Does plastic comb foundation hinder waggle dance communication?

Abstract

In recent years, plastic comb foundation has become widely used by beekeepers but it has not been studied to see if it hinders recruitment communication by reducing the transmission of the comb vibrations produced by bees performing waggle dances. We used laser vibrometry to compare combs built with beeswax foundation vs. plastic foundation in terms of transmission of dance vibrations. We also used behavioral experiments to compare the recruitment effectiveness of dances performed on combs built with beeswax foundation vs. plastic foundation. We found that combs built with plastic foundation are markedly poorer at transmitting the 250 Hz vibrations produced by dancing bees. Nevertheless, we found no evidence of reduced effectiveness of dances performed on combs built with plastic foundation vs. combs built with beeswax foundation. Evidently, a comb built with plastic foundation provides a fully suitable substrate for waggle dance communication.

Full text

513
Apidologie 36 (2005) 513–521
© INRA/DIB-AGIB/ EDP Sciences, 2005
DOI: 10.1051/apido:2005037
Original article
Does plastic comb foundation hinder waggle dance
communication?1
Thomas D. SEELEYa*, Adrian M. REICHa, Jürgen TAUTZb
a Department of Neurobiology and Behavior, Cornell University, Ithaca, NY 14853, USA
b bee group, Biocenter, Universität Würzburg, Am Hubland, 97074 Würzburg, Germany
Received 8 February 2005 – revised 16 March 2005 – accepted 16 March 2005
Published online 13 September 2005
Abstract – In recent years, plastic comb foundation has become widely used by beekeepers but it has not
been studied to see if it hinders recruitment communication by reducing the transmission of the comb
vibrations produced by bees performing waggle dances. We used laser vibrometry to compare combs built
with beeswax foundation vs. plastic foundation in terms of transmission of dance vibrations. We also used
behavioral experiments to compare the recruitment effectiveness of dances performed on combs built with
beeswax foundation vs. plastic foundation. We found that combs built with plastic foundation are markedly
poorer at transmitting the 250 Hz vibrations produced by dancing bees. Nevertheless, we found no evidence
of reduced effectiveness of dances performed on combs built with plastic foundation vs. combs built with
beeswax foundation. Evidently, a comb built with plastic foundation provides a fully suitable substrate for
waggle dance communication.
Apis mellifera / comb vibration / plastic foundation / recruitment communication / waggle dance
1. INTRODUCTION
Modern beekeeping is based on four key
inventions from the 1800s: movable frame
hive, bellows bee smoker, honey extractor, and
comb foundation (Crane, 1990). Comb foun-
dation – thin sheets of material embossed with
the hexagonal pattern of worker cells on which
the bees build their combs – ensures that the
bees build planar combs, saves them much wax
synthesis during comb building, and inhibits
them from rearing drones, by doing away with
most of a colony’s drone comb. Over the past
20 years, the technology of comb foundation
has changed markedly, expanding the types of
foundation available to beekeepers. Previ-
ously, comb foundation consisted simply of
sheets of beeswax that were fitted to wooden
frames. Now beekeepers have the options of
plastic foundation fitted to wooden frames, and
one-piece plastic frames and foundation. Under
favorable conditions, bees will build comb on
plastic foundation, and many beekeepers have
adopted this technology. To date, however, no
one has investigated whether frames of comb
built with plastic foundation, which is less flex-
ible and more massive than beeswax founda-
tion, provide the bees with a fully suitable
substrate for waggle dance communication. If
not, then such combs may have negative effects
on a colony’s recruitment communication and
on its honey production. The aim of this study
was to see if plastic comb foundation hinders
waggle dance communication.
