Microbial mediation of pollinator performance

Background/Question/Methods - Floral nectar, a resource important for mediating plant-pollinator interactions, is often colonized by microorganisms. Their presence and metabolic activity is capable of altering plant-pollinator interactions through changes in nectar traits. Effects of nectar-inhabiting microorganisms (NIMs) may not be limited to plant-pollinator interactions and plant fitness however. As consumers of NIMs and NIM-modified nectar, the effects of NIMs may extend to pollinator performance, as variation in resource quality and quantity can affect pollinator behavior and reproduction. Using both choice and no-choice assays, we tested this hypothesis by manipulating the presence and activity of the common NIM Metschnikowia reukaufii in artificial nectar and measuring its effects on the foraging behavior and reproduction of the common eastern bumble bee, Bombus impatiens. To assess effects on behavior, we presented foragers with different levels of foraging experience mono- or di-chromatic arrays of flowers containing nectar with or without NIMs and measured preference and foraging time. To assess effects on reproduction, we fed microcolonies diets varying in NIM presence and viability and/or high and low pollen availability. By manipulating pollen (protein) availability, we could test the costs and benefits of consuming NIMs and NIM-modified nectar when resources were limiting. 

Results/Conclusions - We found that NIMs altered pollinator preference but not performance. Yeasts influenced the behavior of both naïve and trained B. impatiens foragers. Nectar yeasts may serve as an honest signal for the presence of nectar in artificial flowers, as naïve bees foraged on yeast-inoculated flowers more frequently during initial visits to monochromatic arrays containing either yeast- and no-yeast flowers. Trained foragers also responded positively to yeasts, making a greater proportion of visits to yeast-inoculated flowers vs. controls, suggesting a preference for yeast. In our no-choice performance assay, B. impatiens microcolonies actively utilized artificial diet solutions containing yeasts. Preliminary analyses however suggest that pollen availability may be a more important determinant of reproductive success than yeast. Microcolonies with low-pollen availability had reduced fecundity in comparison to the high pollen-availability treatment, and we failed to detect a significant main or interactive effect of yeast treatment. Taken together, our results suggest that yeasts may serve as honest signals for the presence of floral nectar with important consequences for plant fitness, but consumption of yeasts and yeast-modified nectar does not have direct or indirect effects on pollinator reproduction.