Content uploaded by Sebastian Stehle
Author content
All content in this area was uploaded by Sebastian Stehle on Dec 23, 2015
Content may be subject to copyright.
Content uploaded by Sebastian Stehle
Author content
All content in this area was uploaded by Sebastian Stehle on Aug 18, 2015
Content may be subject to copyright.
RESEARCH ARTICLE
Pesticide authorization in the EU—environment unprotected?
Sebastian Stehle
1
& Ralf Schulz
1
Received: 20 April 2015 /Accepted: 3 August 2015 /Published online: 15 August 2015
#
Springer-Verlag Berlin Heidelberg 2015
Abstract Pesticides constitute an integral part of high-
intensity European agriculture. Prior to their authorization, a
highly elaborated environmental risk assessment is mandatory
according to EU pesticide legislation, i.e., Regulation (EC)
No. 1107/2009. However, no field data-based evaluation of
the risk assessment outcome, i.e., the regulatory acceptable
concentrations (RACs), and therefore of the overall protec-
tiveness of EU pesticide regulations exists. We conducted here
a comprehensive meta-analysis using peer-reviewed literature
on agricultural insecticide concentrations in EU surface waters
and evaluated associated risks using the RACs derived from
official European pesticide registration documents. As a re-
sult, 44.7 % of the 1566 cases of measured insecticide con-
centrations (MICs) in EU surface waters exceeded their re-
spective RACs. It follows that current EU pesticide regula-
tions do not protect the aquatic environment and that insecti-
cides threaten aquatic biodiversity. RAC exceedances were
significantly higher for insecticides authorized using conser-
vative tier-I RACs and for more recently developed insecti-
cide classes, i.e., pyrethroids. In addition, we identified higher
risks, e.g., for smaller surface waters that are specifically con-
sidered in the regulatory risk assessment schemes. We illus-
trate the shortcomings of the EU regulatory risk assessment
using two case studies that contextualize the respective risk
assessment outcomes to field exposure. Overall, our meta-
analysis challenges the field relevance and protectiveness of
the regulatory environmental risk assessment conducted for
pesticide authorization in the EU and indicates that critical
revisions of related pesticide regulations and effective mitiga-
tion measures are urgently needed to substantially reduce the
environmental risks arising from agricultural insecticide use.
Keywords Pesticide
.
Surface water
.
Europe
.
Risk
assessment
.
Regulation (EC) No. 1107/2009
.
Regulatory
acceptable concentration
.
Meta-analysis
Introduction
Agricultural areas cover 40 % (174.1 million hectares) of the
total land area of the EU-28, and two thirds (65.8 %) of these
farmlands are used for the cultivation of arable and permanent
crops (Eurostat 2013). In 2013, pesticides with an approxi-
mate input value of 11 billion Euros were applied to
European arable lands (European Comm ission 2014). The
widespread and intentional release of these highly biologically
active substances poses threats to non-target aquatic and ter-
restrial ecosystems across the EU. Surface waters are especial-
ly at risk as systems that are likely to receive agricultural non-
point source inputs due to their often close proximities to
arable lands (Stehle and Schulz 2015; Davies et al. 2008;
Schulz 2004). We focus here on insecticides, as this particu-
larly toxic group of pesticides exhibits a high toxicity potential
for aquatic organisms that are crucial for ecosystem structure
and functions (Schulz 2004; Schäfer et al. 2012 ;USEPA
2014).
Stehle and Schulz (2015) showed that insecticides threaten
aquatic biodiversity on a global scale, but did they not specify
results, e.g., for the highly regulated EU. Although large-scale
Responsible editor: Philippe Garrigues
Electronic supplementary material The online version of this article
(doi:10.1007/s11356-015-5148-5) contains supplementary material,
which is available to authorized users.
* Sebastian Stehle
stehle@uni-landau.de
1
Institute for Environmental Sciences, University Koblenz-Landau,
Fortstrasse 7, 76829 Landau, Germany
Environ Sci Pollut Res (2015) 22:19632–19647
DOI 10.1007/s11356-015-5148-5
investigations of insecticide exposure of EU surface waters
are lacking (Stehle and Schulz 2015), a recent study (Malaj
et al. 2014) using governmental monitoring data and standard
toxicity data derived from a single species laboratory test
showed that, out of various organic pollutants, insecticides
particularly jeopardize the integrity of EU freshwater ecosys-
tems. In addition, several additional small-scale field studies
conducted in the EU reported that pesticide exposure pro-
duced adverse effects on the aquatic ecosystem structure and
function (e.g., Bereswill et al. 2013; Schäfer et a l. 2012;
Beketov et al. 2013; Schulz 2004). However, no scientific
study exists that has evaluated pesticide, or specifically insec-
ticide, field concentrations in EU agricultural surface waters
using the regulatory acceptable concentrations (RACs) de-
fined by the environmental risk assessment conducted for of-
ficial EU authorization. The present study thus particularly
addresses for the first time the essential question of whether
the fundamental assumption of this pre-authorization EU pes-
ticide risk assessment, i.e., that RACs are not exceeded in the
field, is indeed met.
The new EU Regulation (EC) No. 1107/2009 (European
Commission 2009a), which replaced Directive 91/414/EEC
(European Commission 1991), together with the recently up-
dated guidance document on tiered risk assessment for plant
protection products (EFSA 2013, taking effect 01. January
2015), form the basis of the environmental regulatory risk
assessment, which is mandatory for the authorization of active
substances in the EU. Generally, the EU regulatory risk as-
sessment is based on a single active ingredient toxicity assess-
ment concept, and it follows a tiered a pproach, in which
higher tiers are less conservative but more complex and are
meant to be more realistic than lower tiers (EFSA 2013). Tier I
of the aquatic effect assessment consists of acute and chronic
laboratory toxicity tests using standard test organisms and the
application of large assessment factors (AFs, i.e., 100 for acute
and 10 for chronic toxicity tests) for RAC derivation (see
EFSA (2013) for details). In cases in which there is an unac-
ceptable risk indicated in this first tier, higher tier studies, such
as species sensitivity distributions and a quatic micro-/
mesocosm tests, are performed to derive an RAC, which is
considered more realistic and less conservative (EFSA 2013).
In particular, micro-/mesocosm studies are often conducted
for the refined risk assessment of insecticides (Table 1). The
AFs applied to these higher-tier studies are substantially lower
than tier-I AFs and are set on a case-by-case basis (see EFSA
(2013) for details). RAC comparisons with the predicted en-
vironmental concentration s (PECs) derived from exposure
modeling (see FOCUS (2001) and EFSA (2013) for details)
thus indicate either an acceptable risk for aquatic ecosystems
or the need for a specific application prescription (e.g., no-
spray buffer zones close to surface waters) that becomes part
of the registration procedure as legally binding label amend-
ments for the farmer. Overall, the EU Regulation (EC) No.
1107/2009 claims that a high level of environmental protec-
tion is required (e.g., in article 1 and 4.3). In detail, this direc-
tive states that Bno unacceptable effects on the environment^
shall result from pesticide use and particularly refers in this
context to biodiversity (European Commission 2009a). In ad-
dition to these general protection goals, Nienstedt et al. (2012)
and EFSA (2010) defined specific protection goals for main
groups of aquatic organisms (algae, aquatic plants, aquatic
invertebrates, aquatic vertebrates, aquatic microbes) covering
ecosystem services potentially affected by pesticides. In gen-
eral, to maintain ecosystem services and thus to adhere to
these specific protection goals, aquatic taxa need to be
protected at the population level (see Nienstedt et al. (2012)
and EFSA (2010) for details).
Therefore, after a pesticide is authorized and in use, field
concentrations exceeding their RACs must not occur in order
not to compromise pre-authorization risk assessment outcome
an
d to adhere to the general and specific protection goals
outlined in EU pesticide legislation (EFSA 2010, 2013;
European Commission 2009a; Nienstedt et al. 2012). Based
on a meta-analysis of field studies conducted by Beketov et al.
(2013), Stehle and Schulz (2015) argued that aquatic biodi-
versity is reduced by 29 % at agricultural stream sites with
insecticide concentrations only slightly (i.e., a factor of 1.12)
above regulatory threshold levels relative to uncontaminated
sites. It follows that insecticide concentrations exceeding their
RAC in the field in fact lead to unacceptable effects on aquatic
biodiversity. Consequently, the extent of RAC exceedances in
EU surface waters reveals two important details: (i) the actual
protectiveness and effectiveness of pre-authorization regula-
tory risk assessment schemes and thus EU pesticide legisla-
tions and (ii) the significance of insecticide exposure as a
threat to aquatic biodiversity in EU surface waters. However,
despite extensive decades-long pesticide application in
European agricultural areas, this information has never been
analyzed on a European scale. Such an analysis is urgently
needed considering that a recent study (Knäbel et al. 2012)
revealed substantial failures in the European regulatory pesti-
cide exposure assessment due to insect icide surface water
concentrations. However, this study did not provide any infor-
mation on the relationship between insecticide surface water
concentrations and the RACs.
Therefore, the present study had the following three
objectives:
i. To evaluate the overall protectiveness of the official EU
regulatory pesticide risk assessment conducted for pesti-
cide authorization using the agriculturally related insecti-
cide exposure of EU surface waters and RACs;
ii. To contextualize the different risk assessment tiers and the
associated protection levels with the insecticide risks in
the field and to validate the field relevance of the EU
regulatory risk assessment by considering, e.g., different
Environ Sci Pollut Res (2015) 22:19632–19647 19633
types of water bodies and the pesticide mixture toxicity;
and
iii. To relate the ecotoxicological significance of insecticide
surface water exposure to those of other pesticide groups,
to analyze the aquatic risks for different insecticide clas-
ses, specifically those of EU Water Framework Directive
(WFD) priority substances.
