectiveness over several seasons. In order to study factors that affect the evolution of reproductive characters like inflorescence design in this large genus, it is desirable to employ a phylogenetic hypothesis of the relationships among Asclepias species. I present such an analysis based on morphological data. My results suggest that some of the North American species of Asclepias are more closely related to African species in other genera than to other North American species of Asclepias. Also, most infrageneric groups currently recognized are not monophyletic. Several novel infrageneric groupings are well-supported by my results. I also present analyses of the evolution of growth form and flower "horns" based on an explicit phylogenetic hypothesis.