ArticlePDF AvailableLiterature Review

Figures

Content may be subject to copyright.
Review
A review of the ecological effects of radiofrequency electromagnetic elds (RF-EMF)
S. Cucurachi
a,
, W.L.M. Tamis
a
, M.G. Vijver
a
, W.J.G.M. Peijnenburg
a,b
, J.F.B. Bolte
b
, G.R. de Snoo
a
a
Institute of Environmental Sciences (CML), Leiden University, P.O. Box 9518, 2300 RA Leiden, The Netherlands
b
National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands
abstractarticle info
Article history:
Received 18 June 2012
Accepted 24 October 2012
Available online 20 December 2012
Keywords:
Radiofrequencies
EMF
Biodiversity
Electromagnetic eld exposure
Ecological effect
Mobile telecommunication
Objective: This article presents a systematic review of published scientic studies on the potential ecological
effects of radiofrequency electromagnetic elds (RF-EMF) in the range of 10 MHz to 3.6 GHz (from amplitude
modulation, AM, to lower band microwave, MW, EMF).
Methods: Publications in English were searched in ISI Web of Knowledge and Scholar Google with no restriction
on publication date. Five species groups were identied: birds, insects, other vertebrates, other organisms, and
plants. Not only clear ecological articles, such as eld studies, were taken into consideration, but also biological
articles on laboratory studies investigating the effects of RF-EMF with biological endpoints such as fertility,
reproduction, behaviour and development, which have a clear ecological signicance, were also included.
Results: Information was collected from 113 studies from original peer-reviewed publications or from relevant
existing reviews. A limited amount of ecological eld studies was identied. The majority of the studies were
conducted in a laboratory setting on birds (embryos or eggs), small rodents and plants. In 65% of the studies,
ecological effects of RF-EMF (50% of the animal studies and about 75% of the plant studies) were found both
at high as well as at low dosages. No clear doseeffect relationship could be discerned. Studies nding an effect
applied higher durations of exposure and focused more on the GSM frequency ranges.
Conclusions: In about two third of the reviewed studies ecological effects of RF-EMF was reported at high as well
as at low dosages. The very low dosages are compatible with real eld situations, and could be found under
environmental conditions. However, a lack of standardisation and a limited number of observations limit the
possibility of generalising results from an organism to an ecosystem level. We propose in future studies to con-
duct more repetitions of observations and explicitly use the available standards for reporting RF-EMF relevant
physical parameters in both laboratory and eld studies.
© 2012 Elsevier Ltd. All rights reserved.
Contents
1. Introduction .............................................................. 117
1.1. Scope ............................................................. 117
1.2. Problem denition ....................................................... 117
1.3. Research focus ......................................................... 118
2. Review method ............................................................ 118
2.1. Criteria of literature search ................................................... 118
2.2. Description of the literature search ................................................ 118
2.2.1. Main search strategy .................................................. 118
2.2.2. Related-references search ................................................ 118
3. General overview of results ....................................................... 119
4. Ecological effects of RF-EMF ...................................................... 119
4.1. Birds .............................................................. 119
4.1.1. Laboratory studies ................................................... 119
4.1.2. Field studies ...................................................... 122
4.1.3. Summary ....................................................... 122
Environment International 51 (2013) 116140
Abbreviations: ELF-EMF, extremely low eld electromagnetic eld; CW, continuous wave; MW, microwave; PW, pulsed wave; GSM, global system for mobile communications;
UHF, ultra-high frequency; VHF, very-high frequency; DECT, digital enhanced cordless telecommunications; UWB, ultra wide band; AM, amplitude modulation; FM, frequency
modulation; GTEM, gigahertz transverse electromagneticcell; UMTS, universal mobile telecommunications system;CDMA, code division multiple access; TDMA, time division multiple
access; WCDMA, wideband code division multiple access; Wi-Fi, Wireless Fidelity; WLAN, wireless local area network; WiMAX, worldwide interoperability for microwave access.
Corresponding author. Tel.: +31 71 527 1478.
E-mail address: cucurachi@cml.leidenuniv.nl (S. Cucurachi).
0160-4120/$ see front matter © 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.envint.2012.10.009
Contents lists available at SciVerse ScienceDirect
Environment International
journal homepage: www.elsevier.com/locate/envint
4.2. Insects ............................................................. 122
4.2.1. Honey bees (A. mellifera)................................................ 123
4.2.2. Fruit ies (D. melanogaster)............................................... 123
4.2.3. Effect on other insects ................................................. 129
4.2.4. Summary ....................................................... 129
4.3. Other vertebrates ........................................................ 129
4.3.1. Summary ....................................................... 131
4.4. Other organisms ........................................................ 131
4.4.1. Summary ....................................................... 131
4.5. Plants and yeasts ........................................................ 131
4.5.1. Summary ....................................................... 134
5. Synthesis ............................................................... 135
5.1. General ............................................................ 135
5.2. Doseeffect relationships .................................................... 135
5.3. From biological to ecological mechanisms and effects ....................................... 135
5.4. Differences between effect and no-effect studies: a possible bias? ................................. 135
5.5. Minimum requirements for studies on ecological effects of RF-EMF ................................ 135
6. Conclusions and recommendations ................................................... 136
Acknowledgements ............................................................. 137
Appendix A. .............................................................. 137
Keywords for literature screening ....................................................... 137
References ................................................................. 137
1. Introduction
1.1. Scope
Anthropocene is a term which has been proposed for the current
epoch, due to the global environmental effects of increased human pop-
ulation, and the economic and industrial development and to the deep
overall domination and contamination of humans over the environment
(Crutzen and Stoermer, 2000; Zalasiewicz et al., 2010). Amongst the
many changes, a radical modication has also taken place in the exposure
of beings to man-made electromagnetic elds. A continuous, chronic, ex-
posure to a wide range of modulated radiofrequency electromagnetic
elds (RF-EMF) burdens all species and groups across the globe.
In terms of mechanisms, the WHO conrms that to date the accepted
health effects ascribable to RF-EMF are caused by temperature elevation
(van Deventer et al., 2011). Though, several studies have identied pos-
sible effects of RF-EMF on organisms, no alternative effect mechanisms
have been conrmed to date. Most of the literature has focused on
human and occupational health, largely based on animal model studies
under laboratory conditions and test subjects exposed to lower frequen-
cies of the spectrum (i.e. extremely low eld, ELF-EMF). From the avail-
able studies, it became clear that, especially under higher dosages, effects
of RF-EMF may be observed. As a response, occupational and human
health threshold values and guidelines, proposed by international orga-
nisations (ICNIRP, 2010), have been increasingly incorporated into na-
tional regulations of states (EU, 2011). However, results are still not
conclusive and there is still some uncertainty about the low dosages
and non-thermal effects applied in some studies which did nd an effect,
and the overall quality of the setup of research in the eld. The ever
increasing use of RF-EMF in the cellular phone ranges (e.g. GSM and
UMTS) and the newer forms of wireless communication (e.g. WiFi,
WLAN and WiMAX), which are rarely present in the available studies, re-
quire new investigations which will look at possible short and long-term
effects.
Over time several monographs and reviews have been compiled
as to the biological effects of RF-EMF on humans, and on animals
(see among others: Michaelson and Dodge, 1971; NCRP, 1986;
Bryan and Gildersleeve, 1988; Adair, 1990; Verschaeve and Maes,
1998; Juutilainen, 2005; Balmori, 2009; Pourlis, 2009; ICNIRP,
2010). While of great relevance for the understanding of the phenom-
enon, these studies lack in the consideration of potential effects
which may directly affect other organisms or ecosystems, because of
the very limited attention which is usually received by the adverse
ecological effects of RF-EMF.
1.2. Problem denition
Limited research and reviews have focused on investigating the pos-
sible ecological effects of RF-EMF. It can be argued that many human-
related biological studies using animal models (e.g. rats and rabbits)
may provide also relevant information about potential ecological effects.
Many ecological endpoints (e.g. fertility, reproduction and growth)
studied at the level of the individual animal, are also crucial from an eco-
logical point of view. Ecology is, one of the sub-disciplines of biology,
which studies all living organisms (including human beings), at all
organisational levels (i.e. from the smallest molecular system to the larg-
est ecosystem levels). Ecology is the scientic study of the distribution
and abundance of organisms and the interactions that determine distri-
bution and abundance (Begon et al., 2005). Those interactions refer to
the abiotic and the biotic environment. By denition ecology focuses
on the higher organisational levels of populations, communities and eco-
systems. Despite the lack of information of the ecological effects of
RF-EMF, following this denition, it is then plausible to link biological
studies with ecological endpoints at the individual animal level to eco-
logical interpretations at a higher organisation level.
This eld of research is of crucial importance for the understanding of
mechanisms of interaction between complex ecosystems and the envi-
ronment. Animal studies have still been identied as a major research
agenda point by the WHO (Van Deventer et al., 2011). The WHO stated
that high priority in the eldshouldbegiventoresearchontheeffects
of RF-EMF on development and behaviour, on ageing and reproduction
of animal subjects. The result of these studies might be ecologically
interpreted, because they include ecologically relevant endpoints.
As far as strictly ecological research has been conducted, it was
mostly presented in the form of non-peer-reviewed grey literature.
