Environ. Sci. Technol. 2006, 40, 2448-2454 New Risk Assessment Approach for
Quotient (HQ ) field application rate/oral or contact LethalDose (LD50)). When the calculated value of HQ is higher than
Systemic Insecticides: The Case of
a threshold of 50, further studies are required. This thresholdis derived from data which only consider spray applications
Honey Bees and Imidacloprid (Gaucho)
In the case of plants treated by systemic insecticides,
honey bees may be at risk via contaminated pollen and nectar(4). The contamination of nectar by sprayed systemic
M A R I E - P I E R R E H A L M , * , † A . R O R T A I S , * , ‡
insecticides has been long documented (5), whereas little
G . A R N O L D , ‡ J . N . T A S EÄ I , § A N D
information is available on systemic formulations applied in
soils and on seeds. Published data deal mainly with aldicarb,a carbamate substance used for the protection of various
Centre d’Etude et de Recherche du Me´dicament deNormandie, 5 rue Vaubenard, 14032 Caen cedex, France,
cultures (6). More recently, several authors supplied data on
Laboratoire Evolution, Ge´nomes, Spe´ciation, CNRS-UPR9034,
the presence of imidacloprid, a neonicotinoid systemic
Baˆtiment 13, Avenue de la Terrasse, F-91198 Gif-sur-Yvette,
insecticide, in nectar and pollen of treated plants (7). France, and Laboratoire de Zoologie, INRA,
Generally, systemic insecticides provide the treated plant
with a permanent protection from soil invertebrates andsucking insects (8). Applied in soils and on seeds, they degradeslowly over time and disperse in all the plant tissues duringits growth. Therefore, using the field application rate of active
The procedure to assess the risk posed by systemic
substance as an exposure parameter to assess the risk posed
insecticides to honey bees follows the European Directives
by systemic insecticides to honey bees is not sensible. Unlike
and depends on the determination of the Hazard Quotient
sprayed insecticides, which have a short-lasting action on
(HQ), though this parameter is not adapted to these
plants, systemic insecticides are persistent. Moreover, thesemolecules, detected at low concentrations in the pollen and
molecules. This paper describes a new approach to assess
nectar of treated plants, are more likely to affect honey bees
more specifically the risk posed by systemic insecticides
by acute, chronic, and sublethal intoxications (9) rather than
to honey bees with the example of imidacloprid (Gaucho).
This approach is based on the new and existing chemical
In this paper, we propose a new approach to determine
substances Directive in which levels of exposure (PEC,
the risk posed by systemic insecticides to honey bees. It is
Predicted Exposure Concentration) and toxicity (PNEC,
based on the European Technical Guidance Directive (TGD)
Predicted No Effect Concentration) are compared. PECs
that assesses the impact of new (793/93 and 1488/94/CE
are determined for different categories of honey bees in
legislations) and existing chemical substances (EC-67/54/8
relation to the amounts of contaminated pollen and nectar
and EEC-93/67 Directives) on ecosystems (10). This approach
they might consume. PNECs are calculated from data on
is applied to imidacloprid, which is a good study case because
acute, chronic, and sublethal toxicities of imidacloprid to
it has been extensively studied and presents a lot ofexperimental data.
honey bees, to which selected assessment factors areapplied. Results highlight a risk for all categories of honey
Materials and Methods
bees, in particular for hive bees. These data are discussed
A group of experts, namely the Scientific and Technical
in the light of field observations made on honey bee
Committee (CST), was nominated in 2001 by the French
mortalities and disappearances. New perspectives are
Ministry of Agriculture to assess the risk posed by imida-
given to better determine the risk posed by systemic
cloprid to honey bees. This committee examined all studies,
delivered up to July 2004 by the Ministry of Agriculture, onthe toxicity of imidacloprid to honey bees (7). This paper
refers to some of the work achieved by this committee.
In the European Union, formulated pesticides are registered
For many wildlife species, the standard practice in
by the European Council Directive (EC-91/414) and the risk
pesticide regulation (91/414 EEC) is to determine a toxicity
posed by these molecules to honey bees is directly assessed
exposure ratio (TER) and to compare it to a threshold (a
by the European and Mediterranean Plant Protection
safety factor) that aims at protecting these species. In this
Organization (EPPO) guidelines No. 170 (1). These guidelines
paper, we used the PEC/PNEC ratio (predicted environmental
propose methods for evaluating side effects of agrochemical
concentration/predicted no effect concentration) which aims
products on honey bees. The approach is based on a 3 tier
at protecting ecosystems (10). Honey bees (unlike most other
assessment scheme comprising early studies in laboratory
species) live in colonies and depend on each other for survival.
conditions, followed by semi-field studies, and completed
Such interdependent relationships define the honey bees’
by field studies. According to this Directive, and to the
colony as a superorganism (11). The functioning of a
decision making scheme attached to the EPPO guidelines
superorganism is similar to that of an ecosystem in the sense
(2), moving from tier 1 (laboratory studies) to tier 2 (semi-
that each unit (temporal castes in a colony and species in
field studies) depends on a trigger criterion, the Hazard
an ecosystem) is essential to sustain the system as a whole(12). Moreover, considering their role as pollinators (13),
* Corresponding authors. Phone: (33/0) 2 31 56 59 10 (M.-P.H.),
honey bees represent a good model to assess the risk of
(33/0-1) 69 82 37 17 (A.R.); fax: (33/0) 2 31 93 11 88 (M.-P.H.),
insecticides to pollinators and to protect many plant species
(33/0-1) 69 82 37 36 (A.R.); e- mail: [email protected] or
that rely on these organisms, through pollination, to repro-
Centre d’Etude et de Recherche du Me´dicament de Normandie.
