PS 17-242 - Bumble bee olfactory preference and VOC emissions dependent on pollination status of Mimulus flowers

Abstract: Approximately 87% of angiosperms and 75% of all staple crops are pollinated by animals, with the majority of these (60-70%) pollinated by insects. Flowers are typically a source of carbohydrates, proteins, and lipids to pollinators in the form of nectar and pollen, while pollinators transfer pollen to ensure plant reproduction. Flowers provide visual and olfactory signals to attract pollinators, and pollinators may use these signals to evaluate rewards. For olfactory signals, flowers release volatile organic compounds (VOCs) that may correlate with the quantity or quality of rewards in a flower. In this study, we examine whether olfactory signals are altered after a flower has been pollinated. This study was conducted in the laboratory at Blandy Experimental Farm in Clarke County Virginia in summer 2021. We examined the response of pollinators to olfactory signals from Mimulus guttatus and M. lewisii. We used commercial Bombus impatiens hives from Koppert. Prior to behavioral tests, bees were allowed to forage ad lib on either M. guttatus or M. lewisii. After one week of training, single B. impatiens workers were introduced into a plywood arena (80 cm l x 60 cm w x 40 cm h) for a pairwise choice test. We connected a pair of artificial flowers at one end of the arena to an airflow system that provided scent from live flowers encased in glass chambers. One artificial flower emitted scent from fresh flowers that had been hand-pollinated while the other emitted scent from fresh, unpollinated flowers. We registered the choice to 30 workers per Mimulus species based on which flower they visited first.

In the M. guttatus trials, bumble bee workers did not demonstrate a significant preference for either unpollinated or pollinated flowers (60% vs. 40%, respectively; χ² = 1.2, p = 0.273). We found that bumble bees had a significant preference for unpollinated M. lewisii (31% vs. 69%; χ² = 4.17, p = 0.041). Our findings suggest a volatile chemical shift in M. lewisii but not in M. guttatus. Earlier studies of pollinator preferences for VOC emissions from M. guttatus flowers demonstrated a strong pollinator sensory bias for b-trans-bergamotene, overpowering learned preferences for “honest” VOC signals. Emission of this VOC may also drive pollinator preference after pollination. This compound is absent in M. lewisii, and pollinators may be responding to changes in more “honest” signals. We are using FID and GCMS to characterize VOC emission differences between pollinated and unpollinated flowers.