A review of Balmori (2009) is the only oriented one at the effects of
RF-EMF on wildlife. However, the contribution by Balmori (2009)
has some methodological issues. The criteria for the screening of the
literature or the rationale for the inclusion or exclusion of relevant
articles have, in fact, not been presented. The review is also missing
a detailed analysis of the selected papers (e.g. of the duration of
117S. Cucurachi et al. / Environment International 51 (2013) 116140
exposure and of the physical parameters) and it includes only studies
nding a signicant effect of RF-EMF.
1.3. Research focus
Evidence suggests that there is a large lacuna in research regard-
ing the ecological effects of RF-EMF. The aim of this contribution is
to conduct a scientically sound review of potential ecological effects
of RF-EMF. Using the denition and guidelines provided in the clinical
sector by Higgins and Green (2006), a systematic review on potential
ecological effects of RF-EMF was performed.
The study focuses on the range from 10 MHz to 3.6 GHz (i.e. from AM
to the lower band MW EMF), using a transparent, comprehensive and
objective substantive review approach and analysis of the available sci-
entic literature on the ecological effects of RF-EMF. The literature search
was based on a clear and objective research strategy (see Section 2)
performed which used two databases: ISI Web of Knowledge and Google
Scholar. The experimental, physical and biological parameters, which
were provided by the selected papers were classied and analysed to
look for trends and possible links between dosages and effects.
Papers evaluating ecological endpoints as part of biological investi-
gations were selected with a focus on higher organisational biological
levels: ecosystem, community, and species. As much as possible also bi-
ological studies, present in biological reviews or in relevant papers, if
relevant from an ecological point of view, were included in this review
and analysed.
A complete review of the biological literature was beyond the scope
of this paper. However, laboratory studies on animals and plants which
investigated biological endpoints can still provide information relevant
for the ecological level.
First in Section 2 the methods are described, in Section 3 the general
results are presented, and in Section 4 the specic results are given for
each of the analysed groups (i.e. birds, insects, other vertebrates, other
organisms, and plants). The nal sections (Sections 5 and 6) provide a
synthesis, with possible links between doseresponse relationship,
the setup and dosage of the studies, together with general conclusions
and recommendations.
2. Review method
2.1. Criteria of literature search
The literature research was conducted, in the second half of 2011,
using ISI Webof Knowledge and Google Scholar databases. Publications
on ecological effects of RF-EMF on all relevant endpoints on non-human
organisms and parts of organisms (e.g. tissues and cells) were taken
into consideration Additional scientic articles published after December
2011 were added upon indication and suggestion of experts.
In order to maintain a high scientic standard for this review paper,
only publications which were peer reviewed were considered. As crite-
rion for peer review, the presence of the publication in the ISI Web of
Knowledge was used. As for papers present only on Google Scholar an
expert selection was made based on the ecological relevance and qual-
ity of the studies. The criteria used were based on quality criteria
dened by relevant methodological reviews (Repacholi and Cardis,
1997; Stam, 2010). Repacholi and Cardis (1997) suggest that reviews
should take into consideration only literature published in scientic
peer-reviewed journals to guarantee a selection of articles free from
methodological deciencies and with rigorous analysis and conclusions.
They also suggest care when dealing with peer-reviewed reports not
published in scientic journals as well as conference abstracts, which
are usually not peer-reviewed. In this review, only peer-reviewed pa-
pers have been selected. In a limited number of cases peer revision
could not be guaranteed: the case of a study conducted by Harst et al.
(2006) on honey bee (Apis mellifera), where no sufcient information
could be found on the review procedure of the relative journal, and
the studies by Van Ummersen (1961, 1963),Carpenter et al. (1960),
and Clarke (1978) which were reported by the (peer-reviewed) review
by Bryan and Gildersleeve (1988).
The literature search was limited to the range of frequencies from
10 MHz to 3.6 GHz. Papers on the biological and ecological effects of
ELF-EMF in the range of 1 Hz100 kHz (e.g. power line elds) were
not considered. Date of publication was not used as a restriction and
all publications falling withinthe selection criteria above were analysed,
including those which did not nd signicant effects.
The keywords used in the literature research process are reported
in the appendix to this review. Two main categories were dened:
RF-EMF specickeywords (e.g. GSM, DECT and 1800 MHz) and ecological
keywords (e.g. growth, population and eco*).
2.2. Description of the literature search
2.2.1. Main search strategy
A step-wise search strategy was conducted to nd the most relevant
articles in the RF-EMF range selected.
As a rst step, the literature research was conducted on the ISI Web of
Knowledge website, which provided 451,031 hits. Since this number of
articles was too large to handle, a selection process was started. The
collection was further rened by selecting only articles, reviews and
proceeding papers as document types (440,528 hits). Then specic
categories were selected: applied physics, cell biology, plant sciences, en-
vironmental sciences, biophysics, zoology, ecology, biology and microbi-
ology. The number of hits was so reduced to 98,620.
In order to reduce the number of hits, all the results clearly outside
the RF-EMF eld of research, or beyond the scope of this review were
excluded. This process reduced the number of hits further to 90,408
hits. A further screening was conducted selecting keywords from
the RF-EMF specic and from the ecological dened groups, using
one or two of RF-EMF keywords singularly or in combination with a
single keyword from the ecological group. The obtained results
ranged from 10 hits to 600. Titles were then screened one by one to
select papers that could be of interest.
An analogous pattern of searches was performed on Google Schol-
ar and only articles that had not yet been found on ISI Web of Knowl-
edge were added. The number of hits for the initial combination of
keywords was 3,600,000, and then reduced with an analogous proce-
dure as described in ISI, but with a more attentive look at the content
and the source of the selected papers.
After this rst step of thesearching process, 709 presumably relevant
articles were identied. A one-by-one screening of titles and abstract
was performed to investigate which papers would meet the dened
criteria (e.g. frequency range and biodiversity exposure to RF-EMF).
This second screening led to a new selection of 307 papers.
A closer analysis of the content of these 307 selected papers re-
vealed that most of them regarded highly specic and strictly technical
biological studies (e.g. rat tissues, cell-line studies, neuronal studies and
calcium signalling), which were difcult to link directly to ecological
effects, and, therefore, discarded. The nal selection was reduced to 55
clearly relevant papers.
2.2.2. Related-references search
As a second step, it was decided to proceed by using a selected
number of the 57 available articles to create a search based on related
referencesto the ones used by their authors. The rst articles used
were those that clearly met the scope of the review in terms of focus
and content: e.g. Balmori (2005),Panagopoulos et al. (2010) and ve
others. The screening of a total of 4000 hits provided 32 additional
relevant hits.
Also a selection of the relevant references was conducted from the
four relevant reviews (Bryan and Gildersleeve, 1988; Juutilainen, 2005;
Pourlis, 2009; Verschaeve and Maes, 1998) and this resulted in 15 addi-
tional articles.
118 S. Cucurachi et al. / Environment International 51 (2013) 116140
Regular updates were conducted until October 2012 to also in-
clude the most recently published relevant literature. After a careful
analysis of all gathered information a total of 113 articles was selected
and described in detail in the following sections. The total number of
experiments carried out in these articles was 152.
3. General overview of results
The biggest share of the articles (c. 90%) involves laboratory stud-
ies with biological endpoints with a clear ecological relevance. The
remaining part were ecological eld studies (Table 1).
Most of the laboratory studies included had growth, development,
behaviour and reproduction/fertility as biological endpoints. The end-
points analysed in eld studies were behaviour, shift in populations
and fertility. In circa 65% of the studies a statistical signicant effect of
RF-EMF on ecological relevant endpoints has been found (Table 1).
There were no clear differences in percentage effects between articles
included in reviews or not included in reviews. Development seemed
to be less signicantly affected in percentage than growth and fertility.
The most represented groups include vertebrates, other than birds
(i.e. predominantly rats, mice and rabbits), then birds and plants. Ar-
ticles which found signicant effects of RF-EMF were found more fre-
quently in the case of birds, insects (i.e. mostly honey bees and fruit
ies) and plants. The group of other vertebrates (Table 1) was equally
distributed among signicant and non-signicant effects. Effects were
signicant in all the articles on other organisms.
The type of endpoints studied differed across groups. Fertility was
the mostly analysed endpoint for the birds. Growth was affected in all
the experiments conducted on plants and other organisms, while it
was affected in 25% of the studies on other vertebrates and ca. 40%
on the birds. The effects of RF-EMF on behaviour were found in thir-
teen of the twenty of the studies on other vertebrates and in 85% ca.
of the studies on insects.
4. Ecological effects of RF-EMF
4.1. Birds
Birds have been widely used to analyse the environmental signif-
icance of exposure to nonionizing radiation. The ability of birds to de-
tect magnetic stimuli has been documented by several studies (see
Keeton, 1971; Thalau et al., 2005; Wiltschko and Wiltschko, 1996;
Wiltschko et al., 2001). A total of 26 articles was selected from the
screened literature with 38 relevant endpoints. With the exception
of ve eld studies, all studies were conducted in a laboratory setting.
Of the 26 studies, 70% have been signicantly related to the effect of
RF-EMF (Table 1). In most cases the effects studied were growth and
fertility and were conducted, until the early nineties, under a continu-
ous microwave system of exposure (i.e. 2450 MHz). The physical pa-
rameters usually reported regarded the measured level of power ux
density and specic absorption rate (SAR). These parameters were
either measured using probes or specic detectors or were based on
the information of the manufacturers of the exposure devices.