‡ Laboratoire Evolution, Ge´nomes, Spe´ciation, CNRS.
According to the PEC/PNEC approach and with the
example of imidacloprid, we determined PECs with the
2448 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006 TABLE 1. Estimated PECs for Different Categories of Honey Bees to Imidacloprida imidacloprid imidacloprid imidacloprid imidacloprid imidacloprid categories of bees
a Estimated amounts of food (sugar and pollen in mg) and imidacloprid (in pg) consumed per bee over d days, with different levels of food
contamination (%). b Data from ref 21. c N. A.: No available data.
known concentrations of imidacloprid found in the con-
bees. Nevertheless, a few studies have investigated the impact
taminated pollens (sunflower and maize) and nectar (sun-
of imidacloprid on honey bees by chronic and sublethal
flower) consumed by honey bees, and we determined PNECs
intoxications in laboratory, semi-field, and field conditions.
with the data derived from studies on acute, chronic, and
To assess these studies, we referred to the EPPO guidelines
sublethal toxicities of imidacloprid to honeybees.
because they present guidelines for semi-field and field
Criteria for the Validation of Data. To determine the
concentration of imidacloprid in pollens and nectar issued
PEC Estimates. By definition, a PEC corresponds to the
from imidacloprid seed-dressed plants and the honey bees’
amount of pesticides a honey bee might be exposed to, either
exposure to imidacloprid, the CST validated the data of all
by ingestion or contact. In this paper, with the example of
the studies that met the following requirements in terms of
imidacloprid and honey bees, we only considered oral
sampling procedure, chemical analyses, and toxicity testing.
exposures because data on topical exposures are scant.
For the sampling procedure, studies needed to describe
We can estimate honey bees’ exposure to both contami-
thoroughly the methods used (sampling location, pesticide
nated pollens (sunflower and maize) and nectar (sunflower)
treatments history) and gather sufficient samples, both
with (i) the known and validated concentrations of imida-
qualitatively and quantitatively. Qualitatively, data obtained
cloprid found in contaminated pollens and nectars, and (ii)
from pollens collected directly on anthers of treated flowers,
the amount of contaminated pollen and nectar consumed
rather than in pollen traps, were kept and validated because
by different categories of honey bees (21).
the concentration of imidacloprid in pollens of traps is highly
(i) The amount of imidacloprid present in the food of
variable and depends on the environment (i.e., the amount
honey bees is directly related to the environement. For
of plants treated by systemic insecticides) (14). Quantitatively,
example, if a hive is located near extensive cultures of maize
the CST retained the value of a minimum of 10 samples to
and sunflower plants treated by imidacloprid, the proportions
enable statistics (means and standard deviations). Samples
of pollen and/or nectar that might be contaminated by
coming from different experiments and locations, but
imidacloprid are expected to be high. Since the relative
presenting similar protocols, were grouped together to get
proportions of contaminated food, versus uncontaminated
food, consumed by honey bees are unknown, we considered
For the chemical analyses, given the high toxicity of
5 different levels of contamination ranging from 20% (a low
imidacloprid, we validated studies that detected the molecule
level of contamination) to 100% (the highest level of
the most accurately as possible, that is by high performance
contamination) (Table 1), although the latter case might rarely
liquid chromatography (HPLC) coupled to mass spectrometry
(MS) (15) and using an appropriate limit of quantification
(ii) The amount of contaminated food consumed by
(LOQ ) 1 µg/kg), limit of detection (LOD < 0.5µg/kg), and
different categories of honey bees depends on the amount
sample weight (10 g). The only study that used a radioactivity
of food the bees require to achieve particular tasks within
method coupled with thin-layer chromatography (TLC) and
the colony. Among them, Rortais et al. (21) considered the
automated multiple development (AMD) techniques (16) was
categories that are potentially the most exposed to imida-
validated too because it allowed a clear identification of
cloprid: those that achieve the most costly tasks in terms of
imidacloprid, unlike less specific methods such as the
energy and which consume the highest amounts of pollen
derivation and gas chromatography (GC) (17, 18).
for their development. Therefore, for the calculation of the
To determine the toxicity of imidacloprid to honey bees,
PEC, the following categories of honey bees were consid-
studies apply standardized tests designed by the OECD
ered: the worker larvae which consume pollen and nectar
guidelines (19, 20). Such tests are developed in laboratory
for their development over about 5 days; the drone larvae
conditions to assess oral (19) and topical (20) acute toxicities
which consume pollen and nectar for their development over
of pesticides (and other chemicals) to adult worker honey
about 6.5 days; the nurses which consume pollen over a
bees. These laboratory tests follow the EPPO guidelines No.
