1Laboratoire de Neurobiologie Comparée des Invertébrés, INRA, BP 23, La Guyonnerie, 91440 Bures-sur-Yvette, France.
2Department of Biological and Ecological Chemistry, Rothamsted Experimental Station, Institute for Arable Crops Research, Harpenden, Hertfordshire AL5 2JQ, UK.
To assess the effect of a genetical transformation of oilseed rape on the behaviour of one of its major pollinator, we studied how honey bees learn the odour from transformed (OCI line) and untransformed (Drakkar line) plants, using combined behavioural and chemical analysis methods. At the behavioural level, we used the conditioned proboscis extension (CPE) bioassay, where restrained bees learn to associate an odour (here the odour from oilseed rape flowers) to a sugar reward. Bees learned rapidly and with the same efficiency volatiles from transformed and untransformed oilseed rape. When conditioned to the aroma of one type, they generalized their responses to the odour from the other type. Furthermore, in a differential conditioning procedure, where bees are stimulated alternately with volatiles from each types of plants, one being rewarded and the other being unrewarded, they responded indifferently to the rewarded and the unrewarded odour. Finally, we sought to characterize the individual compounds used by bees to recognize the complex blend odour from transformed or untransformed oilseed rape. Volatile entrainments were carried out on tenax polymeres and were separated by optic gas chromatography (GC). Bees conditioned to the oilseed rape aroma, were tested in a combined GC-CPE procedure, the effluents of the GC being directed to the bees. For both types of plant, two compounds (linalool and phenylacetaldehyde) elicited most of the activity of conditioned bees. Our study suggests that bees do not differentiate the odour of transformed and untransformed oilseed rape and rely on the same key compound to recognize these odours.