MOF, Alm.del - 2018-19 (1. samling) - Endeligt svar på spørgsmål 445: Spm. om, hvorvidt Miljøstyrelsens vurdering i forbindelse med dispensation fra brug af neonikotinoider til sukkerroer udelukkende baserer sig på Aarhus Universitets landbrugsfaglige vurdering, til miljø- og fødevareministeren

Miljø- og Fødevareudvalget 2018-19 (1. samling)
MOF Alm.del
Offentligt

CONCLUSION ON PESTICIDES PEER REVIEW

ADOPTED: 1 February 2018

doi: 10.2903/j.efsa.2018.5178

Peer review of the pesticide risk assessment for bees

for the active substance imidacloprid considering the

uses as seed treatments and granules

European Food Safety Authority (EFSA)

Abstract

The EFSA was asked by the European Commission to perform an updated risk assessment of

neonicotinoids, including imidacloprid, as regards the risk to bees, as a follow-up of previous mandates

received from the European Commission on neonicotinoids. The context of the evaluation was that

required by the European Commission in accordance with Article 21 of Regulation (EC) No 1107/2009

to review the approval of active substances in light of new scientiﬁc and technical knowledge and

monitoring data. In this context and in accordance with Article 31 of Regulation (EC) No 178/2002,

EFSA has been previously asked by the European Commission to organise an open call for data in

order to collect new scientiﬁc information as regards the risk to bees from the neonicotinoid pesticide

active substances clothianidin, thiamethoxam and imidacloprid applied as seed treatments and

granules in the EU. The conclusions were reached on the basis of the evaluation of the supported uses

as an insecticide of imidacloprid applied as seed treatments and granules, on the new relevant data

collected in the framework of the open call organised by EFSA and on the updated literature search

performed by EFSA. The reliable endpoints, appropriate for use in regulatory risk assessment derived

from the submitted studies and literature data as well as any other relevant data available at national

level and made available to EFSA, are presented. Concerns are identiﬁed.

2018 European Food Safety Authority.

EFSA Journal

published by John Wiley and Sons Ltd on behalf

of European Food Safety Authority.

Keywords:

imidacloprid, peer review, risk assessment, pesticide, insecticide, neonicotinoid

Requestor:

European Commission

Question number:

EFSA-Q-2015-00772

Correspondence:

[email protected]

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

MOF, Alm.del - 2018-19 (1. samling) - Endeligt svar på spørgsmål 445: Spm. om, hvorvidt Miljøstyrelsens vurdering i forbindelse med dispensation fra brug af neonikotinoider til sukkerroer udelukkende baserer sig på Aarhus Universitets landbrugsfaglige vurdering, til miljø- og fødevareministeren

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Suggested citation:

EFSA (European Food Safety Authority), 2018. Conclusion on the peer review of

the pesticide risk assessment for bees for the active substance imidacloprid considering the uses as

seed treatments and granules. EFSA Journal 2018;16(2):5178, 113 pp.

https://doi.org/10.2903/j.efsa.

2018.5178

ISSN:

1831-4732

2018 European Food Safety Authority.

EFSA Journal

published by John Wiley and Sons Ltd on behalf

of European Food Safety Authority.

This is an open access article under the terms of the

Creative Commons Attribution-NoDerivs

License,

which permits use and distribution in any medium, provided the original work is properly cited and no

modiﬁcations or adaptations are made.

The EFSA Journal is a publication of the European Food

Safety Authority, an agency of the European Union.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Summary

Imidacloprid was included in Annex I to Directive 91/414/EEC on 1 August 2009 by Commission

Directive 2008/116/EC and has been deemed to be approved under Regulation (EC) No 1107/2009, in

accordance with Commission Implementing Regulation (EU) No 540/2011, as amended by Commission

Implementing Regulation (EU) No 541/2011. A speciﬁc conclusion has been issued by European Food

Safety Authority (EFSA) on the risk assessment for bees as regards the authorised uses applied as

seed treatments and granules.

The speciﬁc provisions of the approval were amended by Commission Implementing Regulation (EU)

No 485/2013, to restrict the uses of clothianidin, thiamethoxam and imidacloprid, to provide for speciﬁc

risk mitigation measures for the protection of bees and to limit the use of the plant protection products

containing these active substances to professional users. In particular, the uses as seed treatment and

soil treatment of plant protection products containing clothianidin, thiamethoxam or imidacloprid have

been prohibited for crops attractive to bees and for cereals except for uses in greenhouses and for

winter cereals. Foliar treatments with plant protection products containing these active substances have

been prohibited for crops attractive to bees and for cereals with the exception of uses in greenhouses

and uses after

ﬂowering.

Furthermore, the European Commission requested EFSA to provide conclusions

concerning an updated risk assessment for bees for clothianidin, thiamethoxam and imidacloprid, taking

into account all uses other than seed treatments and granules, including foliar spray uses as mentioned

in recital 7 of Commission Implementing Regulation (EU) No 485/2013. EFSA

ﬁnalised

its conclusion on

the risk assessment for bees as regards all uses other than seed treatments and granules in July 2015.

It was a speciﬁc provision of the Commission Implementing Regulation (EU) No 485/2013 that the

applicant was also required to submit to the European Commission further ecotoxicological studies by

31 December 2014. The outcomes of the peer review of the conﬁrmatory data assessment were

reported in a Technical Report and a conclusion published in 2016.

Furthermore, according to recital 16 of Regulation (EU) No 485/2013, within 2 years from the date

of entry into force of that Regulation, the European Commission foresees to initiate without undue

delay a review of the new scientiﬁc information available.

For this purpose, with reference to Article 31 of Regulation (EC) No 178/2002 and in accordance

with Article 21 of Regulation (EC) No 1107/2009, the European Commission requested EFSA to

organise an open call for data in order to collect new scientiﬁc information as regards the risk to bees

from the neonicotinoid pesticide active substances clothianidin, thiamethoxam and imidacloprid applied

as seed treatments and granules in the European Union (EU).

The European Commission requested EFSA to provide conclusions concerning an updated risk

assessment for bees for the three neonicotinoids (namely clothianidin, imidacloprid and

thiamethoxam), taking into account:

•

the new relevant data collected in the framework of the speciﬁc open call for data

any other new data from studies, research and monitoring activities that are relevant to the

uses under consideration

the EFSA Guidance Document on the risk assessment of plant protection products on bees

(Apis

mellifera, Bombus

spp. and solitary bees)

EFSA also considered the data available from a systematic literature review performed in June

2016, in order to collect all published scientiﬁc literature relevant for the current evaluation.

Risk assessments were performed for honeybees, bumblebees and solitary bees according to the

EFSA guidance document on the risk assessment of plant protection products on bees. For exposure

via residues in pollen and nectar, where a higher tier (Tier-3) risk assessment could be performed, a

low risk was concluded for some crops for honeybees. However, when all the bee groups (honeybees,

bumblebees and solitary bees) are considered, a high risk was concluded or it was concluded that a

low risk was not demonstrated for all the uses assessed.

For the exposure via residues from dust drift, a low risk was concluded for some crops for honeybees.

However, when all the bee groups (honeybees, bumblebees and solitary bees) are considered, a high risk

was concluded or it was concluded that a low risk was not demonstrated for all the uses assessed.

For exposure via water consumption, a low risk to honeybees was concluded for all uses via residues

in puddles or via surface water. A low risk to honeybees was concluded for residues in guttation

ﬂuid

for

the uses to winter cereals, sugar beet and potatoes. A high risk was concluded for all other uses.

Refer to Table

in the main text of the conclusion for crop-speciﬁc conclusion achieved at each

assessment tier.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Table of contents

Abstract.................................................................................................................................................

Summary...............................................................................................................................................

Background

...........................................................................................................................................

The active substance and metabolites......................................................................................................

Assessment............................................................................................................................................

Uses assessed

.........................................................................................................................

Summary of the data considered in this conclusion

....................................................................

Principles and assessment criteria.............................................................................................

3.1.

Aim of the assessment.............................................................................................................

Tier-1

.....................................................................................................................................

3.2.

3.2.1.

Selection of the endpoints........................................................................................................

3.3.

Reﬁnement of the exposure

.....................................................................................................

Residues in pollen and nectar...................................................................................................

3.3.1.

3.3.1.1.

Data evaluation and selection...................................................................................................

3.3.1.2.

Calculation of reﬁned shortcut values........................................................................................

3.3.1.3.

Estimation of the exposure assessment goal..............................................................................

3.3.2.

Dust drift and deposition..........................................................................................................

Weeds in the

ﬁeld

...................................................................................................................

3.3.3.

3.3.4.

Residues in water sources........................................................................................................

3.4.

Reﬁnement with higher tier experiments

...................................................................................

Building up the lines of evidence

..............................................................................................

3.4.1.

3.4.2.

Integrating the lines of evidence

..............................................................................................

Outcome of the risk assessment: toxicity endpoints

...................................................................

Standard endpoints

.................................................................................................................

4.1.

4.2.

Additional sublethal laboratory data

..........................................................................................

Outcome of the risk assessment

...............................................................................................

Risk assessments for seed treatment products...........................................................................

5.1.

5.1.1.

Risk via systemic translocation in plants

–

residues in nectar and pollen (treated crop scenario

and succeeding crop scenario)

.................................................................................................

5.1.1.1.

Tier-1 risk assessment

.............................................................................................................

5.1.1.2.

Exposure assessment for the treated and succeeding crop scenarios...........................................

5.1.1.3.

Tier-2 risk assessment

.............................................................................................................

5.1.1.4.

Tier-3 risk assessment (weight of evidence)

..............................................................................

5.1.1.4.1.

Weight of evidence higher tier risk assessment for honeybees

....................................................

5.1.1.4.2.

Weight of evidence higher tier risk assessment for bumblebees

..................................................

5.1.1.4.3.

Solitary bees

...........................................................................................................................

5.1.2.

Risk from contamination of adjacent vegetation via dust drift (ﬁeld margin and adjacent crop scenario)

5.1.2.1.

Tier-1 risk assessment

.............................................................................................................

5.1.2.2.

Exposure assessment for the

ﬁeld

margin and adjacent crop scenario

.........................................

5.1.2.3.

Tier-2 risk assessment

.............................................................................................................

5.1.2.4.

Tier-3 risk assessment

.............................................................................................................

5.1.3.

Risk via water consumption......................................................................................................

5.1.3.1.

Guttation water.......................................................................................................................

5.1.3.2.

Puddle water

..........................................................................................................................

5.1.3.3.

Surface water

.........................................................................................................................

5.2.

Risk assessments for granule

...................................................................................................

Risk via systemic translocation in plants

–

residues in nectar and pollen

......................................

5.2.1.

5.2.1.1.

Tier-1 risk assessment for treated crop scenario, weed scenario and succeeding crop scenario

.....

5.2.1.2.

Exposure assessment for the treated crop, succeeding crop scenarios and for

ﬂowering

weeds.....

5.2.1.3.

Tier-2 risk assessment

.............................................................................................................

5.2.1.4.

Tier-3 risk assessment

.............................................................................................................

Risk from contamination via dust drift.......................................................................................

5.2.2.

5.2.2.1.

Tier-1 risk assessment

.............................................................................................................

5.2.2.2.

Exposure assessment for the

ﬁeld

margin and adjacent crop scenario

.........................................

5.2.2.3.

Tier-2 risk assessment

.............................................................................................................

5.2.2.4.

Tier-3 risk assessment

.............................................................................................................

5.2.3.

Risk via water consumption......................................................................................................

5.2.3.1.

Guttation water.......................................................................................................................

5.2.3.2.

Puddle water

..........................................................................................................................

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

5.2.3.3.

Surface water

.........................................................................................................................

Overall conclusion

...................................................................................................................

6.1.

Seed treatment uses

...............................................................................................................

6.2.

Granular uses..........................................................................................................................

Overview of the concerns identiﬁed for each representative use considered.................................

References.............................................................................................................................................

Abbreviations

.........................................................................................................................................

Appendix A

–

List of supported uses

........................................................................................................

Appendix B

–

Overview of endpoint types and related relevance class assigned within the scope of the

present risk assessment

..........................................................................................................................

Appendix C

–

Tier-1 risk assessment based on EFSA (2013c)

.....................................................................

Appendix D

–

Measured residue values and RUD values used for calculation of exposure assessment goals

Appendix E

–

Residue intake in higher tier studies providing reliable Class 1 and Class 2 endpoints

..............

Appendix F

–

Tier-3 lines of evidence

.......................................................................................................

Appendix G

–

List of study references

......................................................................................................

Appendix H

–

Used compound codes

.......................................................................................................

108

113

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Background

Imidacloprid was included in Annex I to Directive 91/414/EEC

on 1 August 2009 by Commission

Directive 2008/116/EC

and has been deemed to be approved under Regulation (EC) No 1107/2009

in accordance with Commission Implementing Regulation (EU) No 540/2011

, as amended by

Commission Implementing Regulations (EU) No 541/2011

. A speciﬁc conclusion has been issued by

European Food Safety Authority (EFSA) on the risk assessment for bees as regards the authorised uses

applied as seed treatments and granules (EFSA, 2013a).

The speciﬁc provisions of the approval were amended by Commission Implementing Regulation

(EU) No 485/2013

, to restrict the uses of clothianidin, thiamethoxam and imidacloprid, to provide for

speciﬁc risk mitigation measures for the protection of bees and to limit the use of the plant protection

products containing these active substances to professional users. In particular, the uses as seed

treatment and soil treatment of plant protection products containing clothianidin, thiamethoxam or

imidacloprid have been prohibited for crops attractive to bees and for cereals except for uses in

greenhouses and for winter cereals. Foliar treatments with plant protection products containing these

active substances have been prohibited for crops attractive to bees and for cereals with the exception

of uses in greenhouses and uses after

ﬂowering.

Furthermore, the European Commission requested

EFSA to provide conclusions concerning an updated risk assessment for bees for clothianidin,

thiamethoxam and imidacloprid, taking into account all uses other than seed treatments and granules,

including foliar spray uses as mentioned in recital 7 of Commission Implementing Regulation (EU)

No 485/2013. EFSA

ﬁnalised

its conclusion on the risk assessment for bees as regards all uses other

than seed treatments and granules in July 2015 (EFSA, 2015).

It was a speciﬁc provision of the Commission Implementing Regulation (EU) No 485/2013 that the

applicant was also required to submit to the European Commission further ecotoxicological studies by

31 December 2014. The outcomes of the peer review of the conﬁrmatory data assessment were

reported in a Technical Report (EFSA, 2016a) and a conclusion (EFSA, 2016b)

ﬁnalised

in 2016.

Furthermore, according to recital 16 of Regulation (EU) No 485/2013, within 2 years from the date

of entry into force of that Regulation, the European Commission foresees to initiate without undue

delay a review of the new scientiﬁc information available.

For this purpose, with reference to Article 31 of Regulation (EC) No 178/2002

and in accordance

with Article 21 of Regulation (EC) No 1107/2009, in February 2015, the European Commission

requested EFSA to organise an open call to collect new scientiﬁc information as regards the risk to

bees from the neonicotinoid pesticide active substances clothianidin, thiamethoxam and imidacloprid

applied as seed treatments and granules in the European Union (EU) (EFSA, 2014b) and then,

following a second mandate received in November 2015, EFSA was requested to provide conclusions

concerning an updated risk assessment for bees for the three neonicotinoids (namely clothianidin,

imidacloprid and thiamethoxam).

The new relevant data collected in the framework of the open call for data and any other new data

from studies, research and monitoring activities relevant for the uses under consideration were taken

into account. To address the mandate, EFSA also considered the data available from a previous

systematic literature review, outsourced in 2013 (Fryday et al., 2015). Furthermore, an update of this

systematic review was performed in June 2016, in order to collect all published scientiﬁc literature

Council Directive 91/414/EEC of 15 July 1991 concerning the placing of plant protection products on the market. OJ L 230,

19.8.1991, p. 1–32, as last amended.

Commission Directive 2008/116/EC of 15 December 2008 amending Council Directive 91/414/EEC to include aclonifen,

imidacloprid and metazachlor as active substances. OJ L 337, 16.12.2008, p. 86–91.

Regulation (EC) No 1107/2009 of the European Parliament and of 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. OJ No L 309,

24.11.2009, p. 1–50.

Commission Implementing Regulation (EU) No 540/2011 of 25 May 2011 implementing Regulation (EC) No 1107/2009 of the

European Parliament and of the Council as regards the list of approved active substances. OJ L 153, 11.6.2011, p. 1–186.

Commission Implementing Regulation (EU) No 541/2011 of 1 June 2011 amending Implementing Regulation (EU) No 540/2011

implementing Regulation (EC) No 1107/2009 of the European Parliament and of the Council as regards the list of approved

active substances. OJ L 153, 11.6.2011, p. 187–188.

Commission Implementing Regulation (EU) No 485/2013 of 24 May 2013 amending Implementing Regulation (EU) No 540/

2011, as regards the conditions of approval of the active substances clothianidin, thiamethoxam and imidacloprid, and

prohibiting the use and sale of seeds treated with plant protection products containing those active substances. OJ L 139,

25.5.2013, p. 12–26.

Regulation (EC) No 178/2002 of the European Parliament and of the Council of 28 January 2002 laying down the general

principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in

matters of food safety. OJ L 31, 1.2.2002, p. 1–24.

EFSA Journal 2018;16(2):5178

www.efsa.europa.eu/efsajournal

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

relevant for the current evaluation (EFSA, 2018a). The EFSA guidance document on the risk

assessment of plant protection products on bees (EFSA, 2013c) was used for the current evaluation as

requested in the mandate.

A consultation on the evaluation and preliminary conclusions of EFSA on the risk assessment for

bees was conducted with Member States via a written procedure in September 2017. The preliminary

draft conclusion drawn by EFSA, together with the points that required further consideration in the

assessment, as well as the speciﬁc issues raised by Member States following the consultation were

discussed at the Pesticides Peer Review Experts’ Meeting 166 on ecotoxicology in October 2017. Details

of the issues discussed together with the outcome of these discussions were recorded in the meeting

report. After the expert meeting, EFSA

ﬁnalised

the conclusions and launched a second written

procedure on the

ﬁnal

draft in December 2017–January 2018 in order to provide their comments on

those parts of the conclusions and supporting documents that have been amended following the Peer

Review Meeting. The compiled comments were considered by EFSA and are published as part of the

background documents to the conclusions.

In addition, key supporting documents to this conclusion are the Technical Report on the evaluation

of data (EFSA, 2018a) and the Peer Review Report (EFSA, 2018b).

The Technical Report provides the methodology developed by EFSA relating to the evaluation of the

available data concerning their relevance for the current risk assessment and their scientiﬁc reliability.

It is composed as follows:

•

Technical Report on the evaluation of data (EFSA, 2018a)

Study Evaluation Notes (Appendices D–O) to the Technical Report (EFSA, 2018a).

The Peer Review Report is a compilation of the documentation developed to evaluate and address

all issues raised in the peer review, it comprises the following documents, in which all views expressed

during the course of the peer review, including minority views where applicable, can be found:

•

the comments received on the preliminary draft EFSA conclusion,

the report of the scientiﬁc consultation with Member State experts,

the comments received on the

ﬁnal

draft EFSA conclusion.

It is recommended that this conclusion report and its background documents would not be

accepted to support any registration outside the EU for which the applicant has not demonstrated that

it has regulatory access to the information on which this conclusion report is based.

The active substance and metabolites

Imidacloprid is the ISO common name for (E)-1-(6-Chloro-3-pyridinylmethyl)-N-nitroimidazolidin-2-

ylideneamine (IUPAC).

Imidacloprid belongs to the group of neonicotinoid compounds which are used as insecticides. They

interact with the receptor protein of nicotinic acetyl choline receptors in the nerve

ﬁbre

membrane of

insects.

The available earlier EU evaluations (EFSA, 2008, 2013a, 2015) have identiﬁed those metabolites of

imidacloprid which require consideration in a risk assessment for bees. Many of these metabolites are

several orders of magnitude less toxic to honeybees than the parent substance or had comparable

acute toxicity to imidacloprid. The single additional endpoint (which was considered as relevant and

reliable) submitted for this review has conﬁrmed this (I.2020). However, their occurrence in pollen and

nectar is signiﬁcantly lower. In the data set considered for this conclusion, some residue data were

available for metabolites imidacloprid oleﬁn and 5-OH-imidacloprid. These new data did neither

indicate higher residue levels of the metabolites in pollen and nectar than the parent imidacloprid.

Therefore, it was considered that the risk from metabolites is covered by the risk assessment of the

parent and no separate risk assessment for metabolites is required.

Assessment

Uses assessed

In accordance with the mandate received in February 2015, EFSA liaised with applicants in order to

collect feedback on the uses they would like to support for the EU market. During the open call for

data, the applicants were requested to submit information on the uses of imidacloprid (Good

Agricultural Practices), applied as a seed treatment or granule that they wish to support. In a second

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

step, in December 2015, Member States were requested to validate the consolidated Good Agricultural

Practices (GAPs) from applicants, providing feedback on the authorised uses in their respective

countries. However, the risk assessment was performed for all uses supported by the applicants. Full

details of the GAPs are given in Appendix

Tables

and

provide a brief summary of the critical

GAPs relevant to the risk assessment for bees. Only the highest and lowest of the maximum

application and seed treatment rates are given in Tables

and

It has to be noted that the potato seed treatment use, as outlined by the Applicant, is an in-planter

tuber treatment. This treatment is done within the planter machine via a spray application just before

the tubers are dropped into the furrow.

Dummy pill use, as outlined by the applicant, is when a treated dead seed is sown together with an

untreated alive seed (ratio 1:1).

Some of the vegetables are harvested before the

ﬂowering

period, unless they are grown for seed

production.

Table 1:

Crop

Spring cereals

Winter cereals

Summary of the seed treatment uses considered in this conclusion

Highest seed

Lowest seed

Highest

Lowest

application rate application rate treatment rate treatment rate Notes

(mg a.s./seed) (mg a.s./seed)

(g a.s./ha)

112

126

0.039

0.015

0.039

–

Four products,

three containing

other active

substances

(a)

–

Applied as a

dummy pill.

Seeds sown in

greenhouse and

transplanted at

BBCH 12 either

in the

ﬁeld

or in

greenhouse

Cotton

Endive

Lettuce

Endive

Lettuce

Brassica,

ﬂowering,

head, leafy (crops

like broccoli,

cauliﬂowers,

brussels sprouts,

head cabbages,

chinese cabbage,

kales)

Maize

Potato

100

104

175

104

120

0.63

0.8

1.5

0.84

0.8

1.2

1.5

120

100

180

No information

–

Two products,

one also

containing

pencycuron

Single product

also containing

beta-cyﬂuthrin

Single product

also containing

beta-cyﬂuthrin

Four products,

three containing

other active

substances

(b)

Spring rape

0.01

Winter rape

0.01

Sugar and fodder

beet

117

0.1

0.9

BBCH: growth stages of mono- and dicotyledonous plants.

(a):

‘FS

373.4’ contains clothianidin, prothioconazole and tebuconazole in addition to imidacloprid.

‘FS

400’ and

‘FS

200’ contains

prothioconazole in addition to imidacloprid.

(b):

‘FS

230’ contains beta-cyﬂuthrin in addition to imidacloprid.

‘FS

280’ contains clothianidin and beta-cyﬂuthrin in addition to

imidacloprid.

‘FS

417.8’ contains teﬂuthrin in addition to imidacloprid.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Table 2:

Crop

Summary of the granular uses considered in this conclusion

BBCH at the

time of

application

Highest

Lowest

application rate application rate Notes

(g a.s./ha)

150

Spread uniformly over the area with

granular application equipment,

immediately followed by sufﬁcient

irrigation to move the active

ingredient through the thatch,

wetting the top inch soil

All crop stages

Managed

amenity turf

(golf courses,

sport grounds,

commercial

and residential

lawns)

BBCH: growth stages of mono- and dicotyledonous plants.

Summary of the data considered in this conclusion

Concerning the effect data, the present conclusion makes use of different sources.

The

ﬁrst

source of data was the open call for data for new scientiﬁc information as regards the risk

to bees from the use of the three neonicotinoid pesticide active substances clothianidin, imidacloprid

and thiamethoxam applied as seed treatments and granules in the EU. EFSA launched this call from

May 2015 to September 2015, as requested by the EU Commission. More details on the open call for

data are available in a dedicated Technical Report (EFSA, 2014b).

Other sources of data were the systematic literature search on the neonicotinoids and the risks to

bees that EFSA outsourced in 2013 (Fryday et al., 2015) and the related update performed by EFSA in

June 2016 (Appendix B to EFSA, 2018a).

The

ﬁrst

systematic literature search comprised 546 (already screened) documents, while the update

of the literature search retrieved 874 documents. In addition, 376 contributions were received during

the open call for data. After duplicate removal, the overall initial list included 1599 documents. A title

and abstract screening step identiﬁed 680 potentially relevant documents which were then subject to

full text screening. During the full text screening, all experiments within the available documents were

identiﬁed and totalled 968. Of these experiments, 588 were critically appraised and the data extracted.

Finally, in accordance with the European Commission mandate, Member States were also further

requested to provide any monitoring data not yet available during the open call data. The data

submitted were already included in the dataset.