The present study thus denotes an important extension of
the work of Stehle and Schulz (2015) as it is the first to report
insecticide RAC exceedance frequencies particularly for EU
surface waters and, among others, as it subsequently
contextualizes insecticide field exposure to the pre-
authorization regulatory risk assessment schemes and EU pes-
ticide legislations.
Materials and methods
Dataset on the insecticide exposure of EU surface waters
We extracted all scientific studies (n=165, published between
1972 and 2012) reporting measured insecticide concentrations
(MICs, i.e., the conce ntrations actually detected and
Table 1 The insecticides included in the meta-analysis, their final
regulatory acceptable concentrations for water (RAC
SW
) and sediments
(RAC
SED
), their respective tiers (higher t iers denote microc osm/
mesocosm studies) of the RAC
SW
setting and their approval status under
Regulation (EC) No. 1107/2009 (DG SANCO 2014)
Insecticide Class Status under Reg.
(EC) No. 1107/2009
RAC
SW
(μg/L) EU risk assessment tier of final RAC
SW
setting (tier-I RAC
SW
a
in μg/L)
RAC
SED
(μg/kg)
Endosulfan OC Not approved 1.3
b
Higher tier (0.02) 0.026
c
Azinphos-methyl OP Not approved 0.32
d
Higher tier (0.011) –
Chlorpyrifos OP Approved 0.1
d
Higher tier (0.001) 1.1
c
Diazinon OP Not approved 2.4
d
Higher tier (0.0041) 0.95
c
Malathion OP Approved 1.25
d
Higher tier (0.0072) 0.9
c
Parathion-ethyl OP Not approved 0.024
d
Tier I 0.13
c
Parathion-methyl OP Not approved 0.073
d
Tier I 0.96
c
Carbofuran Carb Not approved 0.0205
d
Tier I –
Acrinathrin Pyr Approved 0.0087
d
Higher tier (0.00022) –
Bifenthrin Pyr Approved 0.005
d
Higher tier (0.001) –
Cyfluthrin Pyr Approved 0.0068
d
Tier I –
β-cyfluthrin Pyr Approved 0.00068
d
Tier I –
Cypermethrin Pyr Approved 0.025
d
Higher tier (0.003) 1.8
e
α-cypermethrin Pyr Approved 0.015
d
Higher tier (0.003) 1.8
e
Deltamethrin Pyr Approved 0.0032
d
Higher tier (0.0026) 1.3
c
Esfenvalerate Pyr Approved 0.01
d
Higher tier (0.001) 0.41738
f
Fenvalerate Pyr Not approved 0.0022
b
Tier I 0.88
f
λ-cyhalothrin Pyr Approved 0.0021
d
Tier I 10.5
d
Permethrin Pyr Not approved 0.025
b
Tier I 0.87
c
Acetamiprid Neo Approved 0.5
d
Tier I –
Imidacloprid Neo Approved 0.3
d
Higher tier (0.552) –
Thiacloprid Neo Approved 1.57
d
Higher tier (252) –
Thiamethoxam Neo Approved 2.8
d
Tier I –
See Stehle and Schulz (2015) for further details on RAC
SW
and RAC
SED
derivation. B–^ denotes that no sediment concentrations were reported for this
insecticide in the literature; sediment refers to sediment and suspended particle concentrations
OC organochlorine, OP organophosphate, Carb carbamate, Pyr pyrethroid, Neo neonicotinoid
a
RAC
SW
set at the tier I level of the regulatory risk assessment for insecticides, which, however, did not pass at tier 1, meaning that a higher tier RAC
SW
was used for final authorization
b
BBA (2001)
c
Crommentuijn et al. (2000)
d
EFSA (2014); DG SANCO (2014)
e
US EPA (2012)
f
RAC
SED
derived by the application of the modified EPA method according to Crommentuijn et al. (2000) and Akerblom et al. (2008)
19634 Environ Sci Pollut Res (2015) 22:19632–19647
quantified) resulting from the agricultural non-point source
pollution of surface waters for the 28 EU member states from
the global insecticide exposure dataset provided in Stehle and
Schulz (2015; see this publication for detailed information on
the entire literature review process, selection criteria and in-
formation retrieval). The dataset evaluated here thus repre-
sents an exhaustive compilation of Europe-wide insecticide
surface water concentrations. In addition to the insecticide
concentrations in the water (μg/L), sediment or suspended
particles (μg/kg), the scientific studies provided information
on the sampling location (including the distinction between
freshwater and estuarine waters and the hydrology of surface
water bodies), the catchment size, the sampling interval, the
sampling date, the limit of quantification (LOQ), and the
quantity and concentrations of additional pesticides present
in a given sample. Further on, we classified the certainty that
the MIC resulted from an agricultural non-point source entry.
In total, our analysis comprised MICs of 23 insecticide
compounds, and 15 of these 23 insecticides are currently au-
thorized for agricultural uses in the EU under the new pesti-
cide Regulation (EC) No. 1107/2009 (Table 1); the other eight
compounds that are currently not authorized had, however,
formerly been authorized for agricultural uses in the EU. We
classified these compounds for further analyses into four (or-
ganochlorines, organophosphates and carbamates (named
Borganophosphates^ ), pyrethroids, neonicotinoids) genera-
tions of insecticide classes (see Table 1) based on their eco-
toxicological mode of action (Yu 2008) and the time period of
their market introduction (Denholm et al. 2002).
Compilation of European RACs
The derivation and application of the RACs were as follows
(see Stehle and Schulz (2015) for further details): The RAC
SW
(Table 1) were used to evaluate the measured insecticide con-
centrations in the water phase (MIC
SW
). The RAC
SW
were
derived from official European pesticide registration docu-
ments (EFSA 2014; DG SANCO 2014) and denote the final
acute tier-I or higher-tier ecotoxicity endpoints, including the
AF determined within the regulatory aquatic risk assessment
of their respective insecticide compounds. As no official
European pesticide registration documents were available for
the insecticides endosulfan, fenvalerate, and permethrin, we
used the toxicity endpoints and associated AFs provided by
the German Federal Office of Consumer Protection and Food
Safety (BVL) for t heir r espective RAC
SW
(BBA 20 01).
Further details, EU risk assessment tiers of the final RAC
SW
setting, and references for RACs are specified in Table 1.
RAC
SED
(Table 1) are not determined by default for all
pesticide compounds within the official EU regulatory risk
assessment procedure (EFSA 2013; DG SANCO 2002); this
threshold level was thus only available from EU risk assess-
ment documents for the insecticide lambda-cyhalothrin. To
overcome this limitation, we applied the RAC
SED
derived
from the regulatory US EPA pesticide ecological risk assess-
ment (US EPA 2012; available for cypermethrin and
cypermethrin-alpha) or, in cases in which no official EU or
US RAC
SED
was available, maximum permissible concentra-
tions (referred to here also as RAC
SED
;Crommentuijnetal.
2000) to insecticide sediment concentrations (MIC
SED
).
An evaluation of the EU regulatory risk assessment using
MICs
We evaluated the protectiveness and field relevance of the pre-
authorization EU pesticide regulatory risk assessment and un-
derlying EU pesticide legislations and guidance documents in
the following contexts:
First, we assessed the overall protectiveness of the regu-
latory EU pesticide risk assessment and the ecological
significance of insecticide exposure by comparing all
MICs to the respective EU-level RACs for the approval
of active substances.
Second, we evaluated the insecticide exposure of the wa-
ter bodies specifically considered in the EU pesticide reg-
ulatory risk assessment, i.e., the small edge-of-field fresh-
water bodies in close proximity to agricul tural fields
(European Commission 2009a ;EFSA2013;FOCUS
2001). We therefore restricted the evaluation of our
dataset to MICs reported for water bodies with catchment
sizes of up to 1 km
2
(i.e., the water body size used in the
regulatory FOCUS exposure assessment (FOCUS 2001;
EFSA 2013)) and to MICs reported for water bodies with
catchment sizes of up to 10 km
2
in order to be less re-
strictive about the specific focus of the EU regulatory risk
assessment and to include surface waters that are still
typical for agricultural landscapes (Davies et al. 2008)
but are not particularly addressed under the EU WFD.
We further distinguished between different types of sur-
face waters, i.e., we evaluated the MICs separately for
freshwater and estuarine water bodies. As the regulatory
pesticide risk assessment and the resulting RACs are val-
id only for MICs caused by agricultural non-point source
pollution (European Commission 2009a;EFSA2013),
we further restricted our dataset to the MICs definitively
attributable to this source using information, e.g., on land
use, insecticide application schemes, and the routes of
entry provided in the scientific studies (see Stehle and
Schulz (2015) for detailed classification criteria). This
strict classification procedure enabled us to attribute a
specific insecticide concentration to agricultural non-
point source pollution with high confidence and to sub-
sequently analyze those exposure incidences separately.
It is important to note that all these restricted analyses led
to even worse outcomes, i.e., even higher RAC
Environ Sci Pollut Res (2015) 22:19632–19647 19635
exceedance rates; therefore, the evaluation of the EU pes-
ticide risk assessment using the entire dataset indicates
less risk than is actually present.
Third, from the official EU pesticide registration docu-
ments, we determined whether the final RAC
SW
used for
the authorization of a given active substance was derived
from the first tier of the regulatory risk assessment, or, in
cases i n which tier I was not passed, by conducting
higher-tier effect assessment studies. In the latter case,
we also extracted the associated tier-I RAC
SW
from the
respective registration documents (Table 1). Given this
information, we evaluated the MIC
SW
separately for (i)
the compounds finally regulated by tier-I risk assessment
and (ii) the compounds regulated using higher-tier RACs.
In addition, we applied in an additional assessment re-
spective tier-I RAC
SW
to all MIC
SW
, i.e., also to the in-
secticides that were in fact authorized using higher-tier
risk assessments.
