Dr Barnes' Homepage Link



Bee Colony Collapse cell phone radiation, pesticide or pollution?  A geographical correlation   study is presented using existing public domain mapping with incontrovertible evidence of the former.   Dr.  Chris Barnes Bangor Scientific and Educational Consultants



Some potential reasons for bee decline and colony collapse are briefly reviewed. A geographic correlation study  for the phenomenon in the USA is attempted using public domain maps. What is perhaps most striking is that in North Dakota and parts of the central belt both insecticides are in heavy use and let there are regions where there is no colony collapse.  Likewise there are regions with heavy Carbofuran use in the mid-west without Endosufan use and with and without colony collapse. On the other hand there are no areas with heavy mobile frequency penetration which do not exhibit colony collapse.  From this geographic study then it would appear that mobile phone frequency use has a better spatial correlation with colony collapse than does pesticide use.  





Since the new millennium, bee colony collapse has become a world-wide issue which at the very least could impinge of the cost of certain foodstuffs or at worst has the potential to lead to famine or be portending of mass extinction.  Estimates of the percentage of World food production which depends   on insect pollinators vary wildly across Internet Scientific information sites from 1-70%! Thus a good estimate appears to be 35%.  Certainly if bees and all other insect pollinator died out, human kind would face dire consequences. Pesticides and the Scuttle fly are usually blamed for bee decline (1) but neither of these should know geographic boundaries. Yet when colony collapse in the USA is analyzed stark geographic effects are observed.  


Bees and insects in general have survived on our planet for millions of years and indeed survived through the industrial revolution and on into the mid and late 1900’s some of the most polluted times on our planet. Today atmospheric pollutants have often smaller particle size due to catalytic systems, scrubbers and filters in vehicles and chimneys (2).     PM10 has given way to PM2.5 as a way of judging significant particulate pollution.  


PM2.5 pollutants often contain metallic nanoparticles, especially magnetite (3).  As bees utilize biogenic magnetite in homing and navigation and this biogenic magnetite has an opposite crystal size distribution to geologic magnetite (4), it is reasonable to ask if absorption of the latter into bees via such pollution could possibly influence their decline. Mapping of PM2.5 pollution is readily available for the entire USA as it is for bee colony collapse.    



Conflicting studies presently exist on the interaction of bees with microwave radiation. One study using 2.45 GHZ CW (continuous wave) microwave saw no significant effect on bee foraging ability and only a tiny increase in mortality rate after irradiation(5).  However, there is also a newer body of evidence that electromagnetic radiation, particularly from cell phone towers and mobile technology in general may interfere with bee metabolism and behavior and thus could be causing colony collapse (6, 7).     900 MHz GSM FDMA/TDMA is named in some of the literature to be of particular concern.  The present author also wonders if the bee body could resonate at this frequency allowing the formation of a dielectric antenna. This concept will be explored later.  Also the modulation pulse repetition frequencies of GSM are close to the audio emitted when a bee is actually buzzing (signaling)! Even more modern CDMA transmissions contain audio clutter in their modulation associated with timing and critical power adjustment, the latter is an essential feature of their operation.   Some buzzing, so called piping often serves as short range signaling between workers and hive.  If modulation in cell phone transmissions is interfering with bee signaling they must be able to perceive or demodulate it in some way. Involvement of their magnetite component seems a reasonable assumption.  If bees are resonant at 900      MHz, then it will be interesting to see if as G3 and G4 CDMA systems become more popular and 900 MHz is phased out if colonies recover or if evolution starts to favor larger sized honey bees as transmission frequencies are increased?  Indeed also if different species of insect are affected by the new frequencies involved.  Intuitively, from a resonance point of view, one would perhaps expect wasps and large flies to be affected by 1800 and 2600 MHz and large bumble bees to be affected by the proposed use of   the old UHF TV frequencies in the UK for 2nd generation G4.  It is believed that this   is the first wholly geographic correlation study of this very worrying scenario covering an area as wide as the entire USA.




A number of available maps distribution maps of the variables explored have been obtained by internet image searching and are taken form reliable sources such as MODIS satellite, environmental agencies and the like.  




Quite simply whichever factor is most responsible for colony collapse would be expected to show most individual geographic correlation on the available maps.



Results and Conclusions

The results are derived from readily available on –line mapping.  Figure 1 shows colony collapse. Figure 2 shows overall risk of general pesticide exposure.  Figure 3 shows



mobile technology penetrations in the USA.             l闒粀펤闀粀

             Figure 1 Colony Collapse                                                                       Figure 2 Pesticide risk                           Figure3 Mobile penetration

Figure 1

Figure 4.  Average of daily surface PM2.5 concentrations for January 2001 to October 2002 determined from MODIS and MISR measurements of AOD. White space denotes regions with AOD measurements on fewer than either 40% of days for MODIS or 8% of days for MISR.

