Electromagnetic Radiation: Influences on Honeybees (Apis mellifera)
Kimmel, Stefan1*; Kuhn, Jochen2; Harst, Wolfgang3; Stever, Hermann3
1 Institute for Environmental Sciences, University of Koblenz-Landau/Campus Landau, Germany
2 Institute of Science and Science Education (ISSE), Department of Physics, University of Koblenz-Landau/Campus
3 Institute of Educational Informatics, University of Koblenz-Landau/Campus Landau, Germany
* Author for correspondence (e-mail: email@example.com)
Focussing on the influences of non-ionizing radiation towards the behaviour of the honeybee (Apis
mellifera), the here presented study reports partially significant results. Nowadays, there is a certain
increase of radiation impact in today’s environmental ecosystems, and the influence of higher
frequencies on honey bees is analyzed by the workgroup “educational informatics” since 2001
(Stever & Kuhn 2001; Kuhn & Stever 2001; Kuhn & Stever 2002). In ecotoxicology, the honeybee
(Apis mellifera) is of great importance as a tested species for agricultural chemicals, e. g. plant
protection products and pesticides. In this case, significant variations in the behaviour of Apis
mellifera under the influence of non-ionizing radiation were tested. The presented data set is based
on earlier studies from 2005, which showed significant differences in returning, 39.7% of the non-
irradiated bees came back compared to 7.3% of the irradiated ones.
Standard commercial DECT telephones were used as exposition source. Concerning possible
variations in behaviour an experimental setup with irradiated and non-irradiated bee hives was
assembled. The main emphasis of this study was the investigation on significant changes in the
foraging flight under electromagnetic radiation influence.
Keywords: Honeybees, electromagnetic radiation, learning process, changing behaviour,
This study focuses on the effects of an electromagnetic exposition caused by DECT Telephones on
the behaviour of the honeybee. All researches and tests have been carried out at the
Dienstleistungszentrum Ländlicher Raum (DLR), Fachzentrum Bienen und Imkerei, in Mayen
during June/July 2006. There have been several scientific investigations throughout the past years
concerning the electromagnetic radiation and its effects (Greenberg et al., 1981; Hartsgrove et al.,
1987; Eulitz et al., 1998; Rothmana, 2000). In context of the increasing non-ionising radiation, this
study focus on the effects of electromagnetic exposition on the behaviour of the honeybee.
Especially towards crop pesticide testing, Apis mellifera is a confirmed test species in
ecotoxicological researches. Furthermore the honeybee shows an effective learning behaviour,
resulting in olfactory amenities and even forms, structures and faces and also in training abilities on
certain plants (Vareschi & Kaissling, 1970; Hoefer & Lindauer, 1976; Dyer et al 2005). Apis
mellifera is well suited as a bioindicator, because its brain anatomy as well as the learning regions
of the bee brain are well known (Menzel & Müller, 1996; Zhang et al., 1999; Schwärzel & Müller,
2006) and the brain structure of the honeybee concerning associative learning is comparable to
those of vertebrates (Bliss & Collinridge, 1993; Eichenbaum, 2004; Giurfa, 2003; Schwärzel &
Müller, 2006). Concerning the effects of electromagnetic radiation it might be possible to draw
conclusions towards other organisms based on the results according to the monitoring of honeybees.
2. MATERIAL AND METHODS
2.1 Physical aspects
In this case, base stations of everyday used DECT telephones (Digital Enhanced Cordless
Telecommunications) were fixed as radiation sources. Investigating on non-thermal influences of
electromagnetic fields towards the learning behaviour of bees requires an exposition with an
appropriate radiation frequency. The stations send out continually electromagnetic radiation with a
frequency fS ≈ 1900 MHz and an average transmitting power PS of 10 mW. The peak power is 250
mW and the sending signal throughout a talk is frequency modulated and pulsed with a frequency fp
of 100 Hz. For this study the base station is used as radiation source at a permanent standby mode
reached with an average transmitting power of PS = 2.5 mW. To analyze a possible effect of the
radiation intensity, cubic radiation shields made of reed and clay were build around some of the
DECT base stations (experimental group 2, EG2, refer to 2.2), which is completely permeable to the
low-frequency pulse mentioned above, but enables a reduction of the high-frequency sending
radiation about 50% (Moldan & Pauli, 2000). We also installed metal lattices (width 1x1 mm)
between the exposed bee hives (experimental group) in order to avoid possible influences of the
radiation on the non-exposed bee hives (control group, CG).
The stations were put at the bottom of a beehive, right under the honeycombs (Fig. 1).