period of 10 days and nectar and/or honey to maintain the
170 (1) and the recommendations made by the International
nest temperature at 34 °C over the entire brood attendance
Commission for Plant-Bee Relationships (ICPBR). While
period, lasting about 8 days; the wax-producing bees which
these guidelines propose methods to test oral and topical
consume nectar during the period of maximum wax pro-
acute toxicities, there are currently no standardized tests to
duction, lasting about 6 days; the winter bees which consume
study chronic and sublethal toxicities of pesticides to honey
nectar and honey to maintain the nest temperature at viable
VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 2449
temperature during winter, lasting about 90 days in temperate
additional long-term toxicity tests of two or three trophic
regions; and the nectar and pollen foragers which consume
levels of known NOEC, the factors 50 and 10, respectively,
nectar and/or honey to cover their daily flight expenses. As
are selected. In field conditions, an assessment factor is
a forager life span is highly variable (between 1 and 3 weeks),
determined case by case (10).
the amount of food consumed by a forager to collect food
The approach presented in this paper consisted in finding
has been estimated over a minimal period of one week.
appropriate PNECs for honey bees derived from PNECs
Honey Bees’ Exposure to Contaminated Sunflower and/or
designed for ecosystems. These PNEC values were estimated
Maize Pollens. The honey bee’s exposure to contaminated
with the available data obtained from studies on oral acute,
pollens collected on treated plants is determined by the
chronic, and sublethal toxicities of imidacloprid to honey
bees. These values were derived from the lowest validatedtoxicities (LD50, lowest observed effect concentration (LOEC),
PEC (pg) ) Validated concentration of imidacloprid found
or no observed effect concentration (NOEC)) to which
in sunflower and/or maize pollens (µg/kg) ×
assessment factors are applied. This new approach is
Amount of pollens consumed by honey bees (mg) ×
specifically adapted to honey bees because it allows the
Levels of contamination found in pollen (%)
assessment of both colonies and individuals. These factorshad to be determined case by case, following the standard
Honey Bees’ Exposure to Contaminated Sunflower Nectar.
approach used by the TGD. Every time new data enabled us
The nectar brought back to the colony is consumed by honey
to reduce extrapolations (chronic toxicity data in relation to
bees, either rapidly as it is or later on as honey when it is
acute toxicity data), the assessment factors were generally
stored. The relative amounts of nectar and honey consumed
by honey bees are unknown. However, the amounts of sugar
PEC/PNEC Estimates. The hazard posed by new sub-
contained in sunflower honey and nectar are known and
stances to organisms is determined by the PEC/PNEC ratio.
are, on average, 80% and 59%, respectively (22, 23). Therefore,
When this ratio is over 1, it highlights an intoxication risk for
the amounts of sunflower nectar and/or honey consumed
honey bees, whereas when it is below 1, it indicates no risk.
by honey bees can be determined by their sugar consumption,
According to the TGD (10), this ratio is obtained and derived
in relation to their energy requirement (21). As a result, for
from acute toxicity data, but when a risk is found, the ratio
every milligram of sugar required, a honey bee will have to
is re-calibrated with new data obtained in more representative
consume 1.25 mg of sunflower honey or 1.69 mg of sunflower
conditions. For honey bees, the PECs were determined with
nectar. Therefore, a honey bee’s exposure to contaminated
all the available scientific data found on honey bees’ food
sunflower nectar can be determined by the following
consumptions because there were sufficient data, whereas
the PNECs required more data. Therefore, following the TGDprocedure, PNECs were derived from acute toxicity data.
PEC (pg) ) Validated concentration of imidacloprid
When a risk was highlighted, a new PEC/PNEC ratio was
found in sunflower nectar (µg/kg) × Amount of sugar
determined with data obtained from chronic toxicity studies.
consumed by honey bees (mg) × 1.69 Levels of
If the new ratio remained over 1, a final PEC/PNEC ratio was
contamination found in sunflower nectar (%)
then calculated with new data coming from sublethal fieldtoxicity studies. This final ratio is the most representative
Honey Bees’ Exposure to Contaminated Sunflower and/or
ratio of the natural conditions of a honey bees’ colony. Maize Pollens and to Contaminated Sunflower Nectar. Thehoney bee’s exposure to contaminated sunflower and/or
maize pollens and to contaminated sunflower nectar is
PEC Estimates. (i) In pollen collected directly on the anthers
of flowers, the concentrations of imidacloprid found in treatedsunflower and maize plants are 3.3 and 3.5 µg/kg, respectively
(24, 25), or on average 3.4 µg/kg for both pollen types. Theconcentration of imidacloprid found in treated sunflower
PNEC Estimates. By definition, a PNEC corresponds to
nectar is 1.9 µg/kg (25). (ii) Based on the estimated amounts
the amount of substances that will have no impact on
of pollen and nectar consumed by honey bees over several
ecosystems. For numerous substances, the pool of data is
days of activity (21), the potential amounts of imidacloprid
usually too limited to predict their effects on ecosystems. In
ingested by honey bees were determined (Table 1).
