Do alkaloids in nectar enhance pollination?

Plants synthesise an array of chemicals to defend themselves against attack by herbivorous insects, but some of these defence compounds also occur in nectar and pollen – the food reward for insect pollinators. We know surprisingly little about what effect these compounds have on pollinators, or if they have an ecological role, but are beginning to develop some theories for such counterintuitive chemistry. Given that insects pollinate 75% of crops and up to 90% of wild flowering plants (Vanbergen et al., 2013), understanding the ecological functioning of floral reward chemicals has important implications for conserving and managing pollinators and the vital pollination services they provide.

A possible explanation for the presence of these defensive compounds in nectar is that they enhance pollinator behaviour. For example, Phil’s recent research in collaboration with Newcastle University has shown that caffeine occurs in the nectar of Citrus and Coffea species, where it improves honeybee memory for the floral scents associated with food rewards. This increases revisits by the bees to the same plant species and therefore makes pollen transfer more efficient (Wright et al., 2013). It has also been suggested that nectar chemicals might protect bees against disease and pathogens (Manson et al., 2010). However, the same kinds of compounds can also cause mortality in bees - for example, the orchard mason bee (Osmia lignaria) was killed when fed nectar equivalent doses of the compound zygacine from flowers of deathcamas (Zigadenus) (Irwin et al., 2014).

Understanding the effects of floral chemicals on pollinators is therefore important, as they can be detrimental or beneficial to pollinators.These effects could be amplified where they interact with other environmental or anthropogenic pressures. A recent review, co-authored by Phil and collaborators on Insect Pollinators Initiative projects, reported that pollinators are facing multiple threats, in particular habitat loss, pesticides and the spread of pests and diseases.The report concluded that the interactive or additive effects of multiple pressures are more likely to explain pollinator declines and colony collapses than any single factor (Vanbergen et al., 2013).

Combining field and laboratory studies – a new approach using novel motion-detection technology

We combined field and laboratory studies using a new, automated digital monitoring system (called Rana) along with manual observations to record bees visiting flowers, together with chemical analyses of nectar and flowers. This also provided an opportunity to show that the new Rana technology was able to answer an ecological question and introduce it to the wider scientific community.

Rana software activates a small, autofocus camera when it detects concentrated movements and combines consecutive “motion” frames into time-compressed, date- and time-stamped movies. The system is controlled remotely, allowing the user to “stand-off” to control settings, view the live video stream and download data (movies) in the field. It continuously monitors all insect visits to flowers throughout the day – far more convenient than spending days sitting in the field and making continuous observations. In addition, because the movies are time-compressed, the data is easy to analyse – typically, daily monitoring (7am – 9pm) is condensed into minutes of recorded motion footage.   

Is floral chemistry driving observed bee behaviour?

Neighbouring patches of Aconitum lycoctonum and A. napellus growing in established beds at Kew provided the perfect study site for observing bees visiting flowers and for collecting nectar samples for direct comparison of their chemistry with recorded bee behaviours. Over the summer, while plants were flowering and bees were in full swing, we used two Rana motion-detection cameras to capture all insect visits to a marked flowering stem (raceme) of each Aconitum species for a full day. We also carried out manual observations in 2-hour sampling periods during each monitoring day to validate Rana movies. Nectar and flowers were collected from marked racemes at the end of each monitoring period and analysed by liquid chromatography- mass spectrometry (LC-MS). In total, Rana monitored flower-visiting insects to Aconitum species for 244 hours over four weeks and recorded c.500 discrete visits by bees to marked racemes.  

The Aconitum study plants finished flowering in early August and our preliminary analyses from this first season of observations have already given some fascinating insights. As predicted, the principal legitimate visitor to both A. lycoctonum and A. napellus was the long-tongued bumblebee Bombus hortorum, accounting for over 90% of visits to each species. However, the total number of legitimate visits to A. lycoctonum was markedly higher than those to A. napellus. 

Bombus terrestris attempted to steal nectar from both Aconitum species, but bees would frequently give up after partly chewing through the flower. This behaviour was observed more frequently for A. napellus. Based on these observations we were eager to know if floral chemistry could be driving bee foraging behaviour. 

Next steps

A recent study by Phil with collaborators at Trinity College Dublin has found that bees have poor acuity for detecting toxins in nectar at biologically relevant concentrations (Tiedeken et al., 2014). As a result of our recent findings, we now want to find out whether the same is true for aconitine. This work will be done later this year by Prof. Jeri Wright at Newcastle University using Bombus terrestris and following the method described for caffeine in Wright et al. (2013). We are also taking full advantage of Kew’s diverse living collection by repeating the field and laboratory studies using other, late-flowering Aconitum species growing at Kew, so come along to see Rana in action!  Once data analyses are complete we will be able to draw more informed conclusions about the possible ecological role of nectar and flower alkaloids in Aconitum-Bombus interactions.