No abstract

2000. The radiation of Calochortus: generalist flowers moving through a mosaic of potential pollinators. -Oikos 89: 209-222.Few studies have surveyed the insects visiting flowers at many sites and for many related species. Calochortus, a genus of about 60 species with its center of diversity in California, has flowers with a wide range of appearances. Insects visiting Calochortus flowers in 25 species were censused at 40 sites. Visitors at coflowering plants were also censused at each site. Calochortus flowers attracted insects in several orders, especially beetles that feed at glands on the petals and bees that often collect pollen. Species in Calochortus section Calochortus had a somewhat narrower spectrum of visitors than species in section Mariposa. The diversity of visitor species at Mariposa flowers was often greater than the diversity at coflowering species. On the whole, Calochortus flowers seem to be generalists in terms of their pollinators. Differences in visitor assemblage were greater between Calochortus populations at different sites than between visitors to Calochortus versus coflowering species at a site. Nevertheless, there were always significant differences in the proportions of various insects at Calochortus versus coflowering plants. In addition, proportions of visitors also differed between species of Calochortus at the same site. Thus, Calochortus flowers have diverged in the visitors they attract, and evidently they have done so without permanently specializing (except to a limited degree at the sectional level). What we know of the radiation of Calochortus lilies is consistent with an interpretation of adaptive wandering through a spatiotemporal mosaic of pollinator communities.

Fungus gnats (Sciaridae and Mycetophilidae) are the principal pollinators of Listera cordata (L.) R. Br. (Orchidaceae) and Scoliopus bigelovii Torr. (Liliaceae) in coastal redwood forests of northern California. Although primitive diptera have generally been regarded as relatively inefficient pollinators, fruit set for both species is high: 61–78% for L. cordata (1976–1978) and 94.3–98.5% for S. bigelovii (1978–1979). Since probability of pollination per visit is low, we attribute high fruit set to the large number of gnats present at our study sites and corresponding large number of visits to flowers. The relative frequency of geitonogamous vs. xenogamous pollen flow was estimated by emasculating flowers and subsequently comparing pollen reception or fruit set of emasculates with controls. Results for both species indicate that interplant movement of pollen is common. Thus, fungus gnats can be effective pollen vectors, both in terms of overall fruit set and potential for cross‐pollination.

Bumblebees foraging on vertical inflorescences start near the bottom and work upward, behavior commonly interpreted as a response to the greater amounts of nectar available in lower flowers. Lupinus polyphyllus, which produces no nectar, has more pollen available in upper flowers. Although bees are probably unable to detect this gradient, since pollen is hidden from their view, they still start low and forage upward. Therefore, we concluded that the bees' tendency to forage upward on vertical inflorescences is not tied to a reward gradient. In addition, bees use only about 15% of the flowers per inflorescence, although they could be much more efficient by visiting and revisiting every flower systematically. In general, revisits would not be penalized because most flowers contain enough pollen for several visits. Optimal foraging theory may not offer an adequate explanation for such gross inefficiency.