such circumstances, empirically derived assessment factors
PNEC Estimates. There is currently no test and no toxicity
must be applied. These assessment factors allow the predic-
data for larvae. For this category of honey bees, PNECs were
tion of a concentration below which an unacceptable effect
derived from the toxicity data obtained in adult workers.
will most likely not occur. The size of these assessment factors
Table 2 shows the PNECs determined in adults and derived
incorporates various uncertainties due to extrapolations from
from acute, chronic, and sublethal toxicity data, to which
single-species laboratory data to a multi-species ecosystem,
specific assessment factors were applied.
in particular uncertainties due to intra- and inter-laboratory
From acute toxicity data: the lowest validated LD50 (48
variations in toxicity data, intra- and inter-species variations,
h) is 3.7 ng of imidacloprid per bee (26). According to the
short-term to long-term toxicity extrapolations, and from
TGD (10), the assessment factor for acute toxicity data is
laboratory data to field impact studies. For the terrestrial
1000. However, the toxicity of imidacloprid was determined
compartment, the size of the assessment factors depends on
by several studies which tested models belonging to the same
the confidence we have on the representativeness of the
species, and found similar results. As these data present very
toxicity data. For example, the size of these factors is reduced
few uncertainties, an assessment factor of 100 was applied.
when more data become available at various trophic levels
Therefore, the validated PNEC becomes 3.7/100 ) 37 pg/
Based on these parameters and in relation to the
From Chronic Toxicity Data. In laboratory conditions, the
experimental conditions, the TGD determines various as-
lowest validated value was LD50 (10 d) ) 0.012 ng/bee (27).
sessment factors. In laboratory conditions, for short-term
As this value was obtained from a long-term experiment, it
toxicity tests (LD50) and for one trophic level, a factor 1000
seemed appropriate to apply the same assessment factor as
is used, for long-term toxicity tests and for several trophic
the one used for a long-term NOEC experiment, which is 100
levels with known NOEC, a factor 100 is applied, and for
(10). However, this factor is used to determine a PNEC for
2450 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006 TABLE 2. PNECs for Oral intoxications of Honey Bees to Imidacloprid experimental conditions assessment toxicities (imidacloprid intakes) doses (ng/bee) observed variables (pg/bee)
a Feeders contained syrup contaminated by imidacloprid. In field conditions, feeders were placed near hives to reinforce the observed effects
of imidacloprid seed-dressed plants on honey bees.
all the taxonomic groups of an ecosystem. To adjust this
and greater in adult hive bees than in any other categories
factor to one taxonomic group (the honey bees), we applied
of bees. Whatever the validated toxicity data are, the
a factor 10. This factor includes all variations found among
determined PNECs are in a limited range of values (between
and within taxonomic groups (inter- and intra-species
1.2 and 50 pg/bee). These estimates are in agreement with
variations). Therefore, the validated PNEC becomes 0.012/
observations made in regions of extensive sunflower and
maize cultures, which report a decrease in honey production
From Sublethal Toxicity Data. In laboratory, semi-field,
since the launching of imidacloprid on sunflower plants in
and field conditions, one or several administered doses might
1994 (32), and several behavioral dysfunctions, foragers
induce behavioral modifications among treated honey bees.
disappearances, and great honey bee mortalities in summer,
When administering a unique oral dose of imidacloprid to
during the blossoming of maize and sunflower plants, and
honey bees for the testing of the knockdown effect, the lowest
after winter, when all sunflower and maize pollens have been
validated NOEC was 0.94 ng/bee (28). As this value was
obtained from a short-term experiment, an assessment factor
In areas of extensive sunflower and maize cultures treated
of 100 should have been applied (TGD). However, this value
by imidacloprid, all categories of honey bees, whatever their
does not correspond to a LD50; it is a dose that has no impact
age is, are at risk of intoxication. In such a situation, honey
on honey bees. Moreover, the measured effect is a sublethal
bees are most likely to bring back food that is contaminated
effect. Therefore, we applied an assessment factor of 50. The
by imidacloprid, and the observed effects might relate to
validated PNEC becomes 0.94/50 ) 18.8 pg/bee. When
either acute, chronic, or sublethal intoxications, all inducing
administering several oral doses of imidacloprid to honey
bees for the testing of the proboscis extension reflex (PER),the lowest validated concentration, after a 10 day experiment,
In areas where sunflower and maize cultures treated by
was 0.2 ng/bee (29). This value corresponds to a NOEC based
imidacloprid are less abundant, honey bees might be less
on the testing of sublethal effects after a long-term intoxica-
intoxicated because they might consume a mixture of
tion. Therefore, we applied an assessment factor of 10. The
contaminated and uncontaminated food. In this situation,
validated PNEC becomes 0.2/10 ) 20 pg/bee.
honey bees are most likely to be intoxicated by sublethal
In semi-field conditions, a LOEC (5 d) of 0.075 ng/bee
doses, rather than by acute or chronic doses, which might
was validated for the testing of the time spent feeding on
have lethal consequences at the individual and colony levels.
contaminated syrup (30). As this study was conducted in the
At sublethal doses, pesticides are known to have profound
natural conditions of foragers, an assessment factor of 10
impacts on the colony, in particular on the honey bees’
was applied (TGD, 10). The validated PNEC becomes 0.075/
longevity (34), the brood production (35, 36), the development
of hypopharyngeal glands (37), and the egg laying (38).