Chicken (Gallus domesticus) and Japanese quail (Coturnix coturnix
subsp. japonica) represented the most studied experimental system in
laboratory studies on birds. Approximately 60%of the laboratory studies
considered a system at the embryo or egg stages of development.
4.1.1. Laboratory studies
4.1.1.1. Embryo and egg. In the eighties and early nineties researchers
focused on the effects of MW EMF. There was a high level of interest
especially in the ranges that would be relevant, at that time, for the
possible implementation of new source of renewable power based
on the collection of solar energy in space by means of solar power sat-
ellites (SPS add to abbreviation list) and its transmission to earth via
MW EMF (Glaser, 1968; Wasserman et al., 1984). The three more re-
cent studies (Table 2) investigated the typical cellular phones range
of frequencies.
All the measured physical parameters varied greatly across studies.
The estimated SARs ranged between 0.001 W/kg and 140 W/kg
(Kleinhaus et al., 1995; Van Ummersen, 1961), while the duration
of the exposure was as little as 9 s (McRee and Hamrick, 1977)
with peak values of 45 days (Grigoryev, 2003). The variation which
was found for the power density ranged from 4.4×10
6
mW/cm
2
as
in Reijt et al. (2007) to 400 mW/cm
2
measured in Van Ummersen
(1961).
Table 1
General overview of effects and no-effects studies across articles types, endpoints and
species groups.
General ndings of articles
Count
Included in review (including 80 articles,
4 reviews and 18 articles from these reviews)
113
Finding an effect 74
Not nding an effect 39
Laboratory studies 106
Field studies 8
Endpoints investigated 152
Effect No effect
Subdivision of articles among species
Birds 18 8
Insects (including bees, fruit ies and ants) 15 2
Other vertebrates (mostly animal models) 25 25
Other organisms (nematodes, bacteria, etc.) 4 0
Plants 12 4
End points studied in screened articles
Birds 20 18
Growth 3 4
Development 4 3
Fertility/reproduction 4 8
Behaviour/stress 3 0
Mutation 4 0
Mortality 0 1
Population decline 2 2
Insects 22 3
Growth ––
Development 4 0
Fertility/reproduction 9 1
Behaviour/stress 6 1
Mutation ––
Mortality 0 1
Population decline 1 0
Other vertebrates 35 27
Growth 4 1
Development 9 5
Fertility/reproduction 7 11
Behaviour/stress 13 7
Mutation 1 1
Mortality 1 2
Population decline ––
Other organisms 4 0
Growth 2 0
Development ––
Fertility/reproduction ––
Behaviour/stress 2 0
Mutation ––
Mortality ––
Population decline ––
Plants 22 2
Growth 12 0
Development 3 0
Fertility/reproduction 1 0
Behaviour/stress 3 1
Mutation 3 1
Mortality ––
Population decline ––
119S. Cucurachi et al. / Environment International 51 (2013) 116140
Table 2
Summary of articles reporting ecological effects of RF-EMF on birds.
Reference Country Species Scientic name Life stage
a
Type of study
b
Number of
subjects
c
Duration of
exposure
Frequency
[MHz]
Wave/
modulation
d
Power density
[mW/cm
2
]
e
SAR
[W/kg]
f
Effect
g
Effect size
h
Carpenter et
al. (1960)
USA Chicken Gallus gallus subsp.
domesticus
Emb Lab n/a
i
115 min 2450 MWCW 200 70 Teratogenic effects on the embryo +
280 98 Idem
400 140 Idem
Van
Ummersen
(1961,
1963)
USA Chicken As above Emb Lab n/a 115 min 2450 MW CW 200 70 Inhibition of growth +
280 98 Idem
400 140 Idem
Hills et al.
(1974)
Canada Chicken As above Emb Lab n/a 20300 s; rst
2 days of
incubation
2450 MW CW 0.2 246 1020 n/a Reduced chicken hatchability + (33%)
Giarola and
Krueger
(1974)
USA Chicken As above Juv Lab n/a 28 days 880 UHF CW 0.5 n/a Reduced growth rate +
Idem 260 VHF CW 0.5 n/a Reduced growth rate +
Hamrick and
McRee
(1975)
USA Japanese
quail
Coturnix coturnix
subsp. japonica
Emb Lab n/a 24 h 2450 MW CW 30 14 Reduced hatchability, altered/
organ development
McRee et al.
(1975)
USA Japanese
quail
As above Emb Lab 57 (4) 4 h for rst
5 days of
incubation
2450 MW CW 30 14 Altered development
Krueger et
al. (1975)
USA Chicken As above Ad Lab 5 (5) 12 weeks 260 VHF 01 n/a Unaltered fertility, reproduction
and hatchability
Idem 915 UHF 1.25 n/a Unaltered fertility, reproduction
and hatchability
Idem 2450 MW CW 1 n/a Unaltered fertility, reproduction
and hatchability
Davidson et
al. (1976)
Canada Chicken As above Juv Lab n/a 4.56 s 2450 MW 1.043 n/a Unaffected egg production
n/a Unaltered growth, reproduction,
mortality
McRee and
Hamrick
(1977)
USA Japanese
quail
As above Emb Lab n/a First 12 days of
incubation
2450 MW CW 5 4.03 Unaltered development
Clarke
(1978)
USA Chicken As above Emb Lab n/a 34th60th hr of
incubation
2450 MW PW (mod.
60 Hz and
12 Hz)
100 n/a Behavioural changes in hierarchy
positioning as adults
+
Fisher et al.
(1979)
Canada Chicken As above Emb Lab n/a 45 days 2450 MW CW 3.5 n/a Early embryonic development +
Cabe and
McRee
(1980)
USA Japanese
quail
As above Emb Lab n/a First 12 days of
incubation
2450 MW CW 5 4.03 Altered response to behavioural
tests as adults
+
Inouye et al.
(1982)
USA Japanese
quail
As above Emb Lab n/a First 12 days of
incubation
2450 MW 5 4.03 Developmental retardation of
Embryos
+ (7%)
No differences after week 8
120 S. Cucurachi et al. / Environment International 51 (2013) 116140
Reference
Country Species Scientic
name
Life stage
a
Type of
study
b
Number of
subjects
c
Duration of
exposure
Frequency
[MHz]
Wave/ modulation
d
Power density
[mW/cm
2
]
e
SAR
[W/kg]
f
Effect
g
Effect size
h
McRee et al.
(1983)
USA Japanese
quail
As above Emb Lab 270 (120) First 12 days of
incubation
2450 MW CW 5 4.03 Reduction in reproductive
capacity
+ (8%)
Wasserman
et al.
(1984)
USA Sparrow;
junco
Zonotrichia
albicollis;Junco
hyemalis
Var Field 12 ocks
(2 ocks)
20 min;
200 min
2450 MW 25 0.850.92 Variation in level of aggression
of birds after exposure
+ (11%)
20 min 2450 100 Idem
710 min 2450 155 Idem
Byman et al.
(1985)
USA Japanese
quail
As above Egg Lab 30 (90) 60 min during
17 days
incubation
2450 MW CW 2050 0.5 Unaltered growth or abnormal
development
Gildersleeve
et al.
(1987)
USA Japanese
quail
As above Emb Lab 468
(468)
12 days
during
incubation
2450 MW CW 5 4.03 Unaltered fertility, reproduction
and hatchability
Kleinhaus et
al. (1995)
Israel Migratory
birds
n/a n/a Sim n/a n/a 426 Broadcast
station
n/a 0.001
0.004
Unaltered development and
population levels
Bastide et
al.(2001)
France Chicken As above Emb Lab 300
(300)
Incubation
period
900 GSM n/a n/a Increased mortality. Inhibition
of normal development
+ (53%)
Grigoryev
(2003)
Russia Chicken As above Emb Lab n/a 21 days 900 GSM n/a n/a Increased mortality +
Balmori
(2005)
Spain White
stork
Ciconia ciconia Pop Field 60 nests 2 months 9001800 GSM base
station
0.001477 (mean
within 200 m);
7.45093× 10
5
(mean farther
than 300 m)
n/a Severe decline in productivity + (46%)
Balmori and
Hallberg
(2007)
Spain Sparrow Passer domesticus Var Field 40 visits
(1200
data
points)
3 years
and
8 months
1 MHz
3000
GSM to MW 0.00325 (max);
4.24403× 10
5
(mean)
n/a Decline in bird population and
doseeffect relationship found
between electric eld strength
and population decline at specic
locations
+ (75%)
Everaert and
Bauwens
(2007)
Belgium Sparrow As above Var Field 150
locations
4 months
during the
breeding
period
925960 GSM base
station
4.34589× 10
6
n/a Signicant relationship between
number of house sparrows and
levels of power density
+ (70%)
Idem 1805
1880
GSM base
station
9.07759× 10
6
n/a Idem
Reijt et al.