Bees recruit nest mates to rich food sources
by performing waggle dances. Each circuit of
* Corresponding author: tds5@cornell.edu
1Manuscript editor: Stan Schneider
Article published by EDP Sciences and available at http://www.edpsciences.org/apidoor http://dx.doi.org/10.1051/apido:2005037

514 T.D. Seeley et al.
the dance contains a waggle run during which
the bee makes lateral body vibrations at
approximately 15 Hz and thorax/wing vibra-
tions at approximately 250 Hz (Esch, 1961;
Wenner, 1962; Michelsen et al., 1986). The
thorax/wing vibrations generate airborne
sounds that dance followers detect when adja-
cent to a waggle dancer (Michelsen, 1993,
2003). Several recent studies indicate that the
vibrations produced by a dancing bee are also
transmitted via the comb and help followers
locate dancers. (1) Dances performed on open
empty cells recruit twice as many bees per wag-
gle run compared to dances performed on
capped brood cells (Tautz, 1996). (2) The dis-
tance from which a dancer attracts followers is
greater if the dance occurs on open empty cells
rather than capped brood cells (Tautz and
Rohrseitz, 1998). (3) Vibrations applied to the
rims of open cells are transmitted across the
comb and are even amplified through reso-
nance if their frequency is approximately 20 Hz
or 250 Hz (Sandeman et al., 1996). (4) The
200–300 Hz comb vibrations produced during
waggle dances have amplitudes above the
detection threshold of the subgenual organ
(Kilpinen and Storm, 1997; Nieh and Tautz,
2000). These findings leave little doubt that
dancing bees rely, at least in part, on substrate-
borne vibrations in their recruitment commu-
nication. They also suggest that if the transmis-
sion of these comb-borne vibrations is hindered,
then recruitment is weakened.
In this study, we used laser vibrometry to
investigate whether combs built with plastic
foundation, relative to combs built with
beeswax foundation, are inferior at transmit-
ting 250 Hz vibrations. In addition, we used
behavioral measurements to investigate whether
waggle dances produced on combs built with
plastic foundation, relative to those produced
on combs built with beeswax foundation, are
less effective in recruiting bees to a food
source.
2. MATERIALS AND METHODS
2.1. Laser vibrometry study
in the laboratory: April 2003
We used 3 frames that were fitted with foundation
and then filled (by bees) with beeswax comb:
(1) BW: beeswax foundation in a wooden frame,
(2) PW: plastic foundation (Snap-in Pierco) in a
wooden frame, and (3) PP: plastic foundation built
into a plastic frame (Pierco 1-Piece Frame/Founda-
tion unit). All were full-depth Langstroth frames. In
June 2002, the three frames were put in one strong,
comb-building colony. It completely filled the
frames with beeswax comb and used them for honey
storage. In mid September 2002, the honey was
extracted from the three frames of comb and in early
October, when no nectar was being collected, the
combs were returned to the comb-builder colony,
which cleaned and repaired them. In mid October,
the combs, empty of honey and perfect in form, were
retrieved for study.
When taking measurements, we fastened each
frame horizontally to a vibration-free table with a C-
clamp at each end. The bees built the cells in each
comb such that two walls of each hexagonal cell
were vertical when the frame was vertical. We dis-
placed the rims of vertical cell walls from the side
and in the plane of the comb face, thereby approxi-
mating the vibration pattern produced by a bee mak-
ing vertically-oriented waggle runs on the comb.
Displacements were made with a bilayered piezo-
electric transducer (BM/ML 60/40/300, Piezo-
mechanik, München) generating 250 Hz vibrations.
Further technical details on the methods of loading
vibrations into the combs are given in Sandeman
et al. (1996).
The amplitude of cell rim displacement was
measured with a laser vibrometer, first at the rim
where the displacements were applied (input point)
and then at a rim, four cells away (output point, see
Fig. 1). Small reflective flags (0.5 × 0.5 mm) were
attached to the cell rims being measured. Input dis-
placement amplitude was set at 1 µm and the result-
ing output four cells away was measured. Measure-
ments were made only after the signal from the laser
vibrometer had stabilized for 30 s. Each measure-
ment was repeated 5 times.
Each frame of comb contained 45 rows of
cells. Measurements were made on cells in rows 2,
6, 10, …, 22 using cells one quarter of the horizontal
distance across the comb (see Fig. 1). Thus the six
measurement points were spread over the comb
region that corresponds to the “dance floor” (see
Fig. 5.4 in Seeley, 1995). A set of measurements was
made for each comb when the comb was 100%, 50%,
25%, and 0% attached to the frame along the sides
and bottom. Each comb started out 100% attached
to its frame, and was detached to increasing degrees
by melting away 4-mm strips of comb (see Fig. 1).
Even when 100% detached along the sides and bot-
tom of the frame, each comb remained fully attached
along the top.