Finally, we separately evaluated the RAC exceedance
frequencies for the different insecticide substance classes
(i.e., organochlorines, organophosphates, pyrethroids,
neonicotinoids) and for the d ifferent p esticide group s
(i.e., herbicides, fungicides, insecticides). Regarding the
latter, we evaluated the differences in pesticide water-
phase concentration levels, tier-I RAC
SW
values, as de-
termined by the official EU pesticide risk assessment of a
given pesticide compound, and the respective concentra-
tion to tier-I RAC
SW
ratios for fungicides, herbicides, and
insecticides using all the samples analyzed for multiple
pesticide exposure. In detail, we extracted from samples
containing pesticide mixtures in addition to insecticide
concentrations, the concentrations of all further pesticide
compounds detected.
It is worth mentioning that the pesticide registration in the
EU is based on a two-stage registration system, with an initial
assessment of active substances at the EU level (which is
considered in this study) and the subsequent registration of
plant protection p roducts c ontaining approved active sub-
stances by member states. However, member states can only
authorize the use of plant protection products after an active
substance has passed the EU regulatory risk assessment and
has been added to the list of approved active substances eligi-
ble for agricultural uses in the EU.
Water Framework Directive: an assessment of priority
substances
The EU WFD 2000/60/EEC (European Commission 2000)
uses a retrospective risk assessment approach by comparing
chemical monitoring data with environmental quality stan-
dards (EQSs) for EU-wide priority substances. We assessed
the M IC
SW
of the three compounds listed as priority
substances by the WFD (i.e., endosulfan, chlorpyrifos, and
cypermethrin (including isomers)) (European Comm ission
2013), as detected in the water bodies considered in this di-
rective (i.e., catchment sizes > 10 km
2
), by using their respec-
tive maximum acceptable concentration EQS values (MAC-
EQSs) for inland s urface waters. The MAC-EQSs, which
should not be exceeded by a single concentration in the aquat-
ic ecosystem of concern, are as follows: 0.01 μg/L for endo-
sulfan, 0.1 μg/L for chlorpyrifos, and 0.0006 μg/L for
cypermethrin (including isomers) (European Commission
2013).
An evaluation of pesticide mixture toxicity
To evaluate the ecotoxicological significance of mixture tox-
icity for EU surface waters, we compared all water-phase pes-
ticide concentrations quantified in a given sample containing
multiple pesticides (n=516 out of the total of 1140 samples
analyzed) to the respective tie r-I threshold levels (i.e.,
ecotoxicity values including AFs) for the three taxonomic
groups (i.e., fishes, invertebrates, primary producers) consid-
ered in the EU regulatory risk assessment. We calculated tier-I
threshold levels by dividing the lowest acute LC
50
or EC
50
values (compiled from PPDB (2013) and official EU pesticide
registration documents (EFSA 2014)) for fish, Daphnia,green
alga, an additional arthropod species (for substances with an
insecticidal mode of action), and macrophytes (for substances
with a herbicidal mode of action) by their respective AFs (i.e.,
100 in the case of fish, Daphnia, and arthropods and 10 in the
case of primary producers; EFSA 2013).
The mixture toxicity was calculated separately for each
respective taxonomic group by summing up the concentration
to tier-I threshold level ratios for all the pesticides detected in a
surface water sample to obtain the risk quotient of the mixture
(RQ
mix
) for a given taxonomic group:
RQmix ¼
X
n
i¼1
MPCi
TLi
where MPCi is the measured pesticide concentration of the
compound i quantified in a given sample; TLi is the acute
tier-I threshold level for a given taxonomic group of the pes-
ticide i;andwithRQ
mix
<1 indicating an acceptable risk for a
specific taxonomic group.
We used this approach as its modified version (which uses
modeled exposure data instead of MPC), and the underlying
principle of concentration addition (Kortenkamp et al. 2009)
is proposed by the EU Commission for the regulatory risk
assessment of pesticide mixture toxicities for individual taxo-
nomic groups (EFSA 2013). Moreover, this approach is gen-
erally considered as broadly applicable for pesticide mixture
toxicity evaluations (Deneer 2000; Cedergreen et al. 2008).
19636 Environ Sci Pollut Res (2015) 22:19632–19647
Linear model analysis
We conducted a hierarchical linear model analysis to quantify
the influence of different drivers on the outcome variable log-
arithmic MIC
SW
to RAC
SW
ratio. The following independent
variables were entered in the analysis using a complete-case
approach (Pigott 2009): (i) the log sampling interval, (ii) the
log catchment size, (iii) the sampling date, and the dummy-
coded categorical variables for (iv) EU risk assessment tiers of
the RAC
SW
setting (tier I (B0^) vs. higher tier (B1^)), (v) status
under Regulation (EC) No. 1107/2009 (approved (B0^) vs. not
approved (B1^)), and (vi) insecticide substance classes (organ-
ochlorines (B0^), organophosphates and carbamates (B1^), py-
rethroids (B2^)). We excluded the neonicotinoid subst ance
class as only a total of 33 EU surface water concentrations
were documented in the peer-reviewed literature, and this
number was further reduced to only six MIC
SW
available for
complete-case linear model analysis.
Automatic and manual model building were used to iden-
tify independent variables and potential interactions with the
highest explanatory power for the response variable logarith-
mic MIC
SW
to RAC
SW
ratio and best-fit models (see Stehle
and Schulz (2015) for further details). Model checking includ-
ed heteroskedasticity, the normal distribution of residuals and
the influence of single observations using residual-leverage
plots and Cook’s distance. All computations were done with
the open source software R (version 2.15.2 for Mac OS X
10.6.8).
Results
The insecticide exposure of EU surface waters:
an evaluation of the regulatory risk assessment
Overall, 44.7 % (n=700 cases) of the 1566 MICs reported for
EU surface waters exceeded their respective RACs. In partic-
ular, 37.1 % of the 1352 MIC
SW
exceeded their RAC
SW
up to
a factor of 125,750, and 93 % of the 214 MIC
SED
exceeded
their RAC
SED
up to a factor of 31,154 (Fig. 1). Information on
the MICs for the 23 insecticide compounds was available for
385 sites located in 16 of the 28 EU member states (Fig. S1 in
Supplementary Material), with most MICs originating from
southern EU countries: Greece (n=487), Spain (n=415), and
Italy (n=152). Additional summary statisti cs for the EU
dataset are displayed in Table 2.
The temporal analyses of the insecticide exposure data
(Fig. S2; Fig. S3; Table S1) indicates that risks did not de-
crease over time; this is in accordance with the results of the
linear model analysis (Table 3), which predicted significant
increases in MIC
SW
to RAC
SW
ratios over time when consid-
ering the influences of covariates. In total, 546 (38 %) of all
MICs were detected after the year 2000, with 40.5 % of these
exposure incidences exceeded the respective RAC (Table S1).
Approximately 90 % of all MICs (n= 126) exceeded their
RACs in small edge-of-field surface waters with catchment
sizes of up to 1 km
2
,aswellas>75%ofallMICs(n=273)
in the case of water bodies with catchment sizes of up to
10 km
2
(Table S2). The linear model analysis, which predicted
significantly higher MIC
SW
to RAC
SW
ratios for smaller sur-
face waters (Table 3), supports these results. In addition, RAC
exceedance frequencies for freshwater systems (45.4 %, n=
1430) were higher compared with those derived for estuarine
surface waters (37.5 %, n=136) (Table S2). The restriction to
exposure incidences (n=581), which could be linked with
high confidence to agricultural non-point source entries, re-
sulted in RAC exceedance frequencies of 60.9 % (Table S3).
The risk assessment of the three WFD priority substances
(i.e., chlorpyrifos, endosulfan, cypermethrin (including iso-
mers)) included in our meta-analysis showed that 57.5 % of
their MIC
SW
(n=146) exceeded their respective MAC-EQS
values. All cypermethrin concentrations (n=29), as well as
73.3 % of the endosulfan (n=60) and 19.3 % of the chlorpyr-
ifos concentrations (n=57), exceeded their respective MAC-
EQS values.
The risk assessment tiers of RAC
SW
determination
and aquatic risks in the field
Ten of the 23 insecticide compounds considered here gained
authorization for agricultural uses in the EU by passing tier I
Fig. 1 Distribution curves for MICs relative to their respective RACs.
Blue represents the MIC
SW
relative to substance-specific RAC
SW
(n=
1352) and brown represent s MIC
SED
relative to substance-specific
RAC
SED
(n=214). The inlet shows the overall variation of the MIC to
RAC ratios for water and sediment concentrations
Environ Sci Pollut Res (2015) 22:19632–19647 19637
of the regulatory environmental risk assessment for aquatic
organisms, whereas 13 compounds were approved using
higher risk assessment tiers (i.e., RAC
SW
derivation using
microcosms or mesocosms) (Table 1). The tier-I RAC
SW
levels of the 10 insecticides (median 0.02225 μg/L) are no-
ticeably lower than the RAC
SW
levels of the 13 compounds
(median 0.1 μg/L) derived through higher-tier risk assess-
ment (Fig. 2). However, the median toxicity towards tier-I
standard test organisms is approximately one order of mag-
nitude higher (i.e., lower RAC
SW
values) for the latter 13
compounds. Furthermore, the higher-tier RAC
SW
of these
13 compounds are approximately 1.5 orders of magnitude
higher than their associated tier-I RAC
SW
levels (median:
0.003 μg/L; Fig. 2).
The MIC
SW
of the 10 compounds that were approved using
tier-I RAC
SW
led to significantly (Table 3) higher RAC
SW
exceedances (64.9 %; n=576) compared with those of the
13 insecticides that were approved using higher-tier RAC
SW
(16.4 %; n=776; Table 4). However, if we only consider the
tier-I RAC
SW
for all 23 insecticide compounds in the assess-
ment of MIC
SW
,71.4%(n=1352) of the MIC
SW
exceeded the
RAC
SW
(Table 4).