(2007)
Poland Great tit;
blue tit
Parus major;
Cyanistes caeruleus
Ad Field 72 (42) 45 days 1200
3000
Radar 2050 n/a Unaltered fertility and growth
Possible shift in species
distribution
+ (50%)
Batellier et
al. (2008)
France Chicken As above Egg Lab (240) Incubation
period
900 GSM 0.00306
0.04197
n/a Reduced hatchability. Increased
Embryo mortality
+(42%) +
a
Life stage refers to the age of the tested subject at the moment of the experiment. Emb =embryo, Ad= adult and Egg= egg.
b
Studies divided in laboratory and eld studies. Lab =laboratory study and Field=eld study.
c
Number of subjects involved in the experiment or eld study where reported in the study. In brackets information about number of control subjects.
d
Wave/modulation indicates the type of RF-EMF applied/measured in the study. CW =continuous wave, MW =microwave, PW =pulsed wave GSM =Global System for Mobile Communications, UHF =Ultra-High Frequency, and VHF =
Very high frequency.
e
Values of power density are reported as provided by authors or recalculated by conversion of electric eld values (PD= EF
2
/3770) and expressed in mW/cm
2
.
f
Values of SAR are reported as provided by authors and expressed in W/kg.
g
Biological or ecologically relevant endpoints studied.
h
Size of the effect where signicant. It indicates the ration between maximum effect and percentual difference compared to control. A + sign indicates a signicant effect and a sign indicates that no signicant effect was found.
i
n/a indicates that data was not provided by authors.
121S. Cucurachi et al. / Environment International 51 (2013) 116140
The endpoints included growth, hatchability, development based
on evidence of abnormal weight of hatchlings, incidences of abnor-
malities and mortality. Nine of the 15 experiments showed signicant
differences between RF-EMF and controlled/sham-exposed eggs.
It is a common opinion among experts (Baranski and Czerski, 1976;
Bryan and Gildersleeve, 1988) that the results obtained in most of the
studies until the 1980s (i.e. until Inouye et al., 1982 in this selection) re-
late to increases in the temperature ofthe egg due to the consequences
of hyperthermia a few degrees above normal incubation temperature.
An abnormal increase in temperature gradient of 3.5 °C had already
been observed in the early study by Van Ummersen (1961, 1963),
reported in the review conducted by Bryan and Gildersleeve (1988).
In a later study, Byman et al. (1985) found no effect on the growth
and normal development of born chicks of birds nesting in proximity
to antennas. Temperature rise was controlled and the measured
power density was 25 mW/cm
2
. Analogous results were obtained
by Gildersleeve et al. (1987) who kept the internal temperature of ir-
radiated and sham-exposed eggs to a mean of 37.5 °C without detecting
any deciency in the reproductive performance of males and females
allowed to hatch.
Among the three more recent studies, Bastide et al. (2001) and
Grigoryev (2003) found a signicant increase in mortality due to
RF-EMF on chicken (G. gallus subsp. domesticus) embryos exposed to
RF-EMF emitted by a GSM device during the duration of the incubation
period.
Also Batellier et al. (2008) studied the effect of exposure toGSM and
UMTS frequencies on chicken eggs over the entire period of incubation.
Four replicates with a total 240 eggs each were used in the experiment
to assess mortality rates. Results showed an increased mortality of
42.2% for embryos under a regime of controlled temperature, humidity
and external EMF. However, it was not possible to establish a propor-
tional relationship between the intensity of the electric eld and em-
bryo mortality.
4.1.1.2. Juvenile and adult. Five studies focused on the impact of RF-EMF
at a later phase of development of chickens: four studies on juvenile
and only one on adult subjects (Table 2). The endpoints studied were
growth, fertility, rate of egg production, hatchability and mortality.
The only study which found a signicant difference between ex-
posed and control/sham groups is the study by Giarola and Krueger
(1974) on juvenile chickens. The authors examined, exposure to
very-high frequency (VHF) and ultra-high frequency (UHF), together
with investigation of MW EMF. Exposure determined reduced growth
of chicks and consumption. In a follow-up study Krueger et al. (1975),
did not nd effects either on fertility or hatchability with a continuous
exposure period of 12 weeks at a power density (calculated) of
1mW/cm
2
. Experts from the U.S. Department of Energy (1978) attrib-
uted the difference in results to the cage used in the rst study which
may have determined a higher dose of energy absorbed by the target
subjects.
4.1.2. Field studies
There were ve eld studies on the impact of RF-EMF exposure at
various frequencies and physical conditions on populations of birds
living in areas in the vicinity of cellular phone masts or base-stations.
Anomalies and deviations from normality in the behaviour of exposed
subjects and in the level of productivity were found in all these studies.
The values of power density provided by studies ranged from
4.4×10
6
mW/cm
2
in the study on sparrows by Everaert and Bauwens
(2007) to the highest measured value of 155 mW/cm
2
in Wasserman et
al. (1984). In this last case, exposure caused a steady temperature raise
which determined a continuous gaping for the total duration of exposure
of the exposed population of sparrows (Zonotrichia albicollis) and juncos
(Junco hyemalis). Values for the SARs were provided only by the study
of Wasserman et al. (1984) and ranged from 0.85 to 0.92 W/kg. The end-
points studied were density, reproduction, behaviour and community
composition. In all the studies and experiments conducted, effects of the
RF-EMF were found from a variation of 10% to a maximum of 70% com-
pared to control.
Balmori (2005) monitored the variation of a population of
white storks (Ciconia ciconia) in the vicinity of a GSM base station
(i.e. 9001800 MHz with 217 Hz modulation) in search of possible
effects from the exposure. Total productivity within 200 m was on aver-
age 46% less than that found at a distance greater than 300 m from the
emitting station. An analogous signicant difference was found in the
breeding success: in 40% more of the cases no new-born chicks were
found in the nest.
In another study, Reijt et al. (2007) studied the inuence of long term
exposure to RF-EMF from radar (2001300 MHz) on a population of
great tits (Parus major) and blue tits (Cyanistes caeruleus) living around
a military radar station. Possible other sources of co-variance (e.g. from
human interactions with the location of birds and other pollutants)
were not considered in the study. Unlike in the case of Balmori (2005),
the exposure seemed not to have affected the number of nesting tits,
but the distribution of the different species. The authors state that the re-
sults contradict with the study of Balmori (2005), probably because of
the exposure of targets to radar MW (i.e. 12003000 MHz), instead of
mobile phone exposure (i.e. 9001800 MHz with 217 Hz modulation).
Additionally, Reijt et al. (2007) found that exposed nests were oc-
cupied, compared to control, by the less dominant species of tits (blue
tit), which would suggest that birds can perceive high frequency
RF-EMF as a stressful factor and, thus, would try to avoid nesting in
those areas. An average of 50% of the great tits moved from a more ex-
posed section of the study area to a less exposed one: in the interac-
tion with the great tit, the blue tit is usually less dominant according
to behavioural studies by Tanner (1966) and Tanner and Romero-
Sierra (1974). Therefore, the great tit would move to areas where
the power density is lower, and therefore the blue tit would have to
nest elsewhere.
Fig. 1 presents a plot ofthe effect with the relative measured power
density, from studies with a signicant effect (see Table 2 for details on
the studies). It is not possible to dene a clear doseeffect relationship,
but also at low values of power density strong effects of RF-EMF are
found.
4.1.3. Summary
Most studies on birds were laboratory investigations. The target
subjects were in the majority of laboratory studies chicken and Japa-
nese quail. Older laboratory studies exposed targets to high level of
MW EMF which probably determined an uncontrolled raise in tem-
perature which affected the exposed systems. Amongst the more re-
cent laboratory studies, evidence of an effect of RF-EMF on mortality
and development of embryos was in all cases found at both high
and low dosages. In all the ve eld studies found a signicant effect
of RF-EMF on breeding density, reproduction or species composition.
Field observations give a closer representation of real-life exposure,
thus RF-EMF, especially in the 900 MHz GSM band could be a certain
factor inuencing the ecology of birds.
4.2. Insects
Insects are a useful target system for the investigation of RF-EMF be-
cause of the limited size, the short life cycle and the possibility of easily
detecting developmental defects (Schwartz et al., 1985). It has been
demonstrated that insects can sense magnetic elds as a means for nav-
igation and orientation (Abraçado et al., 2005; Kirschvink et al., 2001;
Liedvogel and Mouritsen, 2010; Wajnberg et al., 2010; Winklhofer,
2010). Magneto-reception has been associated with the use of ferro-
magnetic iron oxide particles embedded in tissue or through pairs of
molecules with unpaired electrons (known as radical pairs) that are
associated with a light sensitive photoreceptor (Ritz et al., 2002;
Knight, 2009; Vácha et al. 2009). The exposure to RF-EMF might disrupt
122 S. Cucurachi et al. / Environment International 51 (2013) 116140
this magneto-reception mechanism, which could in turn affect the sur-
vival of insects. The most commonly studied species are the honey bee
(A. mellifera)andthefruity(Drosophila melanogaster).
4.2.1. Honey bees (A. mellifera)
Over the past few years, a phenomenon known as Colony Collapse
Disorder (CCD) has increased the attention of experts on the survival
of colonies of honey bees (Balmori, 2009; Schacker, 2008). The reduc-
tion in population of bees worldwide could have serious ecological, eco-
nomic and, thus, political implications given their role as pollinators. It
has been estimated that 15% of wild plant species in Europe (Kwak et
al., 1998) and 35% of the global crops produced (Klein et al., 2007)are
visited by honey bees. Bees are interesting for this reason from an eco-
nomic perspective: their economic role has been estimated to be
around 153 billion euros in the year 2005 (Gallai et al., 2009). RF-EMF
has been classied as one of the possible causes of honey bee colonies
collapse (Ratnieks and Carreck, 2010). Even though the interest of
media and the public in the effects of exposure of honey bees to mobile
communication RF-EMF has drastically increased, there seem to be no
thorough body of research into their effects in the scientic literature.
