Reproduction of Echinacea angustifolia in fragmented prairie is pollen-limited but not pollinator-limited

Pollen limitation of plant reproduction occurs in many plant species, particularly those in fragmented habitat; however, causes of pollen limitation are often unknown. We investigated the relationship between pollen limitation and pollinator visitation in the purple coneflower, Echinacea angustifolia (Asteraceae), which grows in the extremely fragmented tallgrass prairie of North America. Previous investigations showed that pollen limitation of E. angustifolia increases with plant isolation and decreases with population size. We observed insect visitation to E. angustifolia over two flowering seasons and estimated pollen limitation of observed plants, using seed set as a proxy measure in 2004 and persistence of receptive style rows in 2005. We analyzed spatial patterns of bee visitation and pollination at two spatial scales: individual isolation, as measured by the distance to their fcth nearest flowering neighbors (k = 1-15), and population size. Our results indicate that E. angustifolia is pollinated by over 26 species of native bees, with 70-75% of visits by halictid bees. Surprisingly, in both years, bee visitation increased with isolation of individual plants and did not vary significantly with population size. As expected, plant isolation increased pollen limitation and lowered seed set. There was no effect of population size on seed set in 2004, and pollen limitation decreased nonsignificantly with population size in 2005. We conclude that pollen receipt limits reproduction in E. angustifolia, but pollinator visitation does not. Remarkably, isolated plants simultaneously have increased rates of pollinator visitation by pollinators and decreased reproduction, We discuss alternative explanations of pollen limitation that are consistent with this apparent discrepancy, including a decline in the availability of compatible conspecific pollen with increased plant isolation.