Plastic foundation and recruitment communication 515
2.2. Behavioral study in the field: July
2004
We set up an observation hive, trained 10 bees to
a feeder 350 m from the hive, and then video
recorded the dances of these bees and measured their
recruitment of other bees to the feeder. Data were
collected three times per day, each time with a dif-
ferent frame of comb as the dance substrate. We then
reviewed the video recordings to count the number
of waggle runs produced on each type of comb.
Using these waggle run counts and the recruit
counts, we calculated the waggle dance effective-
ness (measured as recruits per waggle run) for the
three types of dance substrate.
This experiment utilized six frames of comb, two
of each of the three types already mentioned, and
prepared as described above. During data collection,
one comb at a time was placed in the lower position
in a two-frame observation hive which contained
one frame of brood comb (upper) and one frame of
storage comb (lower) (shown in Seeley, 1995,
pp. 71–73). This hive housed a colony consisting of
approximately 4000 worker bees and one queen bee.
A sheet of queen excluder material between the two
frames prevented the queen from laying eggs in the
lower comb. A wedge just inside the hive’s entrance
forced all foragers to enter and exit the hive from one
side of the hive. Because returning foragers perform
their waggle dances shortly after entering the hive,
this wedge established a “dance floor” on one side
of the lower comb.
The observation hive colony was established at
the Cranberry Lake Biological Station (44°09’N,
74°48’W), in the Adirondack State Park, in northern
New York State. This study site is surrounded by for-
est and lakes, hence there are few natural food
sources for bees, which makes it easy to train bees
to forage at a feeder. The hive was mounted in a small
hut (see Fig. 4.3 in Seeley, 1995) and the dances were
recorded with a camera (Panasonic WV-F250B)
whose field of view covered the dance floor. Vide-
otapes were analyzed using a variable-speed video-
deck (JVC BR-S525U). Dances were illuminated by
the daylight that entered the hut through its translu-
cent roof. We used a feeder that provided scented
sucrose solution ad libitum, as described in Seeley
(1995). The sucrose solution was generally a
1.50 mol/L solution, but at times it was adjusted
higher or lower to get the desired level of dancing.
We collected data on 6 days (see Tab. I), using
one set of combs (BW, PW, and PP) on the first
Table I. The arrangement of the 6 trials of the experiment, with the two sets of combs. BW = beeswax
foundation in wooden frame. PW = plastic foundation in wooden frame. PP = plastic foundation in plastic
frame (“one-piece frame/foundation”).
Comb
set Date Time
Tes t
series
Number of
waggle runs
Number of
recruits
1 10 July 04 11:20–14:00 BW-PW-PP 2287 107
1 11 July 04 11:00–13:40 PW-PP-BW 2999 141
1 17 July 04 11:20–14:00 PP-BW-PW 3437 139
2 20 July 04 11:20–14:00 BW-PW-PP 663 24
2 23 July 04 11:00–13:40 PW-PP-BW 1714 69
2 24 July 04 11:40–14:20 PP-BW-PW 2261 111
Figure 1. Diagram of a test comb,
showing the locations of the vibration
input and output measurements, and
the way the comb was detached in three
stages from the sides and bottom of
the frame. The double-headed arrow
between the measurement points indi-
cates the displacement direction for the
input vibrations.

516 T.D. Seeley et al.
3 days and the other set on the second three days. On
each day of data collection, we followed a standard
protocol. We started with the feeder empty and a
standard comb (not a test comb) in the lower position
of the observation hive. At 9:00, we filled the feeder
and soon bees began visiting it and recruiting nest-
mates. (Previously, we had trained 10 labeled bees
to the feeder. Recruits to the feeder were easily
detected as unlabeled bees.) Once recruits began
arriving, we began capturing them in zip-lock
freezer bags, keeping tallies of the number of recruits
captured per 10 min period. This capturing and tal-
lying of recruits continued without interruption until
we ended the experiment in the afternoon. As soon
as our 10 labeled bees were regularly producing
dances for the feeder (by 11:00 or so), we took the
lower frame out of the observation and replaced it
with the first of the 3 test combs. We then shook the
bees off the removed frame onto a board mounted at
the hive entrance, whereupon they crawled back into
the hive. Next, we gave the colony 20 min to recover
and then we started video recording the dances of the
10 labeled bees. We continued this recording for
40 min, all the while capturing recruits at the feeder.
