Iridoid glycosides in floral nectar alter pollinator behavior and pollen transfer dynamics

Background/Question/Methods

Plants produce an array of chemical compounds that may be deterrent or toxic to herbivores, termed plant secondary metabolites. These compounds are also present in plant parts attractive to mutualists, including floral nectar. The presence of secondary metabolites in nectar could be an unavoidable consequence of plant chemical defense, or these compounds could have functional roles in plant reproduction. For example, they could deter consumers that do not pollinate, such as microbes or nectar robbers. In addition, they could be attractive to pollinators, especially if consumption confers some benefit to pollinators. A variety of secondary metabolites, including iridoid glycosides, are known to reduce parasite infection of bumble bee pollinators, but it is not known whether bees self-medicate by tailoring their foraging behavior to the presence of these compounds in nectar. We measured concentrations of two iridoid glycosides, aucubin and catalpol, in nectar and other tissues of a bee-pollinated, obligately outcrossing plant, Chelone glabra (Plantaginaceae). We then manipulated nectar iridoids to create flowers with low and high iridoid glycoside concentrations relative to population means. We recorded visits to these flowers by healthy and parasitized bees and estimated pollen export to stigmas of other plants using powdered fluorescent dyes as pollen analogues.

Results/Conclusions

Aucubin and catalpol concentrations were variable among sites and differed strongly between plant parts. Levels of aucubin and catalpol in nectar were not correlated with those of either compound in other tissues, suggesting that plants may control expression of these compounds in floral tissues separately from leaves. Bees foraged longer at flowers with the high iridoid glycoside concentrations, exporting significantly more dye to conspecifics than when they visited flowers with low iridoid glycoside concentrations. After visiting flowers in the high iridoid glycoside treatment, bees were more likely to fly to other plants in the population rather than visit second flowers, which could increase outcrossing success of plants. Bees with intestinal parasites (Crithidia bombi; Trypanosomatidae) and parasitoid fly larvae (Physocephala spp.; Conopidae) made significantly shorter flower visits than unparasitized bees, especially when both parasites were present. However, parasitism was not associated with preference for or deterrence from nectar iridoid glycosides, providing no evidence for self-medication behavior in the wild. Taken together, these results demonstrate that nectar SMs can manipulate pollinator behavior and subsequent pollen transfer.