, Volume 141, Issue 3, pp 379-388

Luminance dependence of pigment color discrimination in bees

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access


  1. A system is described in which a behavioral test is used to match color shades to an arbitrarily chosen level to the compound of brightness and saturation. This system was used with honeybees (Apis mellifera carnica) to develop a series of color marks which were equal to the compound of brightness and color saturation to bees.

  2. Freely flying honeybees were trained to a blue color mark on a vertical screen. Discrimination of this blue color from violet, green, blue-green and yellow was tested at various natural luminances. In the majority of the experiments the background screen was painted achromatic grey of the same luminosity as the color marks. Two training procedures were used: one utilized a feeding station, the other the hive entrance.

  3. Bees discriminate colors best at luminances between 101 and 102 cd/m2. Color discrimination is less acute at both higher and lower luminance levels; these deficits at high and low luminances have been termed the bright light and dim light effects, respectively. Color discrimination disappears below 10−1 cd/m2 (Fig. 6). The dependence of color discrimination on available light is the same in both the violet and blue-green regions.

  4. It is concluded that bees have achromatic vision at low light levels (<10−1 cd/m2), and that they can use this achromatic vision during extreme low light situations — even during flight orientation. The bright light effect is discussed with respect to results from single unit electrophysiological recording experiments and appears to result from the specific behavior of light adaptation in the retinula cell — monopolar cell system. The narrow luminance band of optimum color discrimination suggests that color discrimination may be better than previously determined using selfruminant spectral lights (von Heiversen 1972a).

Supported by an NS&B Scholarship to Marine Biological Laboratory, Woods Hole, DAAD Grant 315/4-usa-4-sk and a Luft-Brückendank Stipendium
Recipient of a Rand Professorship to Woods Hole. Additional support provided by a grant from the DFG
The authors thank David Schessel, Johannes Menzel, Ute Kindermann and Gabriela Koske for their help in performing these experiments; Dr. D.R. Stokes and E. Lieke for their helpful criticism of this work; and J.H. Reynolds and B. Cwienczek for their careful