When the 40-min period was over, we repeated the
comb swapping procedure described above, replac-
ing test comb 1 with test comb 2, gave the colony a
20 min period for recovery, and then resumed video
recording the dances, now on a different substrate.
After 40 min of video recording for test comb 2, the
entire procedure was repeated for test comb 3. When
this was finished, we replaced test comb 3 with the
standard comb, and shut off the feeder to end the
experiment for the day. As is shown in Table I, this
protocol was performed on 3 days for each of the two
sets of test combs, but on each day we presented the
test combs in a different order, to control for possible
order effects. When we shifted to the second set of
combs, we switched to a new set of 10 recruiter bees.
Also, part way through the experiment, on 21 July,
we added approximately 1500 bees to the observa-
tion hive to restore its strength.
On each day, and for each of the three frames of
comb, we obtained 4 values of dance effectiveness,
one for each 10-min block of the 40 min of data col-
lection for each type of comb. We used the average
of these 4 values as our estimate of dance effective-
ness for a particular type of comb on a given day. We
tested the three frames of comb in each comb set on
three days, using a different presentation order each
day (see Tab. I), hence we used a complete block
design to look for effects of comb type and presen-
tation order. We analyzed our results for each comb
set with a two-way analysis of variance (ANOVA)
without replication, and so tested for significant dif-
ferences in dance effectiveness in relation to two fac-
tors: comb type and presentation order.
3. RESULTS
3.1. Laser vibrometry study
Figure 2 shows the results of the vibration
amplitude measurements made with the BW
Figure 2. Results of measure-
ments of the amplitudes of cell
wall vibrations. Each value is
the mean vibration amplitude
measured at a point 4 cells from
the input point. The input vibra-
tion was always a 250 Hz vibra-
tion with 1.0 µm amplitude.
Measurements were made when
comb was 100%, then 50%,
then 25%, and finally 0%
attached to the sides and bottom
of the frame (see Fig. 1).

Plastic foundation and recruitment communication 517
and PW combs. In the BW comb, there was
good transmission of the 250 Hz vibrations
from input point to output point 4 cells away,
though the degree of comb attachment to the
frame strongly influenced the pattern of vibra-
tion transmission. When the BW comb was
100% attached to the wooden frame, we found
a detectable level of vibration only in the cell
rows in the middle of the comb. In contrast,
when the BW comb was 0% or 25% attached
to the wooden frame, we found a detectable and
essentially uniform level of vibration in all the
cell rows, from bottom to middle of the comb.
Curiously, when the comb was 50% attached to
the frame, we found a considerable level of
vibration in bottom and middle cell rows, but
not in intermediate cell rows.
The results differ markedly for the PW
comb, which showed poor transmission of the
250 Hz vibrations. We found a detectable level
of vibration only when the PW comb was little
(0 or 25%) attached to the frame, and even then
only in the bottom cell row.
We do not show the results for the PP comb
in a graph because they are easily reported in
words: for all cell rows, and for all degrees of
attachment of comb to frame, we detected no
vibrations in cell walls located 4 cells from the
input point.
3.2. Behavioral study
Table II summarizes our results on the effec-
tiveness of waggle dances performed on differ-
ent types of comb. We found no effect of comb
type on dance effectiveness. The mean dance
effectiveness was similar for the three types of
comb: 0.05 recruits per waggle run (see Fig. 3).
ANOVA tests show that the mean values of
dance effectiveness do not differ significantly
among comb types, both for comb set 1 (F2, 4 =
0.16, P > 0.90) and for comb set 2 (F2, 4 = 2.94,
P > 0.25).
We did find, however, a strong effect of
position in the test series on dance effectiveness
(see Fig. 4). Regardless of comb type, the comb
presented at the start had a lower mean dance
effectiveness (approx. 0.025 recruits per wag-
gle run) than the comb presented in the middle
(approx. 0.050 recruits per waggle run), which
in turn had a lower mean dance effectiveness
than the comb presented at the end (approx.
0.068 recruits per waggle run). ANOVA tests
show that the mean values of dance effective-
ness differ significantly among test positions,
both for comb set 1 (F2, 4 = 21.95, P < 0.01) and
for comb set 2 (F2, 4 = 8.85, P < 0.04).