The risk assessment for different insecticide substance
classes and pesticide groups
The MIC
SW
of pyrethroids (n=108) led to the highest percent-
age of RAC
SW
exceedances (70.4 %; see also Table 3 for a
comparison of insecticide classes in the linear model analysis),
followed by the MIC
SW
of organophosphorus insecticides
(37.5 %; n=1100) and neonicotinoids (24.2 %; n=33); in
contrast, only 3.6 % of the MIC
SW
(n=111) reported for the
organochlorine insecticide endosulfan exceeded the RAC
SW
(Fig. S4). Insecticide sediment exposure led to >90 %
RAC
SED
exceedance frequencies for all substance classes (or-
ganochlorine insecticides (n=32) 100 %; organophosphorus
insecticides (n=124): 90.3 %; pyrethroids (n=58) 94.8 %
RAC
SED
exceedance frequencies), except for neonicotinoids,
for which no MIC
SED
was reported in the scientific literature.
We detected higher absolute field concentrations for fungi-
cides (median 0.96 μg/L) compared with those of herbicides
(median 0.063 μg/L) and insecticides (median 0.034 μg/L)
(Fig. 3a) in the samples containing multiple pesticides (n=
516). However, the risk assessment for these pesticide groups
showed higher tier-I RAC
SW
exceedance frequencies for
Table 2 Summary statistics (the number of measured insecticide concentrations (MICs)) for important parameters of the EU insecticide exposure
dataset
Parameter
a
Minimum 25th percentile Median 75th percentile Maximum
Sampling date (n=1447) 1969 1989 1996 2004 2010
Catchment size (km
2
, n= 1320) 0.02 15 800 3315 180,000
Sampling interval
b
(days, n=1192/1054) 0.0416/0.0416 14/12 30/30 60/60 180/180
RAC exceedances per country (%, n= 1566) Belgium (n=26) 3.9
Bulgaria (n=1) 100
Cyprus (n=3) 66.7
Denmark (n=7) 100
France (n=46) 76.1
Germany (n=138) 83.3
Greece (n= 487) 35.2
Hungary (n=3) 0
Italy (n=152) 54.6
Netherlands (n=60) 33.3
Poland (n=33) 27.3
Portugal (n=94) 21.3
Romania (n=5) 0
Spain (n=415) 33.5
Sweden (n=17) 94.1
UK (n=79) 78.5
Hydrology (n=1419) Lotic surface waters: 1211 (85.3 %); lentic surface waters: 208 (14.7 %)
Type of surface water (n=1566) Freshwater systems: 1430 (91.3 %); estuarine waters: 136 (8.7 %)
Source (n=1566) Non-point source
c
: 1222 (78 %); rainfall-induced runoff: 159 (10.2 %); rice field effluents:
81 (5.2 %); spray drift: 41 (2.6 %); aerial application: 27 (1.7 %); irrigation-induced
runoff: 18 (1.1 %); drainage: 18 (1.1 %)
Insecticide classes (n=1566) Organochlorine insecticide: 143; organophosphorus insecticides: 1224; pyrethroids: 143;
neonicotinoids: 33
a
There are fewer MICs for some parameters due to missing information in studies
b
The first value is for all (water and sediment) MICs, and the second value is for MIC
SW
only
c
The non-point source denotes that the exact diffuse pollution source was not specified
19638 Environ Sci Pollut Res (2015) 22:19632–19647
insecticides (53.1 %) compared with those of fungicides
(31 %) and herbicides (3.8 %); in addition, the insecticide
median concentration to tier-I RAC
SW
ratio (1.25) is approx-
imately one and two orders of magnitude higher compared
with those of fu ngicides (0 .13) and herbicides (0.019)
(Fig. 3c).
Risk assessment for pesticide mixtures in EU surface
waters
Overall, 135 different pesticides (66 insecticides; 42 herbi-
cides; 27 fungicides) were detected in the 608 samples ana-
lyzed in total (i.e., water and sediment samples) for pesticide
mixture occurrence in EU surface waters. Mixtures of pesti-
cides occurred in 90 % (n=462 cases) of the insecticide water-
phase samples with information on additional pesticides (n=
516 out of the total of 1140 samples analyzed); these samples
contained up to 13 pesticide compounds (Table S4). The re-
sults for sediment samples were comparable, i.e., 87 % of all
samples with information on additional compounds (n=92)
contained up to 11 pesticides.
The RQ
mix
of the water-phase samples containing multiple
pesticides ( n=462) indicated the highest risks for in verte-
brates, as 82.7 % of these samples showed RQ
mix
exceedances
for this taxonomic group of up to a factor of 1,840,805
(Fig. 4). In relation to fish, 39.6 % of the samples had a
RQ
mix
>1 up to a factor of 18,377, whereas only 8.2 % of
the water-phase samples led to a RQ
mix
>1 for algae/macro-
phytes, with 829 as the highest RQ
mix
.
Discussion
The insecticide exposure of EU surface waters:
the protectiveness of EU pesticide legislation
Our meta-analysis shows that approximately 45 % of all MICs
at >215 sites (i.e., >55 % of all (n=385) sites with MIC data)
across the EU excee ded their respective RACs (Fig. 1;
Fig. S1). It follows that insecticides are an important threat
to European freshwater biodiversity, as insecticide levels >
RACs lead to severe biodiversity reductions (Stehle and
Schulz 2015). This conclusion is in line with smaller-scale
field studies reporting pesticide-induced adverse effects on
ecosystem function and aquatic biodiversity in small agricul-
tural surface waters (e.g., Schäfer et al. 2012; Berenzen et al.
2005; Bereswill et al. 2013) and a study conducted on organic
pollutants in larger EU surface waters (Malaj et al. 2014).
However, for the first time, the present meta-analysis uses
empirical evidence based on scientific data and official
RACs to illustrate the extent of the risk for European aquatic
ecosystems. In terms of regulatory implications, the risk as-
sessment findings presented here question the fulfillment of
Fig. 2 A comparison of the RAC
SW
levels derived from the different
tiers of the official EU pesticide risk assessment (n (insecticides) tier-I risk
assessment: 10, median RAC
SW
=0.02225 μg/L; n (insecticides) higher-
tier risk assessment: 13, median RAC
SW
=0.1 μg/L). The tier-I RAC
SW
associated with higher-tier RAC
SW
(n (insecticides): 13, median
RAC
SW
=0.003 μg/L) denote RAC
SW
derived from the first tier risk
assessment for insecticides, which were finally approved using higher-
tier studies (microcosms/mesocosms, see Table 1)
Table 3 The results of linear model analyses predicting logarithmic
MIC
SW
to RAC
SW
ratios (R
2
=0.612; adjusted R
2
=0.609; p<0.001; n=
942)
Estimate t value p value
Intercept −22.270 −2.992 0.00285
Catchment size −0.262 −9.076 <0.001
Sampling interval −0.274 −6.566 <0.001
Sampling date 0.012 3.160 0.00163
SC (OP) 0.108 0.474 0.6355
SC (Pyr) 1.349 5.267 <0.001
RA tier (higher tier) −2.017 −28.142 <0.001
The substance class (SC) (reference category: organochlorine insecti-
cides) and risk assessment tier of RAC
SW
derivation (RA tier) (reference
category: tier I) were entered as dummy-coded variables, and the catch-
ment size and sampling interval were entered as log-transformed vari-
ables. The same main effects analysis was also performed using the
organophosphates/carbamates insecticide substance class as the reference
category for calculating the significance level of pyrethroids vs.
organophosphates/carbamates (B=1.241; t value=9.646; p<0.001). The
insecticide substance class neonicotinoid was excluded due to the small
number of cases (n=6) available for statistical analysis. The categorical
variable authorization status under Regulation (EC) No. 1107/2009 did
not show significant explanatory power for the outcome variable
OP organophosphates/carbamates, Pyr pyrethroids
Environ Sci Pollut Res (2015) 22:19632–19647 19639
the general protection goals outlined in Regulation (EC) No.
1107/2009 and of the specific protection goals defined by
Nienstedt et al. (2012) and the EFSA PPR Panel (EFSA
2010) based on the ecosystem service concept for the regula-
tory risk assessment of pesticides in the EU. Regarding the
latter, Nienstedt et al. (2012) argued that the protection of
ecosystem services for the fulfillment of the specific protec-
tion goals requires the protection of biodiversity in agricultural
landscapes; our data, however, indicate clear biodiversity im-
pairments (see also Stehle and Schulz 2015) in agricultural
surface waters due to insecticide exposure. Importantly, not
only the endpoints of the regulatory effect assessment (i.e.,
RAC) are exceeded in the field but also those of the regulatory
exposure assessment (i.e., PEC; Knäbel et al. 2012;Knäbel
et al. 2014); it must therefore be concluded that the current
pre-authorization regulatory risk assessment schemes includ-
ing associated risk mitigation obligations (i.e., pesticide appli-
cation prescriptions) and underlying EU pesticide regulations,
do not protect the aquatic environment. In addition, the insec-
ticide field exposure data presented here do not provide a final
conclusion on the reasons for RAC exceedances in the field,
i.e., the failure of the prospective regulatory exposure and risk
assessment or of farmers’ adherence to regulatory risk mitiga-
tion obligations such as no-spray buffers; however, Knäbel
et al. (2012) suggest both factors’ contributions to insecticide
risks for EU surface waters.
In addition to its overall protectiveness, our data also chal-
lenge the field relevance and focus of the EU pesticide regu-
latory risk assessment. Interestingly, we found the highest
RAC exceedances for the MICs detected in small edge-of-
field wat er bodie s (Table S2)andforthosedefinitively
resulting from agricultural non-point source entries
(Table S3). Although this finding can be explained (Schulz
2004;Stehleetal.2013), one would nevertheless expect lower
risks in surface waters and for exposure sources that are the
specific focus of the aquatic regulatory risk assessment. On
the contrary, surface waters not specifically targeted by regu-
latory risk assessment schemes, such as estuarine ecosystems
(EFSA 2013), are also heavily affected by insecticide pollu-
tion; 37.5 % of the MICs exceeded their RACs, even though
estuaries often are not located in close proximity to agricul-
tural areas and non-contaminated seawater dilutes insecticide
exposure (Steen et al. 1999).