In field conditions, a lowest NOEC (10 d) of 0.25 ng/bee
Imidacloprid is known to affect the honey bees’ cognitive
was validated for the testing of dances (31). Studies conducted
behaviors such as the proboscis extension reflex PER (33).
in field conditions present similar conditions to those found
Learning and memorization in honey bees’ tasks are very
in the natural environment of honey bees. Therefore, an
important. For example, a forager that is disorientated might
assessment factor of 1 should be applied (TGD, 10), but we
get lost and eventually die. In the case of massive foragers’
selected an assessment factor of 5 because the setting of the
intoxications, the colony is likely to be greatly affected. In an
feeders is artificial. Therefore, the validated PNEC becomes
experiment under tunnels, Vandame et al. (39) exposed honey
bees to deltamethrin at a sublethal dose that is 20-fold lower
PEC/PNEC Estimates. According to the previously defined
than the registered dose at which foragers are expected to
assessment factors, and whatever the level of food contami-
be exposed to in the environment. They found that 54% of
nation is, all the investigated categories of honey bees
the treated bees were disoriented and took flight toward the
presented an intoxication risk to imidacloprid (Figure 1).
sun. The authors concluded that such sublethal effects may
The PEC/PNEC ratio was the highest for winter bees and
be the cause of the symptom called the “disappearance bee
nurses (between 10 and 100) and the lowest for pollen foragers
disease” by beekeepers who observed colonies’ weakening
without finding dead bees close to the hives. This hypothesiswas formerly raised by other scientists (40, 41). Discussion
Imidacloprid can also affect honey bees by chronic
The PEC/PNEC derived from the calculation of honey bees’
intoxications. In the long run, a repeated ingestion of low
exposure to which appropriate assessment factors were
doses of imidacloprid could cause immunodeficiency and
applied show that the risk posed by imidacloprid is alarming
diseases in honey bees. The impairment of the bees’
for all categories of honey bees. These ratios are all over 1,
immunity system is a nonspecific mechanism (42). For
VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 2451 FIGURE 1. Hazard posed by imidacloprid to different categories of honey bees feeding on various proportions of contaminated food: estimated PEC/PNEC ratios derived from (A) acute toxicity data, (B) chronic toxicity data, and (C) sublethal toxicity data obtained in field conditions for foragers and in laboratory conditions for all the other categories of honey bees (a risk is highlighted when ratios are greater than 1).
example, sublethal concentrations of malathion result in
In some cases, no honey bee troubles were observed by
higher invasions of treated colonies by the wax moth (43).
beekeepers, but no scientific study has ever confirmed these
2452 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006
observations. The presence of untreated or very little treated
impacts of systemic insecticides on honey bees (9, 49, 50),
areas near hives, and the presence of compensatory phe-
as well as their sublethal effects on the behavior (41, 47,
nomena (increase of brood development, replacement of
51-53) and physiology (38, 54, 55) of these organisms.
dead foragers) with no visible harmful consequences for the
Based on the risk assessment method used for terrestrial
colony may occur and explain the absence of any observed
organisms, this method is original. It includes the assessment
of several important parameters such as the following: (i)
When assessing the risk posed by systemic insecticides,
The detection and measurement of the amount of an active
the HQ does not take into account several idiosyncrasic
ingredient present in the various substrates used by honey
parameters such as persistence in soils, presence in pollen
bees. These measures are not statutorily requested. (ii) The
and nectar, and transport in the air. The calculation of the
development of various scenarios of honey bees’ exposure
Toxicity Exposure Ratios (TER) (ratio between a toxicological
to the active ingredient. These scenarios better predict the
end point and a PEC), regularly used in the risk assessment
risk posed by systemic insecticides to honey bees because
of pesticides to organisms (mainly vertebrates and nematods),
they take into account the biology and particular require-
could take into account such crucial parameters. However,
ments of a honey bees’colony (21). (iii) The use of novel and
for social invertebrates such as honey bees, the use of the
validated methods for the assessment of lethal and sublethal
new and existing chemical substances approach (herein the
honey bees’ intoxications. (iv) The use of assessment factors,
PEC/PNEC ratio) should be more appropriate than the use
when experimental designs are tightly related to environ-
of the TER because it enables the protection of the whole
mental conditions. This approach is usually applied to assess
colony. The PEC/PNEC ratio could then be re-calibrated when
more data on imidacloprid and on other systemic insecticidesare available. Acknowledgments
For hive bees (nurses and winter bees), the PNECs could
Authors M.P.H. and A.R. contributed equally to this work.
be refined when more data are available on the mechanisms
The validation of the data used in this study on imidacloprid
of a colony’s regulation (e.g., brood development) in field
was realized in concert with a group of experts nominated
and semi-field conditions. For larvae, exposures were derived
by the French Ministry of Agriculture and known as the
from data obtained on adult toxicities in order to obtain an
“Comite´ Scientifique et Technique (CST) de l’Etude multi-
indicative and comparative value. Given that larvae are more
factorielle des troubles des abeilles”. M.P.H. and A.R. were
or less sensitive than adults to chemicals (4, 44), more studies
financed by the French Ministry of Agriculture to give
need to determine accurately their exposure risk to imida-
cloprid and to other systemic insecticides.