Table II. Mean values of dance effectiveness (recruits per waggle run) for two sets of combs. For both sets,
we show the measured value of dance effectiveness for each combination of comb type and test position.
Comb set 1 results Comb set 2 results
Position in test series Position in test series
1st 2nd 3rd Row means 1st 2nd 3rd Row means
Comb type BW 0.027 0.042 0.071 0.047 0.023 0.048 0.066 0.046
PW 0.022 0.054 0.060 0.045 0.022 0.050 0.088 0.053
PP 0.027 0.058 0.061 0.049 0.032 0.062 0.063 0.052
Column means 0.025 0.051 0.064 0.026 0.053 0.072
Figure 3. Comparison of the three types of comb
with respect to mean dance effectiveness.

518 T.D. Seeley et al.
The results described so far are based on
bees performing dances that were illuminated
by daylight for video recording. Having found
no effect of comb type on dance effectiveness,
we wondered if this was because the dances
were performed in daylight. Perhaps dance fol-
lowers had easily found dancing bees by seeing
their movements rather than by detecting their
comb vibrations. If so, then differences among
the comb types in vibration transmission would
not matter.
To see if we had obscured an effect of comb
type on dance effectiveness by studying dances
performed under lighted conditions, we per-
formed another test with the combs of comb
set 1. For each comb, we checked whether
dances were less effective when performed on
a darkened vs. a lighted dance floor. We rea-
soned as follows. Hypothesis: dances on combs
with plastic foundation were as effective as
dances on combs with beeswax foundation
because dances produced on a lighted dance
floor are unusually easy to find. Prediction:
dances on combs with plastic foundation will
be less effective on a darkened dance floor than
on a lighted one. To check this prediction, we
made recruitment measurements under con-
stant conditions except that part of the time the
observation hive’s cover was off (dance floor
lighted) and part of the time the cover was on
(dance floor darkened). The results are shown
in Figure 5. For all three types of comb, we
found no sign of reduced dance effectiveness
with darkened dance floor.
4. DISCUSSION
Do combs built with plastic foundation
hinder the bees’ ability to communicate with
waggle dances? Evidently, they do not. In our
Figure 4. Comparison of the three positions in the
test series with respect to dance effectiveness.
Figure 5. Results of tests for effect of light on
dance effectiveness. The number of bees recruited
every 10 min by 10 dancers visiting a feeder was
measured for a 150 min period. For the first 50 min
and last 50 min, the observation hive was not
covered, so the dances were performed in daylight.
For the middle 50 min, the observation hive was
covered, so the dances were performed in darkness.
The same result was found for all three types of
comb: no difference in the mean number of recruits
per 10 min between the cover-off and cover-on
times.

Plastic foundation and recruitment communication 519
behavioral tests, we found no sign whatsoever
of reduced dance effectiveness when bees
danced on combs built with plastic foundation
relative to combs built with beeswax founda-
tion. Because we measured the effectiveness of
dances performed in daylight in an observation
hive, we must ask whether our conclusion (no
hindrance of dance communication by plastic
foundation) also holds true for dances per-
formed in darkness in a normal hive.
Two lines of evidence indicate that we did
not get falsely negative results. The first is that
previous investigators of substrate effects on
waggle dance communication (Tautz, 1996;
Tautz and Rohrseitz, 1998) also studied dances
performed in daylight and they found strong
substrate effects, e.g., differences in recruit-
ment success between dances performed on
open cells vs. sealed cells. Clearly, substrate
differences can influence dance effectiveness
even when dances are performed in daylight.
The second line of evidence is the finding in the
present study that when we darkened the dance
floor, we found no drop in recruitment by bees
dancing on combs built with plastic foundation.
If these combs were inferior substrates for
dances performed in darkness, then we should
have seen reduced recruitment when we dark-
ened the dance floor, but we did not. We
believe, therefore, that our main conclusion –
combs built with plastic foundation do not
hinder waggle dance communication – is a true
negative, not a false one.