Wooden frames with bee cells
Cubic radiation shield of reed
around the DECT base
Fig. 1: Position of DECT base station within a bee hive
2.2 test objects and method
Overall, 16 Bee colonies of Apis mellifera carnica were used as test objects. With a permanent
connection establishment between the wireless cells and the DECT base stations, the average
sending power Ps could be estimated. Five of eight exposed hives were under fully electromagnetic
exposure (experimental group EG 1), while in three of the exposed colonies the radiation was
shielded down to 50% (experimental group EG 2, see Fig. 1). The following figure shows the whole
Fig. 2: Experimental set up
For one test run, 15 bees flying out of the hive were trapped with the help of plastic tubes at the hive
entrance. All catched bees were short term paralyzed (using CO2) and got marked with a marker dot
on the thorax. At a distance of about 500 m to the hive all marked bees were set free simultaneously
and got timed from that moment. Concerning the returning behaviour, in every test run irradiated
bees were checked against non-exposed ones (EG 1 vs. CG; EG 1 vs. EG 2; EG 2 vs. CG). Time of
flight for every single bee as well as certain aspects like weather, temperature and hive activity in
common was reported. The returning bees were intercepted at the bee hive's entrance and the
returning time was documented. The observation time lasted 45 minutes, bees that came back
afterwards were disregarded in order to avoid possible mistakes for following test runs.
All results are based on collected data from June, 28.–29., and July, 9.-19., in 2006.
52 paired comparisons had to be taken into consideration, 31 pairs of bee colonies “EG 1 vs. CG”,
15 pairs “EG 2 vs. CG” and finally 6 pairs ”EG 1 vs. EG 2”. In 22 of the 31 tested pairs “EG 1 vs.
CG” more of the non-exposed bees (CG) returned to their colonies. With the total amount of
returned bees (non-exposed 293 = 63.0%, exposed 229 = 49.2%) the tendency of earlier researches
(Stever et al., 2005) could be confirmed.
Overall, 482 (63%) bees of the CG, 203 (56.4%) bees of the EG 2 and 365 (54.1%) bees of the EG
1 returned to their hive. These differences between the groups were not significant (Kruskal-Wallis
One of the main problems of the statistical analysis was to combine the amount of returning bees
with their returning time in one single value (tnr), which reflects the predominate circumstances and
enables a comparison between different testing properties. The following term presents a possible
solution to this problem:
tnR = nR * 46 – Σ tR
In this term the amount of returning bees nR is multiplied by the maximum observation time + 1,
then the sum of the returning time of each bee tR, which were actually returned, is subtracted from
this product. To standardize the tnR-Index its term is related to the maximum value (for this study
(nRmax = 15 bees * [45min + 1]), the tnmax is up to 690):
tn = tnR * 100 / tnmax
CG Non-studied bee hive
Metal lattice EG 1 EG 2
It became obvious that in 29 of 31 tested pairs the tn-index was higher for non-exposed bees, with
tn-index-mean ratio of 48.97 (SD 20.74) for non-exposed bees against a tn-index-mean ratio of
38.48 (SD 16.41) for exposed bees.
The comparison in pairs between bees of the EG 1 with bees of the CG is presented in Fig. 3:
non-exposed - fully exposed
1 2 3 4 5 6 7 8 9 1011121314 151617 181920212223242526 272829 3031
Fig. 3: tn-Index comparison CG (green) vs. EG 1 (red), decreasing ranks
3.2 tn-index mean comparisons for all tested groups
All deviations concerning the mean ratio for each compared group are tested for significant
differences by conducting the t-test for independent variables.
Referring to the results of the t-test, mean differences between non-exposed and exposed honeybees
(CG vs. EG 1) were significant (p = 0.031), whereas the other two tested pairs (CG vs. EG 2; EG 1
vs. EG 2) showed no significant differences.
Furthermore no correlations of uncontrollable factors like weather, temperature and flight frequency
with the tn-Index were found, which shows that there is no influence of these uncontrollable factors
concerning our results.
Obviously certain factors concerning the experimental set up are hard to control, but aspects such as
homogenous bee colonies, the location of the tested hives and the interaction between studied bee
colonies and disregarded neighbour colonies must be observed and controlled before starting a
following study. Also the testing place should be selected as soon as possible, in order to allow the
bees selecting a preferred region for collecting food.
The results of this study are much more heterogenic compared to our examination in 2005. But
despite this aspect, still a significant difference between exposed and non-exposed bee colonies
could be observed. A correlation between the independent factors weather, flight frequency and
temperature on the tn-index could not be determined. A possible influence of the radiation intensity
could not be proven by this study, because no significant differences between the group-pairs CG
and EG 2 as well as EG 2 and EG 1 could be detected. Also, a clear distinction between the low-
frequency pulse of the DECT base station and its high-frequency sending radiation could not be
drawn, despite the fact that a significant difference between the non-exposed bees and the fully
irradiated ones can be counted as a result of the influence of high-frequency electromagnetic
A certain method to improve the experimental set up can be found in automating the testing
intervals, e.g. by using a lock at the hive entrance for automatically collecting the bees. Finally, it
would be also very important to measure the exact radiation intensity within the hives as well as the
concrete character of the used radiation.