Full details on the collection of the available data investigating the effects and exposure of

imidacloprid to bees, together with their assessment for their reliability and relevance, are given in the

Technical Report on the evaluation of data and related appendices (EFSA, 2018a).

Furthermore, for what concern the exposure data, data already used in previous assessments

(EFSA, 2013a) were also considered, as information on residue levels was already systematically

collected and organised by EFSA during such previous assessments.

3.1.

Principles and assessment criteria

Aim of the assessment

The current EU agreed level of protection for bees is to ensure that effects on colonies/populations

are negligible. This means that the exposure of the colonies/populations at the edge of the treated

ﬁelds

should not exceed a level which results in an effect greater than negligible.

As requested by the European Commission mandate, to perform the risk assessment of the three

active substances the EFSA Guidance Document on the risk assessment of plant protection products

on bees (Apis

mellifera, Bombus

spp. and solitary bees), hereafter referred to as EFSA (2013c) was

followed. The basis of the risk assessment according to EFSA (2013c) is to ensure that the speciﬁc

protection goals (SPG) for honeybees, bumblebees and solitary bees are met.

Namely:

•

For honeybees, to ensure that there is not a greater than 7% effect on colony strength,

including after overwintering, and the level of forager mortality does not breach the tolerable

level, for honeybee colonies located at the edge of treated

ﬁelds

which are exposed to the

90th percentile predicted exposure or less.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

•

For bumblebees, to ensure that there is not a greater than 7% impact on the colony for

bumblebee colonies located at the edge of treated

ﬁelds

which are exposed to the 90th

percentile predicted exposure or less.

For solitary bees, to ensure that there is not a greater than 7% effect on the population of

bees located at the edge of treated

ﬁelds

which are exposed to the 90th percentile predicted

exposure or less.

These SPGs deﬁne the problem formulation for the present assessment.

3.2.

Tier-1

According to EFSA (2013c), depending on the product formulation and the application method

under evaluation, different routes of exposure need to be considered to perform the risk assessment

to bees. The exposure from seed treatments and granular formulations in the

‘treated

crop’ and the

‘succeeding

crop’ scenarios derives from residues in pollen and nectar following translocation from

below ground (seeds or soil). The same route of exposure is considered relevant for the

‘weeds’

scenario in the case of granules application.

Concerning the surrounding area (‘ﬁeld margin’ and

‘adjacent

crop’ scenarios), the most relevant

exposure is due to dust drift at the sowing (treated seeds)/application (granules).

Furthermore, a separate risk assessment for exposure via consumption of contaminated water

should be carried out for honeybees.

Details about the entire Tier-1 risk assessment scheme can be found in EFSA (2013c).

The Tier-1 risk assessment was carried out using default exposure values in accordance with EFSA

(2013c), while the selection of the toxicity endpoints is described below (Section

3.2.1).

Due to the

lack of suitable toxicity data for bumblebees and solitary bees, a surrogate endpoint was extrapolated

from the related honeybee data (assuming the endpoint is a factor of 10 lower). In this case,

throughout the present conclusion, we refer to the Tier-1 as

‘screening

Tier-1’.

3.2.1.

Selection of the endpoints

Several endpoints from laboratory studies were obtained from the data considered in this

conclusion and which had not been considered in previous EU assessments. These newer endpoints

have been considered to amend the previously agreed EU endpoints (EFSA, 2015) provided that the

following criteria were fulﬁlled:

•

The endpoint was considered as relevant for a risk assessment according to EFSA (2013c) and

the GAPs under consideration (e.g. the endpoint type, the test species and the test item were

considered as relevant);

The endpoint was assessed to be

‘Fully

reliable’ or

‘Reliable

with minor restrictions’ during the

appraisal exercise (EFSA, 2018a);

The endpoint, from a study with technical active substance, indicated higher toxicity than the

previously agreed EU endpoint for the technical active substance.

Moreover, for endpoints from formulation studies, the following criteria were considered:

•

the previously agreed EU endpoint originating from a formulation study was replaced only if it

was less relevant (e.g. study with a spray formulation) than the newer formulation endpoint

the previously agreed EU endpoint is a surrogate extrapolated endpoint

Where no new endpoints were available, or the criteria above were not fulﬁlled, the previously

agreed EU endpoints were selected for risk assessment. The data available and

ﬁnal

selection of the

endpoints used for the current risk assessment are given in Section

3.3.

Reﬁnement of the exposure

Within EFSA (2013c), no stepwise approach is offered for higher tier risk assessment. Nevertheless,

among the options listed in the guidance, one possibility is to reﬁne the exposure estimate, i.e. replace the

default values with speciﬁc values coming from higher tier exposure studies. Within the scope of this

conclusion, the risk assessment carried out with reﬁned exposure estimates is referred to as

‘Tier-2’

. A further

reﬁnement option given in EFSA (2013c) is to reﬁne the assessment by the use of higher tier effect studies

performed in the

ﬁeld

or under semiﬁeld conditions (see Section

3.4).

Speciﬁc exposure assessment goals

need to be determined in order to use such effect studies in a reﬁned risk assessment, referred to as

‘Tier-3’

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

3.3.1.

Residues in pollen and nectar

3.3.1.1. Data evaluation and selection

The newly available higher tier studies, reporting information on exposure, were evaluated in line

with the validity criteria set in the literature evaluation protocol (EFSA, 2018a) and the protocols

proposed in Appendix G of EFSA (2013c). The valid data on the residue levels occurring in nectar and

pollen for the exposure scenarios for the treated

ﬁeld

and the succeeding crops in line with these

protocols were collated in a table. Residue determination in available

ﬁeld

studies was assessed for

their reliability both in relation to their

ﬁeld

and laboratory phases. For the

ﬁeld

phase, in order to

reﬁne the exposure, higher tier studies from at least

ﬁve

randomly selected locations in the area of use

of the substance should be conducted. This minimum of

ﬁve

randomly selected locations in the area of

use is prescribed by the guidance, to ensure that an estimate can be made of the distribution of

residues that might really be encountered by the bees. This has the aim of accounting for the different

temporal and spatial variability that occurs. In relation to the laboratory phase, the analytical methods

were examined for their adequacy for determining residues at the low levels required. In some

instances, the size of the samples collected in the

ﬁeld

phase was lower than the sample size for which

the method had been validated, in such cases appropriate correction on the method validated limit of

quantiﬁcation (LOQ) (for the target sample size) was applied, i.e. the LOQ was increased to account

for smaller than ideal sample availability of individual sampling events.

Measured residue levels of pollen and nectar were reported for each type of sampling matrix (i.e.

samples from the plant, from the bee, from the bee via pollen traps, from the comb and from soil). In

general, the sampling scheme which aimed to determine residues in the same matrix (either in plant

matrices or bee matrices) during the

ﬁeld

studies was not exhaustive enough to guarantee that the

time dependence of the residue over the period of interest could be captured. This prevents any

analysis aimed to determine a mathematically rigorous percentile exposure value over time. Therefore,

the maximum observed in the available samples was retained as representative of the exposure in

each particular

ﬁeld

experiment. This does not imply that the overall risk assessment has to be

regarded as overly conservative, since the sampling frequency pattern in the studies does not

guarantee that the actual maximum occurrence was picked up by the maximum measured in the

samples taken. Nevertheless, it is expected that the assessment based on these principles may still be

considered to represent a realistic worst-case exposure for the different substances and uses assessed.

Treated crop scenario

Regarding the

ﬁeld

phases, the directly treated crop needed to be the same crop being assessed.

Appendix R of EFSA (2013c) indicates that extrapolation between the residue values from different

crops is inappropriate when substances are systemic, which is interpreted to relate to seed treatment

uses or when granules are placed with seed at the time of drilling. This is because the different

physiology of different crops, including the time from emergence to

ﬂowering

leads to different

translocation and levels of residues in different crops. When assessing the

ﬁeld

phases of the available

experiments, the most critical issues encountered were cross-contamination from

ﬁelds

in the vicinity

and/or due to historical uses in the same

ﬁeld,

i.e. not resulting from the treated seeds at known

application rates. Only data from studies for which there was sufﬁcient certainty that the residues

observed were resulting from the application being investigated as prescribed in the study design,

were retained for the exposure assessment. The presence or absence of residues measured in control

plots was not part of the decision on retention.

For the exposure assessment, the measured residue values (mg analyte/kg pollen or nectar) were

normalised for the seed loading (mg a.s./seed) resulting in Residue per Unit Dose (RUD) (where the

unit dose is 1 mg a.s./seed) to make the residues independent from the application rate used in the

studies. From one study, sometimes more than one RUD value was calculated and included in the

collation table when more than one trial was conducted within the study. A standalone trial was

deﬁned when one or more of the following factors were different from other trials: type of formulation,

plant species, application rate, test site, period of the trial, pretreatment of the soil and test category

(i.e.

ﬁeld

and semiﬁeld trials, where semiﬁeld means bees used to obtain samples were restricted to

foraging on treated plots).

According to EFSA (2013c), in order to reﬁne the exposure, higher tier studies from at least

ﬁve

randomly selected locations in the area of use of the substance should be conducted. Therefore, a

minimum of

ﬁve

RUD values for pollen and nectar were considered necessary to perform a reﬁned

exposure assessment for each exposure scenario for each use under consideration.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Where the residue detected in a trial was reported to be lower than the LOQ (but greater than the

limit of detection (LOD)), as a worst-case assumption, the residue was considered to be equal to the

LOQ for the RUD calculation. In the cases that no residues were detected, the residue was considered

to be equal to the LOD for the RUD calculation.

According to EFSA (2013c), in order to perform an exposure assessment, it is preferable to use

measured RUD values for pollen and nectar collected from bees (speciﬁc for honeybees, bumblebees

and solitary bees), e.g. using pollen traps attached to honeybee hives or sampling nectar by extracting

the honey stomach from forager bees. Using the RUD values for pollen and nectar directly from the

bees aims to give a better representation of the likely exposure to bees and bee colonies by

accounting for dilution by non-contaminated pollen and nectar. Considering each bee taxon separately

is needed to account for differences in their foraging behaviour that would be expected to mean that

dilution was different between the categories. Alternatively, RUD values for pollen and nectar taken

directly from the plant can be used in the exposure assessment. However, RUD values for plant pollen

and nectar are considered to be an overestimation of the exposure to bees as dilution is not accounted

for. Therefore, if there are a sufﬁcient number of RUD values for bee nectar and/or pollen from

ﬁeld

trials, only these values were used for the exposure assessment. RUD values for pollen and nectar

from bees taken from semiﬁeld studies were considered to be representative of situations where there

was no dilution and therefore were considered together with the RUD values for plant pollen and

nectar. In the cases where RUD values were available on both bee pollen/nectar and plant pollen/

nectar from the same semiﬁeld study, the values for bees only were retained. Where less than

ﬁve

RUD values for bee pollen/nectar were available, these were combined with the RUD values for plant

pollen/nectar and bee pollen/nectar from semiﬁeld, i.e. to obtain sufﬁcient data to perform the

exposure assessment. Figure

summarises the process for selecting the RUD values for the reﬁned

exposure assessment for the treated crop scenario.

≥ 5 RUD values

from field trials with specific

bee taxa?

Refine exposure for specific

bee taxa

Yes

RUD values from plants

(fields trials)?

Yes

Combine all available RUD

values*

Yes

≥ 5 RUD values in total?

and/or

RUD values from plants

(semifields trials)?

and/or

RUD values from any

bee species from

semifields trials?

Cannot refine exposure for

specific bee taxa

*If RUD values from both plants and bees are available from the same semi-field trial, only retain RUDs from bees to avoid double counting.

Figure 1:

A summary of the selection process for RUD values for the reﬁned exposure assessment for

treated crop scenario

Succeeding crop scenario

A different approach from the treated crop scenario was used for the succeeding crop scenario.

This is because the residues in the succeeding crop scenario are less dependent on the physiology of

the treated crop but are instead mainly driven by the active substance concentration in soil and by the

physiology of the successive crop. For this reason, in the residue trials, residues in topsoil/the following

crop root zone before the planting of succeeding plant species attractive to bees had to have been

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Yes

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

measured. For trials to be retained in the assessment, the residues in soil needed to be equivalent to

or higher than that estimated to occur (Predicted Environmental Concentrations (PEC)) in soil from the

uses being assessed. How such PEC were calculated is outlined in Sections

5.1.1.2

and

5.2.1.2.

Maximum residues in pollen and nectar from the retained trial sites were used to estimate exposure in

the risk characterisation, whilst following the approach for the selection of residues directly from bees

(either in open

ﬁeld

or semi

ﬁeld

trials) or via plant sampling as already discussed above for the

treated crop scenario. As the residues trial site selection was based just on measured soil residues, it

was not necessary for the agricultural practice or product formulation type that had been used at any

individual trial to match the uses being assessed. As trials were retained only for sites where measured

residues in soil were equivalent to or higher than the PEC soil, RUD values were not calculated.

In order to reﬁne the exposure, residues from at least

ﬁve

trials are needed.

3.3.1.2. Calculation of reﬁned shortcut values

The residue values selected for the reﬁned exposure assessment for the treated crop scenario and

the succeeding crop scenarios were used to calculate new shortcut values (SVs), which represent

active substance intake per day (adults) or per developmental period (larvae). Such calculation was

performed by means of the SHVAL tool (EFSA, 2013a, 2014a). This R-based tool

ﬁts

theoretical

distributions to the available data (e.g. residue levels, consumption rates, sugar concentration in

nectar) and then it runs Monte Carlo simulations with 1,000 iterations (see EFSA, 2014a for details).

The result of such simulation is a distribution of intake values per day (or per developmental period for

larvae). Finally, the 90th percentile of this distribution is selected as the relevant crop/substance-

speciﬁc SV. Separate simulations were carried out for each caste of each bee group (honeybee,

bumblebee and solitary bees).

No data were available to reﬁne consumption rates or sugar concentration in nectar. Hence, for

these variables, default values as presented in Appendix J of EFSA (2013c) were used in the

simulations.

3.3.1.3. Estimation of the exposure assessment goal

Treated crop scenario

To consider the higher tier effect studies in the context of the risk assessment, the exposure within

those effects studies were compared to the expected exposure for the GAPs under consideration. For

the treated crop scenario, speciﬁc

‘exposure

assessment goals’ were estimated by transforming the

reﬁned shortcut values used in the Tier-2 assessment. To transform the reﬁned shortcut values to an

exposure assessment goal, the shortcut values were multiplied by the seed loading rate (in terms of

mg a.s./seed) for each use.

Figure

presents a general overview of the stepwise approach followed for the reﬁnement of

exposure assessment described under Section

3.3.1

for the treated crop scenario.

SV Forag. Ac.

SV Forag. Ch.

SV Nurse

SV Larva

Study 2

Study 3

Study-specific

App Rate 2

App Rate independent

Combined intake

Residue nectar 3

Residue pollen 3

App Rate 3

RUDn 3

RUDp 3

SV Adult Ac.

SV Adult Ch.

SV Larva

GAP-Specific

…

Study

…

Residue nectar

Residue pollen

…

App Rate

…

RUDn

RUDp

Figure 2:

General overview of the reﬁned exposure assessment for the treated crop scenario

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

SV Adult

SV Larva

Forager Ac.

Forager Ch.

Nurse

Larva

Adult Ac.

Adult Ch.

Larva

Adult

Larva

Exposure assessment goals

Residue nectar 2

Residue pollen 2

RUDn 2

RUDp 2

Study 1

Residue nectar 1

Residue pollen 1

App Rate 1

RUDn 1

RUDp 1

App Rate

GAP

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Succeeding crop scenario

For the succeeding crop scenario, reﬁned SVs were calculated by using actual residue values,

without any further normalisation for the application rate. As such, the reﬁned SVs obtained as

described in Section

3.3.1.2

represent as well the exposure assessment goals.

Figure

presents a general overview of the stepwise approach followed for the reﬁnement of

exposure assessment described under Section

3.3.1

for the succeeding crop scenario.

Study 1

Study 2

Study 3

Study-specific

Residue nectar 1

Residue pollen 1

…

Study

…

Residue nectar

Residue pollen

Combined intake

SV Adult

SV Larva

Residue nectar 2

Residue pollen 2

SV Adult Ac.

SV Adult Ch.

SV Larva

Adult Ac.

Adult Ch.

Larva

Adult

Larva

Figure 3:

General overview of the reﬁned exposure assessment for the succeeding crop scenario

3.3.2.

Dust drift and deposition

According to EFSA (2013c), exposure to dust drift in the

ﬁeld

margin or in adjacent crops is considered

relevant for seed treatment uses and granular formulations. In addition, for granular uses, the exposure to

dust drift is relevant for the treated crop and for the weed scenario if granules are used after emergence.

Field experiments measuring dust deposition to the horizontal ground outside the treated area at the

time of drilling seeds were available. They were considered reliable when: the quality of the treated

seed (in terms of dust content and active substance in the dust) had been measured, the drilling

machinery used was adequately described, the application rate in terms of mass of active substance per

unit treated area was adequately measured, dust deposition at different distances downwind of the

treated area was adequately determined and wind speed and direction measurements were available.

In many of these experiments, dust drift outside the boundary of the treated

ﬁeld

was also measured

using vertical gauze netting. These vertical gauze results were not used further, as it was not clear how

the results reported as g a.s./ha were derived and what they represent. Also, agreed methodology is

not available on how to use or interpret such values that may have utility in estimating exposure to

ﬁeld

margin vegetation or adjacent crops when measured at the individual trial sites.

No studies were available for granules.

3.3.3.

Weeds in the

ﬁeld

According to EFSA (2013c), exposure to

ﬂowering

weeds within the treated

ﬁeld

is considered

relevant for uses applied as granules. Some options to reﬁne the exposure to bees from weeds are

recommended by EFSA (2013c) (e.g. considering the proportion of the treated

ﬁeld

which is covered

ﬂowering

weeds, considering measured residues from crops exposed to dust from granular uses).

However, no such data were available for the granular use on amenity turf.

3.3.4.

Residues in water sources

Where the crops being assessed had experimental data where the seed loading rate was measured

and residue levels were adequately determined in sampled guttation

ﬂuid

exuded from the treated

plants, EFSA collated the residues values.

In accordance with the recommendation of EFSA (2013c), measured residues in the guttation

ﬂuids

exuded from the treated plant were considered for reﬁning the assessment related to this route of

exposure relevant for honeybees. EFSA (2013c) speciﬁes that the guttation concentration used in the

risk assessment needs to cover the 90th percentile in guttation

ﬂuid

for the crop of concern

(considering location, growth stage and environmental conditions). For seed treatments and granules

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Exposure assessment goals

Residue nectar 2

Residue pollen 2

SV Forag. Ac.

SV Forag. Ch.

SV Nurse

SV Larva

Forag. Ac.

Forag. Ch.

Nurse

Larva

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

buried with seeds, it is proposed to reﬁne the exposure estimate by conducting (at least)

ﬁve ﬁeld

studies and to measure the concentrations in guttation water.

Therefore, following EFSA (2013c), a reﬁnement of the exposure can only be performed if at least

ﬁve ﬁeld

studies are available for the same crop considering all available values added to the table of

residues values. Residues from

ﬁve ﬁeld

studies were not available for any of the crops being assessed

here. However, for consistency with the conﬁrmatory data conclusion for imidacloprid (EFSA, 2016b),

Tier-2 risk assessments have been completed using the highest guttation

ﬂuid

residues from the

available

ﬁeld

investigations for the crops where these were available.

No exposure reﬁnement was necessary for assessing residues in puddles as using the Tier-1

exposure assessment indicated low risk (see Section

5.1.3.2).

3.4.

3.4.1.

Reﬁnement with higher tier experiments

Building up the lines of evidence

Another approach offered by EFSA (2013c) to reﬁne the risk assessment is to perform higher tier

effect experiments. These experiments are normally carried out under

ﬁeld

or semiﬁelds conditions,

and aim at a higher environmental realism when compared to standard laboratory test.

These experiments present a wide variety of set-ups, designs and investigated endpoints.

Therefore, a weight of evidence (WoE) scheme has been developed to integrate the relevant

information from all available experiments. In order to perform a WoE risk assessment, it is

ﬁrst

necessary to set the problem formulation and then identify lines of evidence which address the

problem. In the case of honeybee, bumblebee and solitary bee, risk assessments performed in

accordance with EFSA (2013c), the problem formulation is already deﬁned by the SPGs.

Within the WoE, it was considered that each

‘line

of evidence’ corresponds to the whole set of

homogeneous endpoints measured in all available experiments. An endpoint in this context is deﬁned

as a parameter which could be informative of a potential effect caused by an exposure to an active

substance (and its metabolites).

Within each experiment, the endpoint is identiﬁed by four dimensions:

•

The

magnitude

of the observed deviation from the control. For endpoints measured as time

series, the extremes of such deviation were recorded in both directions, together with a mean

deviation. In case of such endpoints like forager mortality, this dimension should also account for

the duration of a consistent deviation (e.g. increase of X% in forager mortality observed for Y

consecutive days). Deviations in both directions were classiﬁed as: no deviation, negligible, small,

medium and large deviation from the control. For this classiﬁcation, the scales presented in

Table

were used. These scales were adapted from Appendix B of EFSA (2013c) (Protection

goals), except the scale for homing success, where the categories were arbitrarily chosen. An

example for using these scales: if the average colony strength in a honeybee study at an

observation time was 6% less in the treated group compared to the control, this was classiﬁed as

a negative negligible deviation. If, at another observation day, the colony strength in the treated

group was 16% more than in the control, this was classiﬁed as a positive medium deviation. It

has to be noted that pending on the availability of the data on the relevant endpoints (i.e.

reported details), the deviation from the control was either calculated or only estimated (e.g.

when only graphical presentation was available for the endpoint).

Scale of deviation from the control used for the weight of evidence exercise

All endpoints except mortality Forager mortality and mortality

Homing success

and homing success

in front of the hive

0% to

7% to

15%

15% to

35%

≥

35%

See examples in Table B1 of

Appendix B of EFSA (2013c)

See examples in Table B1 of

Appendix B of EFSA (2013c)

See examples in Table B1 of

Appendix B of EFSA (2013c)

See examples in Table B1 of

Appendix B of EFSA (2013c)

0% to

10%

10% to

20%

20% to

50%

≥

50%

Table 3:

Deviation class

No deviation

Negligible

Small

Medium

Large

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

•

The

reliability

of the endpoint: this was established on the basis of the appraisal exercise

(EFSA, 2018a) and giving a score to each endpoint from 0 (not reliable) to 3 (fully reliable).

The reliability was used to weigh the results obtained in different experiments, and to

estimate, together with the level of consistency of the results, the level of certainty associated

with the line of evidence.

The

level of exposure:

this information is necessary to check where the level of exposure in

the experiment stands compared to the exposure assessment goal(s). Furthermore, this

information can be used to check whether a sort of exposure–response relationship can be

identiﬁed. For oral exposure to residues in pollen and nectar, residue intake (RI) values were

calculated for each caste of bee using the mean residue value on nectar and/or pollen

obtained in the effects study. A sugar content of 15% was assumed for nectar for honeybees

and bumblebees (EFSA, 2013c). In case of colony feeder studies, the sugar content of the

sugar solution speciﬁed in the study was used. If this was not available, then a sugar content

of 50% was assumed. The daily consumption values, for pollen and nectar, for each bee caste

were taken from EFSA (2013c). Where a range of consumption values were available, a range

of RI values was obtained.

The length of exposure:

this is deﬁned as the time period in which there could have been

exposure to residues of the active substance. It is noted, that the

‘length

of exposure’ referred

to in this conclusion does not account for the subsequent consumption of food stores within

colonies/nests. In

ﬁeld

and semiﬁeld studies, this corresponds to the time period the bees

could be exposed to the crop during

ﬂowering.

For colony-feeder studies, the length of

exposure is deﬁned by the time period during which the spiked sugar solution or pollen was

given to the bees. This information is needed to check whether the length of exposure is

realistic to that expected for the GAPs under consideration.

In order to visually illustrate these four dimensions of the endpoints and in order to help the

interpretation of a

‘line

of evidence’, graphical representations were prepared. A graphical

representation of a

‘dummy’

example (invented example for illustrative purpose) is included in Figure

with an explanation of each element of the

ﬁgure.

Figure 4:

Lines of evidence for a

‘dummy’

example: colony strengths of honeybees

The

ﬁgures

include the following information:

•

Each row in the

ﬁgure

represents a higher tier effects study. The numbers given on the left

hand side are the reference identiﬁcation codes of the individual studies which can be traced

back to the reliability assessment provided in the Technical Report on evaluation of data and

EFSA Journal 2018;16(2):5178

www.efsa.europa.eu/efsajournal

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

related Appendices from D to O (EFSA, 2018a). In Appendix

of this conclusion, an overview

is given of all reference identiﬁcation codes of the individual studies cited in the conclusion. A

summary of the reference identiﬁcation codes according to the different assessment streams

identiﬁed during the evaluation of the data is summarised below. Note that several studies

investigated a number of exposure levels, and therefore, these studies may be listed more

than once (e.g. the highest exposure level that caused no or small deviation from the control

and the lowest exposure level with apparent deviation from the control).