There are additional issues that alert us to severe problems.
First, approximately 90 % of the MIC
SW
assessed here were
measured using single or fixed-interval sampling strategies,
which considerably underestimate actual insecticide exposure
levels (Stehle et al. 2013; see also the result of the linear model
Table 4 An evaluation of MIC
SW
as a function of the regulatory risk assessment tiers of the RAC
SW
setting
No. (%) of MIC
SW
below RAC
SW
No. (%) of MIC
SW
above RAC
SW
An evaluation of MIC
SW
based on the final RAC
SW
used for the authorization of compounds
Insecticides with a tier-I RAC
SW
(n=10; 576 MIC
SW
) 202 (35.1) 374 (64.9)
Insecticides with a higher-tier RAC
SW
(n=13; 776 MIC
SW
) 649 (83.6) 127 (16.4)
An evaluation of MIC
SW
based on tier-I RAC
SW
for the 23 insecticide compounds
Insecticides authorized by a higher-tier RAC
SW
(n=13; 776 MIC
SW
) 184 (23.7) 592 (76.3)
All MIC
SW
(n=23; 1352 MIC
SW
) 386 (28.6) 966 (71.4)
Fig. 3 Boxplots of the water-phase concentrations detected in EU sur-
face waters (a), the regulatory acceptable concentrations (RAC
SW
)de-
rived from tier I of the European pesticide risk assessment (b)andrelated
field concentration to tier-I RAC
SW
ratios (c, dashed line indicates the
RAC
SW
) for the different pesticide groups. The comparison is based on
fungicide (n=87; 23 compounds), herbicide (n=852; 36 compounds),
and insecticide (n=1408; 59 compounds) water-phase concentrations de-
tected in the 516 samples analyzed for the occurrence of multiple pesti-
cide exposure
19640 Environ Sci Pollut Res (2015) 22:19632–19647
analysis (Table 3), which indicates higher RAC
SW
exceedances for shorter sampling intervals). Second, no sci-
entific knowledge on insecticide surface water exposure exists
for large parts (i.e., approximately 80 %) of European high-
intensity agricultural areas (Fig. S1), which indicates that fu-
ture monitoring studies are needed to further quantify risks
across the EU; this research is of even more importance be-
cause climate change is expected to lead to increasing insec-
ticide application in EU agriculture (Kattwinkel et al. 2011).
Third, our meta-analysis shows that pesticides occur as mix-
tures in 90 % of the samples analyzed for multiple compounds
(Table S4), with nearly 40 % of these samples containing more
than 5 and up to 13 pesticides per sample. Importantly, most of
the studies analyzed surface water samples for selected pesti-
cide compounds only; thereby most likely they potentially
missed compounds that were additionally present (see also
Moschet et al. (2014) on this topic). However, these findings
on pesticide mixture occurrences in the field challenge the
protectiveness of the RAC, which is defined for single active
ingredients only (EFSA 2013) and thus not covering potential
combined or even synergistic effects (e.g., Denton et al. 2003;
Belden and Lydy 2006). Fourth, only a marginal difference in
RAC excee dances exists between the 15 ins ecticide com-
pounds currently authorized in the EU and the eight com-
pounds that are no longer approved (Table S5). This finding
is supported by our linear model analysis, which could not
detect a significant explanatory power for the differentiation
of authorized and non-authorized compounds (Table 3). We
therefore conclude that the cancelation of the authorization of
obsolete active ingredients under Directive 91/414/EEC and
Regulation (EC) No. 1107/2009 did not reduce insecticides’
acute risks for surface waters; this claim, again, challenges the
overall effectiveness of EU pesticide legislations. Within this
context, we identified even higher MIC
SW
to RAC
SW
ratios
after the enforcement of the Directive 91/414/EEC in 1993
(Fig. S2; Fig. S3) and, as opposed to the global MIC data
presented by Stehle and Schulz (2015), for more recent sam-
pling dates independent of the influence of covariates, such as
the increased detection of more toxic pyrethroids in recent
years (Table 3). Moreover, 40.5 % of all MICs detected since
the year 2000 exceeded respective RACs (T able S1), which
challenges the general perception of decreasing environmental
risks (see, e.g., Lamberth et al. 2013; Devine and Furlong
2007) due to the market introduction of newer insecticide
compounds and the enforcement of more stringent environ-
mental regulations. However, other reasons not concerning
aquatic organisms (e.g., high mammalian and avian toxicities
of organophosphates) presumably led to the withdrawal of
hazardous pesticide compounds under Directive 91/414/
EEC, so that the overall environmental risks might nonethe-
less be reduced over time (Cross and Edward-Jones 2011).
Overall, our data and those of Knäbel et al. (2012, 2014)
indicate that a critical reconsideration of the entire EU pesti-
cide regulatory risk assessment approach including enforce-
ment of mandatory risk mitigation obligations is imperatively
needed; these findings must be seriously considered in future
revisions of EU pesticide regulations. In addition, effective
risk mitigation mea sures (e.g., Reichenberger et al. 2007;
Stehle et al. 2011) have to be implemented and enforced, inter
alia within National A ction Plans, as requested by EU
Directive 2009/128/EC (Sustainable Use Directive for Plant
Protection Products (European Commission 2009b)). EU ag-
ricultural policies and subsidies should also be critically
reconsidered, as they currently foster agricultural intensifica-
tion and agrochemical use (Pe`er et al. 2014).
In relation to the WFD, the scientific exposure data pre-
sented here confirm recent findings based on governmental
data (Malaj et al. 2014), which showed that insecticide pollu-
tion is a significant stressor in large EU surface waters. This
confirmation, however, is a crucial finding, as the characteri-
zation of the chemical status of a large proportion of water
bodies is still deficient due to lacking (European Environment
Agency 2012) and often inappropriate (Stehle et al. 2013)
governmental monitoring. Furthermore, our meta-analysis
identified substantially higher RAC
SW
exceedance frequen-
cies (32.6 %; n=763) in large EU surface waters for the 20
non-priority substances included in our meta-analysis com-
pared with those of the three priority substances (15.8 %; n=
146). This finding challenges the WFD priority substance se-
lection criteria (see also Von der Ohe et al. (2011)andSchäfer
et al. (2011)) that currently disregard the high ecotoxicity po-
tential of modern insecticides. Real-world exposure data and
Fig. 4 Pesticide mixture toxicities detected in the water phase of EU
surface water samples (n=462), expressed as risk quotients (RQ
mix
)for
algae/macro phytes (green), fishes (blue), and invertebrates (red). A
RQ
mix
>1 indicates a risk for the respective taxonomic group
Environ Sci Pollut Res (2015) 22:19632–19647 19641
actual ecological risks in the field should trigger the future
identification and prioritization of WFD priority substances.
The protectiveness of the regulatory risk assessment tiers
The MIC
SW
of the compounds authorized using a higher-tier
risk assessment show considerably lower RAC
SW
exceedances (Table 4). This finding is in line with the general
principles underlying the pre-authorization regulatory risk as-
sessment, i.e., the outcomes of higher risk assessment tiers are
less conservative compared with those of lower tiers (EFSA
2013), which consequentially leads to less frequent
exceedances of these higher-tier RACs in the field. Most im-
portantly, tier-I RACs are derived based on the ecological
threshold option (ETO), which accepts only negligible effects,
whereas the derivation of higher-tier RACs based on micro-/
mesocosm studies generally accepts (temporary) clear popu-
lation level effects (i.e., RACs derived based on the ecological
recovery option, ERO-RACs; see EFSA (2013) for details).
However, it is thought-provoking that such liberal higher-tier
RACs drive the final regulatory risk assessment specifically of
extremely toxic insecticide compounds. These insecticides
have a substantially higher intrinsic ecotoxicity potential to-
wards aquatic (standard test) organisms compared with those
of the compounds auth orized using tier-I RAC
SW
(Fig. 2;
Table S6). It follows that the most toxic insecticides are au-
thorized using least conservative RACs, i.e., those based on
ERO. Considering this high toxicity potential and that these
liberal higher-tier RACs are set with hardly any margin of
safety, they should never be exceeded in the field to prevent
unacceptable adverse effects. Our data (Table 4), however,
clearly disprove this assumption.
There are two more critical issues that have to be consid-
ered in this context. First, higher-tier RAC
SW
are considerably
less conservative compared with tier-I RAC
SW
levels (Fig. 2)
due to the substantial reduction of AFs (up to two orders of
magnitude); however, this reduction in conservatisms is not
justified by actually lower ecotoxicity potentials (Table S6).
Although this AF reduction is often reasoned by the higher
complexities and ecological reali sm of the higher-tier
microcosm/mesocosms stud ies (EFSA 2013), the inherent
limitations of these artificial model ecosystem studies (see,
for example, Crane and Giddings (2004) and references there-
in) jeopardize the protectiveness of higher-tier RAC
SW
for
real-world situations in the field, especially in cases in which
an AF of one was employed (Table S6); these limitations and
the resulting uncertainties are therefore not covered by the
regulatory pesticide risk assessment. In addition, the occur-
rence of pesticide mixtures, consecutive exposure events,
and confounding factors (e.g., hydraulic stress, exposure to
nutrients) in the field further challenge the protectiveness of
higher-tier RAC
SW
set with low AFs. However, according to
EFSA ( 2013), an AF of one is not used anym ore si nce
commencement of this guideline. Nevertheless, higher-tier
RAC
SW
should, in consistency with tier-I RACs, generally
be derived using the ETO and thus without already allowing
for clear population level effects (sensu ERO-RACs).
