Evolution and Chemical Ecology of Buzz-Pollinated Flowers and Pollen-Collecting Bees

More than 70% of flowering plants are animal-pollinated, with many of these plant species relying mainly on bee pollinators. Bees provide essential ecosystem services in natural systems, and play a significant role in the production of several agricultural crops. However, bee populations have recently been negatively affected by a number of environmental stressors, causing serious concern for their health and capacity to continue providing pollination services.

The relationship between flowers and their bee pollinators has been shaped through evolutionary time resulting in, sometimes, complex associations between floral traits and the morphology and behaviour of their visitors. One such case is the evolution of buzz pollination, in which nectar-less flowers with specialised morphologies are pollinated by bees that use vibrations to extract pollen grains (Fig. 1). Puzzlingly, the vibrational behaviour used by bees to extract pollen is present in some (e.g., bumblebees) but not in other species (e.g., honeybees). From the plant perspective, the convergence in floral form in unrelated buzz-pollinated plants is a striking example of convergent evolution.

The evolution of buzz pollination is not well understood. Floral morphologies associated with buzz pollination include anthers that "lock" their pollen behind anther walls that remain closed after the flower matures (non-dehiscent), and from which pollen is released only through microscopic apertures or pores (poricidal anthers). Previous work has suggested that such a locking mechanism serves to exclude greedy floral visitors and selects for a subset of pollinators that can unlock this type of anthers, for example by producing high frequency vibrations. Therefore one of the main hypothesis for the evolution of buzz pollination is that it represents the escalating solution of an arms race between pollen-rewarding flowers and pollen-collecting insects. However, few empirical studies to date have addressed this hypothesis.

The collection of pollen from buzz-pollinated flowers presents a unique opportunity to understand more generally how bees assess pollen rewards, and area that compared to nectar rewards is poorly developed. Anthers of buzz-pollinated flowers "hide" reward levels behind opaque anthers, and therefore bees cannot visually establish reward level (amount of pollen left in the flower). In order to assess reward level, the bee may thus rely on other clues including pollen removal rates, and potentially, chemical cues such as scent. The role of chemical cues has been well studied in nectar-producing species, but we now very little about the role of chemical cues in pollen-only flowers. Understanding how pollen cues mediate pollinator visitation is not only relevant for the thousands of species with pollen-only flowers, but may also have practical applications to improve the pollination of crops such as tomatoes which possess nectarless, buzz-pollinated flowers.

The main aim of this project is to investigate the evolution and chemical ecology of buzz-pollinated flowers and their pollen collecting visitors. Specifically, we will address the following questions:

1. What drives the evolution of buzz pollination?
2. How does visitation by inefficient pollinators and pollen thieves affect plant reproduction?
3. How do bees assess pollen rewards?
4. What is the role of visual vs. chemical cues in mediating pollinator visitation of nectarless, buzz-pollinated flowers?
5. How can flowers use these visual and chemical cues to manipulate bee behaviour?

Ultimately our goal is to understand at both ecological and evolutionary levels the association, and potentially co-evolution, between pollen-rewarding flowers and their bee pollinators.