Assessment stream

All* (with interaction)

All

(no interaction)

Clothianidin

Imidacloprid

Thiamethoxam

Clothianidin * Imidacloprid (with interaction)

Clothianidin

Imidacloprid (no interaction)

Clothianidin * Thiamethoxam (with interaction)

Clothianidin

Thiamethoxam (no interaction)

Imidacloprid * Thiamethoxam (with interaction)

Imidacloprid

Thiamethoxam (no interaction)

Not substance-speciﬁc

Reference identiﬁcation code

All*

All+

C*I.

C+I.

C*T.

C+T.

I*T.

I+T.

•

The left hand panel relates to the biological observations whilst the centre and right hand

panels relate to exposure.

For each experiment, the black solid horizontal line represents the range of the observed

deviations from the control (i.e. both negative and positive deviations). The magnitude of

these deviations is categorised as negligible, small, medium or large. When this is not

indicated, the endpoint was measured only once during the study or it was measured multiple

times, but insufﬁcient details were reported to evaluate the variability of the endpoint in time

(e.g. only averages for the entire study duration were reported)

The position of the blue circle gives an estimation of the overall deviation (mean) from the

control for the entire duration of the study or during the year of use for those studies which

extend over the winter.

The size of the blue circle is an indication of the reliability score of the speciﬁc endpoint. A

‘fully

reliable’ endpoint gives a large circle, an endpoint which was

‘reliable

with minor

restrictions’ gives a medium-sized circle and an endpoint which was

‘reliable

with major

restrictions’ gives a small circle.

For transparency, the experiments giving endpoints which were assessed as

‘not

reliable’ have

also been included in the

ﬁgure,

but the overall deviation is represented by an X.

To help interpret the

ﬁgure,

vertical dotted lines have been added to indicate the mean overall

deviation, the mean negative deviation (if it could be calculated) and the mean positive

deviation (if it could be calculated) across all of the reliable experiments (weighted for the

reliability score of the endpoint).

In the centre panel of the

ﬁgure,

a representation of the estimated level of exposure achieved

in the experiment is given for each bee caste (i.e. ng a.s./bee per day for adult bees and ng

a.s./larvae per development period for larvae). When this is not indicated, reliable information

on the exposure level achieved in the study was not available.

The vertical lines in the centre panel of the

ﬁgure

represent the exposure assessment goal, for

each bee caste, for the GAP under consideration (see Section

3.3.1.3).

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

•

On the right hand side of the

ﬁgure,

the length of exposure (in days) in the experiment is

represented by the red bars. The vertical purple column is the expected range of the

ﬂowering

period for the crop under consideration

In order to conclude that the observed deviations are actual representations of a true effect caused

by the exposure to the active substance (and its metabolites), several aspects were considered for

each line of evidence.

•

The presence/absence of a general trend, giving more weight to results with higher reliability.

The level of consistency among experiments (similar results, exposure–response relationship

etc.)

The level of precision offered by the available experiments (width of the effect size ranges)

In principle, each line of evidence should provide a piece of information characterised by a certain

degree of strength (consistency), precision (degree of variability) and reliability.

Furthermore, in order to use the available information to conclude on the risk assessment, it is

pivotal to check the level of the exposure in the effect experiments relative to the speciﬁc exposure

assessment goals and to check whether the length of potential exposure in the effect study is within

the realistic

ﬂowering

period for the crop (or succeeding crop) under consideration.

3.4.2.

Integrating the lines of evidence

In accordance to EFSA (2013c) and the newly available EFSA guidance document on the WoE

(EFSA Scientiﬁc Committee, 2017), the overall process should account for the relevance of the single

lines of evidence, before performing any integration.

In the scheme, there are endpoints (lines of evidence) that are directly linked to the protection

goals and have the potential to provide a straightforward response to the main issue reported in the

problem formulation (see Section

3.1).

For this reason, they are considered to be the higher class of

endpoints in terms of relevance (Class 1). These are colony/population size (valid for all species);

forager mortality (honeybees only) and all endpoints related to the reproductive output (bumblebees

and solitary bees)

Other endpoints have a rather clear conceptual link with one of the previous two (e.g. brood/

cocoon production can clearly inﬂuence colony/population strength), even if this link cannot be

explicitly quantiﬁed. These endpoints (lines of evidence) belong to Class 2 for relevance.

Other endpoints, on the contrary, may play a role in the colony/population health, but such link is

not immediate in conceptual term (e.g. average duration of foraging trips). These endpoints (lines of

evidence) belong to Class 3.

Finally, there are endpoints that do not offer any explicit link with the protection goals (e.g.

measurement of enzymatic levels at subindividual level). These endpoints are considered not relevant

(Class 4) for addressing the protection goal according to EFSA (2013c).

A proper summary of all endpoints considered in this assessment and a detailed description of their

relationship with the SPGs are available in Appendix

A less detailed summary is reported in Table

A crop-speciﬁc time window for a reasonable

ﬂowering

period was estimating by considering: (i) feedback from Member

States; (ii) assumptions within the FOCUS scenarios; (iii) collating information reported in the present data set. For winter

oilseed rape, a reasonable

ﬂowering

period was estimated to last between 10 and 42 days. The selected maximum is longer

than the assumptions contained in the FOCUS scenarios (maximum 28 days). During the expert meetings, some experts

pointed out that 6 weeks of

ﬂowering

may happen, but only in exceptional years. Nevertheless, within the present data set,

ﬂowering

between 35 and 43 days was recorded in several experiments carried out in France, the UK, Germany and Hungary.

Similarly, data were collected for spring oilseed rape and maize. The succeeding crop scenario is not represented by any

speciﬁc crop. Therefore, on the basis of all of the available information on the

ﬂowering

period of agricultural crops, it was

decided that a reasonable

ﬂowering

period would range from 5 to 42 days.

EFSA Journal 2018;16(2):5178

www.efsa.europa.eu/efsajournal

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Table 4:

Bee group

Honeybees

A summary of the endpoint types and the related relevance class assigned within the

scope of the present risk assessment

Relevance class

Family of endpoint

Colony strength

Forager mortality

Overwintering assessment

General mortality of individuals

Brood production

Homing success

(a)

Behavioural endpoints

Comb building

Weight of the hive

Disease

Food storage

Queen

Behaviour inﬂuencing exposure

Subindividual mass

Suborganism endpoints

Bumblebees

Thermoregulation capacity

Reproductive output

Colony strength

Indirect reproduction

Behaviour

Weight of the nest (colony)

Food storage

General mortality

Individual mass

Homing success

Solitary bees

Behaviour inﬂuencing exposure

Reproductive output

(b)

Indirect reproductive output

Behaviour

General mortality

(a): For the purposes of this conclusion, the endpoint

‘homing

success’ is deﬁned as the proportion of bees returning to the hive/

colony after they were captured and subsequently released at a distance from the hive/colony.

(b): The number of solitary bee offspring emerging after winter was considered to represent the accumulation of several

preceding endpoints related to reproductive success (e.g. number of completed nests, tubes with brood, cocoon

production). Therefore, the weight of evidence focussed primarily on the number of offspring emerging after the winter.

In order to account for this hierarchical structure within the current risk assessment scheme, a

stepwise procedure was followed.

As already mentioned, the

ﬁrst

step was focussing on endpoints belonging to Class 1. If the

available data are sufﬁcient to provide a conclusive answer to the main risk assessment question, the

assessment could stop. If, on the contrary, the available information is not sufﬁcient and/or

appropriate to provide a conclusive answer, the WoE will be extended to other levels of relevance, in

order to get a more comprehensive picture of the available data.

If the evidence in the

ﬁrst

two levels of relevance (Classes 1 and 2) is not sufﬁcient/appropriate to reach a

conclusion, it is considered unlikely that less relevant endpoints will help achieving a conclusive assessment.

4.1.

Outcome of the risk assessment: toxicity endpoints

Standard endpoints

In the data set considered, there were several laboratory studies available for assessing the effects

of imidacloprid, or formulated products containing imidacloprid, on honeybees, bumblebees and

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

solitary bees. Following the selection procedure given in Section

3.2.1,

it was considered whether any

of the newly available data should replace the previously EU agreed endpoints (EFSA, 2016a,b) and be

used for the Tier-1 and Tier-2 risk assessments.

Three acute contact toxicity endpoints for honeybees performed with the technical active substance

were available and resulted in lethal dose (LD

) values which were slightly lower than the previously

agreed endpoint. Furthermore, there were two additional studies performed with formulated products

containing imidacloprid. The lowest of the available endpoints, from study All+.1084 with the technical

active substance, was selected for risk assessment (LD

0.0251

lg/bee;

3.2 times lower than the

previously agreed EU endpoint). In this experiment, the LD

was only reported for an observation

period of 120 h, which is longer than the time window recommended in the OECD 214. Nevertheless,

both validity criteria reported in OECD 214 were considered respected: the mortality in the control was

still 5.8% after 120 h, and the toxicity of other tested substances (i.e.

k-cyhalothrin,

deltamethrin,

esfenvalerate) was in the expected range, providing an indication that sensitivity of the system was

appropriate, in lack of a formal positive control. Therefore, the LD

was considered suitable for

being used in the risk assessment. This issue was discussed and agreed during the expert meeting

(October 2017).

There were four reliable endpoints investigating the acute oral toxicity to honeybees performed with

the technical active substance and one endpoint was available from a study performed with a

formulated product. The endpoints from these studies did not indicate a higher toxicity than the

previously agreed acute oral honeybee endpoint, and therefore, this value was retained for risk

assessment. No reliable laboratory data were available to derive a chronic oral lethal dietary dose

(LDD

) or a no observed effect level (NOEL) value for larval development of honeybees. Therefore, the

previously EU agreed endpoints were retained for risk assessments. No NOEL value could be derived

from the available data for the development of hypopharyngeal glands (HPG) in honeybees (such

endpoint is neither available from previous assessments). It is noted, however, that an approximate

lowest observed effect level (LOEL) of 0.243 ng/bee per 10 days was estimated from one of the studies

(I.545). No data were available to assess whether the use of imidacloprid results in accumulative

effects in honeybees.

There was an acute contact toxicity endpoint (LD

value) for bumblebees performed with the

technical active substance and another one performed with a formulated product. The endpoints from

these studies did not indicate a higher toxicity than the previously agreed acute contact bumblebee

endpoint, and therefore, this value was retained for risk assessment. There was a single study giving

an acute oral LD

value for bumblebees for the technical active substance. Again, this endpoint did

not indicate a higher toxicity than the previously agreed acute oral bumblebee endpoint, and

therefore, this value was retained for risk assessment. No reliable laboratory data were available to

derive a chronic oral LDD

or a NOEL value for larval development of bumblebees. These values were

neither available from previous assessments.

There were no reliable and relevant toxicity data available for solitary bees nor were any data

available from previous assessments.

In accordance with EFSA (EFSA, 2013c), where data are missing, surrogate endpoints can be

calculated using toxicity data for honeybees divided by 10. Surrogate endpoints were therefore

calculated for the acute contact and oral toxicity to solitary bees and for the chronic oral toxicity to

bumblebees and solitary bees. As the available honeybee larvae endpoint had some shortcomings, it

was previously concluded that this endpoint should only be considered as

‘provisional’

(EFSA, 2016b).

Therefore, it was not considered appropriate to use this endpoint to derive surrogate endpoints for

bumblebee and solitary bee larvae.

This selection procedure resulted in the change of the surrogate acute contact endpoint for solitary

bees. As a consequence of no change in other honeybee endpoints, the other surrogate endpoints for

solitary bees and the surrogate adult chronic endpoint for bumblebee are the same as the previously

EU agreed surrogate endpoints.

On the basis of the above considerations, Table

summarises the toxicity endpoints selected for

the Tier-1 and Tier-2 risk assessment.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Table 5:

Toxicity endpoints selected for lower tier risk assessments

Endpoint

(lg a.s./bee)

10-day LDD

(lg a.s./bee per day)

NOEL (lg a.s./larva per

developmental period)

Honeybee

0.0251 (120 h)

0.0037 (48 h)

0.00282

(a)

0.00528

as provisional

(b)

Bumblebee

0.218 (96 h)

0.038 (96 h)

0.000282

(c)

No endpoint available

or extrapolated

Solitary bee

0.00251

(c)

0.00037

(c)

0.000282

(c)

No endpoint available

or extrapolated

Risk

assessment

type

Acute contact

Acute oral

Chronic oral

Larval

: lethal dose, median; LDD

: lethal dietary dose, median;

(a): Endpoint set at the highest concentration tested.

(b): Endpoint determined at 7 days but only 3-day exposure during the study. Endpoint is the highest dose tested. Endpoint is

based on nominal amount of food offered to the larvae.

(c): Extrapolated from the endpoint for honeybee by using a factor of 10.

Note: from the previously EU agreed endpoints, only the acute contact endpoint for honeybees and the acute contact endpoint

for solitary bees were changed.

4.2.

Additional sublethal laboratory data

Several laboratory experiments testing sublethal effects of imidacloprid on bees were available in

the data set. The endpoints investigated encompassed a wide variety of sublethal effects. The effects

investigated in the available data set are listed in Table

Table 6:

Organism

Apis cerana

Apis mellifera

Sublethal endpoints for honeybees, bumblebees and solitary bees investigated in

laboratory studies of the available data set for imidacloprid

Effect

Proboscis extension reﬂex (PER)

–

learning capacity

AChE activity

Kenyon cell depolarisation

Expression of genes

Deformed wing virus replication

Apoptosis of neurons

Hemolymph collection

Total haemocyte count

Encapsulation response

Antimicrobial activity of the haemolymph

Body weight

Food attractiveness (food preference)

Intoxication symptoms

Larvae development

Locomotion activity

Number of Nosema spores

Parameter for

ﬂight

capacity

Parameters for individual immunity

Parameters for social immunity

Proboscis extension reﬂex (PER)

–

learning capacity

Protein content of bee head

Sperm viability

Sucrose consumption

Time spent for feeding (potential feeding)

Time spent for interacting

Wing length

Accumulation in the brain

Cytotoxicity of neurons

Mitochondrial depolarisation

Food attractiveness

Locomotion activity

Sucrose consumption

Bombus terrestris

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Organism

Melipona quadrifasciata

Effect

Locomotion activity

Parameter for

ﬂight

capacity

Respiration

Larvae development

Osmia lignaria

Megachile rotundata

AChE: acetylcholinesterase.

In addition, an endpoint from a Proboscis Extension Reﬂex (PER) tests was available for 5-OH-

Imidacloprid on honeybees.

For most of the cases, only a single endpoint was available. However, a number of the available

experiments focussed on the so-called PER, which investigated how exposure to imidacloprid impairs

the responsiveness to a stimulus and affect memory performances. This response was studied with

different test design, such as acute contact, acute oral or chronic oral. Moreover, some of these

experiments were conducted with

Apis mellifera,

while others with

Apis cerana.

In a number of those

experiments, a no observed effect dose (NOED) (acute) or a no observed effect concentration (NOEC)

(chronic) value could be established, while in others, no effects (i.e. statistically different response

from the control) was seen at the tested doses/concentrations. The endpoints of the available acute

and chronic oral tests were rather inconsistent. However, from the results of the acute contact tests, it

may be concluded that the NOED would be between 0.5 and 1.25 ng/bee.

Another sublethal effect that was studied in several experiments was impairment of locomotion.

Again this response was studied with different test designs, such as acute contact, acute oral,

subacute oral or chronic oral. No consistent response was seen between these studies.

In several studies, it was discussed that the investigated sublethal effects at individual or

subindividual level may result in a colony/population level effect. It is acknowledged that an evident

linkage (direct or indirect) between certain sublethal endpoints and colony/population level effects

might exist. However, no appropriate information was available to establish or further describe these

links. Therefore, these endpoints could not be linked to the protection goal and they were not

considered further in the risk assessments.

5.1.

5.1.1.

Outcome of the risk assessment

Risk assessments for seed treatment products

Risk via systemic translocation in plants

–

residues in nectar and pollen

(treated crop scenario and succeeding crop scenario)

5.1.1.1. Tier-1 risk assessment

The Tier-1 risk assessment for the representative GAPs were performed using the EFSA’s BeeTool

(v.3.) (Appendix Y of EFSA (2013c)) for honeybees and bumblebees, where suitable toxicity data were

available. A screening Tier-1 assessment was carried out for solitary bees and for the chronic adult

assessment for bumblebees as only surrogate endpoints were available. Since no toxicity data was

available for HPG development or for the larvae toxicity for non-Apis bees, no assessment was

performed for these scenarios. For potato, the application rate expressed as mass/tuber was not

available. The outcome of these calculations is summarised in Table

High risk or low risk not

demonstrated is indicated for all cases since one or more combinations (categories of acute, chronic

and larva combined with the treated crop scenario and following crop scenarios) indicated a high risk.

The detailed results are included in Appendix

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Table 7:

Summary of the outcome of Tier-1 and screening Tier-1 risk assessment

Honeybee

Bumblebee

‘Low’

application

rate

High risk

‘High’

application

rate

High risk

Solitary bee

‘Low’

application

rate

‘High’

application

rate

Use

‘Low’

application

rate

‘High’

application

rate

High risk

Spring and winter

cereals, cotton,

endive, lettuce,

brassicas (ﬂowering,

head, leafy), maize,

potato, spring and

winter rape, sugar

and fodder beet

Low risk not Low risk not

demonstrated demonstrated

It is noted that endive, lettuce, brassicas, sugar and fodder beet may be harvested before they

ﬂower.

In that case, the treated crop scenario is not relevant for those crops (low risk for that

scenario). Nevertheless, the following crop scenario indicated a high risk for those crops. For the uses

when crops are transplanted and grown in permanent greenhouses (but not transplanted in the

ﬁeld

or non-permanent greenhouse), none of the scenarios are relevant (low risk for the treated crop and

for the following crop scenario). It is also noted that for potato only, the following crop scenario was

considered as no calculations were possible for the treated crop scenario.

As presented above, the Tier-1 dietary risk assessment for the treated crop and succeeding crop

scenario for all seed treatment uses under consideration indicated a high risk to honeybees and

bumblebees. By the screening assessment for solitary bees, a low risk was not demonstrated. No risk

assessment could be performed for honeybee HPG development or bumblebee and solitary bee larvae.

It is noted that for endive, lettuce and brassicas, a low risk could be concluded for uses when the

treated crop is grown continuously (including

ﬂowering)

in permanent greenhouse.

5.1.1.2. Exposure assessment for the treated and succeeding crop scenarios

Treated crop scenario

Several reliable studies giving measured residue values in nectar and pollen originating from crops

grown from seeds treated with imidacloprid were available. For the seed treatment uses under

consideration, sufﬁcient relevant data were available for oilseed rape only.

According to EFSA (2013c) in order to perform an exposure assessment, it is preferable to have

measured residues in pollen and nectar collected from bees. Taking the pollen and nectar directly from

the bees aims to give a better representation of the likely exposure to bees and bee colonies by

accounting for dilution by non-contaminated pollen and nectar. As regards pollen, residues from pollen

traps attached to the hive entrance are also considered. Alternatively, residues of pollen and nectar

taken directly from the plant can be used in the exposure assessment. However, residues in plant

pollen and nectar are considered to be an overestimation of the exposure to bees as dilution is not

accounted for. Residue data in pollen and nectar from bees taken from semiﬁeld studies were

considered to be representative of situations where there was no dilution and therefore were

combined with the residues from

ﬁeld

studies. In case of nectar, the available residue values from bees

were combined with residue values taken directly from the plant, since a sufﬁcient number of

measured residue values taken from the bees was not available. As regards pollen, all the available

values originated either from bees or from pollen traps. In the cases where data were available from

both bee pollen and from pollen trap, the values from bees only were taken. The residue values from

bees (including when pollen trap was used) were all from pollen and nectar collected from honeybees.

Therefore, exposure reﬁnements were performed only for honeybees. The available data set included

two RUD values for nectar and two RUD values for pollen from studies from winter oilseed rape. As

regards spring oilseed rape, four RUD values for nectar and six RUD values for pollen were available.

The separate data for winter and spring oilseed rape were considered insufﬁcient for separate risk

assessments for winter and spring oilseed rape according to EFSA (2013c). Therefore, the two data

sets were combined resulting in a total of six RUD values for nectar and eight RUD values for pollen.

The actual residue values used for establishing the exposure assessment goals are reported in

Appendix

It is acknowledged that it is likely that RUD values for spring oilseed rape are conservative

relative to those for winter oilseed rape. However, the RUD values for winter oilseed rape do not

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

necessarily cover the residue levels for spring oilseed rape as from the available data it appeared that

residues in spring oilseed rape were higher. Therefore, the reﬁned risk assessment was performed only

for winter oilseed rape. It is noted that in several cases (four of the six RUD values for nectar and six

of the eight RUD values for pollen), insufﬁcient information was available to work out the application

rate used in the study in terms of seed loading (mg/seed). However, assuming a thousand grain

weight (TGW), it was possible. For these cases, a TGW of 5 g was considered.

The measured residue values were normalised for the seed loading (mg a.s./seed) to give RUD

values (mg a.s./kg pollen or mg a.s./kg nectar).

Such RUD values were log-transformed before being used as input for the EFSA SHVAL tool (EFSA,

2014a). A 90th percentile SV for exposure, in terms of RI, is given as output of this tool. Simulations

were run for each bee species and each caste. Tier-1 data for pollen and nectar consumption and

sugar content in nectar were assumed. To transform the reﬁned shortcut values to an exposure

assessment goal, the shortcut values are multiplied by the seed loading rate (in terms of mg a.s./kg

seed) for each use listed in the GAP (Appendix

A).

The resulting reﬁned shortcut values and exposure

assessment goals for the GAP for winter oilseed rape are presented in Table

Table 8:

Reﬁned shortcut values and exposure assessment goals for honeybees for winter oilseed

rape applied as seed treatment

Reﬁned shortcut value

(lg/bee per day or

lg/larva

per developmental period)

Acute forager

Chronic forager

Nurse

Larva

0.21

0.16

0.09

0.12

Exposure assessment goal for 0.01 mg

a.s./seed (ng/bee per day or ng/larva

per developmental period)

2.1

1.6

0.9

1.2

Succeeding crop scenario

As imidacloprid has the potential to accumulate in soil from use over successive years, this has to be

accounted for in the succeeding crop assessment. This potential for accumulation was accounted for by

calculating plateau PEC in soil following FOCUS (1997, 2006). Accordingly, the longest

ﬁeld

dissipation

study single

ﬁrst-order

of 288 days (EFSA, 2008) was used to calculate an accumulated

concentration in soil assuming a soil mixing (tillage) depth of 20 cm and soil bulk density of 1.5 g/cm

. To

cover the representative uses under assessment, annual soil application rates of 175, 139.6 and 34.5 g/ha

were used for the calculations. The value of 175 g/ha is from continuous cropping of cotton; 139.6 g/ha

((180

126)/4) represents a rotation of the use on potatoes (highest dose rate 180 g/ha) followed

by 3 years of use on winter cereals (highest dose rate 126 g/ha). The value of 34.5 g/ha ((9

43)/4)

represents a rotation of the use on oilseed rape (lowest dose rate 9 g/ha) followed by 3 years of use on

winter cereals (lowest dose rate 43 g/ha). Accumulated soil exposure concentrations from the other uses

assessed would fall between the lower two of these three annual application rates. The accumulated PECs

in soil resulting from the assumptions outlined above are 0.0415 mg/kg, 0.033 mg/kg and 0.008 mg/kg,

respectively.

From the seed treatment uses under consideration, relevant data for the exposure assessment for

the succeeding crop were available when the succeeding crops were phacelia, winter oilseed rape,

mustard (semiﬁeld study designs) and maize (ﬁeld studies, pollen collected from plants).

In these studies, the potential exposure of bees to residues in succeeding crops was investigated

based on two different approaches. In a series of studies concentrations of imidacloprid, 5-OH-

imidacloprid and imidacloprid oleﬁn in nectar and pollen from these crops/plants attractive for bees

were measured under conditions of

‘forced’

soil residues, i.e. succeeding crops grown on soils treated

over their entire area with imidacloprid to obtain a theoretical plateau concentration of imidacloprid in

soil. In other

ﬁeld

studies, the untreated succeeding crops were sown in soil with a history of several

years of use of imidacloprid, and thus exposed to so called

‘naturally

aged’ residues in the soil, the

residues analysed were the same as those in the forced studies. Apart from one new

‘forced’

study,

the data set available for this conclusion was identical to the conﬁrmatory data package evaluated in

the EFSA conclusion in 2016 (EFSA, 2016b). In the previous peer review assessment, it was concluded

(Pesticides Peer Review Meeting 145) that the

‘forced

exposure’ is less representative of the exposure

situation under

ﬁeld

conditions, where the imidacloprid residues in soil had already undergone natural

ageing processes, that make them potentially less available for plant uptake. Therefore, studies with

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

‘forced’

soil residues of imidacloprid in soil were not considered further and the highest residue values

measured for pollen and nectar from the

‘natural

aged’ soil residue studies were used to reﬁne the risk

assessment. This approach was also used in this conclusion.