The finding that combs built with plastic
foundation provide a fully suitable substrate for
waggle dances was, at first, surprising because
our laser vibrometry study revealed much
poorer transmission of 250 Hz vibrations in
plastic-based vs. beeswax-based combs. How
can we resolve this seeming conflict between
the two parts of our study? The most likely
explanation is that dance followers used fre-
quencies other than the 250 Hz that we used in
our laser vibrometry study. When performing
a waggle dance, a bee produces not only 250 Hz
vibrations by activating her thoracic muscula-
ture, but also 15 Hz vibrations by swinging her
body side to side. We suggest that dance fol-
lowers use mainly the 15 Hz vibrations to orient
to dancers. We chose 250 Hz displacements for
our laser vibrometry study because prior inves-
tigations found that unframed combs have a
striking peak in transmission effectiveness
(Fig. 7 in Sandeman et al., 1996) and a sharp
drop in impedance (Fig. 25 in Rohrseitz, 1998)
for 250 Hz vibrations. These same studies also
found, however, that framed combs have poor
transmission and high impedance of 250 Hz
vibrations, but have good transmission and low
impedance of 15–20 Hz vibrations (Fig. 10 and
12 in Sandeman et al., 1996, and Fig. 26 in
Rohrseitz, 1998). If a framed comb is freed
from the frame on bottom and sides, then there
is improved transmission effectiveness of
250 Hz vibrations (a finding confirmed by our
results, see Fig. 2), but even so it remains below
the transmission effectiveness of 15–20 Hz
vibrations. In hindsight, it is regrettable that we
studied the effects of plastic foundation on
vibration transmission using only 250 Hz stim-
uli. This work needs to be extended using 15 Hz
stimuli. We predict that the transmission of
15 Hz vibrations will not differ between
framed combs built with beeswax vs. plastic
foundation.
Besides testing the effects of plastic founda-
tion on dance communication, our behavioral
study also yielded a curious finding about the
waggle dance system of bee communication:
its effectiveness changed markedly over the
course of a day. As is shown in Figure 4, we
found that dances performed early in the day
were only one third as effective in getting
recruits to the feeder compared to dances per-
formed later in the day. This change in dance
effectiveness appears robust, for we found the
same pattern on all 6 days of data collection
(Tab. II). How did this change in dance effec-
tiveness arise? Perhaps it reflects a gradual acti-
vation of the study colony’s foragers over the
course of a day, so that more bees were ready
to follow dances later in the day. Another pos-
sibility, however, is that bees that had followed
a dance for the feeder and left the hive to search
for it somehow found it easier to find the feeder
later in the day. Future studies are needed to see
if the daily increase in dance effectiveness that
we found is a typical, and if so, to determine
whether it reflects a gradual “waking up” of the
foragers in a honey bee colony.
ACKNOWLEDGMENTS
This project was supported by the National
Research Initiative of the USDA Cooperative State
Research, Education and Extension Service, grant

520 T.D. Seeley et al.
number 2004-35302-14838, and by the Alexander
von Humboldt Foundation, which has provided
TDS with a Research Award. We are also grateful
to Dr. Stephen A. Teale, director of the Cranberry
Lake Biological Station, for continuing to provide
us with space and facilities at this lovely field sta-
tion.
Résumé – Les fondations en plastique gênent-
elles la communication dansée ? Ces dernières
années les fondations en plastique se sont largement
répandues chez les apiculteurs, mais aucune étude
n’a été faite pour savoir si elles gênaient le recrute-
ment en réduisant la transmission des vibrations du
rayon produites par les abeilles qui exécutent la
danse frétillante. On sait que les danseuses se fient,
au moins partiellement, aux vibrations du rayon pour
communiquer les sources de nourriture, probable-
ment parce que ces vibrations aident les abeilles à
localiser les danseuses dans l’obscurité de la ruche.
Nous avons étudié trois types de rayon : (i) rayon
construit sur une fondation en cire d’abeilles fixée à
un cadre en bois (BW), (ii) rayon construit sur une
fondation en plastique fixée à un cadre en bois (PW),
(iii) rayon construit sur une pièce unique fondation/
cadre en plastique (PP).
Nous avons utilisé la vibrométrie laser pour compa-
rer les trois types de rayons en termes de
transmission des vibrations produites par les danses.