As a result, the screening conducted in this contribution identied
only eight studies which matched the dened criteria (Table 3), for a
total of 12 experiments. Six of the studies focused on the frequency
ranges specic to mobile communication and in all cases found a signif-
icant relationship between the exposure to the eld and the effects
studied. Only two of the studies found were not produced in the last de-
cade (Westerdahl and Gary, 1981a,b). These studies were the only ones
which did not nd any signicant effect on ight, orientation of behav-
iour of bees exposed to CW microwaves (i.e. 2450 MHz) at power den-
sities from 3 to 50 mW/cm
2
.
Among the studies that did nd an effect, Sharma and Kumar
(2010),Kumar et al. (2011) and Sahib (2011) found a critical reduc-
tion of all studied parameters of the exposed colonies of bees as a re-
sponse to RF-EMF. In all cases, an acute decrease in the breeding
performance or even a collapse of the entire colony resulted as a con-
sequence of exposure to RF-EMF. However, the studies provide limited
statistical information on the scale of the effect found and did not take
into account other confounding parameters (e.g. the placement of the
emitting device inside the hive).
The work by Harst et al. (2006) and Kimmel et al. (2007) from a
German research group seems to support the previously described
ndings, but do not provide any statistical measure of the effects
found and did not report any system of control or sham-exposure.
Clearer conclusions can be drawn from the study by Favre (2011),
which seems to be the mostcomplete and qualitatively interesting con-
tribution. Using sound-analysis techniques, the author investigated the
changes that were triggered in the behaviour of a population of honey
bees of the carnicagroup (Apis melliferasubsp. carnica). The sounds pro-
duced by the bees from ve healthy and unexposed hives were used as
a negative control and compared with recordings made when the same
hives were exposed to a mobile phonehandset in a calling position. An-
other inactive mobile phone was placed, at an earlier stage, to investi-
gate the possible disturbing inuence of the sheer presence of the tool
in the hives. The analysis of the recorded sounds revealed that the
bees produced sounds at higher frequency and amplitude after about
25 to 40 min after the communication had started and became quiet
when the handset was switched off.
No particular difference in behaviour and sounds was found for ex-
posure to the inactive handsets. The analysis of the sound data revealed
that the bees were, in fact, producing the so-called worker piping,
which usually serves as a signal for swarm exodus as a response to dan-
gerorstress,thusRF-EMFdirectlyaffected the community of bees under
exposure.
4.2.2. Fruit ies (D. melanogaster)
The screening of the literature identied ve studies on the fruit
y(D. melanogaster) for a total of nine experiments (see Table 3).
All the available studies found a signicant effect. The RF-EMF applied
focused on the GSM 900 MHz and GSM 1800 MHz (named also DCS
Digital Cellular System) systems.
RF-EMF power density was measured in the range of 0.0002 to
0.0407 mW/cm
2
, several order of magnitudes lower than those mea-
sured in the previously analysed laboratory studies on birds and bees.
All the values can be considered typical for digital mobile telephony
handsets and in most cases fall within the current exposure criteria
(ICNIRP, 1998). Unlike the previous cases, in most studies it was possible
to collect information about the magnetic ux density, which ranged
from to the time-averaged 0.003 μTofPanagopoulos et al. (2004) mea-
sured for a DCS frequency to 0.09 μT in the study by Panagopoulos et al.
(1)
(2)
(3) (4)
(5)
(6)
(7)
(8)
(9)
(16)
(15)
(12)
(13) (11)
(14)(10)
R² = 0.1259
0%
10%
20%
30%
40%
50%
60%
70%
80%
0.000001 0.00001 0.0001 0.001 0.01 0.1 1 10 100
Size of the effect [%]
Power density [mW/cm2]
Fig. 1. Size of the ecological effectsof RF-EMF on birds relatedto the power density of exposure.Articles reportedin graph: (1) Hills et al. (1974);(2)Inouyeet al. (1982);(3)McRee
et al. (1975);(4)Wassermanet al. (1984);(5)Balmori (2005);(6)Balmoriand Hallberg (2007);(7)Everaert and Bauwens (2007);(8)Reijt et al. (2007);(9)Batellier et al.
(2008);(11)McRee et al. (1975);(12) Krueger et al. (1975);(13)Davidson et al.(1976);(14)McRee and Hamrick (1977);(15)Byman et al. (1985); and(16) Gildersleeve
et al. (1987). See Table 2 for a complete description of studies. Data is reported for studies from which information could be extracted. The equation of the regression line is
y=0.0078x+0.2908.
123S. Cucurachi et al. / Environment International 51 (2013) 116140
Table 3
Summary of articles on ecological effects of RF-EMF on insects.
Reference Country Life stage
a
Type of
study
b
Number of subjects
(or distances if
specied)
c
Duration Frequency
[MHz]
Wave/modulation
d
Power density
[mW/cm
2
]
e
SAR [W/kg]
f
Effect
g
Effect size
h
Honeybee (Apis mellifera)
Westerdahl
and Gary
(1981a)
USA Adult foragers Lab 50(50) bees 30 min for 10 days 2450 MW CW 350 0.0751.25 No impact of radiation on ight,
orientation and homing abilities at
any power density
Westerdahl
and Gary
(1981b)
USA Adult Lab 50(50) bees 30 min for 10 days 2450 MW CW 350 0.0751.25 No differences in longevity between
exposed and sham exposed at any
power density
Harst et al.
(2006)
Germany Various Field 25 bees selected
from 4 colonies
n/a
i
1900 DECT base station
(mod. 100 Hz)
n/a n/a Reduced weight of bees. Colony
collapse and abnormalities in
behaviour
+ (21%)
Kimmel et al.
(2007)
Germany Various Field 5 at full exposure, 3
at 50% exposure(8)
4 days, 2 months,
45 min per day
1800 DECT (mod. 100 Hz) n/a n/a Change foraging ight + (14%)
Sharma and
Kumar
(2010)
India Various Field 2(2) colonies Continuous for
15 min. 2×day,
2× week, from Feb.
to Apr. (1115 h)
900 GSM 0.0086 n/a Decline in colony strength and in
the egg laying rate. Decline in the
number of returning bees and total
number of foragers. Decline in the
storing ability of honey
+ (62%) (22%)
(16%)
Favre (2011) Switzerland Various Field 5 hives 12 experiments of
40 min
900 GSM n/a 0.2710.98 Effect on behaviour: worker piping
signal was observed 25 to 40 min
after the onset of the mobile phone
+
Kumar et al.
(2011)
India Adult worker Field 10(20) bees 40 min 900 GSM n/a n/a Decreased lipid level in the
organism of exposed bees.
+
Sahib (2011) India Various Field 3(3) colonies 10 days, 10 min per
day
900 GSM n/a n/a Decreased returning ability bees in
exposed hives; reduced strength;
reduced egg laying rate of queen
+ (58%)
Fruit y (Drosophila melanogaster)
Weisbrot et
al., 2003
USA n/a Lab n/a 2 times for 60 min
with an interval of
4 h, for 10 days
1900 GSM PW n/a 1.4 (human
head)
Irradiation increased the number of
off-springs, enhancing reproductive
success
+ (36% mean;
50% max)
Panagopoulos
et al. (2004)
Greece n/a Lab n/a 6 min/day for
5 days
900 GSM device
(in talk mode)
0.041 n/a Decreased reproductive capacity + (50%)
Panagopoulos
et al., 2007
Greece n/a Lab 2 distances
(1 control)
6 min/day for
6 days
900 GSM PW phone
antenna
0.407 (±0.061) n/a Decrease of reproductive capacity,
seemingly dependent on eld
intensity more than on frequency
+ (41.4% mean;
255.2% max)
1800 DCS PW phone
antenna
0.283 (±0.043) n/a Idem
Greece n/a Lab 900 0.89
124 S. Cucurachi et al. / Environment International 51 (2013) 116140
Country Life stage
a
Type of study
b
Number
of
subjects
(or
distances if
specied)
c
Duration Frequency
[MHz]
Wave/modulation
d
Power density
[mW/cm
2
]
e
SAR [W/kg]
f
Effect
g
Effect size
h
Panagopoulos
et al., 2010
12 distances
(1 control)
6 min/day for
6 days
GSM CW phone
antenna
0.378 (±0.059; max
value at 0 cm from
antenna)
Reproductive capacity decreased at
all distances studied at increasing
pwoximity to the antenna. A win-
dow effect was revealed at dis-
tances of 2030 cm.
+ (11% mean;
40.6% max)
0.0004 (±0.0001;
min value at 100 cm
from antenna)
Idem Idem Idem
1800 DCS CW phone
antenna
0.252 (±0.05; max
value at 0 cm from
antenna)
Idem Idem Idem
0.0002 (±0.0001;
min value at 100 cm
from antenna)
Idem Idem
Panagopoulos
and
Margaritis,
2010
Greece n/a Lab n/a 121 min for 5 days 900 GSM PW phone
antenna
0.01 (time
averaged; ±0.002 at
a distance of 30 cm)
0.795 Almost linear decrease in
reproductive capacity at increasing
durations of exposure.
+ (49.3% mean;
67.4% max)
121 min for 5 days 1800 DCS PW phone
antenna
0.011 (time
averaged; ±0.003 at
a distance of 30 cm)
0.795 Idem Idem
Panagopoulos,
2012
Greece n/a Lab n/a 6 min for 5 times 900 GSM CW phone
antenna
0.063 0.795 Decreased ovarian size after two
exposures.