In the

‘natural

aged’ soil residue experiments, where the highest residue values for pollen (2.5

imidacloprid/kg maize, 1.5

imidacloprid/kg phacelia, 1.3

imidacloprid/kg winter oilseed rape) and

nectar (3.5

imidacloprid/kg phacelia, 0.7

imidacloprid/kg winter oilseed rape) were detected, the

measured soil residues were from 0.035 to 0.059 mg/kg, higher than the calculated accumulated soil

PEC of 0.033 mg/kg or 0.008 mg/kg for the potato and the oilseed rape seed treatment uses,

respectively. These soil residues were comparable to or higher than the calculated accumulated PEC of

0.0415 mg/kg for the cotton seed treatment use (see details of these PEC calculations above). These

PECplateau values cover all the GAPs considered in this conclusion. It is noted that soil residues are

independent of the GAPs for the primary crop(s) and can be used for any GAP, provided that the crop

rotation and the ageing processes are leading to soil residue levels comparable to the calculated

PECplateau values. In these

‘natural

aged’ experiments, 5-OH-imidacloprid residues were always

lg/kg

(LOQ) in pollen (phacelia) and

0.3

lg/kg

(LOD) in nectar. Imidacloprid oleﬁn residues

were always

0.3

lg/kg

(LOD) in both pollen and nectar.

The above residue values of imidacloprid in pollen and nectar for the succeeding crop were inserted

in the EFSA SHVAL tool (EFSA, 2014a). As the reﬁned shortcut values are independent of the GAP,

they also represent the exposure assessment goal for the succeeding crop scenario. The calculated

values are presented in Table

It is noted that these values are in line with the calculated SVs for the

conﬁrmatory data for imidacloprid (EFSA, 2016b).

Table 9:

Reﬁned shortcut values and exposure assessment goals for honeybees and bumblebees

for the succeeding crop scenario

Reﬁned Shortcut value

(lg/bee per day or

lg/larva

per

developmental period)

Honeybee forager acute

Honeybee forager chronic

Honeybee nurse

Honeybee larva

Bumblebee adult acute

Bumblebee adult chronic

Bumblebee larva

0.00244

0.00189

0.00101

0.00139

0.00312

0.00269

0.00065

Exposure assessment goal

(ng/bee per day or ng/larva per

developmental period)

2.44

1.89

1.01

1.39

3.12

2.69

6.5

(a)

(a): The shortcut value for bumblebee larva refers only to 1 day. In order to cover the entire developmental period by the

exposure assessment goal, the value was multiplied by 10, as recommended by EFSA (2013c)

5.1.1.3. Tier-2 risk assessment

Treated crop scenario

As explained above in 5.1.1.2, sufﬁcient data on residues were available only for the representative

GAP for winter oilseed rape. Tier-2 risk assessments performed by using the EFSA’s BeeTool (v.3.)

(Appendix Y of EFSA (2013c)) for honeybees. In these calculations, the Tier-1 shortcut values were

replaced by the reﬁned Tier-2 shortcut values. The outcomes of these calculations are summarised in

Table

10.

Low risk could not be demonstrated for honeybees for the winter oilseed rape use.

Table 10:

Category

Acute adult

Chronic adult

Larva

ETR: Exposure toxicity ratio.

Tier-2 risk assessment for honeybees for winter oilseed rape applied as seed treatments

Reﬁned shortcut value

(lg/bee per day or

lg/larva

per

developmental period)

0.21

0.16

0.12

ETR for the seed

treatment rate of

0.01 mg a.s./seed

0.568

0.567

0.227

Trigger

0.2

0.03

0.2

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Succeeding crop scenario

The Tier-2 risk assessments for the succeeding crop scenarios are reported in EFSA (2016b). Those

risk assessments indicated a high risk or a high risk could not be excluded for honeybees, bumblebees

and solitary bees. The PECplateau for the GAPs considered in this conclusion are lower or comparable

to the measured soil residues considered in succeeding crop trials in EFSA (2016b). Therefore, the risk

assessments presented in EFSA (2016b) cover the risk resulting from the GAPs of this conclusion.

5.1.1.4. Tier-3 risk assessment (weight of evidence)

As discussed in Section

3.4,

a WoE approach was developed to utilise the information from the

diverse range of higher tier effect experiments that were available.

The WoE risk assessment could only be performed for the GAPs for which an exposure assessment

goal was calculated (Section

5.1.1.2).

For the use of imidacloprid as a seed treatment, exposure

assessment goals could be determined only for the treated crop scenario for winter oilseed rape

foraged by honeybees. In addition, an exposure assessment goal has also been determined for the

succeeding crop scenario for all uses of imidacloprid under consideration. These assessments are

relevant for all bee species. Consequently, the WoE approach was applied to these combinations only.

As previously discussed, the WoE exercise has two fundamental steps:

ﬁrstly,

the identiﬁcation/

consideration of the lines of evidence and secondly, the integration of the lines of evidence.

5.1.1.4.1. Weight of evidence higher tier risk assessment for honeybees

There was a number of higher tier endpoints (dominated by colony feeder type studies) which had

been assessed and integrated in a WoE exercise. First, each line of evidence (i.e. Class 1 and Class 2

endpoints) were assessed and concluded for the treated crop scenario for winter oilseed rape and for

the succeeding crop scenario. The graphical representations with the interpretation of each line of

evidence speciﬁc to these two situations are presented in Appendix

The second step of the WoE exercise is the integration of the lines of evidence. Tables

and

present the integration of the lines of evidence for the treated crop scenario for winter oilseed rape

and for the succeeding crop scenario, respectively.

In addition, below Table

11,

there is a brief consideration for the use on spring oilseed rape

(treated crop scenario).

It is noted that no reliable data were available for forager mortality. Therefore, no consideration for

this Class 1 endpoint is included in the integrations of the lines of evidence presented below.

Treated crop scenario

Table 11:

The integration of the lines of evidence for honeybees for the treated crop scenario for

winter oilseed rape

Refer to Appendix

Section 1.1.1

A relatively high number of experiments with a realistic or severe exposure regime

indicated an overall negligible deviation from the control, while a fewer experiments

with suitable exposure indicated a larger than negligible negative deviation from the

control. Also, there was a considerable biological variability within and between the

endpoints and the endpoints indicating negligible deviation were, in general, much

more reliable

Overall, this indicated a

moderate evidence for negligible effect

Overwintering

assessment

Refer to Appendix

Section 1.2.1

When the deviations from the controls of the effects on honeybee populations after

overwintering in combination with the exposure estimations of the relevant

experiments were considered, it was concluded that the available data set was rather

contradictory. Nevertheless, more weight was attributed to the available evidence

indicating more than negligible negative effects

However, some experiments that indicated negative effects had considerably longer

exposure duration than the realistic

ﬂowering

period of oilseed rape. When the

severity of these endpoints was taken into consideration, the available evidences

become balanced and no clear trend was apparent

Overall,

this line of evidence was inconclusive

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Class 1 endpoints

Colony strength

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Class 2 endpoints

Mortality in front of the

hive

Refer to Appendix

Section 1.3.1

Relatively low number of reliable endpoints was available. All of these endpoints had

low reliability, but the biological variability of them was low

Only one experiment with appropriate exposure and indicating negligible effect was

available, while in two experiments with mild exposure regimes temporal, larger than

negligible negative deviations were indicated

Overall, this indicated a

weak evidence for larger than negligible effect

Brood abundance

Refer to Appendix

Section 1.4.1

There was a considerable biological variability within and between the endpoints and

all the endpoints had low reliability

A slightly higher number of experiments with realistic or severe exposure regime

indicated an overall negligible (or even positive) deviation from the control, while

slightly fewer experiments with realistic or mild exposure regime indicated more than

negligible negative deviation from the control

Overall, this indicated a

weak evidence for negligible effect

Homing success

Refer to Appendix

Section 1.5.1

Relatively low number of endpoints was available. All were classiﬁed as reliable with

major restrictions and in all but one case, the estimated exposure levels exceed the

exposure assessment goal for winter oilseed rape. One of these endpoints indicated a

negligible (negative) deviation from the control. The single endpoint that had an

estimated exposure level slightly below, but close to the exposure assessment goal

indicated no deviation from the control

Overall, this indicated a

weak evidence for negligible effect

Integration of lines of

Class 1 endpoints, overall, suggest a moderate evidence for negligible effects. Class 2

evidence for winter

endpoints indicated only weak evidence; overall for a negligible effect. However, data

oilseed rape

also suggested that some temporal negative effects cannot be excluded

Quantiﬁcation of the effects

Uncertainty analysis

in line with EFSA

(2013c)

(À potential to make the

true risk lower

potential to make the

true risk higher)

Many of the available endpoints were assessed to be reliable with major

restrictions; therefore, the overall reliability of the WoE is limited

The reasons for the reliability assessment categorisation differed between the

studies

The consistency in case of many lines of evidence is low

Many of the data were not presented in sufﬁcient detail to derive accurate

deviations from the control

Within some experiments, pre-exposure measurements revealed that some

endpoints did not start at comparable level. This initial difference was

accounted for in the derivation of the deviation from the control, but the

accuracy of the quantiﬁcation in these cases is limited

Exposure in the experiments

For the overwintering assessment, the food consumption of foragers and nurse

bees were considered and the lower food consumption of resting winter bees

was not accounted for when estimating exposure in the experiments. This

might have overestimated the exposure in the experiments

The exposure level of the effect

ﬁeld

experiments on crops with nectar was

calculated considering 15% sugar content of the nectar, which is the low end

value of the realistic range. This may result in an overestimation of the

estimated exposure of those experiments

+/À

The level of the dilution of the residue concentrations of the consumed pollen

and nectar in the colony-feeder experiments with free

ﬂying

bees could not be

estimated from the available data

For some colony-feeder experiments where bees were fed with sugar solution,

+/À

the actual percentage of sugar was unknown and therefore assumed to be

50%. This would have an impact on the assumed consumption and in turn on

the active substance intake

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

The exposure in the higher tier effect studies was estimated using mean

residue measurements. In some studies, there were values reported

LOD,

which were conservatively considered as 0 mg/kg. When a value was reported

to be detected but

LOQ, the value was considered to be between the LOQ

and LOD (average of LOQ and LOD) and the LOD was considered as half of

the LOQ

Exposure assessment goals

In several residue studies used for the exposure assessment goal, the

sampling frequency and pattern did not guarantee that the actual maximum

occurrence had been picked up

The limited number of valid residue studies available for the exposure

assessment goal, restricted the potential for the representativeness to cover

percentile exposure situations

The exposure assessment goals were calculated assuming residues equal to

the LOQ every time measured concentration were

LOQ. In the cases that no

residues were detected the residue was considered to be equal to the LOD

The exposure assessment goals were calculated using the maximum residue

levels measured within each trial

Since insufﬁcient RUD values were available from winter oilseed rape to

perform an exposure assessment, the available RUD values for spring oilseed

rape were also used

For the determination of the exposure assessment goals, a 15% sugar content

of the nectar was assumed, which is the low end value of the realistic range.

This may have resulted in an overestimation of the exposure assessment goal

For overwintering assessment, the exposure assessment goal was based on

consumption from active bees and could therefore be overestimating the

actual exposure of bees during winter

Confounding factors in the experiments

In 4 of 23 higher tier studies, there was a conﬁrmation of control

contamination with imidacloprid or with other neonicotinoids. This could

potentially mask triggering of negative effects.The presence of external

substances was also seen in the treatment, creating uncertainties about

detecting effects not due to the treatment

In half of the higher tier studies where the bees were allowed to free

ﬂying,

there were indications for the use of different pesticides, including insecticides

(including neonicotinoids) in the landscape where the bees could forage. This

may affect both control and treatment.There is also uncertainty that this

practice could have been done also in experiments where this was not clearly

reported

Conclusion

+/À

The weight of evidence exercise with a consideration to the uncertainties of the

assessment is considered to indicate a low risk to honeybees from residues in pollen

and nectar for the treated crop scenario for winter oilseed rape. It has to be noted,

however, that the line of evidence for overwintering assessments was assessed as

inconclusive and the indirectly relevant (Class 2) endpoint of

‘mortality

in front of the

hive’ indicated a weak evidence for more than negligible effect

LOD: limit of detection; LOQ: limit of quantiﬁcation; RUD: residue per unit dose.

In lack of exposure assessment goals, a similar weight of evidence exercise as for winter oilseed

rape (integration of lines of evidences obtained from realistic exposure estimations) could not be

performed for the other crops. For the same reason, it is not possible to extrapolate the risk

assessment of winter oilseed rape to the other crops except for spring oilseed rape.

It is expected, and conﬁrmed by the available (rather limited) data for residues in pollen and nectar

(see Appendix

D),

that the residue levels in spring oilseed rape tend to be higher than in winter

oilseed rape. However, the application rate and the attractiveness of the two crops are the same. The

residue levels from spring oilseed rape studies had been considered in the assessments for winter

oilseed rape. Therefore, overall, the risk to spring oilseed rape is expected to be similar than the risk

to winter oilseed rape (i.e. low risk), noting that this includes some uncertainties due to the expected

higher residue levels.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Succeeding crop scenario

The exposure assessment goal for the succeeding crop scenario is only marginally different from

the exposure assessment goal for winter oilseed rape (see Section

5.1.1.2).

Also, the difference

between the length of realistic

ﬂowering

period of winter oilseed rape and any possible succeeding

crop, as considered in this conclusion (10–42 days for winter oilseed rape and 5–42 days for the

succeeding crop scenario), was not important, since all the reliable higher tier studies had more than

10 days exposure period. Therefore, the integration of the lines of evidence and the conclusion for

winter oilseed rape is equally applicable for the succeeding crop scenario. For the sake of

simpliﬁcation, only the uncertainty analysis and the conclusions that are different from the assessment

for winter oilseed rape are detailed in Table

12.

All the other parts of the integration of the lines of

evidence for winter oilseed rape as presented in Table

are applicable for the succeeding crop

scenario.

Table 12:

The integration of the lines of evidence for honeybees for the succeeding crop scenario

Refer to Appendix

Section 1.1.2

For narrative description, see Table

for winter oilseed rape

Overwintering

assessment

Class 2 endpoints

Mortality in front of the

hive

Brood abundance

Refer to Appendix

Section 1.3.2

For narrative description, see Table

for winter oilseed rape

Refer to Appendix

Section 1.4.2

For narrative description, see Table

for winter oilseed rape

Homing success

Appendix

section 1.5.2

For narrative description, see Table

for winter oilseed rape

Integration of lines of

For narrative description, see Table

for winter oilseed rape

evidence for the

succeeding crop

scenario

Quantiﬁcation of the effects

Uncertainty analysis

in line with EFSA

(2013c)

(À potential to make the

true risk lower

potential to make the

true risk higher)

Most of the available endpoints were assessed to be reliable with major

restrictions; therefore, the overall reliability of the WoE is limited

The reasons for the reliability assessment categorisation differed between the

studies

The consistency in case of many lines of evidence is extremely low

Many of the data were not presented in sufﬁcient detail to derive accurate

deviations from the control. Therefore, only crude estimates could be used in

the lines of evidence

Within some experiments, pre-exposure measurements revealed that some

endpoints did not start at comparable level. This initial difference was

accounted for in the derivation of the deviation from the control, but the

accuracy of the quantiﬁcation in these cases is limited

Exposure in the experiments

For the overwintering assessment, the food consumption of foragers and

nurse bees was considered and the lower food consumption of resting winter

bees was not accounted for when estimating exposure in the experiments.

This might have overestimated the exposure in the experiments

The exposure level of the effect

ﬁeld

experiments on crops with nectar was

calculated considering 15% sugar content of the nectar, which is the low end

value of the realistic range. This may result in an overestimation of the

estimated exposure of those experiments

+/À

Refer to Appendix

Section 1.2.2

For narrative description, see Table

for winter oilseed rape

Class 1 endpoints

Colony strength

+/À

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

The level of the dilution of the residue concentrations of the consumed pollen

and nectar in the colony-feeder experiments with free

ﬂying

bees could not be

estimated from the available data

For some colony-feeder experiments where bees were fed with sugar solution,

+/À

the actual percentage of sugar was unknown and therefore assumed to be

50%. This would have an impact on the assumed consumption and in turn on

the active substance intake

The exposure in the higher tier effect studies was estimated using mean

residue measurements. In some studies, there were values reported

LOD,

which were conservatively considered as 0 mg/kg. When a value was reported

to be detected but

LOQ, the value was considered to be between the LOQ

and LOD (average of LOQ and LOD), and the LOD was considered as half of

the LOQ

Exposure assessment goals

For the determination of the exposure assessment goals, a 15% sugar content

of the nectar was assumed, which is the low end value of the realistic range.

This may have resulted in an overestimation of the exposure assessment goal

For overwintering assessment, the exposure assessment goal was based on

consumption from active bees and could therefore be overestimating the

actual exposure of bees during winter

The PECplateau for the GAPs under consideration are lower than the soil

residue levels of the available residue trials. Therefore, the exposure

assessment goals are conservative for the GAPs under consideration. The only

exception was the use on cotton where the PECplateau was comparable with

the residue levels of the test soils

The exposure assessment goal was based on residue values from semiﬁeld

studies or samples taken directly from the crop, i.e. landscape dilution was not

taken into consideration

Confounding factors in the experiments

In 4 of 22 higher tier studies, there was a conﬁrmation of control

contamination with imidacloprid or with other neonicotinoids. This could

potentially mask triggering of negative effects

The presence of external substances was also seen in the treatment, creating

uncertainties about detecting effects not due to the treatment

In half of the higher tier studies where the bees were allowed to free

ﬂying,

there were indications for the use of different pesticides, including insecticides

(including neonicotinoids) in the landscape where the bees could forage. This

may affect both control and treatment.

There is also uncertainty that this practice could have been done also in

experiments where this was not clearly reported

Conclusion

+/À

The weight of evidence exercise with a consideration to the uncertainties of the

assessment is considered to indicate a low risk to honeybees from residues in pollen

and nectar for the succeeding crop scenario

It has to be noted however that the line of evidence for overwintering assessment

was assessed as inconclusive and the indirectly relevant (Class 2) endpoint of

‘mortality

in front of the hive’ indicated a weak evidence for more than negligible

effect

WoE: weight of evidence; LOD: limit of detection; LOQ: limit of quantiﬁcation; PEC: predicted environmental concentration; GAP:

good agricultural practices.

5.1.1.4.2. Weight of evidence higher tier risk assessment for bumblebees

A number of higher tier endpoints (dominated by colony feeder type studies) has been assessed

and integrated in a weight of evidence exercise. First, each line of evidence (i.e. class 1 and class 2

endpoints) were assessed and concluded for the succeeding crop scenario. The graphical

representations with the interpretation of each line of evidence, which had multiple endpoints, are

presented in Appendix

One of the Class 1 and the available Class 2 endpoints had only one or two

data points; therefore, no graphical representations were performed. These endpoints however are

brieﬂy summarised in Appendix

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

The second step of the WoE exercise is the integration of the lines of evidence. Table

present

the integration of the lines of evidence for the succeeding crop scenario.

Table 13:

The integration of the lines of evidence for bumblebees for the succeeding crop scenario

Refer to Appendix

Section 2.1

Only one experiment with appropriate exposure indicated negligible effect and only

one experiment with rather mild exposure regime resulted in a large-negative

deviation from the control. However, another endpoint with low exposure level but

long exposure duration also indicated negative deviation from the control

Overall, this indicated a

weak evidence for larger than negligible effect

Worker production

Refer to Appendix

Section 2.2

Two experiments with appropriate exposure indicated a clear negative deviation from

the control. One of them had higher reliability than all the other endpoints, while the

other one had some uncertainties. On the other hand, three experiments could be

considered as evidence indicating no negative deviation from the controls. However,

none of them had a clearly representative or severe exposure regime. No clear trend

could be seen when all these positive and negative elements were balanced

Overall, this

line of evidence was inconclusive

Drone production

Refer to Appendix

Section 2.3

Two experiments with low exposure level indicated a clear negative deviation from the

control. One of them had too long exposure length, while the other one had some

uncertainties. On the other hand, two experiments could be considered bearing some

evidences that indicate no negative deviations from the controls. However, none of

them had a clearly representative or severe exposure regime. No clear trend could be

seen when all these positive and negative elements were balanced

Overall, this

line of evidence was inconclusive

Reproductive output of

queenless microcolonies

Refer to Appendix

Section 2.4

A number of endpoints with relatively low exposure levels (and one of them with a

realistic exposure length) indicated clear negative deviations from the control. No

endpoint was available indicating negligible or positive deviation from the control

Overall, this indicated a

moderate evidence for larger than negligible effect

Brood production

Refer to Appendix

Section 2.5

A number of experiments indicated clearly negative deviations from the control. Five

of these experiments had a mild exposure regime. On the other hand, only two

endpoints could be considered as evidence indicating no negative deviation from the

controls. However, none of them had a clearly representative or severe exposure

regime

Overall, this indicated a

moderate evidence for larger than negligible effect

Colony strength

Refer to Appendix

Section 2.6

Only two endpoints were available. The estimated exposure levels associated to these

endpoints were in the range of or lower than the exposure assessment goals for

succeeding crops. The length of the exposure was in the realistic range or only

marginally longer than the

ﬂowering

period of succeeding crops. One of the endpoints

was considered as fully reliable. The deviation from the control was between

negligible to large negative and the mean deviation was assessed as medium

negative.

The other endpoint was considered as reliable with major restrictions. The deviation

from the control was large negative in this experiment

This indicated a

strong evidence for larger than negligible effect

Class 1 endpoints

Queen production

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Class 2 endpoints

Emergence rate

(workers)

Refer to Appendix

Section 2.7

Only one endpoint was available and the estimated exposure level associated to this

endpoint was in the range of the exposure assessment goals for succeeding crops.

The length of the exposure was in the realistic range of the

ﬂowering

period of

succeeding crops. The deviation from the control bridged from large positive to large

negative, and the mean deviation was assessed as medium negative. The endpoint

was considered as fully reliable

This indicated a

moderate evidence for more than negligible effect

Mating ability of the new Refer to Appendix

Section 2.8 Only one endpoint was available originating from a

ﬁeld

study on sunﬂower with a 9-day exposure period. No exposure estimation was

queens

available for this endpoint.

This

line of evidence was inconclusive

Integration of lines of

Class 1 endpoints, overall, suggest a moderate evidence for larger than negligible

effects.

evidence for the

This was also indicated by the single Class 2 endpoint, which was conclusive

succeeding crop

scenario

Uncertainty analysis

in line with EFSA

(2013c)

(À potential to make the

true risk lower

potential to make the

true risk higher)

Quantiﬁcation of the effects

Many of the available endpoints were assessed to be reliable with major

restrictions; therefore, the overall reliability of the WoE is limited

+/À

The reasons for the reliability assessment categorisation differed between the

+/À

studies

For some colony-feeder experiments where bees were fed with sugar solution,

+/À

the actual percentage of sugar was unknown and therefore assumed to be

50%. This would have an impact on the assumed consumption and in turn on

the active substance intake

The level of the dilution of the residue concentrations of the consumed pollen

and nectar in the colony-feeder experiments with free

ﬂying

bees could not be

estimated from the available data

In some colony-feeder studies, the amount of the offered food (i.e. the spiked

sugar solution) was not ad libitum. In one

ﬁeld

study, non-spiked food was

offered to the colonies after the exposure phase. These may lead to excessive

dilution of the residue concentrations

In a few colony-feeder studies, some avoidance to the feeding solution had

been reported

Exposure in the experiments

The exposure in the higher tier effect studies was estimated using mean residue

measurements. In two

ﬁeld

experiments, which were on potato (i.e. relevant

only for pollen), there were values reported

LOD, which were conservatively

considered as 0 mg/kg. When a value was reported to be detected but

LOQ,

the value was considered to be between the LOQ and LOD

Exposure assessment goals

The exposure assessment goal was based on residue values from semiﬁeld

studies or samples taken directly from the crop, i.e. landscape dilution was not

taken into consideration

The PECplateau for the GAPs under consideration are lower than the soil

residue levels of the available residue trials. Therefore, the exposure

assessment goals are conservative for the GAPs under consideration. The only

exception was the use on cotton where the PECplateau was comparable with

the residue levels of the test soils

+/À

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Confounding factors in the experiments

In a few cases of the higher tier studies, there was a conﬁrmation of control

contamination with imidacloprid or with other neonicotinoids. This could

potentially mask triggering of negative effects

In some experiments, pollen collected by honeybees without sufﬁcient

information on potential residue contamination was offered to the colonies.

Potential presence of external substances creates also uncertainties about

detecting effects not due to the treatment

In many of the higher tier studies where the bees were allowed to free

ﬂying,

there were indications for the use of different pesticides, including insecticides

(including neonicotinoids) in the landscape where the bees could forage. This

may affect both control and treatment.

There is also uncertainty that this practice could have happened also in

experiments where this was not clearly reported

Conclusion

+/À

The weight of evidence exercise with a consideration to the uncertainties of the

assessment is considered to indicate a high risk to bumblebees from residues in pollen

and nectar for the succeeding crop scenario

WoE: weight of evidence; LOD: limit of detection; LOQ: limit of quantiﬁcation; PEC: predicted environmental concentration; GAP:

good agricultural practices.

5.1.1.4.3. Solitary bees

No higher tier studies were available for solitary bees.

5.1.2.