We could not investigate topical exposures of imidacloprid
Note Added after ASAP Publication
to honey bees because there are not enough data available
An incorrect reference was cited in Table 1 in the version
on this mode of exposure. However, honey bees’ intoxications
published ASAP March 7, 2006; the corrected version was
by topical exposures should not be discarded. For examples,
foragers might get contaminated by contaminated dustparticles during sowing operations (45).
The impact of systemic insecticides on honey bees is not
limited to the impact of the parent compound; it also includes
(1) OEPP/EPPO. EPPO standards PP1/170 (3). Test methods for
exposures to its metabolites. In the case of imidacloprid,
evaluating the side effects of plant protection products on honeybees. Bull. OEPP/EPPO 2001, 31, 323-330.
some metabolites (e.g., olefin, which is twice more toxic than
(2) OEPP/EPPO. Environmental risk assessment scheme for plant
imidacloprid) are found to be very toxic to honey bees (9, 46,
protection products. Chapter 10. Bull. OEPP/EPPO 2003, 33, 47) and some of them are detected at low concentrations
(between 0.3 and 1 µg/kg) in rape pollen and nectar (48).
(3) Smart, L. E.; Stevenson, J. H. Laboratory estimation of toxicity
However, to investigate in further detail the impact of
of pyrethroid insecticides to honeybees: relevance to hazard
metabolites on honey bees, their concentrations in other
in the field. Bee World 1982, 62, 150-152.
(4) Villa, S.; Vighi, M.; Finizio, A.; Serini, G. B. Risk assessment for
types of pollen and nectar must be determined.
honeybees from pesticide-exposed pollen. Ecotoxicology 2000,
Exposures to imidacloprid were estimated by assuming
that the molecule is stable in the hive because it is stored in
(5) Glynne Jones, G. D.; Thomas, W. D. E. Experiments on the
a dark environment. However, the transformation of pollen
possible contamination of honey with Schradan. Ann. Appl.
and nectar into bee bread and honey, respectively, imply the
Biol. 1953, 40, 546-555.
(6) Knapp, J. L.; Ansonmoye, H. Aldicarb and metabolite residues
action of several enzymes that might change the stability of
in nectar following commercial applications to Florida USA
imidacloprid. Therefore, the concentration of imidacloprid
Citrus. In Proceedings of the 6th International Citrus Congress,
in the stored food (bee bread and honey) should be measured
Tel Aviv, Israel, March 6-11, 1988; pp 1107-1111.
to test its stability in the hive over time. Honey bees’ exposures
(7) Comite´ Scientifique et Technique. Imidaclopride utilise´ en
to contaminated sunflower nectar were determined with data
enrobage de semences (Gaucho) et troubles des abeilles. Rapport
issued from one study (25). To confirm and generalize the
du Comite´ scientifique et Technique de l’e´tude multifactorielledes troubles des abeilles remis au Ministe`re de l’Agriculture:
trend found, it is necessary to conduct more studies (i.e., the
concentration of imidacloprid in nectar coming from other
(8) Elbert, A.; Beckert, B.; Hartwig, J.; Erdelen, C. Imidacloprid -
varieties of sunflower and from other melliferous plants).
a new systemic insecticide. Pflanzenschutz-Nachr. 1991, 44,
The method and the assessment factors proposed in this
paper could be re-calibrated when more data are available.
(9) Suchail, S.; Guez, D.; Belzunces, L. P. Discrepancy between acute
and chronic toxicity induced by imidacloprid and its metabolites
Although the determination of the LD50 (48 h) is readily
in Apis mellifera. Environ. Toxicol. Chem. 2001, 20, 2482-2486.
obtained for the calculation of HQ, its representativeness in
(10) European Commission. Technical Guidance Document (TGD)
testing the survival of a honey bee colony is arguable. To
in support of commission Directive 93/67/EEC on risk assessment
assess the risk posed to honey bees, chronic and sublethal
for new notified substances and commission regulation (EC) N .
toxicity tests must be conducted systematically, especially
1488/94 on risk assessment for existing substances; EC: Brussels,
in the case of systemic insecticides which have a long-lasting
(11) Moritz, R. F. A.; Southwick, E. E. Bees as superorganisms, an
action. To achieve these tests, standardized protocols are
evolutionary reality; Springler-Verlag: Berlin, 1992; 395 pp.
required and could be elaborated on the grounds of existing
(12) Heinrich, B. Bumblebee Economics; Harvard University Press:
experimental studies which have investigated the chronic
VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 2453
(13) Williams, I. H. The dependence of crop pollination within the
(35) Ferguson F. Long-term effects of systemic pesticides on honey
European Union on pollination by honey bees. Agric. Zool. Rev.
bees. Australas. Beekeeper 1987, 49-54. 1994, 6, 229-257.