+ (21% mean;
29.5% max)
Other insects: tobacco hornworm (Manduca sexta), American cockroach (Periplaneta americana), and ant (Myrmica sabuleti)
Schwartz et
al. (1985)
Canada Adults exposed
at larval stage
Lab n/a From larva to
pre-pupal stage
2695 (500
pulse per
second)
Anechoic chamber,
horn antenna PW
4 23 Decreased food consumption and
larval body weight after 20 days.
Deformed adults. Higher mortality.
Lower number of laid eggs.
+ (50%) (2%)
(20%) (23%)
Vacha et al.
(2009)
Czech
Republic
n/a Lab 11(11 non exposed) 3 h 1.27 RF generator n/a n/a Rise in the locomotor activity and
disruptive effect at 1.2 MHz.
+ (14%)
Cammaerts et
al. (2012)
France Various life
stages
Lab 6 large naive
colonies
Three exposure
periods: 4.5 days;
6 days; 1.5 days
900 GSM from vector
signal generator
7.95× 10
5
n/a Diminished acquired association
between food and a olfactory and
visual cues.
+ (40%) (42.5%)
a
Life stage refers to the age of the tested subject at the moment of the performance of the experiment.
b
Studies divided in laboratory and eld studies. Lab=laboratory study and Field= eld study.
c
Number of subjects involved in the experiment or eld study where reported in the study. In brackets information about number of control subjects. Further specications of type of subjects involved in the studies are reported if pro-
vided by authors. In the case of studies regarding the fruit y the distances applied are reported.
d
Wave/modulation indicates the type of RF-EMF applied/measured in the study. CW =continuous wave, MW =microwave, GSM =Global System for Mobile Communications, and DECT =Digital Enhanced Cordless Telecommunications.
e
Values of power density are reported as provided by authors or recalculated by conversion of electric eld values (PD = EF
2
/3770) and expressed in mW/cm
2
.
f
Values of SAR are reported as provided by authors and expressed in W/kg.
g
Biological or ecologically relevant endpoints studied.
h
Size of the effect where signicant. It indicates the ration between maximum effect and percentual difference compared to control. A + sign indicates a signicant effect and a sign indicates that no signicant effect was found.
i
n/a indicates that data was not provided by authors.
125S. Cucurachi et al. / Environment International 51 (2013) 116140
Table 4
Summary of articles on ecological effects of RF-EMF on other vertebrates (than birds).
Reference Country Species
(scientic name)
Life stage
a
Number of
subjects
b
Duration of exposure Frequency
[MHz]
Wave/
modulation
c
Power density
[mW/cm
2
]
d
SAR [W/kg]
e
Effect
f
Effect size
g
Chernovetz et al.
(1975)
USA Rat (Rattus
norvegicus)
n/a
h
n/a 1114 days, 10 min 2450 MW CW 20 38 No effect on development
Berman et al.
(1978)
USA Mouse
(Mus musculus)
Emb n/a 117 days, 100
min/day
2450 MW CW 3.428 222 Reduced foetal weight and hampered
development
+
Berman et al. (1980) USA Rat (as above) n/a n/a 80 h, 4 weeks 2450 MW CW n/a 5.6 Transient reduction in fertility
Jensh et al. (1982) USA Rat (as above) Ad 12 (59; 4) 6 h/day
(pregnancy period)
2450 MW CW 20 5.2 No changes in development
3.6 Idem
Kowalczuk et al.
(1983)
Great
Britain
Mouse (as above) Ad 50 (50) 30 min 2450 MW PW n/a 44 Signicant effect on reproduction:
decreased sperm count, increased
abnormal sperm
+ (35%) (330%)
Lary et al. (1983) USA Rat (as above) Ad n/a 611 days, 24 h/day 100 FM 25 0.4 Unaltered development
Nawrot et al. (1985) USA Rat (as above) Emb n/a 615 days, 8 h/day 2450 MW CW 30 40 Altered development +
Lebovitz et al.
(1987a)
USA Rat (as above) n/a n/a 6 h/day, 9 days 1300 PW
(600 Hz pulse)
n/a 7.7 No effect on reproduction/fertility: sperm
production, sperm morphology
Lebovitz et al.
(1987b)
USA Rat (as above) n/a n/a 8 h 1300 CW n/a 9 No effect on reproduction/fertility:
testicular function
D'Andrea et al.
(1989)
USA Rhesus monkey
(Macaca mulatta)
Juv 5
(same test group,
sham-exposed)
1 session of 60 min
per day per 1 week
1300 MW PW 0.92 mean
(peak of 0.1318)
0.09 mean
in the head
(15 peak in
the head)
No change in behaviour as compared to
sham-exposed sessions
Berman et al.
(1992)
USA Rat (as above) Juv/Ad 119 (0; 129) 22 h/day, 18 days
(from 1st through
19th day of gestation)
970 n/a n/a 0.07 Unaltered development
2.4 Unaltered development
4.8 Foetal development alterations + (7%)
Lai et al. (1994) USA Rat (as above) Juv n/a n/a 2450 PW n/a 0.6 Decreased performance in behavioural
tasks in T-maze. Decit in memory
function
+ (50%)
Sherry et al.
(1995)
USA Rhesus monkey
(as above)
Ad 6 (no control or
sham-exposed group)
2 min (7200 pulses) 1001500 MW UWB 1.65782 ×10
7
0.5 (whole
body
average)
Unaltered behavioural test performance
Klug et al. (1997) Germany Mouse (as above) Emb 53 (65) 36 h 150 AM 0.9549195.4907 n/a Unaltered growth
0.2 Idem
1 Idem
5 Idem
900 GSM 1.33714 0.2 Idem
1 Idem
5 Idem
Jensh (1997) USA Rat (as above) Juv/Ad n/a 6 h/day, 5 days 915 GSM CW 10 n/a Unaltered growth
6 h/day, 5 days 2450 MW CW 20 n/a Idem
Magras and Xenos
(1997)
Greece Mouse (as above) Juv 36 5 pregnancies 88.5950 FM; UHF TV;
GSM
1.053× 10
3
1.936× 10
3
Progressive decrease in the number of
newborns per dam leading to irreversible
infertility Improved prenatal development
parameters
+ (76%)+ (27%)
168× 10
6
Idem Idem
Khillare and
Behari (1998)
India Rat (as above) Ad 18 (18) 2 h/day, 35 days 200 AM (mod.
16 Hz)
1.47 1.652 Decreased fertility observed in exposed
tests. Unaltered development
+ (42%)
Bornhausen and
Scheingraber
(2000)
Germany Rat (as above) Ad 12(12) 20 days
(pregnancy period)
900 GSM (mod.
217 Hz)
0.1 0.75 Unaltered growth
Sienkiewicz et al.
(2000)
UK/USA Mouse (as above) Ad n/a 45 min 10 days 900 PW (mod.
217 Hz)
0.54 0.05 Unaltered learning in the performance
of tasks
5.7
Yamaguchi et al.
(2003)
Japan Rat (as above) Ad 168 1 h/day for 4 days;
45 min daily for
4 days; 1 h/day for
5 days and 2 days of
rest for 4 weeks
1439 PW TDMA n/a 1.7 Unaltered learning abilities in the
performance of tasks
126 S. Cucurachi et al. / Environment International 51 (2013) 116140
Reference
Country Species (scientic name) Life stage
a
Number of
subjects
b
Duration of exposure Frequency
[MHz]
Wave/
modulation
c
Power density
[mW/cm
2
]
d
SAR [W/kg]
e
Effect
f
Effect size
g
Cassel et al.
(2004)
France Rat (as above) Ad n/a 45 min 2450 PW n/a 0.6 Unaltered learning in the performance
of tasks
Cobb et al. (2004) USA Rat (as above) n/a n/a 45 min, 10 days 2450 MW PW n/a 0.6 Unaltered brain development and
performance of spatial tasks
Cosquer et al.
(2005)
France Rat (as above) Juv 48 45 min 2450 PW n/a 0.6 Unaltered performance in spatial tasks
Dasdag et al.
(2008)
Turkey Rat (as above) Ad 14 (10 control; 7
sham-exposed)
2 h/day, 7 days/week,
10 months
900 PW 0.023840.17561 0.070.57 Unaltered fertility
Kumlin et al.
(2007)
Finland Rat (as above) Juv 18(6) 2 h/day, 5 days/week,
5 weeks
900 PW n/a 0.3 (mean
value)
Improvement in learning abilities of rats + (20%)
Ribeiro et al. (2007) Brasil Rat (as above) Juv 16 (8) 1 h/day, 11 days 1850 PW 1.4 n/a Unaltered fertility
Yan et al. (2007) USA Rat (as above) Ad 16 2 times/day for 3-h
periods for 18 weeks
1900 CDMA n/a 1.8 Higher incidence of sperm cell death + (37%)
Mathur (2008) India Rat (as above) Juv n/a 2 h/day, 45 days 73.5 AM (mod.
16 Hz)
1.33 0.4 Abnormal behavioural response to
noxious stimuli
+ (38%)
Nittby et al.
(2008)
Sweden Rat (as above) Ad 28 (16;
8 sham-exposed)
2 h/week, 55 weeks 900 Lower power
level GSM
3.3× 10
4
0.62× 10
3
Behavioural abnormalities: altered
performance of rats during episodic-like
memory test
+ (75%)
n/a 0.37× 10
3
Idem
GSM 33× 10
4
62× 10
3
Idem
37× 10
3
Idem
Daniels et al.