Risk from contamination of adjacent vegetation via dust drift (ﬁeld

margin and adjacent crop scenario)

5.1.2.1 Tier-1 risk assessment

The Tier-1 risk assessment for the representative GAPs were performed by using the EFSA’s BeeTool

(v.3.) (Appendix Y of EFSA, 2013c) for honeybees and bumblebees, where suitable toxicity data were

available. A screening Tier-1 assessment was carried out for solitary bees and for the chronic adult

assessment for bumblebees as only surrogate endpoints were available. Since no toxicity data were

available for HPG development nor for the larvae toxicity for non-Apis bees, no assessment was

performed for these scenarios.

It was assumed that a deﬂector was used during the seed drilling. The potato seed treatment use

is an in-planter tuber treatment via a spray application. Therefore, no dust formation was considered

for this use.

The outcome of these calculations is summarised in Tables

and

15.

A low risk is indicated only if

all categories (acute, chronic and larva) for all the relevant scenarios resulted in low risk. When one or

more combinations (categories of acute, chronic and larva combined with the relevant scenarios)

indicated a high risk than high risk or low risk not demonstrated (for Tier-1 screening assessments) is

indicated in the table below. The detailed results are included in Appendix

Table 14:

Summary of the outcome of Tier-1 and screening Tier-1 risk assessment for the contact

route of exposure

Honeybee

Use

‘Low’

application

rate

High risk

‘High’

application

rate

High risk

Bumblebee

‘Low’

application

rate

High risk

‘High’

application

rate

High risk

Solitary bee

‘Low’

application

rate

Low risk not

demonstrated

‘High’

application

rate

Low risk not

demonstrated

Spring

cereals,

cotton,

endive,

lettuce,

brassicas

(ﬂowering,

head, leafy)

and maize

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Honeybee

Use

‘Low’

application

rate

‘High’

application

rate

High risk

Low risk with

deﬂector

Low risk with

and without

deﬂector

Not relevant

(Low risk)

Bumblebee

‘Low’

application

rate

Low risk with

deﬂector

Low risk with

deﬂector

Low risk with

and without

deﬂector

Not relevant

(Low risk)

‘High’

application

rate

High risk

Low risk with

deﬂector

Low risk with

and without

deﬂector

Not relevant

(Low risk)

Solitary bee

‘Low’

application

rate

Low risk not

demonstrated

Low risk not

demonstrated

Low risk with

and without

deﬂector

Not relevant

(Low risk)

‘High’

application

rate

Low risk not

demonstrated

Low risk not

demonstrated

Low risk with

deﬂector

Not relevant

(Low risk)

Winter cereal High risk

Spring and

winter rape

Sugar and

fodder beet

Potato

Low risk with

deﬂector

Low risk with

and without

deﬂector

Not relevant

(Low risk)

Table 15:

Summary of the outcome of Tier-1 and screening Tier-1 risk assessment for the dietary

route of exposure

Honeybee

Bumblebee

‘Low’

application

rate

High risk

‘High’

application

rate

High risk

Solitary bee

‘Low’

application

rate

Low risk not

demonstrated

‘High’

application

rate

Low risk not

demonstrated

Use

‘Low’

application

rate

‘High’

application

rate

High risk

Spring and

winter

cereals,

cotton,

endive,

lettuce,

brassicas

(ﬂowering,

head, leafy),

maize

Spring and

winter rape

Sugar and

fodder beet

Potato

High risk

Low risk with

and without

deﬂector

Not relevant

(Low risk)

High risk

Low risk with

deﬂector

Not relevant

(Low risk)

Low risk not

demonstrated

Low risk not

demonstrated

Not relevant

(Low risk)

Low risk not

demonstrated

Low risk not

demonstrated

Not relevant

(Low risk)

Low risk not

demonstrated

Low risk with

deﬂector

Not relevant

(Low risk)

Low risk not

demonstrated

Low risk not

demonstrated

Not relevant

(Low risk)

Since no dust formation is expected from the use on potato, a low risk was concluded for the

ﬁeld

margin and adjacent crop scenario for this use. Accounting for both contact and oral exposure, a low

risk to honeybees for the

ﬁeld

margin and adjacent crop scenario was indicated to sugar beet and

fodder beet. It is noted that the endpoint for larvae was only considered to be provisional, a reliable

endpoint for HPG is not available and there was no assessment of the potential for accumulative

effects available. Nevertheless, the provisional risk assessment is considered sufﬁcient to conclude a

low risk for this particular crop and scenario, but it would be prudent to update the risk assessment

when further data are available. The screening level assessment performed for bumblebees and for

the high application rate for solitary bees was not sufﬁcient to demonstrate a low risk. A low risk was

indicated for solitary bees for the low application rate of sugar beet and fodder beet. For spring and

winter rape, a high risk to honeybees was indicated and a low risk to bumblebees and solitary bees

was not demonstrated with a screening assessment. For all other seed treatment uses under

consideration, a high risk to honeybees and bumblebees was indicated and a low risk to solitary bees

was not demonstrated with a screening assessment. No risk assessment could be performed for

bumblebee and solitary bee larvae.

It is noted that for endive, lettuce and brassicas, a low risk could be concluded for uses when the

treated crop is sown in permanent greenhouse.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

5.1.2.2. Exposure assessment for the

ﬁeld

margin and adjacent crop scenario

In addition to the data set al.ready used in a previous EFSA conclusion (EFSA, 2016b), results from

dust drift trials measuring imidacloprid in dust deposition outside the drilled

ﬁeld

originating from

treated seed at the time of drilling were available for winter barley (two experimental sites), winter

wheat (two experimental sites) and winter oilseed rape (one experimental site). All these experiments

were carried out with the same type of pneumatic seed driller (Accord DL No. 009190). Information on

the presence or absence of a deﬂector and the deﬂector type was not provided.

In the two winter barley trials, seed dustiness was characterised with mean measured Heubach values

of 0.093 g/10

seeds and 0.076 g/10

seeds. For seeds treated at 0.031 mg/seed with drilling rate

3,312,699–3,317,593 treated winter barley seeds/ha, residues in petri dishes placed on the ground were

12.63–12.81 mg/ha at 3 m and 17.04–22.00 mg/ha at 1 m from the edge of the drilled area. For seeds

treated at 0.024829 mg/seed with drilling rate 3,535,405 treated winter barley seeds/ha, residues in

petri dishes were 10.584 mg/ha at 3 m and 16.589 mg/ha at 1 m.

In the two winter wheat trials, seed dustiness was characterised with mean measured Heubach

values of 0.062 g/10

seeds and 0.064 g/10

seeds. For seeds treated at 0.03096 mg/seed with

drilling rate 4,588,640–4,603,114 treated winter wheat seeds/ha, residues in petri dishes were 8.53–

13.976 mg/ha at 3 m and 13.424–16.047 mg/ha at 1 m. For seeds treated at 0.025619 mg/seed with

drilling rate 4,314,833 treated winter wheat seeds/ha, residues in petri dishes were 8.953 mg/ha at

3 m and 11.202 mg/ha at 1 m.

In a winter oilseed rape trial, seed dustiness was characterised with mean measured Heubach

values of 0.0006 g/10

seeds. For seeds treated at 0.01136–0.01137 mg/seed with drilling rate

649,877–672,332 treated oilseed rape seeds/ha, residues in petri dishes were 3.214 mg/ha at 3 m

and 2.934 mg/ha at 1 m at one trial plot. At the second plot, concentrations were

2.934 mg/ha

(LOQ).

As these trials only represent one seed drill type and two experiments for each crop, with these

seed batch quality measured dustiness (Heubach determinations), it was considered not credible to

use them to replace the Tier-1 values of the guidance. Therefore, no credible reﬁned exposure

assessment for contamination in

ﬁeld

margins and adjacent crops could be performed.

In addition to the dust deposition

ﬁeld

trials, information on dustiness of neonicotinoid treated

seeds was available.

In the open call for data information was provided on the measurement of the dust content of

oilseed rape and maize seeds (Heubach values); and in most cases for oilseed rape, also information

on active ingredient content in the dust (Heubach a.i. values) was provided from seed samples from

seed merchants. Dust was quantiﬁable/present in all seed batches tested. In maize seed, measured

dust was just reported to be signiﬁcantly below the industry standard of below 3 g dust/100 kg seed.

Oilseed rape seeds treated in 2013 from 326 seed treatment sites had a 90th percentile total dust

Heubach value of 0.192 g/700,000 seeds. These values for Heubach a.i. were 6 mg/700,000 seeds for

clothianidin (156 sites), 1.4 mg/700,000 seeds for imidacloprid (52 sites) and 7.8 mg/700,000 seeds

for thiamethoxam (104 sites).

Heubach a.i. values in seed samples from an additional 10 different seed treatment facilities also

where seed was treated in 2013 were 0.21, 0.27, 0.33, 0.46, 0.6, 0.96, 1.29, 1.3, 8.9 and 16.6 mg/

700,000 seeds for clothianidin.

As all these results are from the same year, so they did not provide any information on dustiness of

the seed being supplied to farmers in different years and whether dust levels have reduced in recent

years. It is clear that the Heubach a.i. values can be variable. It is clear that reducing the dust content

of seed to be treated as well as any dust produced during the treatment process as well as any that

might be generated during storage and transport of seed is a good target for improved risk

management. However, with the information available in this review, it was not possible to account for

this in any reﬁned exposure and/or risk characterisation.

5.1.2.3. Tier-2 risk assessment

Since the available higher tier data for the exposure characterisation was not considered to be

sufﬁcient for a reﬁnement, no Tier-2 risk assessment could be performed.

5.1.2.4. Tier-3 risk assessment

Only two higher tier effect studies were available. In the study I.848, winter barley seeds treated

with imidacloprid were drilled in

ﬁelds,

which were surrounded by

ﬂowering

Phacelia tanacetifolia

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

where honeybee colonies were located. Class 1 and Class 2 endpoints (colony strength, brood

abundance, mortality in front of the hive) were investigated. The brood abundance and the colony

strength were very similar in the control and treatment group (with only a few percent deviation on

average) with no statistical difference. The data on mortality in front of the hive indicated a constant

higher mortality rate in the treated

ﬁeld

than in the control (although this was the case in the pre-

exposure period, as well). There were some periods when the mortality in the treated group was

notably higher (up to a factor of 2, and in two cases, the available statistical analysis indicated a

signiﬁcant difference). However, the average number of dead bees was generally low, generally

10 bee/day.

In the study C*I.1324, sugar beet pills treated with clothianidin (0.60 mg/pill), imidacloprid

(0.30 mg/pill) and beta-cyﬂuthrin (0.08 mg/pill) were drilled in the

ﬁeld.

The sugar beet pills were also

treated with the fungicides thiram and hymexazol. Class 1 and Class 2 endpoints (colony strength,

colony strength after overwintering, brood abundance, mortality in front of the hive) were

investigated.

Overall, the colony strength during the season (i.e. until before the winter) and the mortality in

front of the hive had similar trends to the control, although the maximum negative deviation from the

control for the colony strength was marginally above the 7% (7.3%). No negative deviation was

observed on the colony strength after the winter. The data on brood abundance indicated some

ﬂuctuations,

it varied between small-positive to medium-negative deviation from the control.

All the endpoints from both studies were assessed as reliable with major restrictions. Moreover, in

the study on sugar beet, other insecticides, including clothianidin, were used. Overall, the information

available for the Tier-3 assessment was not considered to be sufﬁcient for a robust conclusion on the

risk via dust drift.

5.1.3.

Risk via water consumption

5.1.3.1. Guttation water

The risk assessment based on water solubility indicated a high risk. The detailed results are

included in Appendix

It should be highlighted that the EFSA evaluation of the conﬁrmatory data for imidacloprid and

clothianidin (EFSA, 2016b,c) concluded that the exposure of honeybees from contaminated guttation

ﬂuids

in the crops considered therein (winter cereals, sugar beet and potatoes) was of low relevance.

Such conclusion was conﬁrmed during the expert meeting related to this assessment (Pesticide Peer

Review Meeting 166), despite the experts acknowledged that such an assessment was based on

studies presenting major limitations. On the basis of this, no risk assessment from exposure to

contaminated guttation

ﬂuids

was carried out, and a low risk was concluded for winter cereals, sugar

beet and potatoes. In addition, considering this, a low risk was concluded for fodder beets.

In addition, it is noted that for endive, lettuce and brassicas, a low risk could be concluded for uses

when the treated crop is grown continuously in permanent greenhouse.

Higher tier assessment for crops other than winter cereals, sugar beet and potatoes

Valid reliable

ﬁeld

studies (two trials in Italy) investigating imidacloprid, 5-OH imidacloprid and

imidacloprid oleﬁn residues in guttation water collected from maize were available. As would be

expected, highest residues occurred when plants were small (interval from drilling to guttation

ﬂuid

sampling was short). The highest residues determined were up to 222 mg imidacloprid/L from an

application rate of 1.25 mg/seed and 47 mg imidacloprid/L from an application rate of 0.5 mg/seed.

This small data set was considered not sufﬁcient for selecting a 90th percentile exposure value as

suggested by EFSA (2013c). Therefore, in line with the consideration of the experts at the meeting for

the conﬁrmatory data for imidacloprid (EFSA, 2016b), the maximum measured concentrations were

considered to perform a reﬁned risk assessment for maize. The reﬁned calculations are presented in

Table

16.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Table 16:

Tier-2 risk assessment to honeybees exposed via consumption of contaminated guttation

ﬂuids

Water

consumption

(lL/bee or

lL/larva)

11.4

111

Measured

concentration

in guttation

ﬂuid

(lg/lL)

Bee type

Category

ETR

Trigger

Maize

Honeybee forager

Honeybee larva

Acute

Chronic

Larva

0.222

684

897

4,667

0.2

0.03

0.2

ETR: Exposure toxicity ratio.

On the basis of this Tier-2 assessment, a low risk could not be concluded for maize.

No higher tier effect studies were available for crops other than winter cereals, potato or sugar

beet; therefore, no further assessments were performed.

5.1.3.2. Puddle water

It was not necessary to perform exposure modelling to predict residues of imidacloprid in puddles

as the concentrations in surface runoff calculated by any version of the PRZMsw tool are always

negligible when seeds are drilled below the soil surface. Consequently, a low risk to honeybees from

residues in puddles for the seed treatment uses under consideration is concluded following EFSA

(2013c) that indicates that concentrations in runoff water as calculated by PRZMsw are an appropriate

estimate of puddle water concentrations. Experts from member states noted that the EFSA (2013c)

approach might represent a best case as cultivation following harvesting of the treated crop

redistributes soil residues, such that concentrations at the soil surface will be present to desorb into

puddles. PRZM calculations as prescribed by FOCUS surface water do not account for this as the

FOCUS PRZM tool and FOCUS surface water runoff scenarios do not account for soil cultivation.

5.1.3.3. Surface water

PEC in surface water were calculated at Step 3 following FOCUS surface water (2001) guidance and

the FOCUS tools SWASH 5.3, SPIN 2.2, MACRO 5.5.4, PRZMsw 3.3.2 and TOXSWA 4.4.3. for the active

substance imidacloprid. The PECs were calculated for the GAPs on potatoes and winter cereals.

Substance input values were selected in accordance with generic guidance for FOCUS surface water

scenarios from the list of endpoints of EFSA (2008). The laboratory geomean single

ﬁrst-order

for

soil used was 117.7 days and that for water was 67 days with a value of 1,000 days (default) used for

sediment. The geomean KFoc of 209.5 mL/g (KFom 121.5 mL/g) arithmetic mean 1/n 0.8 were used

as input in the simulations. The option soil incorporation was selected and the application window was

set in line with the planting dates speciﬁed at each scenario for winter cereals and potatoes. For PRZM

CAM 8 was selected, i.e. that the incorporation was at a speciﬁc depth with 4 cm which is applicable

for cereals and represents a conservative approach regarding surface runoff for potatoes. The highest

PEC surface water calculated with this approach was 0.075

lg/L

for the D6 ditch scenario from the use

on winter cereals. From the simulations for potatoes, the highest PEC surface water was 0.042

lg/L

for the D4 stream scenario. For these uses where seeds are drilled and granules are incorporated

below the soil surface, concentrations in surface runoff as calculated by PRZMsw are always negligible.

Therefore, as for puddle water (see

5.1.3.2),

concentrations in the FOCUS surface water bodies are

also always characterised as negligible for the FOCUS surface water runoff scenarios, when the option

soil incorporation is selected.

Exposure toxicity ratio (ETR) calculations with the above-mentioned water concentrations resulted

in a low risk to all honeybee castes (for details of the calculations see Appendix

D).

As no toxicity

endpoint was available for the HPG development, a risk assessment could not be performed for this

scenario. However, considering the notable margin of safety (two orders of magnitude in the chronic

risk assessment) obtained in these assessments, a low risk was concluded for this scenario, as well.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

5.2.

5.2.1.

Risk assessments for granule

Risk via systemic translocation in plants

–

residues in nectar and pollen

This route of exposure is relevant for the treated crop scenario, for the weed and for the

succeeding crop scenario.

5.2.1.1. Tier-1 risk assessment for treated crop scenario, weed scenario and succeeding

crop scenario

The Tier-1 risk assessments for the representative GAPs were performed by using the EFSA’s

BeeTool (v.3.) (Appendix Y of EFSA, 2013c) for honeybees and bumblebees, where suitable toxicity

data were available. A screening Tier-1 assessment was carried out for solitary bees and for the

chronic adult assessment for bumblebees as only surrogate endpoints were available. Since no toxicity

data was available for HPG development or for the larvae toxicity for non-Apis bees, no assessment

was performed for these scenarios.

The outcome of the calculations is summarised in Table

17.

A high risk or low risk not

demonstrated (for Tier-1 screening assessments) is indicated in the Table

since all combinations

(categories of acute, chronic and larva combined with the relevant scenarios) indicated a high risk. The

only exception is for the treated crop scenario, but only when the application is done after the

ﬂowering

period of the crop. The detailed results are included in Appendix

It is however to be noted

that turf is typically a mix of different plant species that may

ﬂower

in different periods. Moreover, turf

may be mowed during the vegetative season and shortly after mowing, turf may

ﬂower

again.

Therefore, this scenario (application is done after the

ﬂowering

of the crop) has a limited applicability

for this use. During the experts meeting (October 2017), some Member States highlighted that

country-speciﬁc authorisations of some products are limited to some speciﬁc categories that are

generally considered as

‘highly

managed amenity turf’. That includes generally low abundance of

weeds and regular mowing of the

ﬁeld,

which makes the

ﬁeld,

in general, unattractive to bee species.

Therefore, in the Member States where granular uses on amenity vegetation are authorised, this issue

should be further considered. Moreover, it is noted that typically amenity vegetation is grown for

several years on the same

ﬁeld.

Therefore, the succeeding crop is typically a real vegetation growing

from the same root system. The Tier-1 calculations for succeeding crop refer to the situations when

the amenity vegetation is removed as a result of the preparation of a seed bed to plant an attractive

following crop.

Table 17:

Summary of the outcome of Tier-1 and screening Tier-1 risk assessment for the dietary

route of exposure

Honeybee

Use

‘Low’

application

rate

‘High’

application

rate

Bumblebee

‘Low’

application

rate

‘High’

application

rate

Solitary bee

‘Low’

application

rate

‘High’

application

rate

Managed amenity turf

High risk

Low risk not demonstrated

5.2.1.2. Exposure assessment for the treated crop, succeeding crop scenarios and for

ﬂowering

weeds

No new data that could be used for an exposure reﬁnement for the treated crop scenario and for

the weed scenario were available for the evaluation of the GAPs for granular formulations.

As the concentrations in pollen and nectar in succeeding crops are considered to be independent of

the GAP and formulation type, the available information as assessed in Section

5.1.1.2

(for the seed

treatment uses) are also applicable to the granular uses under consideration.

5.2.1.3. Tier-2 risk assessment

As no reﬁned exposure assessment for residues in pollen and nectar for the treated crops scenario

ﬂowering

weeds scenario or information on weed coverage was available, no Tier-2 risk assessment

could be performed for these scenarios.

The risk assessments performed under Section

5.1.1.3

(for the seed treatment uses) are also

applicable to the granular use under consideration. A low risk for the succeeding crop scenario could

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

not be demonstrated with the available exposure reﬁnements. It is noted, however, that the

succeeding crop scenario is relevant only for situations when the amenity vegetation is removed as a

result of the preparation of a seed bed to plant an attractive following crop.

5.2.1.4. Tier-3 risk assessment

Treated crop scenario

Only one higher tier study was available (I.462) for granular uses. In this study, the effects on

bumblebee colonies (Bombus

impatiens)

were investigated in semiﬁeld conditions in the USA. The

cages were set up in grassland with

ﬂowering

white clovers. A granular formulation (Merit 0.5 G) was

applied by hand, and shortly after the treatment, the plots were irrigated. From the most relevant

endpoints (Class 1 and Class 2), only colony strength and brood production were studied. At the end

of the study, on average, 26% less workers were counted in the treated group. Contrary, the number

of brood chambers was 43% higher than in the control group. The endpoints were considered as

reliable with major restrictions.

Therefore, the information available for the Tier-3 assessment was not considered to change the

conclusion of the Tier-1 risk assessment.

Weed scenario

No higher tier data were available for the weed scenario. However, the assessments and

conclusions for the treated crop scenario are considered also applicable for the weed scenario.

Succeeding crop scenario

As the concentration in pollen and nectar in succeeding crops is considered to be independent of

the GAP and formulation type, the weight of evidence risk assessment performed under

Section

5.1.1.4

(for the seed treatment uses) is also applicable to the granular uses of imidacloprid

under consideration. It is noted that the succeeding crop scenario is relevant only for situations when

the amenity vegetation is removed as a result of the preparation of a seed bed to plant an attractive

following crop.

5.2.2.

Risk from contamination via dust drift

This route of exposure is relevant for the

ﬁeld

margin and adjacent crop scenarios. In addition, this

route of exposure is relevant for the treated crop scenario and for the weed scenario after emergence

for the use in managed amenity turf.

5.2.2.1. Tier-1 risk assessment

The Tier-1 risk assessments for the representative GAPs were performed by using the EFSA’s

BeeTool (v.3.) (Appendix Y of EFSA, 2013c) for honeybees and bumblebees, where suitable toxicity

data were available. A screening Tier-1 assessment was carried out for solitary bees and for the

chronic adult assessment for bumblebees as only surrogate endpoints were available. Since no toxicity

data was available for HPG development or for the larvae toxicity for non-Apis bees, no assessment

was performed for these scenarios.

The outcome of the calculations is summarised in Table

18.

A high risk or low risk not

demonstrated (for Tier-1 screening assessments) is indicated in Table

18,

since all combinations

(categories of acute, chronic and larva combined with the relevant scenarios) indicated a high risk (for

both the contact route of exposure and the dietary route of exposure). The detailed results are

included in Appendix

Table 18:

Summary of the outcome of Tier-1 and screening Tier-1 risk assessment for the contact

route of exposure and the dietary route of exposure

Honeybee

Use

‘Low’

application

rate

‘High’

application

rate

Bumblebee

‘Low’

application

rate

‘High’

application

rate

Solitary bee

‘Low’

application

rate

‘High’

application

rate

Managed amenity turf

High risk

Low risk not demonstrated

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

5.2.2.2. Exposure assessment for the

ﬁeld

margin and adjacent crop scenario

No new data on dust drift deposits of granular formulations containing imidacloprid were available

for this evaluation. Therefore, no reﬁned exposure assessment for contamination in

ﬁeld

margins and

adjacent crops could be performed.

5.2.2.3. Tier-2 risk assessment

As no reﬁned exposure assessment for dust drift was available, no Tier-2 risk assessment could be

performed.

5.2.2.4. Tier-3 risk assessment

No new higher tier data investigating the effects to bees from dust drift generated during the

application of granules in accordance with the use on amenity turf were available. Therefore, no Tier-3

risk assessment was performed.

5.2.3.

Risk via water consumption

5.2.3.1. Guttation water

The risk assessment based on water solubility indicated a high risk. The detailed results are

included in Appendix

Higher tier assessment

No new valid data were submitted for reﬁning the risk to bees from exposure to guttation

ﬂuid

from the granular uses.

5.2.3.2. Puddle water

It was not necessary to perform exposure modelling to predict residues of imidacloprid in puddles

as the concentrations in surface runoff calculated by any version of the PRZMsw tool are always

negligible when granules are buried or otherwise incorporated (i.e. by watering) below the soil surface.

Consequently, a low risk to honeybees from residues in puddles for the granular uses under

consideration is concluded following EFSA (2013c) that indicates that concentrations in runoff water as

calculated by PRZMsw are an appropriate estimate of puddle water concentrations.

5.2.3.3. Surface water

The assessments and the conclusion as presented in Section

5.1.3.3

(for the seed treatment uses)

are applicable for the granular uses, as well. Therefore, a low risk to honeybees from residues in

surface water for the granular uses under consideration was concluded.

Overall conclusion

The conclusion of the risk assessment to bees for the uses of imidacloprid as seed treatment is

summarised below, considering the different scenarios. It should be highlighted that for the seed

treatment uses, the weed scenario was not considered relevant, in agreement with EFSA (2013c).

For the crop-speciﬁc conclusion achieved at each assessment tier, please refer to Table

19.