The results could not be more visually striking. The correlation with PM2.5 is poor. It is immediately apparent from even a cursory inspection of the maps that there are large swathes of the north western USA which have low PM2.5 but significant colony collapse.  Thus PM2.5 is not thought to be the major cause of colony collapse.  The pesticide risk correlation is equally poor, except in the North East USA. It should be remembered, however, that bees are thought to be far more sensitive to some pesticides than others.  On the other hand there seems   to be an incontrovertible association between regions of the mobile phone coverage map which have low phone coverage and states or regions where colony collapses have not been recorded.   Taking a logical approach it is reasonable to suppose the converse must also be due and that colony collapse appears to be associated, from mapping studies at least, with regions which have a total penetration of mobile phone communications use.  Indeed it would seem that bee colonies are only surviving in any numbers in the regions of low mobile phone technology penetration particularly in the central belt of the USA

Figure 4:  States of the USA

Survival is only happening especially in the more rural states and/or sparsely populated states which logically will be expected to have less mobile phone telephony.  Alaska and North Dakota are of course also mountainous areas and radio propagation is also more limited in mountainous areas.  Strangely, Nevada and Arizona often has very high levels of PM2.5 pollution suggesting that PM2.5 alone   is not a significant factor in bee   colony decline.  It is further instructive to consider population density since one would expect some kind of strong positive correlation between mobile phone use and population density.  Besides mobile phone use on has to consider the possibly of some other as yet under discovered population proportionate anthropogenic factor as a prima fascia cause of colony collapse linked to densely polluted areas.    Take the recent example of ‘round up’ household weed killer. The active ingredient thought to be relatively harmless to humans yet one of the carrier compounds now known to be highly lethal to human cells! Maybe something similar is going on for bees?    If such a scenario was to be the case one should see a better correlation of colony collapse with population density than with mobile phone penetration.


Lower Left Mobile Phone technology penetration
Figure 5 Comparison of correlation by population density and mobile phone technology penetration


The correlation of colony collapse with population density is almost as impressive as the one for mobile phone technology penetration but not quite.  Montanna and Wyoming are the only states which have very low population density and bee colony decline but in these cases it would appear that the state border or boundary regions are heavily penetrated by mobile technology. Perhaps this effects or prevents bee migration in and out of the states?    


Exploring the pesticide risk in more detail.

Pesticide is the general term to cover chemicals used to destroy weeds so called herbicides, mites so called miticides and insects so called insecticides.  Bees are thought to be unaffected by the majority of herbicides but badly affected by certain insecticides.  The following more detailed comparisons ought then to confirm the same and to be further instructive as to the cause of colony collapse.

Figure 6 The correlation between colony collapse and two  herbicides.

Figure 6 shows the correlation of colony decline with Atrazine and Alachlor to be  poor. Indeed bees survive in parts of the central belt where these herbicides are used extensively but where there are patches of poorer mobile technology coverage.  Likewise survival occurs where Benefin is used even in high concentrations.  On the other hand bee colonies are collapsing in areas where none of these herbicides are used.

A similar comparison can be made for insecticides.

Figure 7 The comparison shown above is for general insecticide residues in US water courses.


The correlation in figure 7 is better than with herbicides but not as good as with mobile phone technology, penetration.

It is further instructive to make a comparison with specific  figure 8. Data is available for Endosufan and Carbofuran. The latter of which is highly toxic to bees. map of Endosulfan use

Figure 8

What is perhaps most striking is that in North Dakota and parts of the central belt both insecticides are in heavy use and let there are regions where there is no colony collapse.  Likewise there are regions with heavy Carbofuran use in the mid-west without Endosufan use and with and without colony collapse.


Contrast this with the mobile phone technology penetration correlation where there are no areas whatsoever of heavy mobile technology use without colony collapse.


Thus the conclusion has to be that from at least a geographic correlation perspective mobile phone technology appears to be the main cause of colony collapse.  





The paradox has to be just as for parasites and insecticides, general radio frequency radiation pre-dates colony collapse.     VHF radiation has only been around since the 1950’s and there is some reference in the literature to certain types of bee species declining since that date. This is suggestive that VHF radiation with the then modulation schemes of AM and FM was harmful to bees and/or there have been cumulative effects of pesticides and mites over the last six decades.  

Various mechanisms have been discussed elsewhere by other workers for the interaction of UHF and microwave electromagnetic radiation with honey bees.  Some consider direct interaction of the radio frequency field with the bee’s body and/or biogenic magnetite.  There have also been discussions of the significance of the pulse repetition frequency and modulation schemes in altering bee signaling.  

As stated in the introduction the present author concurs with some of these hypotheses and wonders if via the biogenic magnetite the bee or parts of it resonate at this frequency or allows the formation of a dielectric antenna.  Also the modulation pulse repetition frequencies are close to the audio emitted when a bee is actually buzzing!  Such buzzing often serves as short range signaling between workers and hive.