(2009)
South
Africa
Rat (as above) Juv 12 (12) 3 h/day, 12 days
(2 days after birth)
840 RF signal
generator
2.1247× 10
10
(d= 0.93 m)
n/a Decreased behaviour. Decreased
locomotive activity. Unaltered
performance of memory tasks
+ (60%)
Gathiram et al.
(2009)
South
Africa
Rat (as above) Ad 32 (32) 8 h/day, 10 days 1003000 Unique eld
system
n/a n/a Unaltered fertility of exposed male and
female individuals
Lee et al. (2009) Korea Mouse (as above) Ad 17 (14) 90 min/day (15 min
break) 17 days
(gestation period)
848.5 CDMA 1.41748.2501 0.694.04 Unaltered development
n/a 2 (Power=
30 W)
Unaltered development
20 (20) 90 min/day (15 min
break) 17 days
(gestation period)
1950 WCDMA 1.09237.0043 1.117.13
Mailankot et al.
(2009)
India Rat (as above) Juv n/a 1 h/day, 28 days 9001800 GSM n/a n/a Detrimental effects on fertility + (53%)
Nicholls and Racey
(2009)
UK Bat (Pipistrellus
Pipistrellus)
n/a n/a 20 h (bat activity);
16 h (insect count); 3
elds
n/a PW radar 3.8101 ×10
3
1.7275× 10
1
(peak values at
distance of 10
30 m)
n/a Reduced foraging and activity of bats + (16% in bat
counts; 13% bat
passes)
No effect on the abundance of insects
Sommer et al.
(2009)
Germany Mouse (as above) Multi-gen. 128 male 256
female,3 generations
570 days (chronic
exposure), 30 min/
day break
2000 UMTS 0.135 0.080.144 Unaltered fertility and development
0.68 0.40.72 Idem
2.2 1.32.34 Idem
Fragopoulou et al.
(2010)
Greece Mouse (as above) Juv 12 (12) 4 days, 2 h/day 900 GSM 0.050.2 0.410.98 Decits in consolidation and/or retrieval
of learned spatial information
+ (30%)
Balmori (2010) Spain Frog
(Rana temporaria)
Juv 70 (70) 2 months from egg
phase until prior to
metamorphosis
6482155 Cell-phone base
station
8.5942× 10
4
3.2493× 10
3
n/a Increased mortality rate. Asynchronous
growth of exposed subject; disrupted
behaviour
+ (90%)
Salama et al.
(2010a)
Japan Rabbit (Oryctolagus
cuniculus)
Ad 8 (8;
8 sham-exposed)
8 h/day, 12 weeks 800 PW 6.2910× 10
5
2.2616× 10
3
(mean value
over time at
minimum to
maximum
distance
from the phone)
0.43
(whole
body)
Signicant decrease in sperm
concentration at week 8. Decrease in
motile sperm population at week 10.
Overall effect on testicular function and
reproduction ability
+ (62%) (25%)
(continued on next page)
127S. Cucurachi et al. / Environment International 51 (2013) 116140
Table 4 (continued)
Reference Country Species
(scientic name)
Life stage
a
Number of
subjects
b
Duration of exposure Frequency
[MHz]
Wave/
modulation
c
Power density
[mW/cm
2
]
d
SAR [W/kg]
e
Effect
f
Effect size
g
Salama et al.
(2010b)
Japan Rabbit (as above) Ad 8 (8;
8 sham-exposed)
8 h/day, 12 weeks 800 PW 6.2910× 10
5
2.2616× 10
3
(mean value
over time at
minimum to
maximum
distance
from the phone)
0.43
(whole
body)
Detrimental effects on sexual behaviour:
increased number of mounts, increased
number of mounts without ejaculation
+
Imai et al. (2011) Japan Rat (as above) Juv 24 (24;24) 5 h/day, 7 days/week,
5 weeks
1950 WCDMA CW n/a 0.4 No effects on reproduction and
development
Kesari et al. (2011) India Rat (as above) Juv 6 (6 sham-exposed ) 2 h/day, 35 days 900 n/a 9.2558 ×10
2
(peak value at
20 m);8.2819×
10
2
(peak value
at 30 m)
0.9 (Power=
2mW)
Potential signicant effect on
reproduction (fertilizing potential of
spermatozoa)
+ (41%)
Sarookhani
et al. (2011)
Iran Rabbit (as above) n/a 18 2 h/day, 2 weeks 950 GSM n/a n/a Decreased reproductive capacity + (90%)
Aldad et al. (2012) USA Mouse (as above) Ad 39 pregnant (42
sham-exposed)
0 to 24 h/day from
day 1 to day 17 of
gestation
800
1900
GSM n/a 1.6 Behavioural and neurophysiological
alterations
+ (7%)
Bouji et al. (2012) France Rat (as above) Middle-aged 9 (9 sham-exposed) 15 min 900 GSM PW n/a 6 Altered behaviour and increased stress + (47%)
Hao et al. (2012) China Rat (as above) n/a 16 (16) 2 times/dayfor 3 h/day,
for 5 days/week, for 10
weeks
916 Mobile phone
antenna
1 n/a Altered learning.
Altered memory. Adaptation to eld after
long term exposure
+ (18%) (18%)
Jiang et al. (2012) China Mouse (as above) n/a 5 (5; 5 exposed to
gamma radiation; 5
exposed to combined
RF and gamma
radiation)
4 h/day for 1 to 14
days
900 Wireless
transmitter
120 0.548 No effect on mutation
Lee et al. (2012) Korea Rat (as above) n/a 5 (5; 5 exposed to
gamma radiation; 5
exposed to combined
RF and gamma
radiation)
45 min/day, 5 days/
week, 12 weeks
848.5 CDMA n/a 2 (4
combined
with
WCDMA)
No effect on reproduction
idem idem 1950 WCDMA idem 2 (4
combined
with CDMA)
idem
Ozlem Nisbet
et al. (2012)
Turkey Rat (as above) Juv 11 (11;11) 2 h/day for 90 days 900 GSM n/a 0.003 Increased testosterone level and sperm
motility. Altered morphology
+ (15%) (3%)
idem idem 1800 GSM idem 5.3× 10
5
idem + (14%) (2%)
Poullettier de Gannes
et al. (2012)
France Rat (as above) Various 20 (20;20) 2 h/day, 6 days/week,
18 days
2450 W-LAN
Wi-Fi
n/a 0.08 No abnormalities in reproduction and
development
Yang et al. (2012) China Rat (as above) Ad 12 (12
sham-exposed)
20 min 2450 MW PW 65 6 Stress response elicited in rat
hippocampus
+ (30%)
a
Life stage refers to the age of the tested subject at the moment of the performance of the experiment.
b
Number of subjects involved in the experiment or eld study where reported in the study. In brackets information about number of control subjects. Further specicationsoftypeofsubjectsinvolvedinthestudiesarereportedifprovidedbyauthors.
c
Wave/modulation indicates the type of RF-EMF applied/measured in the study. CW =continuous wave, MW =microwave, GSM =Gl obalSy stemfor Mobile Communications, DECT =Digital Enhanced Cordless Telecommunications, PW =pulsed wave,
UWB= ultra wide band, AM= amplitude modulation, FM= frequency modulation; UMTS= Universal Mobile Telecommunications System; CDMA =Code division multiple access; TDMA =time division multiple access; and WCDMA =Wideband Code
Division Multiple Access.
d
Values of power density are reported as provided by authors or recalculated by conversion of electric eld values (PD= EF
2
/3770) and expressed in mW/cm
2
.
e
Values of SAR are reported as provided by authors and expressed in W/kg.
f
Biological or ecologically relevant endpoints studied.
g
Size of the effect where signicant. It indicates the ration between maximum effect and percentual difference compared to control. A + sign indicates a signicant effect and a sign indicates that no signicant effect was found.
h
n/a indicates that data was not provided by authors.
128 S. Cucurachi et al. / Environment International 51 (2013) 116140
(2010). SAR levels were, when provided, obtained by elaboration of data
provided by the manufacturer (i.e. for the human head) of the system
used for exposure and not directly measured.
The ecologically relevant endpoints analysed in the studies were
growth and reproduction. All of the analysed studies found a signicant
effect compared to the control. With the exception of a study by
Weisbrot et al. (2003), all studies were conducted by a research group
from the University of Athens, Greece. In the study of Weisbrot et al.
(2003) the irradiation determined a benecial effect on the reproduc-
tive success of the exposed system. The number of offsprings even in-
creased by up to 50% compared to control. All the other studies found
asignicant depression of growth and reproduction as a response to
exposure. Several studies performed by Panagopoulos and co-authors
(see Table 4) found a maximum decrease in the endpoints of at least
40% compared to control. Exposure duration lasted for6 min/day orin-
creased over time up to a maximum of 21 min over a period of six or
ve days. The reproduction of experiments performed at several dis-
tances from the emitting system (i.e. a telephone device) suggested in
all cases a quasi-linear decrease at increasing durations of exposure
(Panagopoulos and Margaritis, 2010) and increase in proximity to the
source of the emission (Panagopoulos et al., 2010). In this last study a
window-effect was found at distances of 2030 cm from the device,
which resulted in the highest decrease of the measured values.