Second, recent field studies (Schäfer et al. 2012;Beketov
et al. 2013;Petersetal.2013) reported pesticide-induced ad-
verse effects at concentrations even well below (i.e., 1/10 to
1/100) conservative tier-I RAC
SW
. In addition, based on sta-
tistical analyses, Luttik et al. (2011)arguedthattheAFsof100
used for tier-I RAC
SW
derivation may not adequately cover
interspecies sensitivity variation. These findings provide evi-
dence that even the conservative RAC
SW
are potentially not
protective in the field. An even worse protection level may
thus be expected for the even less conservative higher-tier
ERO-RAC
SW
, although they have been established under con-
ditions that are considered more realistic.
Overall, we conclude that in addition to cases with RAC
exceedances, the occurrence of unacceptable adverse effects
in the field can potentially not be excluded for the 35 % of
MIC
SW
that comply with conservative tier-I ERO-RAC
SW
and
are even more likely for the 83.6 % of MIC
SW
that comply
with higher-tier ERO-RAC
SW
(Table 4). Our findings on the
lack of the protectiveness of higher-tier RACs for insecticide
compounds are in line with a recent study on aquatic ecosys-
tems and fungicides (Zubrod et al. 2015), which also claimed
that the higher-tier regulatory EU risk assessment does not
provide an adequate level of protection. EFSA (2013) ac-
knowledges these regulatory risk assessment shortcomings
by admitting that the RAC
SW
may not be protective for all
cases occurring in the field; the effects not covered by the
prospective risk assessment, the combined effects between
pesticides and en vironmental stressors, the e xposure to
multiple pesticides, and the repeated exposure due to serial
pesticide application are potential reasons for these
uncertainties. As a consequence, EFSA (2013)postulatesfur-
ther strengthening the link between the RACs and real-world
field situations, e.g., by conducting appropriate field studies
that clearly link pesticide exposure to related effects. We be-
lieve our study addresses one aspect of this issue. However,
further targeted studies are urgently needed.
Risk assessment for pesticide groups and insecticide
classes: a proposal for a new hazard-based cut-off
criterion
The comparison of pesticide risks shows that insecticides par-
ticularly threaten EU surface waters (Fig. 3c). This finding is
explained by the substantially higher ecotoxicity potential of
insecticides. The median insecticide tier-I RAC
SW
(0.029 μg/L) is more than two orders of magnitude lower
compared with those of herbicides (4 μg/L) and fungicides
(11.3 μg/L; Fig. 3b). This high ecotoxicity potential of insec-
ticides overcompensates the absolutely higher field
19642 Environ Sci Pollut Res (2015) 22:19632–19647
concentrations of fungicides and herbicides (Fig. 3a), which
result from higher application rates and physicochemical
properties (e.g., large DT
50
values, high water solubilities),
which foster surface water e xposure (Stehle et al. 2011,
2013). Our findings support those of Stehle e t al. (2011),
who also reported lower concentrations and higher ecotoxico-
logical risks for insecticides compared with those of herbi-
cides and fungicides at the inlet and outlet of vegetated treat-
ment systems. The high ecotoxicity of insecticides, particular-
ly for aquatic invertebrates (Devine and Furlong 2007), to-
gether with the overall high sensitivity of this group of organ-
isms to pesticide exposure (US EPA 2014), is also a major
reason that aquatic invertebrates are at risk to the largest extent
when exposed to multiple pesticides (Fig. 4).
Overall, our results provide strong evidence that regulatory
risk assessment and risk management for insecticides particu-
larly needs reconsideration; targeted and more protective risk
assessment concepts, stricter decision criteria, and mandatory
risk mitigation obligations should be defined specifically for
the authorization procedures of insecticides. However, it is
important to note that field data-based meta-analyses are also
needed for herbicides and fungicides to thoroughly evaluate
the protectiveness and field relevance of the EU regulatory
risk assessment for these pesticide groups. For example, the
standard test organisms currently used in the aquatic effect
assessment of pesticides are potentially unsuitable for ade-
quately assessing fungicide effects in the field (Zubrod et al.
2015;Maltbyetal.2009).
Excluding neonicotinoids, for which a valid conclusion is
hindered due to insufficient data, the development and autho-
rization of newer insecticide classes led to an increase in acute
environmental risks for surface waters (Fig. S4; Table 3), with
the pyrethroids outpacing the other insecticide classes due to
their extremely high toxicities for non-target organisms
(Spurlock and Lee 2008) and their fast mode of action
(Schulz and Liess 2000; Forbes and Cold 2005; Solomon
et al. 2001). Pyrethroids’ acute toxicities for fishes and inver-
tebrates are several orders higher than those of other pesticides
(Table S7), which substantially increases the ecotoxicological
risks for aquatic ecosystems. Balderacchi and Trevisan (2010)
showed that authorized pesticides are generally less toxic, less
hydrophilic, and more rapidly degraded than non-authorized
pesticide compounds; this finding, however, does not account
for pyrethroids (Table S7). We therefore propose considering a
new hazard-based cut-off criterion, very Toxic, fast Mode of
Action (vTfMoA), in the regulatory risk assessment of pesti-
cides. This criterion could complement the hazard-based cut-
off criteria int roduce d by the n ew Reg ulation ( EC) No.
1107/2009, which aim to enhance human and environmental
health protection (Table S8). However, the exact classification
schemes for the vTfMoA criterion still have to be defined,
e.g., by using acute toxicity thresholds and time-to-event anal-
yses (Newman and McCloskey 1996). The implementation of
the vTfMoA criterion could substantially reduce the environ-
mental risks caused by extremely toxic and rapidly acting
pesticides, such as pyrethroids, which, despite their high acute
risk potentials and related RAC exceedances, are still predom-
inantly authorized in the EU (Table 1). However, another fact
to consider here is that the introduction of additional hazard-
based cut-off criteria potentially decreases the anticipated
number of active ingredients to be (re-)authorized and there-
fore available for crop protection in Europe (ECPA 2006); this
fact should not be ignored considering the increasing resis-
tance of target pests (Denholm et al. 2002).
Until now, only very limited field data are available for
neonicotinoids (Fig. S4), which, in addition to their entirely
different mode of action, selectivity, plant systemicity, persis-
tence, and resulting delayed effects (Jeschke and Nauen 20
08;
Tennekes and Sanchez-Bayo 2011;Sanchez-Bayo2014), hin-
ders a thorough assessment of their acute risks for EU surface
waters. However, numerous recent studies reporting the sub-
stantial ecological effects of neonicotinoids in aquatic and
terrestrial ecosystems (e.g., van Dijk et al. 2013;Hallmann
et al. 2014; Chagnon et al. 2015; Goulson 2013) strongly
indicate that further research is needed on the ecological con-
sequences of neonicotinoid use. In this context, it is important
to note that insecticide use patterns in the EU have changed
substantially over the past few decades, with the discontinua-
tion of many organochlorine and organophosphate insecti-
cides and recent increases in pyrethroid and neonicotinoid
use (together, with an insecticide market share of approxi-
mately 40 % in 2008; Jeschke et al. 2010). Future monitoring
studies should therefore particularly focus on contemporary
insecticide classes and new ly introduced insecticide com-
pounds; the sampling strategy must be suitable for the com-
pounds of concern (Stehle et al. 2013) and must be conducted
by independent organizations.
Case studies: EU regulatory risk assessments
for bifenthrin and imidacloprid
This study uses the E U authorizations of the insecticides
bifenthrin (EFSA 2011) and imidacloprid (EFSA 2008)to
illustrate the lack of field relevance and margins of safety in
the current EU regulatory risk assessment schemes. The pre-
dicted aquatic exposure concentration of bifenthrin was calcu-
lated to be 0.0049 μg/L, using FOCUS step-4 PEC
SW
(incor-
porating 20 m no-spray buffer and 80 % runoff reduction) and
thus making use of essentially all exposure mitigating assump-
tions that the FOCUS model provides. The effect assessment
for aquatic organisms for this compound defined based on a
higher-tier mesocosms study with a no observed ecologically
adverse effect concentration (NOEAEC) of 0.015 μg/L and an
AF set to 3, a RAC
SW
of 0.005 μg/L. Overall, the final higher-
tier regulatory risk assessment for bifenthin indicated an ac-
ceptable aquatic risk, as the final RAC
SW
of 0.005 μg/L is
Environ Sci Pollut Res (2015) 22:19632–19647 19643
higher than the final PEC
SW
of 0.0049 μg/L. In essence, the
active substance bifenthrin was authorized in the EU using the
highest and therefore least conservative tiers in both the expo-
sure and effect assessment, with a difference of 0.0001 μg/L
(or 0.1 ng/L) between the PEC
SW
and the higher-tier RAC
SW
.
Although this procedure appears formally correct according to
legal requirements, it immediately becomes evident that, from
a scientific point of view, it cannot be ensured that this small
margin of safety is protective considering multifaceted field
conditions. It must be concluded that the field relevance, as
well as the margin of safety, of such an aquatic risk assessment
is considerably questionable. Within this context, it is worth
noting that all bifenthrin concentrations detected in EU sur-
face waters (n=8) exceeded both the PEC
SW
and the RAC
SW
,
which suggests that unacceptable effects occur in the field and
further challenges the protectiveness of the current regulatory
risk assessment approach for real-world situations.
The case study for the neonicotinoid insecticide
imidacloprid reveals further regulatory risk assessment uncer-
tainties. The EU authorization of this compound was based on
FOCUS step-4 PEC
SW
(incorporating 95 % spray drift reduc-
tion and 90 % runoff reduction) ranging between 0.152 and
0.429 μg/L subject to crop and FOCUS scenarios (see EFSA
(2008) for details). The higher-tier RAC
SW
of 0.3 μg/L was
based on a mesocosm study with a NOEC of 0.6 μg/L and an
AF of 2. It follows that the risk assessment already forecasts
surface water concentrations potentially to exceed the RAC
SW
under certain conditions. EFSA (2008)thusadmits:BOverall
it is concluded that a high risk for aquatic organisms is indi-
cated for the representative uses in orchards and tomatoes
requiring substantial risk mitigation measures to reduce spray
drift and runoff. BImidacloprid surface water concentrations
(n=21) were reported for six countries across the EU, with
concentrations reaching up to >200 μg/L (Mohr et al. 2012;
Starner and Goh 2012) and 28.6 % of all MIC
SW
exceeding
the RAC
SW
; these findings, again, challenge the overall pro-
tectiveness of the pre-authorization regulatory risk assessment
in the EU.
