The function of floral orientation in bluebells: interactions with pollinators and rain in two species of Mertensia (Boraginaceae)

Aims Pollinators are traditionally considered to be the primary agent of selection on floral traits. However, floral traits may also be under selection from abiotic agents (e.g. rain), which makes considering the relative importance of pollinators and abiotic selective agents on floral traits essential. The functional significance of floral orientation is often ascribed to pollinator attraction, but orientation can also protect reproductive structures from rain. Therefore, a study that incorporates both factors will enhance our understanding of the ecological roles of floral orientation in plant fitness. Mertensia brevistyla and M. fusiformis are herbaceous species that differ in their floral orientations. A series of field and laboratory experiments was used to investigate the adaptive function of floral orientation in these species, particularly with respect to pollinators and rain. Methods We measured and compared floral orientation and visitor assemblages between M. brevistyla and M. fusiformis populations in western Colorado, USA. We manipulated floral stems and conducted a choice experiment with floral visitors, and also compared orientations of pollinator-visited stems with those of unvisited stems in a natural setting. We examined pollinator- and rain-mediated selection on floral orientation by manipulating orientation, conducting supplemental pollinations, applying watering treatments and measuring subsequent seed set. We also experimentally tested the likelihood of rain contact with anthers, and the effect of rainwater on pollen germinability. Important Findings Mertensia brevistyla had a significantly more upright floral orientation than M. fusiformis, and seed set was highest in upright M. brevistyla and in horizontal/pendant M. fusiformis stems, supporting an adaptive function (via female fitness) of the interspecific difference in orientation. However, floral visitor assemblages did not differ significantly between the two species; visitors did not exhibit significant preference for either orientation; and pollinator-mediated selection on orientation was undetectable. Similarly, there was little effect of water on seed set in either species, regardless of floral orientation. However, pollen germinability was reduced in both species by immersion in water; and water was more likely to contact anthers in M. fusiformis than in M. brevistyla, due to interspecific differences in floral morphology. We conclude that pollinators are likely not the primary selective agent driving differences in orientation in these Mertensia species. Instead, the negative effect of rain on pollen germinability helps explain the more pendant orientation of M. fusiformis, while short anthers in more upright M. brevistyla provide an alternative adaptation to rain. The selective agent driving effects of orientation on seed set remains unclear. This study illustrates the necessity of considering male fitness and abiotic agents in interpreting the functional significance of inflorescence traits.