(36) Bendahou, N.; Fleche, C.; Bounias, M. Biological and biochemical
(14) Charvet, R.; Katouzian-Safadi, M.; Colin, M. E.; Marchand, P.
effects of chronic exposure to very low levels of dietary
A.; Bonmatin, J. M. Insecticides syste´miques: de nouveaux
risques pour les insectes pollinisateurs. Ann. Pharm. Fr. 2004,
(Hymenoptera: Apidae). Ecotoxicol. Environ. Safe. 1999, 44,
(15) Bonmatin, J. M.; Moineau, I.; Charvet, R.; Fle´che´, C.; Colin M.
(37) Lensing, W. Changes in honey bee workers after feeding them
E.; Bengsh, E. R. A LC/APCI-MS/MS method for the analysis of
with sublethal dose of dimethoate. Apidologie 1986, 17, 340-
imidacloprid in soils, in plants, and in pollens. Ann. Chem. 2003,
(38) Lensing, W. Changes in honey bee workers after feeding on
(16) Stork, A. A residue of 14C-NTN 33893 (imidacloprid) in blossoms
sublethal doses of dimethoate. Apidologie 1987, 18, 353-356.
of sunflowers (Helianthus annus) after seed dressing. Bayer A.
(39) Vandame, R.; Meled, M.; Colin, M. E.; Belzunces, L. P. Alteration
G., Crop Protection Development, Institute for Metabolism
of the homing-flight in the honey bee Apis mellifera L. exposed
Research and Residue Analysis: Leverkusen, 1999; p 56.
to sublethal dose of deltamethrin. Environ. Toxicol. Chem. 1995,
(17) MacDonald, L.; Meyer, T. J. Determination of imidacloprid and
triadimefon in white pine by gas chromatography/mass spec- trometry. Agric. Food. Chem. 1998, 46, 3133-3138.
(40) Faucon, J. P.; Flamini, C.; Colin, M. E. Evaluation de l’incidence
(18) Uroz, F. J.; Arrebola F. J.; Egea-Gonzales, F. J.; Martinez-Vidal,
de la deltamethrine sur les proble`mes du cheptel apicole: Essais
J. L. Monitoring of 6-chloronicotinic acid in human urine by
en plein champ. Bull. Lab. Ve´te´rinaires 1985, 18, 33-45.
gas chromatography-tandem mass spectrometry as indicator
(41) Cox, R. L.; Wilson, W. T. Effects of Permethrin on the behavior
of exposure to the pesticide imidacloprid. Analyst 2001, 126,
of individually tagged honey bees, Apis mellifera L. (Hy-
menoptera: Apidae). Environ. Entomol. 1984, 13, 375-378.
(19) OECD. Honey bees, Acute oral toxicity test, OECD’s guidelines
(42) Glinski, Z.; Kauko, L. Immunosuppression et immunotoxicolo-
for the testing of chemicals, Section 2, Effects on biotic systems;
gie: Aspects lie´s a` la protection de l’abeille mellife`re contre les
agents microbiens et parasitaires. Apiacta 2000, 35, 65-76.
(20) OECD. Honey bees, Acute contact toxicity test, OECD’s guidelines
(43) Nation, J. L.; Robinson, F. A.; Yu, S. J.; Bolten, A. B. Influence
for the testing of chemicals, Section 2, Effects on biotic systems;
upon honey bees of chronic exposure to very low levels of
selected insecticides in their diet. J. Apic. Res. 1986, 25, 170-
(21) Rortais, A.; Arnold, G.; Halm, M. P.; Touffet-Briens, F. Modes
of honeybees exposure to systemic insecticides: estimated
(44) Atkins, E. L.; Kellum D. Comparative morphogenic and toxicity
amounts of contaminated pollen and nectar consumed by
studies on the effect of pesticides on honeybee brood. J. Apic.
different categories of bees. Apidologie 2005, 36, 71-83. Res. 1986, 25, 242-255.
(22) Crane, E. Honey: A Comprehensive Survey; Heinemann: London,
(45) Greatti, M.; Sabatini, A. G.; Barbattini, R.; Rossi, S.; Stravisi, A.
Risk of environmental contamination by the active ingredient
(23) Pham-Dele`gue, M. H. Etude par conditionnement associatif
imidacloprid used for corn seed dressing. Preliminary results.
des parame`tres olfactifs qui de´terminent le comportement
Bul. Insect. 2003, 56, 69-72.
alimentaire se´lectif de l’abeille. Ph.D. Thesis, University Pierre
(46) Nauen, R.; Ebbinghaus-Kintscher, U.; Schmuck, R. Toxicity and
nicotinic acetylcholine receptor interaction of imidacloprid and
(24) Bonmatin, J. M.; Moineau, I.; Colin, M. E.; Bengsch, E. R.;
its metabolites in Apis mellifera (Hymenoptera: Apidae). Pest.
Lecoublet, S. Effets des produits phytosanitaires sur les abeilles.Manag. Sci. 2001, 57, 577-586. Analyse de l’imidaclopride dans les pollens; Rapport de re´sultats
(47) Decourtye, A.; Lacassie, E.; Pham-Dele`gue, M. H. Learning
n . 10; CNRS-CBM, INRA: Orle´ans, 2001.
performance of honeybees (Apis mellifera L.) are differentially
(25) Stork, A. Uptake, translocation and metabolism of imidacloprid
affected by imidacloprid according to season. Pest Manage. Sci.
in plants. Bull. Insect. 2003, 56, 35-40. 2003, 59, 269-278.