The

assessments included in this conclusion considered the risk to bees from imidacloprid as active

substance only. It should be noted that formulation products containing imidacloprid may also contain

other insecticides including clothianidin, as shown in the GAP table in Appendix

6.1.

Seed treatment uses

Risk via systemic translocation in plants

–

residues in nectar and pollen

Treated crop scenario

A risk assessment was carried out for all the uses with the exception of the use on potato. For

potato, no information was available on the amount of imidacloprid sprayed to the individual tubers.

Therefore, the assessment could not be

ﬁnalised

for the treated crop scenario for potato. A high risk

at the Tier-1 was concluded for all the other crops and for all bee groups. It should be noted that, in

lack of speciﬁc toxicity data, a high risk for solitary bees was identiﬁed using only surrogate toxicity

estimation based on honeybee data. It should also be noted that vegetable crops and sugar and

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

fodder beets may be harvested before they

ﬂower.

In that case, the treated crop scenario is not

relevant for those crops (low risk for the treated crop scenario). Vegetable crops may be grown

continuously in permanent greenhouse. A low risk could be concluded for those situations.

The availability of residue data for winter oilseed rape allowed carrying out a Tier-2 risk assessment

for honeybees. The Tier-2 risk assessment resulted in a high risk.

A Tier-3 risk assessment was also carried out for the use on winter oilseed rape, using data from

colony-feeder, semiﬁeld and

ﬁeld

experiments. The Tier-3 risk assessments resulted in a low risk. This

was extrapolated to spring oilseed rape; therefore, a low risk was concluded for both winter and spring

oilseed rape for the treated crop scenario.

Succeeding crop scenario

A high risk at Tier-1 was concluded for all crops and all bee groups. It should be noted once again

that, in lack of speciﬁc toxicity data, a high risk for solitary bees was identiﬁed using only surrogate

toxicity estimation based on honeybee data. Vegetable crops may be grown continuously in permanent

greenhouse. A low risk could be concluded for those situations.

The availability of residue data allowed carrying out a Tier-2 risk assessment for the succeeding

crop scenario. The Tier-2 risk assessment resulted in a high risk for all the uses. As for the Tier-1

assessments, in lack of speciﬁc toxicity data, a high risk for solitary bees was identiﬁed using only

surrogate toxicity estimation based on honeybee data.

A Tier-3 risk assessment was also carried out for honeybees and bumblebees, using data from

colony-feeder, semiﬁeld and

ﬁeld

experiments that were available for honeybees and bumblebees. The

Tier-3 risk assessments resulted in a low risk for honeybees. For bumblebees, a high risk was indicated

by the risk assessments.

Risk from contamination of adjacent vegetation via dust drift

Field margin and adjacent crop scenarios

A risk assessment was carried out for all the uses with the exception of the use on potato. The use

on potato was an in-planter tuber spraying; therefore, it was considered that there is no dust emission

from this use. A high risk at the Tier-1 was concluded for all the other crops and for all bee groups

with the exception of the uses on sugar and fodder beets. For the use with the lowest application rate

on sugar and fodder beets, a low risk was indicated by the Tier-1 risk assessment for solitary bees

considering that a deﬂector was used. The use with the highest application rate resulted in a high risk.

A low risk was indicated for both the lowest and the highest application rates for honeybees. In case

of the highest application rate, the use of a deﬂector was taken into consideration. It should be noted,

however, that toxicity data for the HPG development and data on accumulative toxicity were not

available.

It is noted that for endive, lettuce and brassicas, a low risk could be concluded for uses when the

treated crop is sown in permanent greenhouse.

As already mentioned above, in lack of speciﬁc toxicity data, a high risk for solitary bees was

identiﬁed using only surrogate toxicity estimation based on honeybee data. The high risk identiﬁed for

bumblebees for the uses on spring and winter oilseed rape and on sugar and fodder beets was based

on surrogate toxicity data.

The available data did not allow performing any reﬁned risk assessment for exposure via dust drift.

Risk via consumption of contaminated water

Guttation

ﬂuids

A low risk to honeybees was concluded for the uses on winter cereals, potato and sugar beet in

agreement with the evaluation of the conﬁrmatory data for imidacloprid (EFSA, 2016b) and

clothianidin (EFSA, 2016c) and conﬁrmed during the expert meeting related to this assessment. The

conclusion on sugar beet was extrapolated to fodder beet; therefore, a low risk was also concluded for

fodder beet.

For all other crops, a low risk to honeybees could not be demonstrated using the screening

assessment based on the solubility of imidacloprid.

However, it is noted that for endive, lettuce and brassicas, a low risk could be concluded for uses

when the treated crop is grown continuously in permanent greenhouse.

An exposure reﬁnement could be considered for maize. The reﬁned risk assessments indicated a

high risk to honeybees.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Puddle water

A low risk is concluded to honeybees from residues in puddles for the seed treatment uses under

consideration.

Surface water

PECs in surface water considering FOCUS surface water modelling at Step 3 were calculated for the

uses under evaluation. The risk assessments considering these PECs resulted in a low risk to

honeybees. It is noted that no toxicity endpoint was available for the HPG development and data on

accumulative toxicity were not available.

6.2.

Granular uses

Only one use on granular formulation was considered in this assessment: the use on amenity turf.

Risk via systemic translocation in plants

–

residues in nectar and pollen

Treated crop scenario

A high risk at the Tier-1 was concluded. It should be noted that, in lack of speciﬁc toxicity data, a

high risk for solitary bees was identiﬁed using only surrogate toxicity estimation based on honeybee

data. It should also be noted that a low risk could be concluded for highly managed amenity turf

where regular mowing of the crop takes place.

The available data did not allow performing any reﬁned risk assessment.

Weed scenario

A high risk at the Tier-1 was concluded. It should be noted that, in lack of speciﬁc toxicity data, a

high risk for solitary bees was identiﬁed using only surrogate toxicity estimation based on honeybee

data. It should also be noted that a low risk could be concluded for highly managed amenity turf,

which, in general, has a low abundance of attractive weeds and where regular mowing takes place.

The available data did not allow performing any reﬁned risk assessment.

Succeeding crop scenario

A high risk at the Tier-1 was concluded. It should be noted that, in lack of speciﬁc toxicity data, a

high risk for solitary bees was identiﬁed using only surrogate toxicity estimation based on honeybee

data. Moreover, it is noted that typically amenity vegetation is grown for several years on the same

ﬁeld.

Therefore, the succeeding crop is typically areal vegetation growing from the same root system.

The Tier-1 calculations for succeeding crop refer to the situations when the amenity vegetation is

removed as a result of the preparation of a seed bed to plant an attractive following crop.

The availability of residue data allowed carrying out a Tier-2 risk assessment for the succeeding

crop scenario. The Tier-2 risk assessment resulted in a high risk. As for the Tier-1 assessments, in lack

of speciﬁc toxicity data, a high risk for solitary bees was identiﬁed using only surrogate toxicity

estimations based on honeybee data.

A Tier-3 risk assessment was also carried out for honeybees and bumblebees, using data from

colony-feeder, semiﬁeld and

ﬁeld

experiments that were available for honeybees and bumblebees. The

Tier-3 risk assessments resulted in a low risk for honeybees. For bumblebees, a high risk was indicated

by the risk assessments.

Risk from contamination of adjacent vegetation via dust drift

Field margin and adjacent crop scenarios

A high risk at the Tier-1 was concluded.

The available data did not allow performing any reﬁned risk assessment for exposure via dust drift.

Risk via consumption of contaminated water

Guttation

ﬂuids

A low risk to honeybees could not be demonstrated using the screening assessment based on the

solubility of imidacloprid.

The available data did not allow performing any reﬁned risk assessment.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Puddle water

A low risk is concluded to honeybees from residues in puddles for the seed treatment uses under

consideration.

Surface water

The risk assessment considering FOCUS surface water PECs resulted in a low risk to honeybees. It

is noted that no toxicity endpoint was available for the HPG development and data on accumulative

toxicity were not available.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Table 19:

A summary of the conclusions for the tiered risk assessment

Honeybee

Treated crop

scenario

Treated crop

scenario

Succeeding

crop

Bumblebee

Treated crop

scenario

Succeeding

crop

Solitary bee

Succeeding

crop

Guttation

ﬂuid

Adjacent

crop

Adjacent

crop

Surface

water

Field

margin

Field

margin

Spring cereals

(112–126 g

a.s./ha,

0.039 mg

a.s./seed)

Winter cereals

(43–126 g

a.s./ha,

0.015–0.039 mg

a.s./seed)

Cotton (100–

175 g a.s./ha,

0.63–0.84 mg

a.s./seed)

Endive and

lettuce

–

ﬁeld

use

(104 g a.s./ha,

0.8 mg a.s./seed)

Harvested after

ﬂowering

Endive and

lettuce

–

ﬁeld

use

(104 g a.s./ha,

0.8 mg a.s./seed)

Harvested before

ﬂowering

Tier 1

Tier 2

Tier 3

Tier 1

Tier 2

Tier 3

Tier 1

Tier 2

Tier 3

Tier 1

Tier 2

Tier 3

N/R

Field

margin

Puddle

water

Use

Tier

N/R

Tier 1

Tier 2

Tier 3

N/R

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Adjacent

crop

Weed

scenario

Weed

scenario

Weed

scenario

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Honeybee

Treated crop

scenario

Treated crop

scenario

Succeeding

crop

Bumblebee

Treated crop

scenario

Succeeding

crop

Solitary bee

Succeeding

crop

Guttation

ﬂuid

Adjacent

crop

Adjacent

crop

Surface

water

Field

margin

Field

margin

Endive, lettuce

and Brassica

(a)

transplanted from

non-permanent

greenhouse to

ﬁeld

or to non-

permanent

greenhouse,

harvested after

ﬂowering

Endive, lettuce

and Brassica

(a)

transplanted from

non-permanent

greenhouse to

ﬁeld

or to non-

permanent

greenhouse,

harvested before

ﬂowering

Endive, lettuce

and Brassica

(a)

transplanted from

non-permanent

greenhouse to

permanent

greenhouse,

harvested before

or after

ﬂowering

Tier 1

Tier 2

Tier 3

N/R

Field

margin

Puddle

water

Use

Tier

Tier 1

Tier 2

Tier 3

N/R

Tier 1

Tier 2

Tier 3

N/R

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Adjacent

crop

Weed

scenario

Weed

scenario

Weed

scenario

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Honeybee

Treated crop

scenario

Treated crop

scenario

Succeeding

crop

Bumblebee

Treated crop

scenario

Succeeding

crop

Solitary bee

Succeeding

crop

Guttation

ﬂuid

Adjacent

crop

Adjacent

crop

Surface

water

Field

margin

Field

margin

Endive, lettuce

and Brassica

(a)

transplanted from

permanent

greenhouse to

ﬁeld

or to non-

permanent

greenhouse,

harvested after

ﬂowering

Endive, lettuce

and Brassica

(a)

transplanted from

permanent

greenhouse to

ﬁeld

or to non-

permanent

greenhouse,

harvested before

ﬂowering

Endive, lettuce

and Brassica

(a)

transplanted from

permanent

greenhouse to

permanent

greenhouse,

harvested before

or after

ﬂowering

Maize (60–100 g

a.s./ha, 1 mg

a.s./seed)

Tier 1

Tier 2

Tier 3

N/R

Field

margin

Puddle

water

Use

Tier

Tier 1

Tier 2

Tier 3

N/R

Tier 1

Tier 2

Tier 3

N/R

Tier 1

Tier 2

Tier 3

N/R

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Adjacent

crop

Weed

scenario

Weed

scenario

Weed

scenario

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Honeybee

Treated crop

scenario

Treated crop

scenario

Succeeding

crop

Bumblebee

Treated crop

scenario

Succeeding

crop

Solitary bee

Succeeding

crop

Guttation

ﬂuid

Adjacent

crop

Adjacent

crop

Surface

water

Field

margin

Field

margin

Potato (120–

180 g a.s./ha)

Spring rape (9–

12 g a.s./ha,

0.01 mg a.s./

seed)

Winter rape (9–

12 g a.s./ha,

0.01 mg a.s./

seed)

Sugar and fodder

beet (13 g a.s./

ha, 0.1 mg a.s./

seed)

Harvested after

ﬂowering

Sugar and fodder

beet

(13 g a.s./ha,

0.1 mg a.s./seed)

Harvested before

ﬂowering

Sugar and fodder

beet

(117 g a.s./ha,

0.9 mg a.s./

seed), harvested

after

ﬂowering

Tier

N/R

Tier 1

Tier 2

Tier 3

Tier 1

Tier 2

Tier 3

N/R

Field

margin

Puddle

water

Use

Tier

N/R

Tier 1

Tier 2

Tier 3

N/R

Tier 1

Tier 2

Tier 3

N/R

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Adjacent

crop

Weed

scenario

Weed

scenario

Weed

scenario

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Honeybee

Treated crop

scenario

Treated crop

scenario

Succeeding

crop

Bumblebee

Treated crop

scenario

Succeeding

crop

Solitary bee

Succeeding

crop

(c)

Guttation

ﬂuid

Adjacent

crop

Adjacent

crop

Surface

water

Field

margin

Field

margin

Sugar and fodder

beet

(117 g a.s./ha,

0.9 mg a.s./

seed), harvested

before

ﬂowering

Managed amenity

turf (golf courses,

sport grounds,

commercial and

residential lawns)

(150 g a.s./ha)

‘Highly’

managed

amenity turf (golf

courses, sport

grounds,

commercial and

residential

lawns)

(b)

(150 g a.s./ha)

Tier 1

Tier 2

Tier 3

N/R

Field

margin

Puddle

water

Use

Tier

Tier 1

Tier 2

Tier 3

(c)

Tier 1

Tier 2

Tier 3

(c)

L: A low risk is concluded for the risk assessment (for the

ﬁeld

margin and adjacent crop scenario for the seed dressing uses, the use of a deﬂector was considered).

R1: A high risk is concluded on the basis of the assessment.

R2: A low risk cannot be demonstrated as a result of the assessment (screening-type risk assessment or incomplete conclusion at Tier-3).

X: Assessment not

ﬁnalised

(lack of exposure or endpoint for effects).

Empty grey box: no assessment.

N/R: scenario not relevant.

(a): The application rates for endive and lettuce are 89–120 g a.s./ha, 0.8–1.2 mg a.s./seed and for brassicas (ﬂowering, head, leafy) 90 g a.s./ha, 1.5 mg a.s./seed). Brassicas (ﬂowering, head,

leafy) includes crops like broccoli, cauliﬂowers, Brussels sprouts, head cabbages, Chinese cabbage and kales.

(b): Member States highlighted during the peer review that country-speciﬁc authorisations of some products are limited to some speciﬁc categories that are generally considered as

‘highly

managed amenity turf’. That includes generally low abundance of weeds and regular mowing of the

ﬁeld,

which makes the

ﬁeld,

in general, unattractive to bee species.

(c): Amenity vegetation is typically grown for several years on the same

ﬁeld.

Therefore, the succeeding crop is typically areal vegetation growing from the same root system. The risk assessment

for succeeding crop refers to the situations when the amenity vegetation is removed as a result of the preparation of a seed bed to plant an attractive following crop.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Adjacent

crop

Weed

scenario

Weed

scenario

Weed

scenario

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Overall appraisal of the uncertainty related to the risk assessment

In order to reach the aforementioned conclusions on the risk assessment of imidacloprid,

clothianidin, and thiamethoxam, EFSA has considered a large number of documents, reporting very

diverse experiments, where many heterogeneous endpoints were measured under different conditions

and using different methodologies.

One of the most relevant outputs of this complex exercise is to account for the uncertainty related to

the overall assessment. At the lower tier (e.g. Tier-1 and screening), this is accounted for by the use of

conservative estimates which is particularly important when standard Tier-1 parameters have been

extrapolated from more worse-case situations (e.g. in cases where data were lacking for a particular

crop). On the contrary, as acknowledged in EFSA (2013c), there are several routes of exposures which

are not covered by the risk assessment scheme (e.g. insect honeydew, exposure via soil).

At higher tiers (Tier-2 and Tier-3), the uncertainty starts to act in two opposite ways, and it is

worth breaking it up in different factors, whose relative importance can be investigated more in detail.

Several factors were identiﬁed as source of uncertainty when establishing the revised shortcut values

and exposure assessment goals. Some of them indicated that the estimated exposure assessment goals

might be overestimated with respect to the actual exposure to bees (e.g. calculation of the exposure

assessment goals using the maximum value from each trial, assuming residues equal to the LOQ every

time they were

LOQ), consequently have the potential to decrease the actual risk in comparison with

the present assessment. On the contrary, other factors may act in the opposite way (e.g. the sampling

frequency was insufﬁcient to ensure that the peak residue was captured, limited number of residue trials

resulting in a lower capacity to ensure that the 90th percentile determination was captured).

Similar factors were identiﬁed as source of uncertainty for the estimates of exposure within the

effect experiments. For some sources of uncertainty which were applicable to both the calculation of

the exposure assessment goals and the estimated exposure within the experiments, it was ensured

that the same assumption was equally applied to both. In this way, the uncertainty is balanced, e.g.

the same percentage of sugar content in nectar was assumed for both the exposure assessment goal

and the estimated exposure in the experiments.

Other sources of uncertainty are related to the quantiﬁcation of the effects. In this case, the

direction of the uncertainty is rarely identiﬁable, as the uncertainty itself is linked to low reliability of

the experimental design/methodology and to the lack of precision in reporting the results.

Finally one of the most important sources of uncertainty is related to the presence of

‘confounding

factors’ in most of the higher tier experiments, particularly those performed under

ﬁeld

conditions. As

an example, other chemicals (i.e. herbicides, fungicides, acaricides or other classes of insecticides)

were often applied to both the treatments and the control plots in line with standard

ﬁeld

practises.

Nevertheless, the relative inﬂuence that exposure to these substances might have on the bees in the

control and in the treatment is unknown.

Furthermore, putting together the information from all

ﬁeld

experiments considered for the present

risk assessment review (encompassing imidacloprid, clothianidin, and thiamethoxam), EFSA noted that

in more than 40% of the cases (15 experiments of 35), some matrices collected from the controls

(e.g. from hives, plants, or soils) were contaminated with at least one neonicotinoid substance.

Contamination of controls was sometimes even indicated in experiments where bee colonies were

exposed via contaminated sugar solutions.

It is worth noting that, in the large majority of the cases, the residue analysis only focused on the

substance used in the treatment and on its metabolites. There were only six studies where residues for

a wider range of neonicotinoid substances were investigated. Five of these experiments reported

residues of substances not included in the study design at quantiﬁable concentrations. Cross-

contamination from substances other than the test item resulted in some cases in residue levels

comparable to those due to the applied treatment.

Similar issues had been already pointed out by EFSA in relation to other studies not included in the

present review (EFSA, 2013a,b).

This

ﬁnding

can be explained considering that neonicotinoids substances have been largely used in

Europe for several years and on a wide range of crops. Furthermore, neonicotinoids insecticides are

persistent in the environment, particularly in soil. EFSA (2008) reported

ﬁeld

value ranging from

104 to 228 days for imidacloprid. For the other two substances, some DT

values are reported in the

respective EU review reports (European Commission, 2006a,b). The mean/median DT

values

reported therein are 156 days for clothianidin and 174 days for thiamethoxam. It might be worth

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

noting that the main soil metabolite of thiamethoxam is clothianidin, so the DT

of the active

substance alone is not fully representative of the whole exposure time-variable proﬁle.

It is important to note that this

ﬁnding

has implications on different aspects of the present Tier-3

risk assessment for the treated crop and succeeding crops scenarios. Firstly, it impaired the reliability

of some experiments where contamination of controls was recorded. Furthermore, it creates high

uncertainty around the reliability of the results for those studies where either residue measurements

were not available or, as in the vast majority of the studies, where substances other than the test item

were not properly investigated. In general, this

ﬁnding

highlights a general disadvantage about the use

ﬁeld

studies for addressing the risk assessment. It exposed a source of uncertainty related to the

biological observations from

ﬁeld

studies, particularly for their interpretation and their reliability when

used in the risk assessment.

It is very likely that one cause of the control contamination/cross-contamination recorded in the

available studies was due to applications performed during previous years on the control plots.

Other sources may be from other treated crops or contaminated plants in the landscape. It is

acknowledged that the same mechanism had the potential to artiﬁcially increase the exposure in the

‘treated’

groups of the experiments, thus potentially amplifying effects expected from the treatment alone.

Nevertheless, widespread use of these substances makes this situation likely to occur in the environment,

and the data should not necessarily be disregarded as uninformative for the present risk assessment.

Overview of the concerns identiﬁed for each representative use

considered

Summary of concerns for each scenario according to the risk assessment scheme in EFSA

(2013c)

Honeybee

Bumblebee

Solitary bee

Table 20:

Use

Spring cereals

Winter cereals

Cotton

Endive and lettuce,

ﬁeld

use,

Harvested before or after

ﬂowering

Endive, lettuce, Brassica, sown in or transplanted to non-

permanent greenhouse,

Harvested before or after

ﬂowering

Endive, lettuce, Brassica, sown and grown in permanent

greenhouse,

Harvested before or after

ﬂowering

Maize

Potato

Spring rape

Winter rape

Sugar and fodder beet

Harvested after

ﬂowering

Sugar and fodder beet

Harvested before

ﬂowering

Managed amenity turf,

‘highly’

managed amenity turf

References

EFSA (European Food Safety Authority), 2008. Conclusion regarding the peer review of the pesticide risk

assessment of the active substance imidacloprid. EFSA Journal 2008;6(5):148, 120 pp.

https://doi.org/10.

2903/j.efsa.2008.148r

EFSA (European Food Safety Authority), 2013a. Conclusion on the peer review of the pesticide risk assessment for

bees for the active substance imidacloprid. EFSA Journal 2013;11(1):3066, 58 pp.

https://doi.org/10.2903/j.

efsa.2013.3066

EFSA (European Food Safety Authority), 2013b. Evaluation of the FERA study on bumblebees and consideration of

its potential impact on the EFSA conclusions on neonicotinoids. EFSA Journal 2013;11(6):3242, 20 pp.

https://doi.org/10.2903/j.efsa.2013.3242

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

EFSA (European Food Safety Authority), 2013c. EFSA Guidance Document on the risk assessment of plant

protection products on bees (Apis

mellifera, Bombus

spp. and solitary bees). EFSA Journal 2013;11(7):3295,

268 pp.

https://doi.org/10.2903/j.efsa.2013.3295

EFSA (European Food Safety Authority), 2014a. A small application developed in R for the estimation of the

residue intake rate for certain bee species under given conditions: the SHVAL tool. EFSA supporting publication

2014;11(7):EN-623, 15 pp.

https://doi.org/10.2903/sp.efsa.2014.EN-623

EFSA (European Food Safety Authority), 2014b. Technical report on the open call for new scientiﬁc information as

regards the risk to bees from the use of the three neonicotinoid pesticide active substances clothianidin,

imidacloprid and thiamethoxam applied as seed treatments and granules in the EU. EFSA supporting

publication 2014;11(10):EN-903, 8 pp.

https://doi.org/10.2903/sp.efsa.2014.EN-903

EFSA (European Food Safety Authority), 2015. Conclusion on the peer review of the pesticide risk assessment for

bees for the active substance imidacloprid considering all uses other than seed treatments and granules. EFSA

Journal 2015;13(8):4210, 77 pp.

https://doi.org/10.2903/j.efsa.2015.4210

EFSA (European Food Safety Authority), 2016a. Technical report on the outcome of the consultation with Member

States, the applicant and EFSA on the pesticide risk assessment for imidacloprid in light of conﬁrmatory data.

EFSA Supporting Publication 2016;13(6):EN-1038, 62 pp.

https://doi.org/10.2903/sp.efsa.2016.EN-1038

EFSA (European Food Safety Authority), 2016b. Conclusion on the peer review of the pesticide risk assessment for

the active substance imidacloprid in light of conﬁrmatory data submitted. EFSA Journal 2016;14(11):4607, 39 pp.

https://doi.org/10.2903/j.efsa.2016.4607

EFSA (European Food Safety Authority), 2016c. Conclusion on the peer review of the pesticide risk assessment for

the active substance clothianidin in light of conﬁrmatory data submitted. EFSA Journal 2016;14(11):4606, 34 pp.

https://doi.org/10.2903/j.efsa.2016.4606

EFSA (European Food Safety Authority), 2018a. Evaluation of the data on clothianidin, imidacloprid and

thiamethoxam for the updated risk assessment to bees for seed treatments and granules in the EU. EFSA

supporting publication 2018:EN-1378, 31 pp.

