, Volume 76, Issue 6, pp 277–280 | Cite as

Honeybees can be recruited by a mechanical model of a dancing bee

  • A. Michelsen
  • B. B. Andersen
  • W. H. Kirchner
  • M. Lindauer
Kurze Originalmitteilungen


Mechanical Model 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    von Frisch, K.: Tanzsprache und Orientierung der Bienen. Berlin-Heidelberg: Springer 1965Google Scholar
  2. 2.
    von Frisch, K.: Science 183, 663 (1974)Google Scholar
  3. 3.
    Michelsen, A., et al.: J. Comp. Physiol. A 161, 633 (1987)Google Scholar
  4. 4.
    Michelsen, A., Kirchner, W. H., Lindauer, M.: Behav. Ecol. Sociobiol. 18, 207 (1986)Google Scholar
  5. 5.
    Kirchner, W. H., Lindauer, M., Michelsen, A.: Naturwissenschaften 75, 629 (1988)Google Scholar
  6. 6.
    The first mechanical models of dancing bees were built by H. Steche (Naturwissenschaften 44, 598 (1957)) and by H. Esch (Z. vergl. Physiol. 45, 1 (1961)). These models did not emit sound, but in a later experiment by Esch a wax-coated microphone was made to waggle and emit sounds. J. L. Gould (Quart. Rev. Biol. 51, 211 (1976)) experimented with moving and vibrating paralyzed bees. In all cases, the bees showed great interest in the models, but the models did not elicit recruitment.Google Scholar
  7. 7.
    The dancing bees vibrate their wings with a frequency of 280 Hz and a peak velocity of approximately 1 m/s, thus causing a sound pressure of approximately 1 Pa at the wings' surfaces [3]. The sound pressure decreases much with distance, however, since the vibrating wings act as a dipole emitter: the sound pressure at the upper and lower surfaces of the wings are totally out-of-phase, and the large pressure gradient causes oscillating air currents close to the dancing bee's abdomen (but not around the head and thorax). Most of the follower bees hold their antennae near the dancer in the zone of maximum oscillation of air currents [3]. Honeybees can be conditioned to respond to oscillating air currents similar to those produced by dancers (W. F. Towne, W. H. Kirchner: Science, submitted). Wagging and sound emission occur in both wagging dances and round dances, and both types of dances carry information on direction and distance [5].Google Scholar
  8. 8.
    Some of the dance parameters were independent of the aimed distance: The dance sound had an amplitude of 1 Pa (peak, measured above the wing) and was presented as 40 pulses per second during the wagging run, each pulse having a duration of 11 ms (three waves of the 280-Hz carrier). During wagging the model was rotated around the center of the “thorax” with a frequency of 13 Hz and a peak amplitude of 30°. Other dance parameters varied with distance: The total durations of the figure-eight paths were 4, 5, and 6.5s at 250, 500, and 1000m, respectively. The durations of the wagging runs at these distances were 510, 710, and 1110ms, respectively. These values were very close to those observed when live dancers in the same bee family communicated the presence of food at these distances. The length of the wagging run was 9 mm in the dances aimed at 250 and 500 m and 10 mm in the 1000-m dances (cf. [1]).Google Scholar
  9. 9.
    Silent dances by living bees are fairly common, and such dances are ineffective in transmitting information (Esch, H.: Verb. Dtsch. Zool. Ges. 1962, 302 (1963)). However, silent dances are also less energetic; thus their ineffectiveness does not prove that the dance sounds are essential ([1], p. 59). In our experiments, the vigor of the dance was kept constant when we removed or changed the sounds.Google Scholar
  10. 10.
    A slight preference for north (the direction opposite to that indicated by the model) is apparent in Fig. 2C, D and was also observed in the experiment with continuous sound and in a control experiment, in which the model offered small amounts of scented sugar water but did not dance (a total of eight bees was observed in the four directions). The slight preference for north appeared to be due to the presence of some flowers approximately 800–1000m to the north.Google Scholar
  11. 11.
    These control experiments with live dancers were performed 1 week after the experiments with the model. In each of the experiments (Figs. 2B, 3D) only two honeybees were allowed to dance (the other bees visiting the feeding station were captured). The total duration of their dances was approximately 35 % of the experimental periods (2.5 and 3h, respectively). So, although the total duration of the model's dances was more than twice this figure, the live dancers recruited more follower bees, and the follower bees obtained a more precise indication of the direction.Google Scholar

Copyright information

© Springer-Verlag 1989

Authors and Affiliations

  • A. Michelsen
    • 1
  • B. B. Andersen
    • 1
  • W. H. Kirchner
    • 2
  • M. Lindauer
    • 2
  1. 1.Institute of BiologyOdense UniversityOdense M.
  2. 2.Zoologisches Institut II der UniversitätWürzburg

Personalised recommendations