Les rayons construits sur une fondation en plastique
transmettaient mal les vibrations de 250 Hz. Par con-
tre ceux construits sur une fondation en cire d’abeille
les transmettaient très bien, en particulier lorsque le
rayon était partiellement détaché du cadre en bois
(Fig. 2).
Nous avons réalisé une étude comportementale pour
comparer l’efficacité du recrutement (ER) des dan-
ses sur les différents types de rayons. L’ER a été
mesurée en termes de nombre d’abeilles recrutées
sur un nourrisseur à 350 m par un trajet frétillant pro-
duit pour le nourrisseur. Nous n’avons trouvé aucun
effet (Fig. 3). Les trois types de rayon ont donné
la même valeur moyenne d’efficacité : environ
0,05 recrue par trajet frétillant. Quel que soit le type
de rayon, l’ER différait notablement selon que
c’était le premier, le second ou le troisième rayon uti-
lisé dans la journée. Les danses exécutées en fin de
journée avaient une plus grande efficacité que celles
exécutées plus tôt (Fig. 4). Nous suggérons que
l’augmentation de l’efficacité au cours de la journée
reflète une activation graduelle des butineuses de la
colonie, si bien que les ouvrières sont plus nombreu-
ses à suivre les danses plus tard dans la journée.
Que nous n’ayons pas trouvé un effet du type de
rayon sur l’ER peut être dû au fait que notre expéri-
mentation a été réalisée en lumière du jour. Cela a
pu aider les abeilles à trouver les danseuses en
voyant leurs mouvements plutôt qu’en sentant leurs
vibrations. Nous avons testé cette hypothèse en cher-
chant si les danses exécutées sur les rayons PW et
PP, qui ont une faible transmission des vibrations de
250 Hz, voyaient leur efficacité diminuer lorsque le
plancher des danses était assombri. Nous n’avons
trouvé aucun signe montrant une réduction de l’ER
(Fig. 5). Nous en concluons que l’absence d’effet du
type de rayon sur l’efficacité des danses est un résul-
tat négatif vrai et non un faux. Il est clair qu’un rayon
construit sur une fondation en plastique constitue un
substrat tout à fait adapté à la communication par la
danse frétillante.
Apis mellifera / vibration du rayon / fondation en
plastique / communication / recrutement / danse
frétillante
Zusammenfassung – Behindert eine Kunststoffmit-
telwand der Bienewabe die Tanzkommunikation?
In den letzten Jahren werden von Imkern zunehmend
Mittelwände aus Kunststoff für den Ausbau von
Waben eingesetzt. Es war bisher nicht untersucht, ob
solche Mittelwände die Rekrutierung zu Futterplät-
zen beeinträchtigen, etwa indem sie die von den
Bienen beim Schwänzeltanz erzeugten Schwingun-
gen reduzieren. Es ist bekannt, dass tanzende Bienen
für ihre Futterkommunikation zumindest teilweise
auf über die Wabe weitergeleiteten Schwingungen
angewiesen sind. Vermutlich helfen diese Schwin-
gungen tanzenden Bienen in der Dunkelheit des
Nestes zu lokalisieren. Wir untersuchten die drei fol-
genden Wabenarten: Auf Mittelwänden oder auf
Plastikmittelwänden gebaute und in ein Holzrähm-
chen eingepasste Waben (BW bzw. PW), und eine
aus einem Stück bestehend Einheit aus Mittelwand
und Rähmchen ausgebaute Wabe (PP).
Wir benutzten Laservibrometrie, um die drei
Wabentypen auf ihre Weiterleitungseigenschaften
für die beim Tanz entstehenden Vibrationen zu
untersuchen. Auf Plastikmittelwänden gebaute
Waben leiteten Vibrationen mit einer Frequenz von
250 Hz schlecht weiter. Waben mit Wachsmit-
telwänden leiteten diese Frequenz dagegen gut
weiter, besonders wenn sie teilweise nicht mit dem
umgebenden Rähmchen verbunden waren (Abb. 2).
Anhand einer Verhaltensuntersuchung verglichen
wir die Effizienz der Rekrutierung von Bienen durch
die Tanzkommunikation auf den drei Wabentypen.