EFSA (European Food Safety Authority), 2018b. Peer review report to the conclusion regarding the peer review of

the pesticide risk assessment for bees of the active substance imidacloprid. Available online:

www.efsa.

europa.eu

EFSA Scientiﬁc Committee, Hardy A, Benford D, Halldorsson T, Jeger MJ, Knutsen HK, More S, Naegeli H,

Noteborn H, Ockleford C, Ricci A, Rychen G, Schlatter JR, Silano V, Solecki R, Turck D, Benfenati E, Chaudhry

QM, Craig P, Frampton G, Greiner M, Hart A, Hogstrand C, Lambre C, Luttik R, Makowski D, Siani A,

Wahlstroem H, Aguilera J, Dorne J-L, Fernandez Dumont A, Hempen M, Valtuena Mart

ınez

S, Martino L,

Smeraldi C, Terron A, Georgiadis N and Younes M, 2017. Scientiﬁc Opinion on the guidance on the use of the

weight of evidence approach in scientiﬁc assessments. EFSA Journal 2017;15(8):4971, 69 pp.

https://doi.org/

10.2903/j.efsa.2017.4971

European Commission, 2006a. Review report for the active substance clothianidin. Finalised in the Standing

Committee on the Food Chain and Animal Health at its meeting on 27 January 2006 in view of the inclusion of

clothianidin in Annex I of Directive 91/414/EEC. SANCO/10533/05 - Final 18 January 2005.

European Commission, 2006b. Review report for the active substance thiamethoxam. Finalised in the Standing

Committee on the Food Chain and Animal Health at its meeting on 14 July 2006 in view of the inclusion of

thiamethoxam in Annex I of Directive 91/414/EEC. SANCO/10390/2002 - rev.

ﬁnal

14 July 2006.

FOCUS (Forum for the Co-ordination of Pesticide Fate Models and their Use), 1997. Soil persistence models and EU

registration. The

ﬁnal

report of the work of the Soil Modelling Work group of FOCUS. February 1997.

FOCUS (Forum for the Co-ordination of Pesticide Fate Models and their Use), 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, 245 pp., as updated by Generic guidance for

FOCUS surface water scenarios, v. 1.1, March 2012.

FOCUS (Forum for the Co-ordination of Pesticide Fate Models and their Use), 2006. Guidance document on

estimating persistence and degradation kinetics from environmental fate studies on pesticides in EU

Registration Report of the FOCUS Work Group on Degradation Kinetics. EC Document Reference SANCO/

10058/2005-v. 2.0, 434 pp.

Fryday S, Tiede K and Stein J, 2015. Scientiﬁc services to support EFSA systematic reviews: lot 5 Systematic

literature review on the neonicotinoids (namely active substances clothianidin, thiamethoxam and imidacloprid)

and the risks to bees. EFSA supporting publication 2015:EN-756, 656 pp.

Abbreviations

1/n

a.s.

AChE

BBCH

slope of Freundlich isotherm

wavelength

active substance

acetylcholinesterase

growth stages of mono- and dicotyledonous plants

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

EEC

ETR

FOCUS

GAP

HPG

ISO

IUPAC

Foc

LDD

LOD

LOEL

LOQ

NOEC

NOED

NOEL

OECD

PEC

soil

PER

ppm

RUD

SMILES

SPG

SVs

TGW

WoE

body weight

period required for 50% dissipation (deﬁne method of estimation)

effective concentration

European Economic Community

exposure toxicity ratio

Forum for the Co-ordination of Pesticide Fate Models and their Use

Good Agricultural Practice

hypopharyngeal glands

hazard quotient

International Organization for Standardization

International Union of Pure and Applied Chemistry

Freundlich organic carbon adsorption coefﬁcient

lethal dose, median; dosis letalis media

lethal dietary dose; median

limit of detection

lowest observed effect level

limit of quantiﬁcation

millimetre (also used for mean measured concentrations)

no observed effect concentration

no observed effect dose

no observed effect level

Organisation for Economic Co-operation and Development

predicted environmental concentration

predicted environmental concentration in soil

Proboscis Extension Reﬂex

parts per million (10

À6

)

residue intake

residue per unit dose

simpliﬁed molecular-input line-entry system

speciﬁc protection goal

shortcut values

thousand grain weight

weight of evidence

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Appendix A

–

List of supported uses

Appendix

can be found in the online version of this output (‘Supporting information’ section):

https://doi.org/10.2903/j.efsa.2018.5178

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Appendix B

–

Overview of endpoint types and related relevance class

assigned within the scope of the present risk assessment

Specie

Honeybee

Endpoint as deﬁned in the

study/study evaluations

Family of

endpoint

Colony strength

Relevance to the

protection goal

Directly relevant. These

endpoints have a direct link

to the protection goal as the

correct entity was studied

Relevance

Class

•

Colony strength

Number of frames with

bees present

Forager mortality (far

from the hive)

Overwinter success

Overwintering

Overwintering (colony

strength)

Overwintering survival

Forager mortality

Overwintering

assessments

Mortality (in front of the

General mortality of Indirectly relevant. These

individuals

endpoints have an obvious,

hives)

but not quantiﬁed

Worker longevity

conceptual link to the

Foraging lifespan

protection goal (e.g.

Queen survival

Brood production

Brood development

(time)

Brood production (comb

area with eggs, larvae,

pupae)

Brood termination rate

Capped brood

Drone cells

Larvae mortality

Worker production

(yes/no)

mortality measured in front

of the hives consist of

forager mortality, but also

includes an unknown

proportion of mortality of in-

hive bees and do not include

forager mortality occurred in

the

ﬁeld)

Homing failure

Homing success

Number of bees

returning to the hive

Number of returning

foragers

Homing success

Behavioural

Activity index

endpoint

Behaviour

Queen

ﬂight

activity

Duration of foraging trip

Duration of the foraging

bouts

Flight activity (bees

exiting the hive)

Flight behaviour

Foraging/ﬂight intensity

Learning ability/capacity

Mean time between

ﬂights

Number of

ﬂights

per

bee per day

Number of foraging

bouts

Indirectly relevant. These

endpoints likely have a link

to the protection goal, but

this link is not clear and not

quantiﬁed (e.g. a

detrimental effect on the

foraging behaviour may

contribute to the weakening

of the colony, but the link

between the two entities is

unknown)

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Specie

Endpoint as deﬁned in the

study/study evaluations

Family of

endpoint

Relevance to the

protection goal

Relevance

Class

•

Number of visits to the

feeder

Time to

ﬁrst

foraging

Time to

ﬁrst

foraging

trip

Total

ﬂight

time per bee

Swarming events

Pollination efﬁciency

Comb cell production

Comb construction

Comb building

Hive weight (including

Weight of the hive

hive, combs, food etc.)

Hive weight (including

hive, wax, food stores

etc.)

Weight gain of the hives

Disease

Varroa mite count

(prevalence)

Viral load of Black

queen cell virus (BQCV)

Viral load of Deformed

wing virus (DWV)

Viral load of Israeli

acute paralysis virus

(IAPV)

Food storage

Honey storage (area/

weight)

Pollen storage

Queen

rearing success

supersedure

survival

ﬂight

activity

Food storage

Queen

Behaviour

Food consumption

inﬂuencing

Guttation

ﬂuid

use

Quantity of taken syrup

exposure

Suborganism

Expression of AChE

endpoint

gene

Expression of genes

Apoptosis of neurons

Assessment of HPG

Lipid content of royal

jelly

Total haemocyte count

Encapsulation response

Not relevant. These

endpoints have no explicit

link to the protection goals

Ovary weight of queens

Subindividual mass

Queen subindividual

reproductive system

Thermoregulation

capacity

Thermoregulation

capacity

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Specie

Bumblebee

Endpoint as deﬁned in the

study/study evaluations

Family of

endpoint

Colony strength

Relevance to the

protection goal

Directly relevant. These

endpoints have a direct link

to the protection goal as the

correct entity was studied

Relevance

Class

•

Colony strength

Number of foragers

Brood production (comb

Reproductive

area with eggs, larvae,

output

pupae)

Male vs. queen

production

Number of brood

Drone production

Egg production

No. of drones

No. of young queens

and queen brood cells

Produced queens (BB)

Produced workers (BB)

Queen production

Queens produced

Worker production

Worker/males produced

Reproductive output

[queenless

microcolonies]

Emergence rates

Mating ability

Total cast biomass

Queen mortality

Indirect

reproduction

•

Foraging/ﬂight intensity

Behaviour

Learning ability

Duration of

ﬂower

visit

(mean)

Duration of the foraging

bouts

Duration of the foraging

bouts for pollen only

Flower visitation rate

(ﬂowers/bee per

minute)

Foraging time

Nectar load size

Number of forager

carrying pollen

Number of foraging

bouts

Number of foraging

days

Number of switches

between

ﬂower

varieties

Pollen load size

Indirectly relevant. These

endpoints have an obvious,

but not quantiﬁed

conceptual link to the

protection goal (e.g.

emergence contributes to

the reproductive success,

but alone does not

determine the reproductive

output)

Indirectly relevant. These

endpoints likely have a link

to the protection goal, but

this link is not clear and not

quantiﬁed (e.g. a

detrimental effect on the

foraging behaviour may

contribute to the weakening

of the colony, but the link

between the two entities is

unknown)

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Specie

Endpoint as deﬁned in the

study/study evaluations

Family of

endpoint

Relevance to the

protection goal

Relevance

Class

•

Solitary bee

Proportion of visit to a

certain

ﬂower

variety

Time between

ﬂower

visits

Visited

ﬂowers

Learning ability (PER)

Hive weight (including

Weight of the nest

hive, wax, food stores

(colony)

etc.)

Weight gain of the hives

Honey storage (area/

weight)

Pollen storage

Food collection

Food storage

General mortality

Worker longevity

Dead bees (in the nest)

Death rate

Forager age

Forager mortality (far

from the hives)

Mortality (in front of the

hives)

Worker mortality

Queen mortality (old

queens)

Queen longevity

Mortality

Dead larvae

Homing success

Number of returning

foragers

BW (of individuals)

Weight of newly

emerged bees

Average bee weight

Cocoon weight

Worker size

Homing success

Individual mass

Quantity of taken syrup

Behaviour

Sugar water collection

inﬂuencing

exposure

Not relevant. This endpoint

has no explicit link to the

protection goals

Directly relevant. These

endpoints have a direct link

to the protection goal as the

correct entity was studied

•

Emergence after winter

Reproductive

output

Hatching success

Offspring failure

Overwinter success

Overwintering

(hatching)

Cell production

Completed nests

Nest occupation

Number of tubes with

brood cells

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Specie

Endpoint as deﬁned in the

study/study evaluations

Family of

endpoint

Relevance to the

protection goal

Relevance

Class

•

Offspring (cocoon)

production

Reproduction rate

Undeveloped eggs and

larvae

Cocoon weight

Offspring weight

Sex ratio after

emergence

Cocoon sex

Indirect

Indirectly relevant. These

reproductive output endpoints have an obvious,

but not quantiﬁed

conceptual link to the

protection goal

Indirectly relevant. These

endpoints likely have a link

to the protection goal, but

this link is not clear and not

quantiﬁed (e.g. a

detrimental effect on the

foraging intensity may have

an indirect effect on the

reproductive success, but

the link between the two

entities is unknown)

Foraging/ﬂight intensity

Behaviour

Forager mortality

General mortality

AChE: acetylcholinesterase; BW: body weight.

Notes: Colony survival length was assessed in a single experiment C.1290. As it is a Class 1 endpoint, in principle it should have

been considered further in the assessment. However, as the endpoint was assessed to be unreliable, no further consideration

was given to this endpoint type.

Non-Apis bee monitoring was assessed in a single experiment C.1184. As it is a Class 1 endpoint, in principle it should have been

considered further in the assessment. However, as the endpoint was assessed to be unreliable, no further consideration was

given to this endpoint type.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Appendix C

–

Tier-1 risk assessment based on EFSA (2013c)

Contact exposure and exposure to consumption of contaminated pollen and nectar

1) Seed treatment

Spring cereals 112 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

HQ: hazard quotient.

Bumblebee

5.1

Trigger

2.3

Solitary bee

441.8

Trigger

2.6

44.2

Trigger

Spring cereals 112 g a.s./ha, 0.039 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

Field margin

Adjacent crop

Chronic

Succeeding crop

Treated crop

Field margin

Adjacent crop

Larva

Succeeding crop

Treated crop

Field margin

Adjacent crop

Succeeding crop

ETR: exposure toxicity ratio.

Bumblebee

ETR

0.03

0.19

0.16

2.65

4.15

23.20

18.87

309.79

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

1.05

6.89

8.28

148.32

1.38

9.04

10.87

194.61

–

Trigger

0.04

0.0054

0.2

ETR

0.13

1.11

1.10

21.19

0.17

1.14

1.11

21.45

0.01

0.46

0.45

8.48

Trigger

0.2

0.03

0.2

Winter cereals 43 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

17.0

Trigger

Bumblebee

2.0–19.5*

Trigger

2.3

Solitary bee

169.6

Trigger

2.6

HQ: hazard quotient.

*: The higher value reﬂects a scenario sowing without deﬂector as considered in EFSA, 2013c.

Winter cereals 43 g a.s./ha, 0.015 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

Field margin

Adjacent crop

Succeeding crop

Chronic

Treated crop

Field margin

ETR

0.05

0.43

0.42

8.14

0.06

0.44

Trigger

0.2

0.03

Bumblebee

ETR

0.01

0.07

0.06

1.02

1.60

8.91

Trigger

0.036

0.0048

Solitary bee

ETR

0.41

2.65

3.18

56.95

0.53

3.47

Trigger

0.04

0.0054

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Honeybee

Category

Scenario

Adjacent crop

Succeeding crop

Larva

Treated crop

Field margin

Adjacent crop

Succeeding crop

ETR

0.42

8.23

0.01

0.18–1.77*

0.17–1.72*

3.26

Trigger

0.03

0.2

Bumblebee

ETR

7.25

118.94

–

Trigger

0.0048

0.2

Solitary bee

ETR

4.17

74.72

–

Trigger

0.0054

0.2

ETR: exposure toxicity ratio.

*: The higher value reﬂects a scenario sowing without deﬂector as considered in EFSA (2013c).

Spring and Winter cereals 126 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

HQ: hazard quotient.

Bumblebee

5.7

Trigger

2.3

Solitary bee

497.0

Trigger

2.6

49.7

Trigger

Spring and Winter cereals 126 g a.s./ha, 0.039 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

Field margin

Adjacent crop

Chronic

Succeeding crop

Treated crop

Field margin

Adjacent crop

Larva

Succeeding crop

Treated crop

Field margin

Adjacent crop

Succeeding crop

ETR

0.13

1.25

1.24

23.84

0.17

1.28

1.24

24.13

0.01

0.52

0.50

9.55

Trigger

0.2

0.03

0.2

Bumblebee

ETR

0.03

0.21

0.18

2.98

4.15

26.10

21.23

348.51

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

1.05

7.75

9.32

166.86

1.38

10.17

12.22

218.94

–

Trigger

0.04

0.0054

0.2

Cotton 100 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

ﬁeld

margin

HQ: hazard quotient.

Bumblebee

7.8

Trigger

2.3

Solitary bee

677.3

Trigger

2.6

67.7

Trigger

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Cotton 100 g a.s./ha, 0.63 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

Field margin

Adjacent crop

Succeeding crop

Chronic

Treated crop

Field margin

Adjacent crop

Succeeding crop

Larva

Treated crop

Field margin

Adjacent crop

Succeeding crop

ETR: exposure toxicity ratio.

Bumblebee

ETR

14.92

0.29

0.24

2.37

1742.55

35.15

28.44

276.60

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

834.32

10.44

12.48

132.43

1094.68

13.70

16.37

173.76

–

Trigger

0.04

0.0054

0.2

ETR

119.19

1.68

1.66

18.92

120.64

1.73

1.67

19.15

47.73

0.70

0.68

7.58

Trigger

0.2

0.03

0.2

Cotton 175 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

HQ: hazard quotient.

Bumblebee

13.6

Trigger

2.3

Solitary bee

1185.3

Trigger

2.6

118.5

Trigger

Cotton 175 g a.s./ha, 0.84 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

Field margin

Adjacent crop

Chronic

Succeeding crop

Treated crop

Field margin

Adjacent crop

Larva

Succeeding crop

Treated crop

Field margin

Adjacent crop

Succeeding crop

ETR: exposure toxicity ratio.

Bumblebee

ETR

19.89

0.50

0.42

4.14

2323.40

61.51

49.76

484.04

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

1112.43

18.28

21.84

231.76

1459.57

23.98

28.65

304.08

–

Trigger

0.04

0.0054

0.2

ETR

158.92

2.94

2.91

33.11

160.85

3.02

2.92

33.51

63.64

1.23

1.18

13.26

Trigger

0.2

0.03

0.2

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Endive 104 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

HQ: hazard quotient.

Bumblebee

8.1

Trigger

2.3

Solitary bee

704.4

Trigger

2.6

70.4

Trigger

Endive 104 g a.s./ha, 0.8 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop*

Field margin

Adjacent crop

Succeeding crop

Chronic

Treated crop*

Field margin

Adjacent crop

Succeeding crop

Larva

Treated crop*

Field margin

Adjacent crop

Succeeding crop

ETR

151.35

1.75

1.73

19.68

153.19

1.80

1.73

19.91

60.61

0.73

0.70

7.88

Trigger

0.2

0.03

0.2

Bumblebee

ETR

18.95

0.30

0.25

2.46

2212.77

36.55

29.57

287.66

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

1059.46

10.86

12.98

137.73

1390.07

14.25

17.03

180.71

–

Trigger

0.04

0.0054

0.2

ETR: exposure toxicity ratio.

*: Not relevant when the crop is harvested before it

ﬂowers.

Note: The

‘Leafy

vegetable’ scenario was considered in these calculations above.

Lettuce 104 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

HQ: hazard quotient.

Bumblebee

8.1

Trigger

2.3

Solitary bee

704.4

Trigger

2.6

70.4

Trigger

Lettuce 104 g a.s./ha, 0.8 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop*

Field margin

Adjacent crop

Chronic

Succeeding crop

Treated crop*

Field margin

Adjacent crop

Larva

Succeeding crop

Treated crop*

Field margin

ETR

2.59

1.75

1.73

19.68

3.40

1.80

1.73

19.91

0.30

0.73

Trigger

0.2

0.03

0.2

Bumblebee

ETR

18.95

0.30

0.25

2.46

85.11

36.55

29.57

287.66

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

1059.46

10.86

12.98

137.73

28.37

14.25

17.03

180.71

–

Trigger

0.04

0.0054

0.2

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Honeybee

Category

Scenario

Adjacent crop

Succeeding crop

ETR

0.70

7.88

Trigger

0.2

Bumblebee

ETR

–

Trigger

0.2

Solitary bee

ETR

–

Trigger

0.2

ETR: exposure toxicity ratio.

*: Not relevant when the crop is harvested before it

ﬂowers.

Maize 60 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

ﬁeld

margin

HQ: hazard quotient.

Bumblebee

4.7

Trigger

2.3

Solitary bee

406.4

Trigger

2.6

40.6

Trigger

Maize 60 g a.s./ha, 1.0 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

Field margin

Adjacent crop

Succeeding crop

Chronic

Treated crop

Field margin

Adjacent crop

Succeeding crop

Larva

Treated crop

Field margin

Adjacent crop

Succeeding crop

ETR: exposure toxicity ratio.

Bumblebee

ETR

0.79

0.17

0.14

1.42

106.38

21.09

17.06

165.96

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

27.03

6.27

7.49

79.46

35.46

8.22

9.82

104.26

–

Trigger

0.04

0.0054

0.2

ETR

3.24

1.01

1.00

11.35

4.26

1.04

1.00

11.49

0.38

0.42

0.41

4.55

Trigger

0.2

0.03

0.2

Maize 100 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

HQ: hazard quotient.

Bumblebee

7.8

Trigger

2.3

Solitary bee

677.3

Trigger

2.6

97.7

Trigger

Maize 100 g a.s./ha, 1.0 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

Field margin

Adjacent crop

succeeding crop

ETR

3.24

1.68

1.66

18.92

Trigger

0.2

Bumblebee

ETR

0.79

0.29

0.24

2.37

Trigger

0.036

Solitary bee

ETR

27.03

10.44

12.48

132.43

Trigger

0.04

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Honeybee

Category

Chronic

Scenario

Treated crop

Field margin

Adjacent crop

Larva

Succeeding crop

Treated crop

Field margin

Adjacent crop

Succeeding crop

ETR: exposure toxicity ratio.

Bumblebee

ETR

106.38

35.15

28.44

276.60

–

Trigger

0.0048

0.2

Solitary bee

ETR

35.46

13.70

16.37

173.76

–

Trigger

0.0054

0.2

ETR

4.26

1.73

1.67

19.15

0.38

0.70

0.68

7.58

Trigger

0.03

0.2

Potato 120 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

N/R

Trigger

N/R

Bumblebee

Trigger

2.3

N/R

Solitary bee

Trigger

2.6

N/R: not relevant; HQ: hazard quotient.

Potato 120 g a.s./ha

(application rate expressed as mg a.s./tuber was not available)

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

scenario

Treated crop

Field margin

Adjacent crop

Succeeding crop

Chronic

Treated crop

Field margin

Adjacent crop

Succeeding crop

Larva

Treated crop

Field margin

Adjacent crop

Succeeding crop

N/R: not relevant; ETR: exposure toxicity ratio.

Bumblebee

ETR

–

N/R

2.84

–

N/R

331.91

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

–

N/R

158.92

–

N/R

208.51

–

Trigger

0.04

0.0054

0.2

ETR

–

N/R

22.70

–

N/R

22.98

–

N/R

9.09

Trigger

0.2

0.03

0.2

Potato 180 g a.s./ha

Acute contact exposure

–

Honeybee

scenario

Field margin

N/R

Trigger

N/R

Bumblebee

Trigger

2.3

Solitary bee

N/R

Trigger

2.6

N/R: not relevant; HQ: hazard quotient.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Potato 180 g a.s./ha

(application rate expressed as mg a.s./tuber was not available)

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

Feld margin

Adjacent crop

Chronic

Succeeding crop

Treated crop

Field margin

Adjacent crop

Larva

Succeeding crop

Treated crop

Field margin

Adjacent crop

Succeeding crop

N/R: not relevant; ETR: exposure toxicity ratio.

Bumblebee

ETR

–

N/R

4.26

–

N/R

497.87

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

–

N/R

238.38

–

N/R

312.77

–

Trigger

0.04

0.0054

0.2

ETR

–

N/R

34.05

–

N/R

34.47

–

N/R

13.64

Trigger

0.2

0.03

0.2

Spring and winter rape 9 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

2.4–23.7*

Trigger

Bumblebee

0.3–2.7*

Trigger

2.3

Solitary bee

23.7

Trigger

2.6

HQ: hazard quotient.

*: The higher value reﬂects a scenario sowing without deﬂector as considered in EFSA, 2013c.

Spring and winter rape 9 g a.s./ha, 0.01 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

Field margin

Adjacent crop

Succeeding crop

Chronic

Treated crop

Field margin

Adjacent crop

Succeeding crop

Larva

Treated crop

Field margin

Adjacent crop

Succeeding crop

ETR

1.89

0.06–0.59*

0.06–0.61*

1.70

1.91

0.06

1.72

0.76

0.02

0.68

Trigger

0.2

0.03

0.2

Bumblebee

ETR

0.24

0.01–0.1*

0.01–0.09*

0.21

27.66

1.24

1.04

24.89

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

13.24

0.37

0.46

11.92

17.38

0.48

0.60

15.64

–

Trigger

0.04

0.0054

0.2

ETR: exposure toxicity ratio.

*: The higher value reﬂects a scenario sowing without deﬂector as considered in EFSA, 2013c.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Spring and winter rape 12 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

3.2–31.6*

Trigger

Bumblebee

0.4–3.6*

Trigger

2.3

Solitary bee

31.6

Trigger

2.6

HQ: hazard quotient.

*: The higher value reﬂects a scenario sowing without deﬂector as considered in EFSA, 2013a.

Spring and winter rape 12 g a.s./ha, 0.01 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

Field margin

Adjacent crop

Chronic

Succeeding crop

Treated crop

Field margin

Adjacent crop

Larva

Succeeding crop

Treated crop

Field margin

Adjacent crop

Succeeding crop

ETR

1.89

0.08–0.79*

0.08–0.81*

2.27

1.91

0.08

2.30

0.76

0.03–0.33*

0.91

Trigger

0.2

0.03

0.2

Bumblebee

ETR

0.24

0.01–0.14*

0.01–0.12*

0.28

27.66

1.66

1.39

33.19

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

13.24

0.49

0.61

15.89

17.38

0.65

0.80

20.85

–

Trigger

0.04

0.0054

0.2

ETR: exposure toxicity ratio.

*: The higher value reﬂects a scenario sowing without deﬂector as considered in EFSA, 2013c.

Sugar and fodder beet 13 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

0.0–0.2*

Trigger

Bumblebee

0.00–0.002*

Trigger

2.3

Solitary bee

0.2–1.6*

Trigger

2.6

HQ: hazard quotient.

*: The higher value reﬂects a scenario sowing without deﬂector as considered in EFSA, 2013c.

Note: The

‘Sugar

beet’ scenario was considered for both the sugar beet and the fodder beet crops.

