Pollen transfer efficiency and its effect on inflorescence size in deceptive pollination strategies.

Pollination systems differ in pollen transfer efficiency, a variable that may influence the evolution of flower number. Here we apply a comparative approach to examine the link between pollen transfer efficiency and the evolution of inflorescence size in food and sexually deceptive orchids. We examined pollination performance in nine food-deceptive, and eight sexually deceptive orchids by recording pollen removal and deposition in the field. We calculated correlations between reproductive success and flower number (as a proxy for resources allocated during reproductive process), and directional selection differentials were estimated on flower number for four species. Results indicate that sexually deceptive species experience decreased pollen loss compared to food-deceptive species. Despite producing fewer flowers, sexually deceptive species attained levels of overall pollination success (through male and female function) similar to food-deceptive species. Furthermore, a positive correlation between flower number and pollination success was observed in food-deceptive species, but this correlation was not detected in sexually deceptive species. Directional selection differentials for flower number were significantly higher in food compared to sexually deceptive species. We suggest that pollination systems with more efficient pollen transfer and no correlation between pollination success and number of flowers produced, such as sexual deception, may allow the production of inflorescences with fewer flowers that permit the plant to allocate fewer resources to floral displays and, at the same time, limit transpiration. This strategy can be particularly important for ecological success in Mediterranean water-deprived habitats, and might explain the high frequency of sexually deceptive species in these specialised ecosystems.

Bees use honest floral signals as indicators of reward when visiting flowers.

Pollinators visit flowers for rewards and should therefore have a preference for floral signals that indicate reward status, so called 'honest signals'. We investigated honest signalling in Brassica rapa L. and its relevance for the attraction of a generalised pollinator, the bumble bee Bombus terrestris (L.). We found a positive association between reward amount (nectar sugar and pollen) and the floral scent compound phenylacetaldehyde. Bumble bees developed a preference for phenylacetaldehyde over other scent compounds after foraging on B. rapa. When foraging on artificial flowers scented with synthetic volatiles, bumble bees developed a preference for those specific compounds that honestly indicated reward status. These results show that the honesty of floral signals can play a key role in their attractiveness to pollinators. In plants, a genetic constraint, resource limitation in reward and signal production, and sanctions against cheaters may contribute to the evolution and maintenance of honest signalling.

Sex pheromone mimicry in the early spider orchid (ophrys sphegodes): patterns of hydrocarbons as the key mechanism for pollination by sexual deception.

We investigated the female-produced sex pheromone of the solitary bee Andrena nigroaenea and compared it with floral scent of the sexually deceptive orchid Ophrys sphegodes which is pollinated by Andrena nigroaenea males. We identified physiologically and behaviorally active compounds by gas chromatography with electroantennographic detection, gas chromatography-mass spectrometry, and behavioral tests in the field. Dummies scented with cuticle extracts of virgin females or of O. sphegodes labellum extracts elicited significantly more male reactions than odorless dummies. Therefore, copulation behavior eliciting semiochemicals are located on the surface of the females' cuticle and the surface of the flowers. Within bee and orchid samples, n-alkanes and n-alkenes, aldehydes, esters, all-trans-farnesol and all-trans-farnesyl hexanoate triggered electroantennographic responses in male antennae. Most of the alkanes and alkenes occurred in similar patterns both in the bees and orchids. O. sphegodes leaf extracts contained mostly the same compounds but in different proportions. In behavioral tests with synthetic compounds, blends of alkenes triggered significantly more approaches and pounces of the males whereas alkanes were not more attractive than odorless dummies. Since alkanes and alkenes together were most attractive, we conclude they constitute the bees' sex pheromone as well as the pseudocopulation-behavior releasing orchid-odor bouquet.