Multifunctional bracts in the dove tree Davidia involucrata (Nyssaceae: Cornales): rain protection and pollinator attraction.

Although there has been much experimental work on floral traits that are under selection from mutualists and antagonists, selection by abiotic environmental factors on flowers has been largely ignored. Here we test whether pollen susceptibility to rain damage could have played a role in the evolution of the reproductive architecture of Davidia involucrata, an endemic in the mountains of western China. Flowers in this tree species lack a perianth and are arranged in capitula surrounded by large (up to 10 cm x 5 cm) bracts that at anthesis turn from green to white, losing their photosynthetic capability. Flowers are nectarless, and pollen grains are presented on the recurved anther walls for 5-7 days. Flower visitors, and likely pollinators, were mainly pollen-collecting bees from the genera Apis, Xylocopa, Halictus, and Lasioglossum. Capitula with natural or white paper bracts attracted significantly more bees per hour than capitula that had their bracts removed or replaced by green paper. Experimental immersion of pollen grains in water resulted in rapid loss of viability, and capitula with bracts lost less pollen to rain than did capitula that had their bracts removed, suggesting that the bracts protect the pollen from rain damage as well as attracting pollinators.

Pollinator response to female and male floral display in a monoecious species and its implications for the evolution of floral dimorphism.

Pollinator-mediated selection has been hypothesized as one cause of size dimorphism between female and male flowers. Flower number, ignored in studies of floral dimorphism, may interact with flower size to affect pollinator selectivity. In the present study, we explored pollinator response, and estimated pollen receipt and removal, in experimental populations of monoecious Sagittaria trifolia, in which plants were manipulated to display three, six, nine or 12 female or male flowers per plant. In this species, female flowers are smaller but have a more compressed flowering period than males, creating larger female floral displays. Overall, pollinators preferred to visit male rather than female displays of the same size. Both first visit per foraging bout and visitation rates to female displays increased with display size. However, large male displays did not show increased attractiveness to pollinators. A predicted relationship that pollen removal, rather than pollen receipt, is limited by pollinator visitation was confirmed in the experimental populations. The results suggest that the lack of selection on large male displays may affect the evolution of floral dimorphism in this species.