Wine-scented flower draws in fruit flies with yeasty tones

ByEd Yong
October 13, 2010
4 min read

In a German lab, Johannes Stokl is wafting a series of fruity and yeasty smells in front of a panel of restrained testers. As the chemical cocktail tickles their senses, electrodes and brain scanners record their every reaction. This bizarre wine-tasting event is all part of a study into the bizarre deception of a flower – the Solomon’s lily. And Stokl’s subjects aren’t humans – they’re fruit flies.

Solomon’s lily is one of the arum lilies, a group that specialises in manipulating flies. They attract these unusual pollinators by giving off odours of urine, dung and rotting meat, repugnant smells that seem completely at odds with their attractive appearance. Solomon’s lily is an exception – it smells rather pleasant, a bit like a fruity wine. But this fragrance, like the fouler ones of other arum lilies, is also a trick. Solomon’s lily uses it to draw in flies that eat decaying fruit.

The lilies grow in Israel, Syria and Lebanon and if you cut them open, you can find flies in their hundreds. Stokl counted more than 400 individuals in each of two different flowers. The trapped insects included 8 different drosophilids – the fruit-eating species that are such darlings of geneticists.

The lily’s aroma of fermenting fruit certainly seems like the type of scent that would draw in such insects, but Stokl wanted to be sure. He collected the plants’ fragrances and ran them through equipment that separated them into their constituent chemicals. Each of these components was individually wafted over tethered flies, whose antennae had been hooked up to electrodes. Through this clever design, Stokl could identify the exact chemicals in the lily’s milieu that roused the fly’s senses.

He found six. Each of these is mildly attractive to a fruit fly but in combination, mixed according to the flower’s own recipe, they were just as enticing as powerful commercially-available traps. Among flowers, these chemicals are rare; two of the set have only ever been detected once before within a floral scent. But you inhale them whenever you take a whiff of overripe or rotting fruit, wine, or vinegar. Balsamic vinegar is an exceptionally rich source. All of these chemicals are given off by yeasts during the process of fermentation.

Yeast is the staple food of fruit flies – it’s what they’re after when they seek out rotting fruit. And the lily’s chemical ruse is so exact that it’s unlikely that the fly can separate the flower’s smell from the real deal. To demonstrate that, Stokl also deconstructed the scents of several rotting fruits, balsamic vinegar and a bottle of red wine (a “fruity Lambrusco variety” apparently), and wafted these in front of his tethered flies. The recordings show that the fly perceives all of these odours in much the same way as it does the lily’s scent, with the wine and vinegar providing the closest matches.

This deception is a deep one, for the lily exploits a sense that the flies have been using for millions of years. Using detailed brain scans, Stokl found that the six critical chemicals tickle a set of proteins that are conserved throughout the drosophilid group. As the flies evolved and diverged, these stalwart proteins changed very little, retaining their ancestral role as yeast detectors. As a result, even drosophilid species that have been separated by 40 million years of evolution respond to the smell of the Solomon’s lily in virtually the same way.

Like a good wine-tasting, Stokl’s thorough experiments have revealed something that is far more subtle than a casual sniff would suggest. It would be say that a wine-scented lily attracts flies that like fermenting fruit and call it a day. But by bringing the tools of neuroscience and genetics to the table, Stokl showed that Solomon’s lily produces a smell that taps into a sense embedded in the evolutionary history of the entire drosophilid line. It has evolved an all-purpose lie that dupes all manner of drosophilid flies, drawing in pollinators in droves.

Reference: Current Biology http://dx.doi.org/10.1016/j.cub.2010.09.033

More on mimicry:

Go Further