Unique-colors in honeybees?

Abstract

Neuronal color coding and color choice behavior in ordinary color training experiments, in which a bee is trained to one stimulus, are very well described by the color theory for the honeybee on the basis of the excitations of two types of tonic color opponent coding opponent (see Backhaus, 1991, 1992b, 1993 for review). Under certain conditions, color perception in bees was already suspected of being not always related directly to the excitations of color opponent coding neurons but to the amounts of unique-colors which should be linearly related to the excitations of these neurons (Backhaus, 1988). These unique-colors might be comparable to the six unique-colors (Urfarben, elemental, primary, basic, fundamental, or irreducible colors) blue, yellow, red, green, black and white in humans (Hering, 1878, 1905). Since color vision in honeybees has no ‘brightness’ dimension, only five different unique-colors, four ‘chromatic’ and one ‘achromatic’, were postulated. Furthermore, it was expected that the bee learns in double color training experiments the unique-color which two alternately rewarded colors have in common and to choose the stimuli in unrewarded tests according to the amounts of these unique-colors. The results of the first double color training experiments (Backhaus and Kratzsch, 1993) indeed did not agree with the predictions of the common color theory for the honeybee, and thus could not be explained on the basis of the electrical properties of neurons alone. But the results of these experiments agreed well with the predictions from the color theory extended by the unique-color hypothesis for the bee. It obviously depends on the training method used, whether the color choice behavior of the bee is directly related to the excitations of the color opponent neurons or to the amounts of unique-colors which are steered by the excitations of the color opponent coding neurons. Comparable relations between the amounts of unique-colors, the axes of the color space, and the excitations of color coding neurons arehypothesized to exist in human color vision.