Pollinators are widely regarded to drive the evolution of attractive floral traits, like flower colour and scent. Subtle variations in flower shape and size, however, are also likely to experience selection by pollinators, albeit through a different mechanism. This mechanism, mechanical fit, involves a tight match between the size and shape of flowers and their pollinators. Tailoring flowers to match their pollinators may be expected, because it can optimise pollen transfer by increasing the contact time between a flower’s reproductive organs and the pollinator.
In a new paper Marinus de Jager and Rod Peakall investigate mechanical fit and show that the labellum of a sexually deceptive orchid experiences strong stabilizing selection from its pollinators. Pollinators usually land on the labella of orchids when foraging for nectar, but in the sexually deceptive orchid Chiloglottis trapeziformis, they do something else as well: they try to mate with it (Figure 1). This interspecies mishap occurs because C. trapeziformis accurately mimics the females of their male wasp pollinator by copying the female sex pheromones. As if such deception isn’t misleading enough, the orchids also produce these sex pheromones in amounts ten-fold that of actual females.
The outcome of this trickery is that when a hopeful male wasp lands on the labellum of C. trapeziformis, he grabs hold of a callus outgrowth resembling a female wasp (see Figure 2) and attempts copulation by probing the labellum edge with his genital claspers (see video). Eventually, the male departs, defeated. It is during this pseudocopulation that the flower glues its pollinia to the back of the male wasp, who unknowingly transfers its pollen to the stigma of the next flower he tries to mate.
de Jager and Peakall reveal that pollinia removal only occurs during this frantic pseudocopulation behaviour. The longer the male tries to mate, the more likely he is to pick up a pollinium. The distance between the insect-looking callus and the edge of the labellum (Callus-tip distance, see Figure 2) determines how long a male wasp displays pseudocopulation behaviour. We know this, because de Jager and Peakall experimentally elongated and shortened the labellum on manipulated flowers (Figure 2), which received much shorter mating attempts than control flowers with a normal sized labellum. Flowers with smaller, or larger labella, should thus receive shorter mating displays from their pollinators, leading to less pollen transfer and a reduction in plant fitness.
Investigating the patterns of pollinia removal and pollen deposition on stigmas in two natural C. trapeziformis populations, de Jager and Peakall confirm this prediction by documenting a greater incidence of pollen transfer in flowers with intermediate labellum lengths than flowers with extreme labellum lengths, be it too short or too long. Taken together, these lines of evidence serve as a comprehensive investigation and confirmation of stabilizing selection in an enigmatic plant, mediated by their tireless, and ever optimistic pollinators.
Marinus de Jager is a postdoctoral fellow at Stellenbosch University in South Africa who regularly publishes about pollination by sexual deception in orchids and daisies (e.g., Gorteria diffusa). He is passionate about the ecology and evolution of all plant-pollinator interactions. You can find him on Twitter @mldejager.