4.2.3. Effect on other insects
The remaining studies in this section focus on the tobacco horn-
worm (Manduca sexta), on the American cockroach (Periplaneta
americana) and on a species of ant (Myrmica sabuleti;Table 3). The
study by Schwartz et al. (1985) analysed differences in development,
reproduction and mortality in tobacco hornworms exposed during
their larval stage to PW RF-EMF at a frequency of 2695 MHz and a
power density of 4 MW/cm
2
. All the measured parameters were af-
fected and effect size was as high as 50% lower compared to control.
The studies on the American cockroach (Vacha et al. 2009) and the
ant (Cammaerts et al., 2012) focused on the effects of RF-EMF on the
magneto-reception of the insects. In the study by Vacha et al. (2009),
it was found that, during and after the rotation of the natural geomag-
netic eld, the insects turned around, as a response of the detection of
the eld. However, their ability to detect the geomagnetic eld was
disrupted after exposure to a eld at 1.2 MHz with a magnetic ux
density between 12 and 18 nT.
Cammaerts et al. (2012) investigated the impact of RF-EMF on the
acquisition and loss of olfactory and visual cues of six experimental
colonies of the ant Myrmica sabuleti. The exposure to a GSM-generated
signal determined a loss in the acquired association between food and a
visual cue (40% worse than control), a decreased retention of acquired
knowledge, and a total loss of visual memory.
The representation of the size of the effect compared to the power
density (Fig. 2) shows that signicant effects are found both at high
and low dosages, revealing no clear doseresponse relationship. In one
of the analysed studies, no effects were found at high levels of power
density.
4.2.4. Summary
A limited set of articles regarding the possible impact of RF-EMF on
honey bees is available in literature. Most of the analysed studies found
an effect on the target colonies. The most affected endpoints seemed to
be behaviour and orientation of exposed bees, which lead to disruptive
consequences in the colonies. The majority of the studies did not provide
statistical analysis and did not use clear control measures to analyse re-
sults. One exception is the study conducted by Favre (2011), in which
thebehaviourofthebeesseemstobecomparabletothatexperienced
by colonies exposed to extreme danger and stress.
The studies analysing the effects of RF-EMF on fruit ies found in
all cases a signicant effect. Results of one study show an increased
reproductive success after exposure. The remaining studies, which
were conducted by the same research institute in Greece, found in
all cases a signicant depression of growth and reproduction at both
900 and 1800 MHz. Two studies on the American cockroach and a
species of ant analysed the effects of exposure to RF-EMF on the
magneto-reception and orientation of the insects. The behaviour of
target systems was disrupted by the exposure to RF-EMF.
4.3. Other vertebrates
The impossibility of conducting laboratory experiments into the
effects of RF-EMF on humans steadily increased the number of scientic
studies on laboratory vertebrate models. As suggested by the WHO
(2006), studies conducted on immature animals can, for instance, pro-
vide a useful indicator of possible cognitive and behavioural effects on
children.The vast majority of studies focused on the analysis of intracel-
lular pathways, for instance through changes in calcium permeability
across membranes (e.g. Maskey et al., 2010); or on gene expression,
(2)
(1)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
R² = 0.5909
-10%
0%
10%
20%
30%
40%
50%
60%
0.00001 0.0001 0.001 0.01 0.1 1 10 100
Size of the effect
Power density [mW/cm2]
Fig. 2. Size of the effects of RF-EMF on insects compared to the power density of exposure. Articles reported in graph: (1) Westerdahl and Gary (1981a); (2) Westerdahl and Gary
(1981b); (3) Sharma and Kumar (2010); (4) Panagopoulos et al. (2004); (5) Panagopoulos et al. (2007) (6) Panagopoulos et al., 2010; (7) Panagopoulos and Margaritis,
2010; (8) Panagopoulos (2012); (9) Schwartz et al. (1985); and (10) Cammaerts et al. (2012). See Table 3 for a complete description of studies.
129S. Cucurachi et al. / Environment International 51 (2013) 116140
namely on the neurons of rats exposed to RF-EMF (e.g. Salford et al.,
2003; Zhao et al., 2007); or on possible chromosomal damage in mice
cells (e.g. Nikolova et al., 2005).
A total of 50 scientic articles were selected for a total of 62 relevant
ecological experiments (Table 4). The endpoints analysed which were
of interest were fertility, growth, behaviour and mortality (Table 1).
With the exception of one study on bats (Pipistrellus pipistrellus,
Pipistrellus pygmaeus,Myotis daubentonii,andMyotis nattereri) breeding
nearby a wind turbine and one study on the tadpoles of frogs (Rana
temporaria), all studies were conducted in a laboratory setting. The ani-
mal systems underinvestigationwere rats commonly used in laboratory
studies (Wistar albino rat and Sprague Dawley rat), mice (Balb/cand
Balb/c/f), rabbits (White New Zealand Rabbit), rhesus monkeys (Macaca
mulatta). Of the total of 50 articles, 50% of the studies were conducted
on rats. A total of 27 experiments (43%) showed no signicant results
of an impact of RF-EMF under the physical and experimental settings
used. The power density ranged from 0.6 ×10
6
to 20 mW/cm
2
, which
was the maximum value measured for MW CW exposures (Table 4).
The SARs values measured ranged from 0.00194 to 44 W/kg, with a
peak value measured at 2450 MHz for MW PW exposure. In the studies
in which higher level of exposure to RF-EMF were applied and tempera-
ture was not controlled, results may be related to an increase in body
temperature as a consequence of exposure.
A large share of the studies on vertebrate animal models focused
on changes in behaviour as a result of exposure. This choice may be
related to investigating of possible inuences of RF-EMF on the be-
haviour and cognitive performance of humans, who use mobile
phone devices in close proximity to their heads. Some commonalities
between human and rat response to noxious substances have been
explored by other elds of science (Hammond et al., 2004). Lai et al.
(1994) suggested that rats suffer from a decit in spatial working
memory function when exposed to RF-EMF (50% decreased perfor-
mance compared to control). The repetition of the experiment with
similar conditions of exposure by Cassel et al. (2004),Cobb et al.
(2004), and Cosquer et al. (2005) found no effects on learning
abilities of rats in the performance of spatial tasks and no evidence
of altered brain development.
Another example in this direction is the work of Daniels et al.
(2009), who investigated the effect of RF-EMF in the mobile phone
range on the behaviour of the rat with controversial results. Spatial
memory was tested using the Morris water maze test (Morris, 1984),
and mood disturbances and anxiety-like behaviour were tested in an
open eld test, for twelve radiated and twelve control subjects. Results
showed no signicant differences between groups in the Morris test,
suggesting no signicant difference in the behaviour of exposed and
control rats. However, male rats performed signicantly worse (60%)
in the open eld test.
The articles by Lee et al. (2009, 2012) and Imai et al. (2011) are the
only studies focusing on the impact of the frequencies network stan-
dards found in 3 G mobile communication (Collins and Smith, 2001),
working with protocols like wideband code division multiple access
(W-CDMA) or CDMA. All experiments, on mice and rats, did not have
any observable adverse effect on development, reproduction or mutation
of tested subjects. No effects on the development of rats were also ob-
served by the study of Poullettier Poulletier de Gannes et al. (2012),
where Wireless Fidelity (Wi-Fi) signal at 2450 MHz was applied on
rats, and by the study of Jiang et al. (2012),wheremicewereexposed
to a wireless transmitter at 900 MHz. These studies represent the rst at-
tempt to investigate the effects of wireless communication on health.
The eld experiment of Balmori (2010) on the behaviour and
growth of the tadpoles of frogs (Rana temporaria)placed140mfroma
eld station provides evidence of the effect of RF-EMF. The exposed
group showed low coordination of movements, an asynchronous growth
and a high mortality (90%). The control group was exposed to the same
environmental conditions, but placed inside a Faraday cage. As a result,
the coordination of movements was normal, the development was syn-
chronous, and the mortality rate was 4.2%. The research goal of the eld
study by Nicholls and Racey (2009) was to test whether PW RF-EMF
emitted by a radar could be used as a method of preventing bats from
death caused by collisions with wind turbines. The authors analysed 20
(1)
(2)
(3)
(4)
(5)
(6) (7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(18)(17)
(23)
(19)
(20) (21)
(22)
(26) (24)
(16)
(27) (25)
R² = 0.079
-20%
0%
20%
40%
60%
80%
100%
1E-10 1E-08 0.000001 0.0001 0.01 1 100
Size of the effect
Power density [mW/cm2]
Fig. 3. Size of the effectsof RF-EMF compared to the power density of exposedvertebrate animal models. Articles reported in graph: (1)Berman et al. (1992);(2)Magras and Xen os
(1997);(3)Khillare and Behari(1998);(4)Nittbyet al. (2008);(5)Mathur (2008);(6)Salama et al. (2010a);(7)Kesari et al. (2011);(8)Lai e t al. (1994);(9)Daniels et al.
(2009); (10) Nicholls and Racey (2009); (11) Fragopoulou et al. (2010);(12)Balmori (2010);(13)Hao et al. (in press); (14) Yang et al. (2012);(15)Jiang et al.
(2012); (16) Chernovetz et al. (1975); (17) Jensh et al. (1982);(18)Lary et al. (1983); (19) D'Andrea et al. (1989); (20) Sherry et al. (1995);(21)Klug et al. (1997);
(22) Jensh (1997);(23)