Frequency dependent selection in a deceptive orchid

Populations of deceptive plants are considered to be more variable than populations of rewarding species in traits associated with pollinator attraction. This variability in traits increases the difficulty for pollinators to recognize and avoid deceptive plants. Hence, rare phenotypes are thought to have a higher fitness. The evolutionary process by which the fitness of a phenotype depends on its frequency relative to other phenotypes in a given population, is called frequency dependent selection. In rewarding plants, positive frequency dependent selection is expected, in deceptive plants, negative frequency dependent selection. Floral scent is a key mediator in plantpollinator interactions. It is known to vary not only among plant species, but also within species among and within populations. It is still unknown, whether such variability is the result of divergent selective pressures, exerted by pollinators or of other scenarios. A prominent deceptive plant is Cypripedium calceolus L., the only lady`s-slipper orchid species in Europe. Its successful pollination depends on bees trapped in the labellum and escaping the slippery cavern via an exit of similar height as the insect itself, thereby passing the stigma and the anthers to first depositing and second collecting pollen. In this planed project, we aim to show, by experimental investigations at natural and artificial populations, that negative frequency dependent selection processes are active in Cypripedium calceolus L. on floral scent. Key methods will be collection and analysis of floral scent, artificial rewarding with nectar and the application of plant-specific pigment color codes to track pollen flow in populations of up to 200 flowers. Specifically, we will test the following hypotheses: a) father- mother pairs of C. calceolus plants (pollen donors and receivers of that pollen) are more different in scent than randomly chosen pairs, b) rare phenotypes within populations of C. calceolus have a higher reproductive success than more abundant phenotypes, and c) artificially rewarding plants of C. calceolus experience a different type of selection than deceiving ones. C. calceolus is a good model to study frequency dependent selection, because the species is deceptive, pollen limited, dependent on insects for pollen transfer and emits a strong and variable scent involved in communication with pollinators. The consistency of the pollen smear and how it is offered as well as the shape of the flower make the planned experiments possible. The unique setup with pigment color codes allows us to track the pollen of individual flowers of an entire population. Our study aims to elucidate evolutionary processes in a rare native orchid species. This will also help to understand similar processes in other organisms.