Conclusion and recommendations for risk assessment
amendments
For the first time, we evaluated the protectiveness and field
relevance of the regulatory EU pesticide risk assessment on a
continental scale. As a result, our meta-analysis shows that
MICs frequently exceed the RACs set for the authorization
of active substances at the EU level. This finding reveals the
critical failures of the EU pesticide regulations and the sub-
stantial and widespread ecological risks for the aquatic biodi-
versity. Moreover, even compliance, especially with higher-
tier RACs, may not provide sufficient protection for aquatic
ecosystems. The lack of consideration of pesticide mixtures
and significantly increasing risks due to the market
introduction of newer insecticide compounds poses further
challenges to the overall protectiveness of EU pesticide legis-
lation; the latter are also important for the future selection of
WFD priority substances. Overall, we conclude that the
European pre-authorization regulatory risk assessment for in-
secticides (and pesticides in general) must be substantially
improved in terms of field relevance and environmental pro-
tectiveness. We therefore propose the following five risk as-
sessment amendments:
(i) The conservatism of the regulatory exposure assessment
must be increased, e.g., by only considering step 1 PECs
or by applying safety factors to step 3 and 4 PECs (see
also Knäbel et al. (2012) and Knäbel et al. (2014) for
further information); in addition, the scope of the expo-
sure assessment must be extended to larger surface waters
and estuarine systems. Alternatively, the entire FOCUS
exposure assessment approach must be completely re-
vised and the protectiveness of the revised approach must
be validated independently using field data.
(ii) The uncertainties of the overall pre-authorization risk
assessment must be substantially reduced, and its protec-
tiveness must be increased; in particular, a critical recon-
sideration of the ecotoxicity endpoints (including
magnitude and duration of effects considered
acceptable for ERO-RACs) and AF used in higher-tier
risk assessment for the RAC derivation and authorization
of highly toxic compounds must thoroughly be ad-
dressed. In addition, mixture toxicity must be considered
in the prospective regulatory risk assessment, and the
implementation of additional hazard-based cut-off
criteria, e.g., for extremely toxic compounds, should be
considered. Incidences of unacceptable adverse effects at
concentrations below the RACs in the field should be
excluded with high confidence.
(iii) The overall link between the regulatory risk assessment
and the actual situation in the field must be considerably
strengthened, and findings from field studies on pesti-
cide exposure and effects must be used for a retrospec-
tive validation of the current EU regulatory risk assess-
ment, particularly for its future development. The fun-
damental rationale of the risk assessment, i.e., to protect
aquatic biocenoses in the field, not in the computer or
any sort of artificial test system, must be the driver for all
future risk assessment revisions.
(iv) Effective risk management measures (e.g., large non-
cropped buffer zones) should be mandatory for all pes-
ticide approvals.
(v) An obligatory validation of the risk assessment through
targeted chemical and biological post-authorization
moni toring programs must be implemented for com-
pounds of concern to ensure that their application does
not lead to unacceptable effects in the field.
19644 Environ Sci Pollut Res (2015) 22:19632–19647
In addition to these risk assessment amendments, farmers’
knowledge about appropriate pesticide use and environmental
awareness must also substantially be improved through oblig-
atory professional training, and adherence to risk mitigation
obligations (i.e., application prescriptions) should be moni-
tored. Above all, the reliance of EU agriculture on pesticides
should be critically reconsidered and replaced by more envi-
ronmental friendly alternatives, such as truly integrated pest
management and organic farming, wherever possible.
Acknowledgments We thank Walter H. Schreiber and Ralf B. Schäfer
for their statistical advice and Jörg Rapp and Caroline Nägele for their
support with Fig. S1. We are grateful to Jörn Wogram and two anony-
mous reviewers for their valuable comments on the manuscript. We thank
David Imo, Niklas Keck, Bonny Krell, and Koffi Tassou for translating
the foreign-language studies. This study was funded by the German So-
ciety for the Advancement of Sciences (DFG SCHU 2271/6-1).
Conflict of interest The authors declare that they have no competing
interests.
References
Akerblom N, Arbjörk C, Hedlund M, Goedkoop W (2008) Deltamethrin
toxicity to the midge Chironomus riparius Meigen—effects of ex-
posure scenario and sediment quality. Ecotoxicol Environ Saf 70:
53–60
Balderacchi M, Trevisan M (2010) Comments on pesticide risk assess-
ment by the revision of Directive EU 91/414. Environ Sci Pollut Res
17:523–528
BBA (2001) German Federal Office of Consumer Protection and Food
Safety (BVL) database of pesticide risk assessment endpoints
Beketov MA, Kefford BJ, Schäfer RB, Liess M (2013) Pesticides reduce
regional biodiversity of stream invertebrates. PNAS 110:11039–
11043
Belden JB, Lydy MJ (2006) Joint tox icity of chlorpyrifos and
esfenvalerate to fathead minnows and midge larvae. Environ
Toxicol Chem 25:623–629
Berenzen N, Kumke T, Schulz HK, Schulz R (2005) Macroinvertebrate
community structure in agricultural streams: impact of runoff-
related pesticide contamination. Ecotoxicol Environ Saf 60:37–46
Bereswill R, Streloke M, Schulz R (2013) Current-use pesticides in
stream water and suspended particles following runoff: exposure,
effects, and mitigation requirements. Environ Toxicol Chem 32:
1254–1263
Cedergreen N, Christensen AM, Kamper A, Kudsk P, Mathiassen SK,
Streibig JC, Sørensen H (2008) Review of independent action com-
pared to concentration addition as reference models for mixtures of
compounds with different molecular target sites. Environ Toxicol
Chem 27:1621–1632
Chagnon M, Kreutzweise D, Mitchell EAD, Morrissey CA, Noome DA,
van der Sluijs JP (2015) Risks of large-scale use of systemic insec-
ticides to ecosystem functioning and services. Environ Sci Pollut
Res 22:119–134
Crane M, Giddings JM (2004) Risk around the world—Becologically
acceptable concentrations^ when assessing the environmental risks
of pesticides under European Directive 91/414/EEC. Hum Ecol Risk
Assess 10:733–747
Crommentuijn T, Sijm D, de Bruijn J, van Leeuwen K, Van de Plassche E
(2000) Maximum permissible and negligible concentrations for
some organic substances and pesticides. J Environ Manag 58:297–
312
Cross P, Edward-Jones G (2011) Variation in pesticide hazard from arable
crop production in Great Britain from 1992 to 2008: an extended
time-series analysis. Crop Prot 30:1579–1585
Davies BR, Biggs J, Williams PJ, Lee JT, Thompson S (2008) A com-
parison of the catchment sizes of rivers, streams, ponds, ditches and
lakes: implications for protecting aquatic biodiversity in an agricul-
tural landscape. Hydrobiologia 597:7–17
Deneer JW (2000) Toxicity of mixtures of pesticides in aquatic systems.
Pest Manag Sci 56:516–520
Denholm I, Devine GJ, Williamson MS (2002) Insecticide resistance on
the move. Science 297:2222–2223
Denton DL, Wheelock CE, Murray SA, Deanovic LA, Hammock BD,
Hinton DE (2003) Joint acute toxicity of esfenvalerate and diazinon
to larval fathead minnows (Pimephales promelas). Environ Toxicol
Chem 22:336–341
Devine GJ, Furlong MJ (2007) Insecticide use: contexts and ecological
consequences. Agricul Human Val 24:281–306
DG SANCO (2002) Guidance document on aquatic ecotoxicology in the
context of the Directive 91/414/EEC. European Commission Health
& Consumer Protection Directorate-General, Sanco/3268/2001
rev.4 (final)
DG SANCO (2014) EU pesticides database. http://ec.europa.eu/sanco_
pesticides/public/index.cfm2011. Accessed 08 March 2014
ECPA (2006) Proposal for a regulation concerning the placing on the
market of plant protection products—
impact of the criteria for
‘no
n-approval’ and ‘candidates for substitution’. ECP A evaluation,
PP/06/EJ/15809, Brussels
EFSA (2008) Conclusion regarding the peer review of the pesticide risk
assessment of the active substance imidacloprid. EFSA Sci Report
148:1–120
EFSA (2010) Scientific opinion on the development of specific protection
goal options for environmental risk assessment of pesticides, in par-
ticular in relation to the revision of the guidance documents on
aquatic and terrestrial ecotoxicology (SANCO/3268/2001 and
SANCO/10329/2002). EFSA J 8:1821 (55 pp)
EFSA (2011) Conclusion on the peer review of the pesticide risk assess-
ment of the active substance bifenthrin. EFSA J 9:2159 (101 pp)
EFSA (2013) Guidance on tiered risk assessment for plant protection
products for aquatic organisms in edge-of-field surface waters.
EFSA J 11:3290 (186 pp)
EFSA (2014) Pesticides publications. http://www.efsa.europa.eu/en/
pesticides/pesticidesscdocs.htm. Accessed 12 March 2014
European Commiss ion (1991) Council Directive 91/414/EEC of 15
July 1991 concerning the placing of plant protection products on
the market. Off J Eur Union L230:1–32
European Commission (2000) Directive 2000/60/EC of the European
parliament and of the council of 23 October 2000 establishing a
framework for community action in the field of water policy. Off J
Eur Comm L327:1–72
European Commission (2009a) Regulation (EC) No 1107/2009 of the
European parliament and the council of 21 October 2009 concerning
the placing of plant protection products on the market and repealing
Council Directives 79/117/EEC and 91/414/EEC. Off J Eur Union
L309:1–50
European Commission (2009b) Directive 2009/128/EC of the European
parliament and the council of 21 October 2009 establishing a frame-
work for community action to achieve sustainable use of pesticides.