 It is instructive to calculate if a bee could act as a dielectric antenna.  No dielectric data is available for bee tissue so assumptions must be made on the data available for mammalian tissue.   A bee has a keratinous exoskeleton which needs to be taken into account.  At 900 MHz the wavelength in mammalian cartilage is 5cm and in mammalian tissue averaged across all types is 5.4cm (8). Honey bees are 5-15 mm in length.  Thus honey bees of 12.5 -13.5 mm in length would make quarter wave dielectric antennas at 900 MHz, a common GSM frequency.   At 2100 MHz the wavelength in average tissue is 22.4 mm.  Thus a bee 11.2 mm long would make a half wave dielectric resonator.  Both scenarios could be relevant to radio frequency absorption of the bee. If it is in flight the half wave scenario would be favored but if it is at rest on ground plane energy absorption could take place via the quarter wave scenario.   


There is a known study from 1980 where bees were exposed to various power densities of 2.45GHz CW microwave for 30 minute periods (5).  There was no discernible effect on foraging ability but a slightly increased mortality rate after treatment.  It must not be forgotten that mobile communication technologies used pulsed transmissions not CW. There was no major colony collapse in the era of the 1G analogue mobile phones so on has to conclude that the pulsed aspect is critically significant.


The studies of VP Sharma, NR Kumar - Current Science (Bangalore), 2010 - and D Favre - Apidologie, 2011 – Springer both conclude that mobile phone signals lead to colony collapse. The latter elicits induced piping in workers. A possible criticism of both these studies is that their effects are elicited by mobile phone handsets being mounted very close to bee hives. Handsets give out different modulation pulses to base stations and also radiate significant low frequency magnetic fields due to current swings form their batteries for instance when changing form standby to active mode, when handshaking and when power adjusting.       

It is crucial that we find out how much influence these magnetic fields have compared with the pulsed microwave component and its demodulation within the bee. Certainly this study suggests that overall use of cellular radio technology is responsible for bee colony collapse. In the light of recent understanding of electromagnetic wave generation and propagation, quantum mechanical effects and coherent detection may also be relevant in the destruction or alteration of the bee’s magnetic memory map.      

Further work

If, as and when the data becomes available, the author hopes to make a similar study for the UK and/or Europe and/or Australasia. 

The author has found quantum mechanical effects are involved in similar biological paradox, namely that of tree die-back and decline, wherein clusters of trees are more likely to be affected at specific distances from transmitters and also in the magneto-acoustic phenomenon known as the Hum (9,10). Such effects have also been described in cancer epidemiology in relation to TV transmitters (11). Until fairly recently classical physicists and engineers would be very dismissive of such findings but the world of science should take not in the light of supportive evidence in the very nature of electromagnetic radiation itself (12)     Bee keepers and bee keeping associations could provide local data on such a scale that would be required to ascertain if the same is true of electromagnetic damage to bees and their hives.   We are faced with an urgent problem here. Einstein is reputed to have said ‘"If the bee disappears from the surface of the earth, man would have no more than four years to live."   A recent review suggests mobile technologies are injuring all sorts of other wild life (13). 



The author wishes to thank Mr. Peter Van Doorn for interesting and stimulating discussions on the topic of bee and other insect decline.   The author also wishes to express his thanks and gratitude to the Italian National Research Council for their excellent on -line dielectrics calculator. 





1.      The case of the empty hives BEE ALER - 1979 -

2.      Air Pollution Emission Control Devices for Stationary Sources -

3.      Atmospheric Environment 33 (1999) 4565}4569 Magnetic biomonitoring of roadside tree leaves: identification of spatial and temporal variations in vehicle-derived particulates J. Matzka!, B.A. Maher"

4.      Morphology of nanomagnetite crystals: Implications for formation conditions. (2005) Faivre, Damien et al.

5.      DOESPS-HoneyBeesExposedToElectromagneticEnergy.pdf 1980

6.      VP Sharma, NR Kumar - Current Science (Bangalore), 2010

7.      D Favre - Apidologie, 2011 ‘Mobile-phone induced honey bee worker piping’

  1. Italian National Research Council on -line dielectrics calculator. 
  2. A new mechanism for Tree Die-Back by Dr Chris Barnes Bangor Scientific and Educational Consultants
  3. Active Prediction of Sites Prone to the HUM by Aharonov- Bohm   Criteria
  4. Distance Related Effects near Radio and Television Transmitters -

A New Perspective - Cyril W. Smith


  1. ‘Toward a more complete Electro-dynamic Theory’, M. Hively and GC Giak, Oak Ridge National Laboratory,  

Electronic Systems, Control and Artificial Intelligence and Systems Engineering

Volume 5, Number 1/2012

10.1504/IJSISE.2012.046745 Online date May 2012


  1. Electromagnetic pollution from phone masts. Effects on wildlife A Balmori - Pathophysiology, 2009 – Elsevier.