Bliss, T. V. P. & Collingridge, G. L. (1993); A synaptic model of memory: long-term potentiation
in the hippocampus. Nature, Vol. 361 (pp. 31-39).
Dyer, A. G., Neumeyer, C. and Chittka, L. (2005); Honeybee (Apis mellifera) vision can
discriminate between and recognise images of human faces. Journal of Experimental Biology,
Vol. 208 (pp. 4709-4714).
Eichenbaum, H. (2004); Hippocampus: cognitive processes and neural representations that underlie
declarative memory. Neuron, Vol. 44 (pp. 109-120).
Eulitz, C., Ullsperger, P., Freude, G. & Elbert, T. (1998); Mobile phones modulate response
patterns of human brain activity. Neuroreport, Vol. 9, No. 14 (pp. 3229-3232).
Giurfa, M. (2003); The amazing mini-brain: lessons from a honey bee. Bee World, Vol. 84, No. 1
Greenberg, B., Bindokas, V. P. & Gauger, J.R. (1981); Biological effects of a 765-kV transmission
line: exposures and thresholds in honeybee colonies. Bioelectromagnetics, Vol. 2, No. 4 (pp.
Hartsgrove, G., Kraszewski, A. & Surowiec, A. (1987); Simulated biological materials for
electromagnetic radiation absorption studies. Bioelectromagnetics, Vol. 8, No. 1 (pp. 29-36).
Hoefer, I. & Lindauer, M. (1976). Der Einfluss einer Vordressur auf das Lernverhalten der
Honigbiene. Journal of Comparative Physiology, Vol.109 (pp. 249-264).
Kimmel, S., Kuhn, J., Harst, W. & Stever, H. (2007): Effects of Electromagnetic Exposition on the
Behaviour of the Honey Bee (Apis mellifera). ACTA SYSTEMICA - IAAS International
Journal; in press
Kuhn, J. & Stever, H. (2001); Handy-Boom: eine Gefahr für die Imkerei? ADIZ 35/die Biene
137/Imkerfreund 56, Heft 2, S. 12-13
Kuhn, J. & Stever, H. (2002); Einwirkung hochfrequenter elektromagentischer Felder auf
Bienenvölker. Deutsches Bienenjournal, 10, Heft 4, S. 151-154
Locatelli, F., Bundrock, G. & Müller, U. (2005 ); Focal and Temporal Release of Glutamate in the
Mushroom Bodies Improves Olfactory Memory in Apis mellifera. The Journal of Neuroscience,
Vol. 25, No. 50 (pp. 11614-11618).
Menzel R. & Müller, U. (1996); Learning and memory in honeybees: from behaviour to neural
substrates. Annual Review of Neuroscience, Vol. 19 (pp. 379-404).
Moldan, D. & Pauli, P. (2000); Reduzierung hochfrequenter Strahlung im Bauwesen: Baustoffe und
Abschirmmaterialien. Iphofen: Eigenverlag.
Rothmana, K. J. (2000); Epidemiological evidence on health risks of cellular telephones.
The Lancet, Vol. 356, No. 9244 (pp. 1837-1840).
Schmickl, T. (2003). Sammeln, Verteilen und Bewerten von Informationen: Verteilte Intelligenz in
einem Bienenvolk. Wien: Böhlau Verlag.
Schwärzel, M. & Müller, U. (2006); Dynamic memory networks: dissecting molecular mechanisms
underlying associative memory in the temporal domain. Cell. Mol. Life Sci., Vol. 63 (pp. 989-
Stever, H. & Kuhn, J. (2001): Schutz der Bienen vor Handy-Strahlung. Schweizerische Bienen-
Zeitung 124,Heft 9, S. 23-27
Stever, H., Kuhn, J., Otten, C. Wunder, B. & Harst, W. (2005); Verhaltensänderung unter
elektromagnetischer Exposition. Landau: Arbeitsgruppe Bildungsinformatik (http://agbi.uni-
Tautz, J. (1996); Honeybee waggle dance: Recruitment success depends on the dance floor.
The Journal of Experimental Biology, Vol. 199 (pp. 1375-1381).
Vareschi, E. & Kaissling, K.-E. (1970). Dressur von Bienenarbeiterinnen und Drohnen auf
Pheromone und andere Duftstoffe. Zeitschrift vergl. Physiologie, Vol. 66 (pp. 22-26).
Zhang, S. W., Lehrer, M. & Srinivasan, M. V. (1999); Honeybee Memory: Navigation by
Associative Grouping and Recall of Visual Stimuli. Neurobiology of Learning and Memory, Vol.
72 (pp. 180-201).