Off J Eur Union L309:71–86
European Commission (2013) Directive 2013/39/EU of the
European parliament and the council of 12 August 2013
amending Directives 2000/ 60 /EC and 2008 /105/ EC as reg ards
priority substances in the field of water p olicy. Off J Eur Union
L226:1–17
Environ Sci Pollut Res (2015) 22:19632–19647 19645
European Commission (2014) Member states factsheets—European
Union. DG Agriculture and Rural Development, Agricultural
Policy Analysis and Perspectives Unit
European Environment Agency (2012) European waters—assessment of
status and pressures. EEA Report No 8/2012, Office for Publications
of the European Union, Luxembourg
Eurostat (2013) Agriculture, forestry and fishery statistics 2013 edition.
European Union, Luxembourg
FOCUS (2001) FOCUS surface water scenarios in the EU evaluation
process under 91/414/EEC. Report of the FOCUS Working Group
on Surface Water Scenarios, EC Document Reference SANCO/
4802/2001-rev.2
Forbes VE, Cold A (2005) Effects of the pyrethroid esfenvalerate on life-
cycle traits and population dynamics of Chironomus riparius—im-
portance of exposure scenario. Environ Toxicol Chem 24:78–86
Goulson D (2013) An overview of the environmental risks posed by
neonicotinoid insecticides. J Appl Ecol 50:977–987
Hallmann CA, Foppen RPB, van Turnhout CAM, de Kroon H, Jongejans
E (2014) Declines in insectivorous birds are associated with high
neonicotinoid concentrations. Nature 511:341–343
Jeschke P, Nauen R (2008) Neonicotinoids—fromzerotoheroininsec-
ticide chemistry. Pest Manag Sci 64:1084–1098
Jeschke P, Nauen R, Schindler M, Elbert A (2010) Overview of the status
and global strategy for neonicotinoids. J Agric Food Chem 59:
2897–2908
Kattwinkel M, Kühne J-V, Foit K, Liess N (2011) Climate change, agri-
cultural insecticide exposure, and risk for freshwater communities.
Ecol Appl 21:2068–2081
Knäbel A, Stehle S, Schäfer RB, Schulz R (2012) Regulatory FOCUS
surface water models fail to predict insecticide concentrations in the
field. Environ Sci Technol 46:8397–8404
Knäbel A, Meyer K, Rapp J, Schulz R (2014) Fungicide field concentra-
tions exceed FOCUS surface water predictions: urgent need of mod-
el improvement. Environ Sci Technol 48:455–463
Kortenkamp A, Backhaus T, Faust M (2009) State of the art on mixture
toxicity. Final report, study contract number 070307/2007/485103/
ETU/D.1, European Commission, Brussels
Lamberth C, Jeanmart S, Luksch T, Plant A (2013) Current challenges
and trends in the discovery of agrochemicals. Science 341:742–746
Luttik R, Hart A, Roelofs W, Craig P, Mineau P (2011) Variation in the
level of protection afforded to birds and crustaceans exposed to
different pesticides under standard risk assessment procedures.
Integr Environ Assess Manag 7:459–465
Malaj E, von der Ohe PC, Grote M, Kühne R, Mondy CP, Usseglio-
Polatera P, Brack W, Schäfer RB (2014) Organic chemicals jeopar-
dize the health of freshwater ecosystems on the continental scale.
PNAS 111:9549–9554
Maltby L, Brock TCM, Van den Brink PJ (2009) Fungicide risk assess-
ment for aquatic ecosystems: importance of interspecific variation,
toxic mode of action, and exposure regime. Environ Sci Technol 43:
7556–7563
Mohr S, Berghahn R, Schmiediche R, Hübner V, Loth S, Feibicke M,
Mailahn W, Wogram J (2012) Macroinvertebrate community re-
sponse to repeated short-term pulses of the insecticide imidacloprid.
Aqua T oxicol 110–111:25–36
Moschet C, Wittmer I, Simovic J, Junghans M, Piazzoli A, Singer H,
Stamm C, Leu C, Hollender J (2014) How a complete pesticide
screening changes the assessment of surface water quality. Environ
Sci Technol 48:5423–5432
Newman MC, M cCloskey JT (1 996) Time-to-event analyses of
ecotoxicity data. Ecotoxicology 5:187–196
Nienstedt KM, Brock TCM, Van Wensem J, Montforts M, Hart A et al
(2012) Development of a framework based on an ecosystem ser-
vices approach for deriving specific protection goals for environ-
mental risk assessment of pesticides. Sci Total Environ 415:31–38
Pe`er
G, Dicks LV, Visconti P, Arlettaz R, Baldi A et al (2014) EU
agricultural reform fails on biodiversity. Science 344:1090–1092
Peters K, Bundschuh M, Schäfer RB (2013) Review on the effects of
toxicants on freshwater ecos ystem functions. Environ Poll 180:
324–329
Pigott TD (2009) Handling missing data. In: Cooper H, Hedges LV,
Valentine JC (eds) The handbook of research synthesis and
meta-analysis, 2nd edn. Russel Sage Foundation, New York,
pp 399–416
PPDB (2013) The Pesticide Properties Database (PPDB) developed by
the Agriculture & Environment Research Unit (AERU), University
of Hertfordshire, funded by UK national sources and the EU-funded
Footprint project (FP6-SSP-022704). http://sitem.herts.ac.uk/aeru/
ppdb/en/index.htm. Accessed 15 December 2013
Reichenberger S, Bach M, Skitschak A, Frede H-G (2007)
Mitigation strategies to reduce pesticide inputs into ground-
and surface water and their effectiveness; a revi ew. Sci Tota l
Environ 384: 1–35
Sanchez-Bayo F (2014) The trouble with neonicotinoids. Science 346:
806–807
Schäfer RB, von der Ohe PC, Kühne R, Schüürmann G, Liess M (2011)
Occurrence and toxicity of 331 organic pollutants in large rivers of
north Germany over a decade (1994 to 2004). Environ Sci Technol
45:6167–6174
Schäfer RB, von der Ohe PC, Rasmussen J, Kefford BJ, Beketov MA,
Schulz R, Liess M (2012) Thresholds for the effects of pesticides on
invertebrate communities and leaf breakdown in stream ecosystems.
Environ Sci T echnol 46:5134–5142
Schulz R (2004) Field studies on exposure, effects and risk mitigation of
aquatic nonpoint-source insecticide pollution—a review. J Environ
Qual 33:419–448
Schulz R, Liess M (2000) Toxicity of fenvalerate to caddis fly larvae:
chronic effects of 1- vs 10-h pulse-exposure with constant doses.
Chemosphere 41:1511–1517
Solomon KR, Giddings JM, Maund SJ (2001) Probabilistic risk assess-
ment of cotton pyrethroids: I. Distributional analysis of laboratory
aquatic toxicity data. Environ T oxicol Chem 20:652–659
Spurlock F, Lee M (2008) Synthetic pyrethroid use patterns, properties,
and environmental effects. In: Gan J, Spurlock F, Hendley P, Weston
D (eds) Synthetic pyrethroids: occurrence and behavior in aquatic
environments. American Chemical Society, Washington, DC, pp 3–
25
Starner K, Goh KS (2012) Detections of the neonicotinoid insecticide
imidacloprid in surface wate rs of three agricultural regions of
California, USA, 2010–2011. Bull Environ Contamin Toxicol 88:
316–321
Steen RJCA, Leonards PEG, Brinkman UAT, Barcelo D, Tronczynski J,
Albanis TA, Cofino WP (1999) Ecological risk assessment of agro-
chemicals in European estuaries. Environ Toxicol Chem 18:1574–
1581
Stehle S, Schulz R (2015) Agricultural insecticides threaten surface wa-
ters at the global scale. PNAS 112:5750–5755
Stehle S, Elsaesser D, Gregoire C, Imfeld G, Niehaus E, Passeport E,
Payraudeau S, Schäfer RB, Tournebize J, Schulz R (2011 )
Pesticide risk mitigation by vegetated treatment systems: a meta-
analysis. J Environ Qual 40:1068–1080
Stehle S, Knäbel A, Schulz R (2013) Probabilistic risk assessment of
insecticide concentrations in agricultural surface waters: a critical
appraisal. Environ Monit Assess 185:6295–6310
Tennekes HA, Sanchez-Bayo F (2011) Time-dependent toxicity of
neonicotinoids and other toxicants: implications for a new approach
to risk assessment. J Environ Anal Toxicol S4:001. doi:10.4172/
2161-0525.S4-001
US EPA (2012) Pesticide reregistrati on status. http://www
.epa.gov/
pesticides/reregistration/status.htm. Accessed 12 December 2012
19646 Environ Sci Pollut Res (2015) 22:19632–19647
US EPA (2014) ECOTOXicology database system. Version 4.0. http://
www.epa.gov/ecotox/. Accessed 03 February 2013
Van Dijk TC, van Staalduinen MA, van der Sluijs JP (2013) Macro-
invertebrate decline in surface water polluted with imidacloprid.
PLoS One 8, e62374 (10 pp)
Von der Ohe PC, Dublio V, Slobodnik J, De Deckere E, Kühne R et al
(2011) A new risk assessment approach for the prioritization of 500
classical and emerging organic microcontaminants as potential river
basin specific pollutants under the Europea n Water Framework
Directive. Sci Total Environ 409:2064–2077
Yu SJ (2008) The toxicology and biochemistry of insecticides. CRC
Press, Boca Raton
Zubrod JP, Englert D, Feckler A, Koksharova N et al (2015) Does the
current fungicide risk assessment provide sufficient protection for
key drivers in aquatic ecosystem functioning? Environ Sci Technol
49:1173–1183
Environ Sci Pollut Res (2015) 22:19632–19647 19647