¨ ck, R.; Scho¨ning, R.; Stork, A.; Schramel, O. Risk posed
to honeybees (Apis mellifera L, Hymenoptera) by an imida-
(48) Scott-Dupree, C. D.; Spivak, M. S. The impact of Gaucho and
cloprid seed dressing of sunflowers. Pest Manage. Sci. 2001, 57,
TI-435 Seed-treated canola on honey bees, Apis mellifera L.;
University of Guelph, Ontario, Canada, University of Minnesota,
(27) Suchail, S. Etude pharmacodynamique de la le´talite´ induite par
l’imidaclopride et ses me´tabolites chez l’abeille domestique (Apis
(49) Stoner, A.; Wilson, W. T.; Rhodes, H. Carbofuran: effect of long-
mellifera L.). Ph.D. Thesis, University Claude Bernard, Lyon,
term feeding of low doses in sucrose syrup on honey bees in
standard-size field colonies. Environ. Entomol. 1982, 11, 53-
(28) Wilhelmy, H. Substance A. Acute effects on the honeybee Apismellifera (Hymenoptera, Apidae), NON-GLP; Laboratorium fu
(50) Moncharmont, F. X. D.; Decourtye, A.; Hennequet-Hantier, C.;
angewandte biologie: Sarstedt, 2000.
Pons, O.; Pham-Dele`gue, M. H. Statistical analysis of honeybee
(29) Decourtye, A.; Armengaud, C.; Renou, M.; Devillers, J.; Cluzeau,
survival after chronic exposure to insecticides. Environ. Toxicol.
S.; Gauthier, M.; Pham-Dele`gue M. H. Imidacloprid impairs
Chem. 2003, 22, 3088-3094.
memory and brain metabolism in the honeybee (Apis mellifera
(51) Johansen, C. A. Behavior of pollinisators following insecticide
L.). Pestic. Biochem. Phys. 2004, 78, 83-92.
exposure. Am. Bee J. 1984, March, 225-227.
(30) Colin, M. E.; Bonmatin, J. M.; Moineau, I.; Gaimon, C.; Brun,
(52) Taylor, K. J.; Waller, G. D.; Crowder, L. A. Impairment of classical
S; Vermande`re, J. A method to quantify and analyse the foraging
conditioned response of the honey bee (Apis mellifera L.) by
activity of honey bees: relevance to the sublethal effects induced
sublethal doses of synthetic pyrethroid insecticides. Apidologie
by systemic insecticides. Arch. Environ. Contam. Toxicol. 2004, 1987, 18, 243-252.
(53) Thompson, H. M. Behavioural effects of pesticides in bees. Their
(31) Kirchner, W. H. Mad-bee disease? Sublethal effects of imida-
potential for use in risk assessment. Ecotoxicology 2003, 12,
cloprid (Gaucho) on the behavior of honey-bees. Apidologie1999, 30, 421-422.
(54) Bounias, M.; Dujin, N.; Popeskoviae, D. S. Sublethal effects of
(32) Bonmatin, J. M.; Moineau, I.; Charvet, R.; Colin, M. E.; Fle`che,
synthetic pyrethroid, deltamethrin, on the glycemia, the lipemia,
C.; Bengsch, E. R. Behaviour of imidacloprid in fields. Toxicity
and the gut alkaline phosphatases of honeybees. Pestic. Biochem.
for honey bees. In Environmental Chemistry, Green ChemistryPhys. 1985, 24, 149-160. and Pollutants in Ecosystems; Lichtfouse, E., Schwarzbauer, J.,
(55) Papaefthimiou, C.; Theophilidis, G. The cardiotoxic action of
Robert, D., Eds.; Springer: New York, 2005; pp 483-494.
the pyrethroid insecticide deltamethrin, the azole fungicide
(33) Decourtye, A.; Devillers, J.; Genecque, E.; Le Menach, K.;
prochloraz, and their synergy on the semi-isolated heart of bee
Budzinski, H.; Cluzeau S.; Pham-Dele`gue M. H. Comparative
Apis mellifera macedonica. Pestic. Biochem. Phys. 2001, 69, 77-
sublethal toxicity of nine pesticides on olfactory learning
performances of the honey bee Apis mellifera. Arch. Environ. Contam. Toxicol. 2005, 48, 242-250.
(34) Smirle, M. J.; Winston, M. L.; Woodward, K. L. Development of
Received for review July 18, 2005. Revised manuscript re-
a sensitive bioassay for evaluating sublethal pesticide effects
ceived February 2, 2006. Accepted February 3, 2006.
on the honey bee (Hymenoptera: Apidae). J. Econ. Entomol. 1984, 77, 63-67. 2454 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006
Florida Breast Health Specialists Hormone Therapy Information and Questions to Ask Your Doctor What is Hormone Therapy? Hormonal therapy medicines are whole-body (systemic) treatment for hormone-receptor- positive breast cancers. Hormone receptors are like ears on breast cells that listen to signals from hormones. These signals "turn on" growth in cells that have receptors