Als Rekrutierungseffizienz wurde die Anzahl der zu
einer Futterstelle angeworbenen Bienen pro gezeig-
tem Schwänzeltanz bestimmt. Wir fanden keinen
Einfluss des Wabentypes auf die Rekrutierungseffi-
zienz (Fig. 3); auf allen drei Wabentypen war diese
ähnlich und betrug ungefähr 0.05 rekrutierte Bienen
pro Schwänzellauf. Wir fanden allerdings einen
Positionseffekt in der Testserie. Unabhängig von
dem Wabentyp gab es deutliche Unterschiede in der
Effizienz der Tänze je nachdem ob diese auf der ers-
ten, zweiten oder dritten der im Verlauf des Tages
genutzten Waben stattfanden.
Später am Tag durchgeführte Tänze waren effektiver
als früh am Tag durchgeführte (Fig. 4). Wir nehmen
an, dass dieser Anstieg in der Tanzeffektivität im
Tagesverlauf auf die allmähliche Aktivierung der
Sammlerinnen in einem Volk zurückzuführen ist,

Plastic foundation and recruitment communication 521
wonach später am Tag mehr Sammlerinnen bereit
wären den Tänzen zu folgen. Es hätte sein können,
dass ein Wabeneffekt deshalb nicht gefunden wurde,
weil die Experimente unter Tageslichtbedingungen
durchgeführt wurden unter denen es für die Bienen
leicht gewesen sein könnte, Tänzerinnen auch ohne
Wabenvibrationen anhand ihrer Bewegungen zu fin-
den. Wir untersuchten daher bei verdunkeltem
Tanzboden, ob die Tanzeffektivität auf PP Waben
mit ihrer schlechteren Weiterleitung für Vibrationen
von 250 Hz geringer war als bei PW Waben. Hierbei
ergab sich kein Anzeichen einer verringerten Tanz-
effektivität (Fig. 5) woraus sich schließen lässt, dass
das negative Ergebnis über einen Wabeneffekt ein
wahres, nicht falsches Negativergebnis darstellt.
Offensichtlich stellt eine auf einer Plastikmittelwand
gebaute Wabe einen vollständig geeigneten Unter-
grund für die Schwänzeltanzkommunikation dar.
Apis mellifera / Wabenvibration / Plastikmit-
telwände / Futterkommunikation / Schwänzel-
tanz
REFERENCES
Crane E. (1990) Bees and beekeeping, Cornell
University Press, Ithaca, New York.
Esch H. (1961) Über die Schallerzeugung beim
Werbetanz der Honigbiene, Z. Vgl. Physiol. 45,
1–11.
Kilpinen O., Storm J. (1997) Biophysics of the
subgenual organ of the honeybee, Apis mellifera,
J. Comp. Physiol. A 173, 135–141.
Michelsen A. (1993) The transfer of information in
the dance language of honeybees: progress and
problems, J. Comp. Physiol. A 181, 309–318.
Michelsen A. (2003) Signals and flexibility in the
dance communication of honeybees, J. Comp.
Physiol. A 189, 165–174.
Michelsen A., Kirchner W.H., Lindauer M. (1986)
Sound and vibrational signals in the dance
language of the honeybee, Behav. Ecol.
Sociobiol. 18, 207–212.
Nieh J.C., Tautz J. (2000) Behaviour-locked signal
analysis reveals weak 200–300 Hz comb
vibrations during the honeybee waggle dance, J.
Exp. Biol. 203, 1573–1579.
Rohrseitz K. (1998) Biophysikalische und
ethologische Aspekte der Tanzkommunikation
der Honigbienen (Apis mellifera carnica Pollm.),
Ph.D. thesis, Universität Würzburg.
Sandeman D.C., Tautz J., Lindauer M. (1996)
Transmission of vibration across honeycombs
and its detection by bee leg receptors, J. Exp.
Biol. 199, 2585–2594.
Seeley T.D. (1995) The wisdom of the hive, Harvard
University Press, Cambridge, Massachusetts.
Tautz J. (1996) Honeybee waggle dance: recruitment
success depends on the dance floor, J. Exp. Biol.
199, 1375–1381.
Tautz J., Rohrseitz K. (1998) What attracts honeybees
to a waggle dancer? J. Comp. Physiol. A 183,
661–667.
Wenner A. (1962) Sound production during the
waggle dance of the honeybee, Anim. Behav. 10,
79–95.
To access this journal online:
www.edpsciences.org