Floral scent, a key mediator in plant–pollinator interactions, varies not only among plant species, but also within species. In deceptive plants, it is assumed that variation in floral scents and other traits involved in pollinator attraction is maintained by negative frequency-dependent selection, i.e., rare phenotypes are more attractive to pollinators and hence, have a higher fitness than common phenotypes. So far, it is unknown whether the rarity of multivariate and/or continuous floral scent traits influences the pollination success of flowers. Here, we tested in the deceptive orchid Cypripedium calceolus, whether flowers with rarer scent bouquets within a population have a higher chance to getting pollinated than flowers with more common scents. We collected the scent of more than 100 flowers in two populations by dynamic headspace and analyzed the samples by gas chromatography coupled to mass spectrometry (GC/MS). From the same flowers we also recorded whether they set a fruit or not. We introduced rarity measures of uni- and multivariate floral scent traits for single flowers, which allowed us to finally test for frequency-dependent pollination, a prerequisite for negative frequency-dependent selection. Our results do not show rarity has an effect on the likelihood to set fruits in neither of the two populations and in none of the scent characteristics analyzed. Hence, there is no evidence of negative frequency-dependent pollination mediated by the floral scent of C. calceolus. We discuss that our approach to determine rarity of a scent is applicable to any univariate or multivariate (semi)quantitative trait.
Hybrid floral scent novelty drives pollinator shift in sexually deceptive orchids
