FITNESS, POLLEN DISPERSAL, AND FLORAL-DISPLAY SIZE EVOLUTION IN ANIMAL-POLLINATED PLANTS
Howard, Aaron Frederick
Barrows, Edward M.
The great variation in floral-display size (the number of flowers per plant) (FDS) in the thousands of angiosperm species has intrigued biologists and others for centuries, eliciting investigations regarding the ecology, evolution, and development of this variation. Many studies have reported a strong relationship between FDS and reproductive success that indicates there can be considerable natural selection on FDS. My studies continued the exploration of FDS evolution by addressing how pollinators and their interactions with floral displays, from the inflorescence through the population level, affect plant reproduction and the consequent evolution of FDS. To address these questions I used computer simulations and the clonal eudicots Asclepias syriaca (Common Milkweed) and Asclepias incarnata (Swamp Milkweed) as model organisms.I found that the insect pollinator taxa (Apis mellifera, Bombus spp., and a suite of butterflies and moths) of Asclepias syriaca can differ significantly in the proportions of self-pollen that they deposit on its flowers. Further, these rates differ with FDS. Using them, I estimated selection on FDS through plant male and female fitness and found that different pollinator taxa may exhibit differing selection pressures on inflorescence size.The spatial arrangement of a plant's flowers (e.g., inflorescence size versus total number of flowers on a plant) can also influence plant fitness. In Asclepias incarnata, I found that display size, from the patch level through the inflorescence level, affects the magnitude of fruit production and pollen dispersal and that there is a significant trade-off between the size of inflorescences and the number of inflorescences on a stem.Angiosperm clone size varies from a few through thousands of ramets (stems). Clones can sport large floral displays that can increase their rates of geitonogamy (self-pollination). I used an individual-based simulation of pollen dispersal that accounted for the spatial arrangement of clonal plants and found that clone size, radius, and the distribution of ramets (stems) within clones all statistically interacted, and significantly influenced geitonogamy.In conclusion, my studies add to the understanding of pollinators and their interactions with FDS and illustrate the complexities and interdependencies of many of the factors involved in floral-display evolution.