Solving the sexual chemistry between greenhood orchids and male fungus gnats

How does an orchid pretend to be a female fungus gnat and trick a male fungus gnat into copulating with it?

This article is based on research published by myself and colleagues in 2021 in Current Biology. The full details of our published research are given below, including the link to the final article. The article is subscription only, meaning you (or your institution) needs to be subscribed to Current Biology. However, I’ve uploaded a copy of the accepted manuscript here (it is pretty much the same thing but without the journal formatting).

Hayashi, T., Bohman, B., Scaffidi, A., Peakall, R., and Flematti, G.R. (2021) An unusual tricosatriene is crucial for male fungus gnat attraction and exploitation by sexually deceptive Pterostylis orchids. Current Biology, 31:9 p1954-1961. DOI: https://doi.org/10.1016/j.cub.2021.01.095

So, how exactly does Pterostylis orbiculata, the coastal banded greenhood orchid, convince a particular male fungus gnat to attempt to copulate with it, and how does this help with pollination? This is one of the questions which I’ve been trying to answer through my PhD research. 

The video below shows an example of a male gnat copulating with the orchid’s labellum. The orchid is pretty convincing as far as the gnat is concerned!

The answer is nearly all in the chemistry. In other words, the orchid smells like a female fungus gnat which is why the male is attracted to the orchid and attempts to copulate with it. We know from previous research that you can cover Pterostylis orbiculata flowers with a screen and male fungus gnats will still find their way to the flower and attempt copulation. 

How does this help with pollination? Once the horny male fungus gnat lands on the labellum and starts to copulate, the labellum triggers upwards* and traps the fungus gnat inside the top of the flower (the ‘hood’). The fungus gnat must then make its way through a tight passageway at the front of the hood, whereupon it collects pollinia on its thorax and can finally escape through the front of the flower. Pollination is achieved when the male fungus gnat visits another flower while carrying the pollinia and brushes up against the stigmatic surface (located inside the top of the hood). 

*The trigger mechanism on the labellum becomes desensitised after repeated triggering (you can see in the video above that the labellum does not trigger). Unfortunately I don’t have any good video of Pterostylis orbiculata labellum being triggered by its fungus gnat pollinator, but I’ve included two videos below showing this process in the related Pterostylis longifolia which is pollinated in a similar way by a different species of fungus gnat. 

The goal of my research was to find out precisely which chemicals the orchid produces to attract the male fungus gnat. After all, plants don’t normally make insects horny, so how does this orchid evolve to sexually excite male fungus gnats? 

We know from previous research that some other plants (by and large restricted to a few groups of orchids) achieve pollination in a similar way, by fooling male bees, wasps and beetles into mating with the flowers. In many of these cases, the orchids produce exactly the same compounds that the female insects produces to attract the male for mating. In other words, the orchids have evolved to be able to produce insect sex pheromones just to achieve pollination. Perhaps even more remarkably, the compounds produced by the female insects are often biologically unusual compounds that plants don’t normally produce.

No research had yet been done on which compounds greenhood orchids produce to attract male fungus gnats, or for that matter any other fungus gnat-attracting plant. In fact, we knew precisely zero about mycetophilid fungus gnat sex pheromones. This is where my research comes in. 

So how do you go about working out which of the compounds produced by Pterostylis orbiculata are important for attracting male fungus gnats? First, we made solvent extracts of the orchid labellum (the only bit of the orchid attractive to male fungus gnats) by placing the labellum in dichloromethane for 24 hrs. We also did the same with male and female fungus gnats. (In case you’re wondering, its not easy to find female fungus gnats and be sure they are the right species! I ended up spending a lot of time crawling around at ground level and capturing copulating pairs of fungus gnats - I then DNA sequenced the male from the pair and compared it with males captured at orchid flowers, to be sure that the female from the pair was the right species.)

Once we had these solvent extracts, we compared compounds present in orchid labella and female gnats (both of which are attractive to males) with compounds present in other orchid tissue and male gnats (neither of which are attractive to other male gnats). We came up with 5 compounds in common between orchid labellum and female gnat extracts, including 2 promising looking compounds not present in male gnat extracts (1 and 2 in Figure 1E-H below). To confirm that these two compounds were attractive to male gnats, we created four fractions of the orchid extract and found that male gnats only responded to one fraction (the one which contained compounds 1 and 2). 

The next step was to work out exactly what these compounds were. Using Gas Chromatography - Mass Spectrometry (GC-MS) we found a match in the published literature for compound 2 but although compound 1 was clearly similar we couldn’t find a match. After using Nuclear Magnetic Resonance (NMR) and a bit of guesswork, we came up with a tentative identification. Then, in order to confirm we had correctly identified the 2 compounds, we made these compounds synthetically in the lab. This work was done by Adrian Scaffidi, Björn Bohman and Gavin Flematti who are actually chemists, rather than me, a pollination biologist! Once synthesised, we were able to confirm that our identifications were correct, and that compound 1 - (6Z,9Z)-1,6,9-tricosatriene - was a new natural product i.e. it had not been identified in nature before.

The final step in the process was to present synthetic versions of these compounds to male fungus gnats and confirm that they elicited sexual attraction. To do this, we placed the synthetic compounds on little filter paper dummies (Figure 1D), using orchid labellum extract placed on filter paper dummies as a control. Success! A synthetic blend of all 5 compounds in common between orchid labellum and female gnat extracts elicited attraction and sexual behaviour from male gnats. Interestingly, when I removed compound 1 from the blend the male gnats didn’t respond at all. On the flip side, male gnats responded to compound 1 alone, albeit with less sexual behaviour than the full blend.

Like many other sexually deceptive orchids, Pterostylis orbiculata produces compounds which are unusual among plants, in order to attract male insect pollinators. Indeed, compound 1 had not previously been identified in nature, and in general hydrocarbons with >18 carbons and >2 double bonds are unusual natural compounds. The most similar natural compounds published thus far are known as sex pheromones of some species of moth. Interestingly, we detected a known moth sex pheromone in a closely related species of Pterostylis, although it’s unlikely that species is pollinated by moths.

Although compound 1 is crucial for pollinator attraction in Pterostylis orbiculata, we didn’t find any evidence of this compound in the extracts of closely related species P. concava and P. vittata. Instead, we found similar hydrocarbons with 23 carbons and 3 (and 4) double bonds and slightly different double bond positions. We don’t know yet whether these related species are also pollinated by male fungus gnats, but our working hypothesis is that each Pterostylis species produces a similar but slightly different compound which attracts a different species of fungus gnat. 

This pattern of ‘one orchid one pollinator’ is common among sexually deceptive orchids, and can sometimes lead to a situation where two species look identical yet attract different pollinators and so remain biologically separate species. After all, in these cases the chemistry is the most important stimulus the flower produces. 

This article is based on research published by myself and colleagues in 2021 in Current Biology. The full details of our published research are given below, including the link to the final article. The article is subscription only, meaning you (or your institution) needs to be subscribed to Current Biology. However, I’ve uploaded a copy of the accepted manuscript here (it is pretty much the same thing but without the journal formatting).

Hayashi, T., Bohman, B., Scaffidi, A., Peakall, R., and Flematti, G.R. (2021) An unusual tricosatriene is crucial for male fungus gnat attraction and exploitation by sexually deceptive Pterostylis orchids. Current Biology, 31:9 p1954-1961. DOI: https://doi.org/10.1016/j.cub.2021.01.095