Phase reversal of vibratory signals in honeycomb may assist dancing honeybees to attract their audience
J. Tautz1,*,
J. Casas2 and
D. Sandeman3
1 Biozentrum, Zoologie II, Am Hubland, 97074 Würzburg, Germany,
2 Université de Tours, Institut de Recherches sur la Biology and
3 School of Biological Science, University of New South Wales, Sydney, NSW 2052, Australia
View larger version (14K):
[in a new window]
|
Fig. 1. Diagram of the comb showing the location of the stimulus probe and the three rows of cells from the walls of which measurements were taken. The cell rows are oriented along the same horizontal axis as they would be in the hive. The large arrow shows the point of application of the lateral sinusoidal displacement to the top rim of a cell wall. The power stroke of the stimulus is in the direction of the arrow, thus ‘pulling’ the cell walls on the left and ‘pushing’ the cell walls on the right. The return stroke of the probe allows the comb to move back to its starting position. The phase reversal occurs across walls 2 and 3, walls 2b and 3b and walls and 2c and 3c. No phase reversal was found to occur across cells on the ‘pull’ side of the stimulus.
|
|
View larger version (16K):
[in a new window]
|
Fig. 2. Displacement velocity Vwall of wall 7 compared with that of wall 30 measured simultaneously using two identical lasers. The 0.5 ms time delay between the forward displacement of wall 7 and that of wall 30 at the onset of the series provides a measure of the conduction velocity of the signal across the comb. In this case, 230 m s-1.
|
|
View larger version (50K):
[in a new window]
|
Fig. 3. (A–F) Comparison of the simultaneously measured displacement velocities Vwall of cell walls at different distances from the stimulus. (A) Time course of the stimulus and movement of the wall opposite to it (wall 1). There is no phase lag, but the wave form is already distorted. Vstimulus, velocity of the stimulus. (B) A sequence of wall 1 excursions compared with those occurring at wall 2. The displacement of wall 2 (broken line) exhibits a small phase lead over that of wall 1 (solid line). (C). The small phase lead of wall 2 over wall 1 advances suddenly so that the displacements of walls 2 and 3 are about 180° out of phase. (D). One cell further along the line from the stimulus (walls 3 and 4), the cell wall displacements are again in phase with one another. (E). Sixty-two cells away from the stimulus, the cell walls are still in phase with one another. The small phase lag between walls 62 and 63 is introduced by the finite conduction velocity of the signal across the comb. (F) Displacement velocities in the ‘pull’ direction (to the left of the stimulated cell in Fig. 1) between wall -1 and wall -2. No phase reversal was found in this direction at any distance from the stimulus.
|
|
View larger version (33K):
[in a new window]
|
Fig. 4. (A,B) Laser traces of signals from wall 2b to wall 3b (A) and from wall 2c to wall 3c (B) show that the cell walls in rows lying below the point of application of the displacement also exhibit the phase-reversal phenomenon. Vwall, displacement velocity.
|
|
View larger version (15K):
[in a new window]
|
Fig. 5. The dependence of the phase-reversal phenomenon on the frequency of the displacement velocity measured across those cell walls at which the phase reversal occurs. The small phase lead of 30° at 30 Hz increases rapidly with an increase in displacement frequency, with a phase lead of 180° being maintained from 170 to 270 Hz. The arrows indicate the range of frequencies known to occur during the waggle dance. This coincides with the greater part of the range within which the phase reversal is exhibited.
|
|
View larger version (28K):
[in a new window]
|
Fig. 6. (A–F) Histograms showing the recruitment of dance-followers. Each part of the figure represents the results from a single sector, indicated in the inset. Each column in each figure shows the number of dance-followers that turned their heads and ran in to follow the dancer. The distance from which they came is indicated along the abscissa. The data were analysed into 5 mm ‘bin widths’ because this represents the approximate width of a single cell. The distance is, nevertheless, given here as ‘cell widths’ to simplify comparison with the previous figures. (A). Sector 1. Most of the dance-followers in this sector are attracted from a distance of 2–3 cells away. N=34. (B). Sector 2. The attraction in the 30–60° sector is less narrow than that in the 0–30° sector, and dance-followers come from distances of 2–5 cells away. N=30. (C). Sector 3. Recruitment of dance-followers in this sector is also broadly distributed over distances ranging from three to five cells. N=30. (D,E). Sectors 4 and 5. Dance-followers in these two sectors are recruited predominantly from the 2–4 cell distance. N=30 (D), N=22 (E). (F). Sector 6. Most of the dance-followers in this sector are recruited from close behind the dancer, suggesting the presence of an attractant in addition to, or in place of, substratum vibratory signals. N=28. The line parallel to the x axis indicates the range over which phase reversal would be expected to occur.
|
|
View larger version (25K):
[in a new window]
|
Fig. 7. Results of Fig. 6 superimposed in a single graph in which the data from the different sectors have been normalised by expressing the columns in each sector (which record the numbers of dance-followers) as a percentage of the tallest column in that sector. This figure shows that a high percentage of dance-followers from all sectors are recruited from a distance of three cell widths from the dancer. The distribution of percentages over all sectors varies, with sectors 1 and 2 being the broadest. The recruitment of dance-followers from sector 6 suggests that other factors are at work. The line parallel to the x axis indicates the range over which phase reversal would be expected to occur.
|
|
© The Company of Biologists Ltd 2001
