Side-specific reward memories in honeybees.

We report a hitherto unknown form of side-specific learning in honeybees. We trained bees individually by coupling gustatory and mechanical stimulation of each antenna with either increasing or decreasing volumes of sucrose solution offered to the animal's proboscis along successive learning trials. Next, we examined their proboscis extension response (PER) after stimulation of each antenna 1, 2, 3, and 24 h after training. The bees extended their proboscises earlier after stimulation of the antenna that had been coupled with increasing volumes than after stimulation of the antenna that had been coupled with decreasing volumes, thereby revealing short- and long-term side differences in the bees' PE reaction time. The bees' reaction time correlated well with the reaction time of the muscles M17. Long-term side differences in reaction time were prevented by repetitive antennal stimulation. Mechanosensory input was indispensable and sufficient for revealing side differences in reaction time. Such differences were specific to the gustatory input that the bees experienced during training. Our results show that side differences in the bees' PE reaction time depend upon the activation of side-specific reward memories. These memories are formed via the combined effect of a specific property of reward, i.e., that its magnitude increases or decreases over time, and side information seemingly relying on mechanosensory input. We present a learning procedure suitable to study reward learning in honeybees, which includes precise behavioral measures, physiological correlates of behavior, and within-animal controls. This procedure will prove fruitful in pharmacological and electrophysiological analyses of the neural substrates underlying reward memories in honeybees.