Wednesday, August 6, 2008: 8:00 AM
101 A , Midwest Airlines Center
Background/Question/Methods:
The ecological factors driving evolutionary shifts between wind and insect pollination in the seed plants are not yet well understood. Targeted pollen transfer may make pollination by insects more efficient than pollination by wind in sparse plant populations, but suitable study systems for comparison are rare. The vast majority of species in the family Cyperaceae are wind-pollinated, reflecting the ancestral condition. However, several species with brightly coloured flowers are presumed to have shifted to insect pollination. We first tested the hypothesis that one of these, Cyperus obtusiflorus (white, scented inflorescences) is indeed pollinated more effectively by insects than by wind: in the coastal grasslands of KwaZulu-Natal, South Africa, we covered inflorescences in bags that excluded wind- and insect-borne pollen; covered others in bags that excluded insect- but not wind-borne pollen; and left others unbagged. We used the co-flowering, apparently wind-pollinated Pycreus oakfortensis (brown, unscented inflorescences; phylogenetically nested within Cyperus) as a positive control. We then tested the hypothesis that seed set declines less sharply with decreasing density in the putatively insect-pollinated species than in the wind-pollinated one (i.e., that the insect-pollinated species is less susceptible to an Allee effect).
Results/Conclusions:
Inflorescences of C. obtusiflorus from which insects, but not wind, were excluded showed a significant 98 % reduction in seed set compared to open pollinated controls, whereas flowers of P. oakfortensis set seed similarly in these treatments: insect pollination was indeed more important than wind pollination in C. obtusiflorus. This is the first experimental evidence for a shift to insect pollination in Cyperaceae. Bees, beetles and flies visited inflorescences of C. obtusiflorus more often than those of P. oakfortensis. Colour was more important than scent in attracting these insects to artificial inflorescences: white artificial inflorescences attracted more insects than brown ones, but those mimicking the scent of C. obtusiflorus with a blend of ocimene, linalool, benzyl alcohol and phenylethyl alcohol were no more attractive than solvent controls. A wind tunnel transported pollen of P. oakfortensis more effectively than pollen of C. obtusiflorus, suggesting a tradeoff between effectiveness of insect pollination and effectiveness of wind pollination. Seed set of C. obtusiflorus was lower than that of P. oakfortensis at high density, but this reversed at low density: as predicted, seed set declined significantly less steeply with decreasing density in the insect-pollinated taxon. Thus, insect pollination appeared to weaken the Allee effect, potentially making insect pollination selectively advantageous for plants growing at low densities.
The ecological factors driving evolutionary shifts between wind and insect pollination in the seed plants are not yet well understood. Targeted pollen transfer may make pollination by insects more efficient than pollination by wind in sparse plant populations, but suitable study systems for comparison are rare. The vast majority of species in the family Cyperaceae are wind-pollinated, reflecting the ancestral condition. However, several species with brightly coloured flowers are presumed to have shifted to insect pollination. We first tested the hypothesis that one of these, Cyperus obtusiflorus (white, scented inflorescences) is indeed pollinated more effectively by insects than by wind: in the coastal grasslands of KwaZulu-Natal, South Africa, we covered inflorescences in bags that excluded wind- and insect-borne pollen; covered others in bags that excluded insect- but not wind-borne pollen; and left others unbagged. We used the co-flowering, apparently wind-pollinated Pycreus oakfortensis (brown, unscented inflorescences; phylogenetically nested within Cyperus) as a positive control. We then tested the hypothesis that seed set declines less sharply with decreasing density in the putatively insect-pollinated species than in the wind-pollinated one (i.e., that the insect-pollinated species is less susceptible to an Allee effect).
Results/Conclusions:
Inflorescences of C. obtusiflorus from which insects, but not wind, were excluded showed a significant 98 % reduction in seed set compared to open pollinated controls, whereas flowers of P. oakfortensis set seed similarly in these treatments: insect pollination was indeed more important than wind pollination in C. obtusiflorus. This is the first experimental evidence for a shift to insect pollination in Cyperaceae. Bees, beetles and flies visited inflorescences of C. obtusiflorus more often than those of P. oakfortensis. Colour was more important than scent in attracting these insects to artificial inflorescences: white artificial inflorescences attracted more insects than brown ones, but those mimicking the scent of C. obtusiflorus with a blend of ocimene, linalool, benzyl alcohol and phenylethyl alcohol were no more attractive than solvent controls. A wind tunnel transported pollen of P. oakfortensis more effectively than pollen of C. obtusiflorus, suggesting a tradeoff between effectiveness of insect pollination and effectiveness of wind pollination. Seed set of C. obtusiflorus was lower than that of P. oakfortensis at high density, but this reversed at low density: as predicted, seed set declined significantly less steeply with decreasing density in the insect-pollinated taxon. Thus, insect pollination appeared to weaken the Allee effect, potentially making insect pollination selectively advantageous for plants growing at low densities.
