Chemicals involved in honeybee-sunflower relationship.

We present a review of work on the plant chemicals involved in the honeybee-sunflower model system. Combined behavioral and chemical analyses were conducted under natural and controlled conditions. First the distribution of forager bees' visits on two pairs of sunflower genotypes producing a different level of hybrid seed yield was recorded under pollen-proof tunnels. Mirasol parental lines producing high seed yields were visited at random, whereas forager bees visited preferentially the female parental line of Marianne, resulting in low seed yield. Nectar samples collected on the genotypes were analyzed by gas chromatography. Fructose, glucose, and sucrose were identified. Parental lines of Mirasol showed similar sugar profiles, whereas the female line of Marianne contained higher amounts of sucrose than the male line. We assume that the bees' preferences between genotypes might rely on differences in the sugar composition of floral nectars, especially in the amount of sucrose. Aromas from headspace collection were compared between pairs and periodically during the flowering period. Of the 144 components indexed for Marianne lines and 136 components for Mirasol lines, 17 of the components for Marianne lines and 18 for Mirasol lines differed significantly according to flowering stage. Significant differences appeared in eight of the 134 components of Marianne lines and in 20 of the 250 components for Mirasol lines. Such differences, even restricted to a few components, might account for honeybees' discrimination between genotypes or flowering stage. Experiments then were conducted in a flight room using an artificial flower device. A total volatile extract was used as a conditioning scent previous to the test where the total extract was successively compared to several of its subfractions. Fractions significantly less visited than the total extract were discarded, whereas fractions confused with the total extract were kept. From step to step, a restricted fraction of 28 polar components, among which 15 were identified, was shown to be as active as the initial conditioning extract. These data emphasized honeybees' abilities to generalize from simplified to more complex chemical information. Finally, this work considers the possible use of such plant chemicals, from nectars or aromas, either as targets for genetic modification of crop plants or as direct attractants when sprayed on the crop, for the improvement of entomophilous cross pollination.