Sugar and fodder beet 13 g a.s./ha, 0.1 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category Scenario

Acute

Treated crop**

Field margin

Adjacent crop

Succeeding crop

Chronic

Treated crop**

Field margin

Adjacent crop

Succeeding crop

ETR

18.92

0.00–0.00*

2.46

19.15

0.00–0.00*

2.49

Trigger

0.2

0.03

Bumblebee

ETR

2.37

0.00–0.00*

0.00-0.00*

0.31

276.60

0.01

35.96

Trigger

0.036

0.0048

Solitary bee

ETR

132.43

0.00–0.02*

0.00–0.03*

17.22

173.76

0.00–0.03*

0.00–0.04*

22.59

Trigger

0.04

0.0054

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Honeybee

Category Scenario

Larva

Treated crop**

Field margin

Adjacent crop

Succeeding crop

ETR

7.58

0.00–0.00*

0.98

Trigger

0.2

Bumblebee

ETR

–

Trigger

0.2

Solitary bee

ETR

–

Trigger

0.2

ETR: exposure toxicity ratio.

*: The higher value reﬂects a scenario sowing without deﬂector as considered in EFSA, 2013c

**: Not relevant when the crop is harvested before it

ﬂowers

Note: The

‘Sugar

beet’ scenario was considered for both the sugar beet and the fodder beet crops.

Sugar and fodder beet 117 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin

0.1–1.4*

Trigger

Bumblebee

0.0–0.2*

Trigger

2.3

Solitary bee

1.4–14.0*

Trigger

2.6

HQ: hazard quotient.

*: The higher value reﬂects a scenario sowing without deﬂector as considered in EFSA, 2013c

Note: The

‘Sugar

beet’ scenario was considered for both the sugar beet and the fodder beet crops.

Sugar and fodder beet 117 g a.s./ha, 0.9 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category Scenario

Acute

Treated crop**

Field margin

Adjacent crop

Chronic

Succeeding crop

Treated crop**

Field margin

Adjacent crop

Larva

Succeeding crop

Treated crop**

Field margin

Adjacent crop

Succeeding crop

ETR

170.3

0.00–0.04*

22.14

172.3

0.00–0.04*

22.40

68.2

0.00–0.01*

8.86

Trigger

0.2

0.03

0.2

Bumblebee

ETR

21.3

0.00–0.01*

2.77

2489.3

0.07

0.06

323.62

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

1191.9

0.02–0.22*

0.03–0.27*

154.95

1563.8

0.03

0.04

203.30

–

Trigger

0.04

0.0054

0.2

ETR: exposure toxicity ratio;

*: The higher value reﬂects a scenario sowing without deﬂector as considered in EFSA, 2013c

**: Not relevant when the crop is harvested before it

ﬂowers

Note: The

‘Sugar

beet’ scenario was considered for both the sugar beet and the fodder beet crops.

2) Seed treatment as dummy pill uses

Brassica,

ﬂowering,

head, leafy, Endive, Lettuce, 90 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

Field margin*

61.0

Trigger

7.0

Bumblebee

Trigger

2.3

Solitary bee

609.6

Trigger

2.6

HQ: hazard quotient.

*: Not relevant when the application is done in permanent greenhouse.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Brassica,

ﬂowering,

head, leafy, 90 g a.s./ha, 1.5 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

(a),(b)

Field margin

(c)

Adjacent crop

(c)

Succeeding crop

(b)

Chronic

Treated crop

(a),(b)

Field margin

(c)

Adjacent crop

(c)

Succeeding crop

(b)

Larva

Treated crop

(a,b)

Field margin

(c)

Adjacent crop

(c)

Succeeding crop

(b)

Bumblebee

ETR

35.53

0.26

0.21

2.13

4148.94

31.63

25.59

248.94

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

1986.49

9.40

11.23

119.19

2606.38

12.33

14.74

156.38

–

Trigger

0.04

0.0054

0.2

ETR

283.78

1.51

1.50

17.03

287.23

1.55

1.50

17.23

113.64

0.63

0.61

6.82

Trigger

0.2

0.03

0.2

ETR: exposure toxicity ratio.

(a): Not relevant when the crop is harvested before it

ﬂowers.

(b): Not relevant when the crop is grown in permanent structure greenhouse.

(c): Not relevant when the application is done in permanent greenhouse.

Endive, Lettuce 89 g a.s./ha

Acute contact exposure

–

Honeybee

Scenario

ﬁeld

margin*

60.3

Trigger

6.9

Bumblebee

Trigger

2.3

Solitary bee

602.8

Trigger

2.6

HQ: hazard quotient.

*: Not relevant when the application is done in permanent greenhouse.

Endive, 89 g a.s./ha, 0.8 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

(a),(b)

Field margin

(c)

Adjacent crop

(c)

Chronic

Succeeding crop

(b)

Treated crop

(a),(b)

Field margin

(c)

Adjacent crop

(c)

Larva

Succeeding crop

(b)

Treated crop

(a),(b)

Field margin

(c)

Adjacent crop

(c)

Succeeding crop

(b)

ETR

151.35

1.50

1.48

16.84

153.19

1.54

1.48

17.04

60.61

0.62

0.60

6.74

Trigger

0.2

0.03

0.2

Bumblebee

ETR

18.95

0.26

0.21

2.11

2212.77

31.28

25.31

246.17

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

1059.46

9.29

11.11

117.86

1390.07

12.19

14.57

154.65

–

Trigger

0.04

0.0054

0.2

ETR: exposure toxicity ratio.

(a): Not relevant when the crop is harvested before it

ﬂowers.

(b): Not relevant when the crop is grown in permanent structure greenhouse.

(c): Not relevant when the application is done in permanent greenhouse.

Note: The

‘Leafy

vegetable’ scenario was considered in the calculations above.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Lettuce, 89 g a.s./ha, 0.8 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

(a),(b)

Field margin

Adjacent crop

(c)

Chronic

Succeeding crop

(b)

Treated crop

(a),(b)

Field margin

(c)

Adjacent crop

(c)

Larva

Aucceeding crop

(b)

Treated crop

(a),(b)

Field margin

(c)

Adjacent crop

(c)

Succeeding crop

(b)

(c)

Bumblebee

ETR

18.95

0.26

0.21

2.11

85.11

31.28

25.31

246.17

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

1059.46

9.29

11.11

117.86

28.37

12.19

14.57

154.65

–

Trigger

0.04

0.0054

0.2

ETR

2.59

1.50

1.48

16.84

3.40

1.54

1.48

17.04

0.30

0.62

0.60

6.74

Trigger

0.2

0.03

0.2

ETR: exposure toxicity ratio.

(a): Not relevant when the crop is harvested before it

ﬂowers.

(b): Not relevant when the crop is grown in permanent structure greenhouse.

(c): Not relevant when the application is done in permanent greenhouse.

Endive, Lettuce, 120 g a.s./ha

Acute contact exposure

–

Honeybee

scenario

ﬁeld

margin*

81.3

Trigger

9.4

Bumblebee

Trigger

2.3

Solitary bee

812.7

Trigger

2.6

HQ: hazard quotient.

*: Not relevant when the application is done in permanent greenhouse.

Endive, 120 g a.s./ha, 1.2 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Scenario

Treated crop

(a),(b)

Field margin

(c)

Adjacent crop

(c)

Succeeding crop

(b)

Chronic

Treated crop

(a),(b)

Field margin

(c)

Adjacent crop

(c)

Succeeding crop

(b)

Larva

Treated crop

(a),(b)

Field margin

(c)

Adjacent crop

(c)

Succeeding crop

(b)

ETR

227.03

2.02

2.00

22.70

229.79

2.07

2.00

22.98

90.91

0.84

0.81

9.09

Trigger

0.2

0.03

0.2

Bumblebee

ETR

28.42

0.34

0.29

2.84

3319.15

42.18

34.12

331.91

–

Trigger

0.036

0.0048

0.2

Solitary bee

ETR

1589.19

12.53

14.97

158.92

2085.11

16.44

19.65

208.51

–

Trigger

0.04

0.0054

0.2

ETR: exposure toxicity ratio.

(a): Not relevant when the crop is harvested before it

ﬂowers.

(b): Not relevant when the crop is grown in permanent structure greenhouse.

(c): Not relevant when the application is done in permanent greenhouse.

Note: The

‘Leafy

vegetable’ scenario was considered in these calculations above.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Lettuce, 120 g a.s./ha, 1.2 mg a.s./seed

Acute, chronic and larvae oral exposure

–

ETRs

Honeybee

Category

Acute

Chronic

Larva

Water uptake

11.4

lL/bee

per day

11.4

lL/bee

per day

111

lL/larva

per 5 days

ETR

1,889

1,338

9,279

Trigger

0.2

0.03

0.2

ETR: exposure toxicity ratio.

Note: calculations based on the water solubility of 613 mg/L (demineralised water, pH 5.5, 20°C), EFSA, 2008.

Tier-1

–

Risk from residues in surface water

All outdoor

ﬁeld

uses (represented by winter cereals)*

Honeybee

Category

Acute

Chronic

Larva

Water uptake

11.4

lL/bee

per day

11.4

lL/bee

per day

111

lL/larva

per 5 days

ETR

0.0002

0.0003

0.0016

Trigger

0.2

0.03

0.2

ETR: exposure toxicity ratio.

*: Represented by the highest PEC surface water of 0.075

lg/L,

which was calculated for the D6 ditch scenario from the use on

winter cereals.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Appendix D

–

Measured residue values and RUD values used for calculation of exposure assessment goals

Exposure

assessment

goal

Study

type

Field

Semiﬁeld

Field

Semiﬁeld

Field

Maximum

Seed

measured

residue value

mg/seed

mg/kg

0.0191

0.0105

(a)

0.0398

0.0211

(a)

0.01

(a)

0.0488

(a)

0.0488

(a)

0.0191

(a)

0.0105

(a)

0.0398

0.0211

(a)

0.01

(a)

0.0003

0.0008

0.0015

0.005

0.009

0.0008

0.001

0.0076

0.0003

0.0002

0.005

0.01

0.0002

Crop

Matrix

Site

RUD mg/kg

Reference

Honeybee pollen Winter oilseed

rape

for winter

oilseed rape

Spring oilseed

rape

Sorted oilseed rape pollen from foragers

Pollen from pollen trap

Pollen from foragers

Sorted oilseed rape pollen from foragers

Pollen from pollen trap

Sorted canola pollen from pollen trap

Germany

Poland

Germany

Poland

Canada

USA

Germany

Poland

Germany

Poland

0.015707

0.07619

0.037688

0.125628

0.42654

0.08

0.02049

0.155738

0.015707

0.019048

0.125628

0.473934

0.02

I.1094

All+.1080

E 370 1548-8

E 370 1553-4

I.1095

All+.1080

110403 CAN/US

I.1094

All+.1080

E 370 1548-8

E 370 1553-4

I.1095

All+.1080

Honeybee

Winter oilseed

nectar for winter rape

oilseed rape

Spring oilseed

rape

Nectar from foragers

Nectar from

ﬂowers

Nectar from

ﬂowers

Nectar from

ﬂowers

Nectar from foragers

Nectar from

ﬂowers

RUD: residue per unit dose.

(a): Seed loading in terms of mg a.s./seed not provided in the study and could not be calculated from the available information. The RUD was therefore calculated assuming thousand grain weight

(TGW) of 5 g/1,000 seeds.

Reference

Exposure

assessment

goal

Crop

Matrix

No of sampling

events during

ﬂowering

Tot No

samples

LOQ

lg/kg

0.3

0.8

Range

lg/kg

0.3

0.8

1.5

samples

≤

LOQ

samples

No samples

LOQ

–

lg/kg

I.1094

All+.1080

E 370 1548-8

E 370 1553-4

Honeybee pollen Winter

oilseed

for winter

rape

oilseed rape

Sorted oilseed rape

pollen from foragers

Pollen from pollen trap 4, from 3 to 4

day duration

Pollen from foragers

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Reference

Exposure

assessment

goal

Crop

Matrix

No of sampling

events during

ﬂowering

Tot No

samples

Not reported

LOQ

lg/kg

0.3

0.8

0.3

0.2

0.3

0.2

Range

lg/kg

0.3–9

0.8

1–7.6

0.3

0.2

0.3–10

0.2

samples

≤

LOQ

All

No samples

samples

LOQ

–

lg/kg

I.1095

All+.1080

110403 CAN/US

I.1094

All+.1080

E 370 1548-8

E 370 1553-4

I.1095

All+.1080

LOQ: limit of quantiﬁcation.

Spring

oilseed

rape

Sorted oilseed rape

pollen from foragers

Pollen from pollen trap 4, from 3 to 4

day duration

Pollen from pollen trap 2

Sorted canola pollen

from pollen trap

Honeybee nectar Winter

oilseed

for winter

rape

oilseed rape

Spring

oilseed

rape

Nectar from foragers

Nectar from

ﬂowers

Nectar from

ﬂowers

Nectar from

ﬂowers

Nectar from foragers

Nectar from

ﬂowers

‘Several

times’

Not reported

‘Several

times’

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Appendix E

–

Residue intake in higher tier studies providing reliable Class 1 and Class 2 endpoints

Mean residue

values/test

concentration/

test dose

Pollen:

2.7

lg/kg;

nectar

0.75

lg/kg

Forager

ng/bee

per day

Min

0.16

max

0.64

Nurse

ng/bee

per day

Min

0.188

max

0.282

Larva

ng/bee per

developmental period

Min

0.301

max

0.302

The exposure length was

11 days. Mean residue for pollen

was calculated from

ﬁve

samples

(includes positive

ﬁndings,

but

also values

LOD). Mean

residue for nectar was derived

from LOQ and LOD

The exposure length was

18 days. Mean residue was

calculated from

ﬁve

samples

(includes positive

ﬁndings,

but

also values

LOQ)

The exposure length was 15

days. Mean residues for pollen

and nectar were derived from

LOQ and LOD

The exposure length was

11 days. Mean residue for pollen

was calculated from

ﬁve

samples

that consisted of values

LOD

and

LOQ). All values for nectar

was

LOD

The exposure length was

11 days. Mean residue for pollen

was calculated from

ﬁve

samples

(includes positive

ﬁndings,

but

also values

LOD). Mean

residue for nectar was derived

from a positive sample and from

a value

LOD

Notes

Organism Reference Type

Exposure

Honeybee

I.1498

Field

Pollen/nectar

from

sunﬂower

ﬁeld

I.2043

Field

Pollen from

maize

ﬁeld

1.51

lg/kg

0.01

0.018

0.002

0.003

I.2044

Field

I.2045

Field

Pollen/nectar

from

sunﬂower

ﬁeld

Pollen/nectar

from winter

oilseed rape

ﬁeld

Pollen:

2.18

lg/kg;

nectar:

1.05

lg/kg

Pollen:

0.13

lg/kg;

nectar:

0.0

lg/kg

0.224

0.896

0.252

0.376

0.419

0.42

0.0008 0.0016

0.0002

0.0003

I.2046

Field

Pollen/nectar

from spring

oilseed rape

ﬁeld

Pollen:

2.2

lg/kg;

nectar:

5.0

lg/kg

1.067

4.267

1.148

1.693

1.983

1.984

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Organism Reference Type

Exposure

Mean residue

values/test

concentration/

test dose

12.5 ppb

Forager

ng/bee

per day

Min

0.8

max

3.2

Nurse

ng/bee

per day

Min

0.85

max

1.25

Larva

ng/bee per

developmental period

Min

1.485

max

1.485

The exposure period was

42 days. Intake obtained with

default consumption values and

50% sugar content

The exposure period was 42

days. Intake obtained with

default consumption values and

50% sugar content

The exposure period was

42 days. Intake obtained with

default consumption values and

50% sugar content

Exposure period was 84 days.

Intake obtained with default

consumption values and 67%

sugar content (pollen patties

consisted pollen and sugar

syrup)

Exposure period was 84 days.

Intake obtained with default

consumption values and 67%

sugar content (pollen patties

consisted pollen and sugar

syrup)

Exposure period was 84 days.

Intake obtained with default

consumption values and 67%

sugar content (pollen patties

consisted pollen and sugar

syrup)

Notes

I.163

Colony Sugar syrup

feeder

I.163

Colony Sugar syrup

feeder

25 ppb

1.6

6.4

1.7

2.5

2.97

I.163

Colony Sugar syrup

feeder

50 ppb

3.2

12.8

3.4

5.94

I.362

Colony Pollen patties 5

lg/kg

feeder

0.24

0.96

0.29

0.43

0.45

I.2017

Colony Pollen patties 5

lg/kg

feeder

0.24

0.96

0.29

0.43

0.45

I.2017

Colony Pollen patties 100

lg/kg

feeder

4.78

19.10

5.73

8.66

9.02

9.07

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Organism Reference Type

Exposure

Mean residue

values/test

concentration/

test dose

0.5

lg/kg

Forager

ng/bee

per day

Min

0.032

max

0.128

Nurse

ng/bee

per day

Min

0.034

max

0.05

Larva

ng/bee per

developmental period

Min

0.059

max

0.059

The exposure period was

33 days. Intake obtained with

default consumption values and

50% sugar content

The exposure period was

33 days. Intake obtained with

default consumption values and

50% sugar content

The exposure period was

63 days. Intake obtained with

default consumption values and

50% sugar content was

assumed

The exposure period was

63 days. Intake obtained with

default consumption values and

50% sugar content was

assumed

Notes

I.411

Colony Sugar syrup

feeder

I.411

Colony Sugar syrup

feeder

lg/kg

0.32

1.28

0.34

0.5

0.594

I.843

Colony Sugar syrup

feeder

400

lg/kg

25.6

102.4

27.2

40.0

47.52

I.843

Colony Sugar syrup

feeder

lg/kg

1.28

5.12

1.36

2.376

C+I.844

I.178

Colony Sugar syrup

feeder

Colony Sugar syrup

feeder

136

lg/L

100 ppb

8.704

6.4

34.816

25.6

9.248 13.6

16.157

Acute exposure regime. Intake

obtained with default

consumption values and 50%

sugar content

Acute exposure regime. Intake

obtained with default

consumption values and 50%

sugar content

Acute exposure regime. Test

dose provided by the study

authors

I.178

Colony Sugar syrup

feeder

500 ppb

128

C+I.423

Colony Sugar syrup

feeder

7.5 ng/bee

7.5

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Organism Reference Type

Exposure

Mean residue

values/test

concentration/

test dose

11.25 ng/bee

Forager

ng/bee

per day

Min

11.25

max

11.25

Nurse

ng/bee

per day

Min

max

Larva

ng/bee per

developmental period

Min

max

Acute exposure regime. Test

dose provided by the study

authors

Acute exposure regime. Test

dose provided by the study

authors

Acute exposure regime. Test

dose provided by the study

authors

Notes

C+I.423

Colony Sugar syrup

feeder

Colony Sugar syrup

feeder

Colony Sugar syrup

feeder

Colony Sugar syrup

feeder

C+I.1244

1.5 ng/bee

1.5

C+I.1244

6 ng/bee

I.2025

20 ppb

0.889

3.556

0.944

1.389

1.65

I.1455-II

Colony Sugar syrup

feeder

6.1 ppb

0.390

1.562

0.415

0.610

0.725

The exposure length was

39 days feeding. Intake obtained

with default consumption values

and 72% sugar content was

assumed (honey was used)

The exposure length was

84 days feeding. Intake obtained

with default consumption values

and 50% sugar content. Study

authors estimated an adult

consumption of 0.12 ng/bee per

day

The exposure period was

11 weeks (77 days). Intake

obtained with default

consumption values and 50%

sugar content

The exposure period was

28 days. Intake obtained with

default consumption values and

50% sugar content was

assumed

Bumblebee C+I.1248

Colony Sugar syrup

feeder

10 ppb

1.46

2.98

4.76

I.475

Colony Sugar syrup

feeder

10 ppb

1.46

2.98

4.76

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Organism Reference Type

Exposure

Mean residue

values/test

concentration/

test dose

19.2 ppm

(0.0192 mg/g)

Forager

ng/bee

per day

Min

3314

max

6303

Nurse

ng/bee

per day

Min

max

Larva

ng/bee per

developmental period

Min

11117

max

16723

The exposure period was

30 days. Intake obtained with

default consumption values and

with 72% sugar content of the

syrup assumed (honey)

The exposure period was

42 days. Intake obtained with

default consumption values and

40% sugar content

Notes

I.2019

Colony Pollen/sugar

feeder syrup

I.2004

Colony Sugar syrup

feeder

10 ppb

1.825

3.725

5.95

I.787

Colony Sugar syrup

feeder

0.15 ppb

0.0219

0.0447

0.0714

I.787

Colony Sugar syrup

feeder

1.44 ppb

0.21024

0.42912

0.68544

The exposure period was

14 days. Intake obtained with

default consumption values and

50% sugar content was

assumed. The 0.15 ppb is a

calculated EC

values from the

series of test concentrations

0.68544 The exposure period was

14 days. Intake obtained with

default consumption values and

50% sugar content was

assumed. The 1.44 ppb is a

calculated EC

values from the

series of test concentrations

0.60452 The exposure period was

13 days. Intake obtained with

default consumption values and

50% sugar content was

assumed

46.8384 The exposure period was

14 days. Intake obtained with

default consumption values and

50% sugar content was

assumed

0.0714

I.788

Colony Sugar syrup

feeder

1.27 ppb

0.18542

0.37846

0.60452

I.2051

Colony Sugar syrup

feeder

98.4 ppb

14.3664

29.3232

46.8384

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Organism Reference Type

Exposure

Mean residue

values/test

concentration/

test dose

7 ppb

Forager

ng/bee

per day

Min

0.1862

max

0.2121

Nurse

ng/bee

per day

Min

max

Larva

ng/bee per

developmental period

Min

0.721

max

2.765

The exposure period was

84 days. Intake obtained with

default consumption values

The exposure period was 84

days. Intake obtained with

default consumption values

The exposure period was

84 days. Intake obtained with

default consumption values.

The exposure period was

85 days. Intake obtained with

default consumption values and

with 75% sugar content of the

syrup. Study authors calculated a

residue consumption of 2.15

ng/bee per day for adults

The exposure period was

85 days. Intake obtained with

default consumption values and

with 75% sugar content of the

syrup. Study authors calculated a

residue consumption of 4.81

ng/bee/day for adults

The exposure period was

14 days. Intake obtained with

default consumption values and

50% sugar content was

assumed

The exposure period was

14 days. Intake obtained with

default consumption values and

50% sugar content was

assumed

Notes

I.937

Colony Pollen

feeder

Colony Pollen

feeder

Colony Pollen

feeder

Colony Pollen/sugar

feeder syrup

I.2052

7 ppb

0.1862

0.2121

0.721

2.765

I.2052

30 ppb

0.798

0.909

3.09

11.85

I.1386

Pollen: 6 ppb;

syrup: 10 ppb

1.1329

2.1685

3.7913

5.5433

I.1386

Colony Pollen/sugar

feeder syrup

Pollen: 16 ppb;

syrup: 25 ppb

2.8589

5.4515

9.5813

14.2533

I.1505

Colony Pollen/sugar

feeder syrup

Pollen: 6 ppb;

syrup: 0.7 ppb

0.2618

0.3904

0.9512

2.7032

I.1505

Colony Pollen/sugar

feeder syrup

Pollen: 12 ppb;

syrup: 1.4 ppb

0.5236

0.7808

1.9024

5.4064

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Organism Reference Type

Exposure

Mean residue

values/test

concentration/

test dose

10 ppb

Forager

ng/bee

per day

Min

1.46

max

2.98

Nurse

ng/bee

per day

Min

max

Larva

ng/bee per

developmental period

Min

4.76

max

4.76

The exposure period was

77 days. Intake obtained with

default consumption values and

50% sugar content was

assumed

The exposure period was

77 days. Intake obtained with

default consumption values and

50% sugar content was

assumed

The exposure period was

43 days. Intake obtained with

default consumption values and

50% sugar content was

assumed

The exposure length was

17 days. Mean residue for pollen

was calculated from three

samples (includes positive

ﬁndings,

but also values

LOD).

Mean residue for nectar was

derived from a positive sample

and from a value

LOD

The exposure length was

15 days. Mean residue for pollen

was calculated from

ﬁve

samples

(includes positive

ﬁndings,

but

also values

LOQ). Mean

residue for nectar was derived

from a positive sample and from

values

LOQ

Notes

I+T.931

Colony Sugar syrup

feeder

I+T.2015

Colony Sugar syrup

feeder

10 ppb

1.46

2.98

4.76

C+I.2017

Colony Sugar syrup

feeder

2.1 ppb

0.3066

0.6258

0.9996

I.724

Field

Pollen from

potato

ﬁeld

0.37

lg/kg

0.0098

0.0112

0.0381

0.1462

I.725

Field

Pollen from

potato

ﬁeld

0.77

lg/kg

0.0204

0.0232

0.0790

0.3028

LOD: limit of detection; LOQ: limit of quantiﬁcation.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Appendix F

–

Tier-3 lines of evidence

The graphical representation of the lines of evidence for the treated crop scenario for winter oilseed

rape for honeybees and the succeeding crop scenario for honeybees and bumblebees.

F.1

F.1.1.

Honeybees

Colony strength (Class 1 endpoint)

Three

ﬁgures

are presented for the treated crop scenario (winter oilseed rape) and one for the

succeeding crop scenario. All of them are based on the same data set (colony strength for honeybee

colonies), but with slightly different setups in order to help the interpretation of the data. These

differences are explained in the title of the

ﬁgures.

The general interpretation is relevant for all

ﬁgures

for honeybee colony strength.

General interpretation

•

Altogether 20 endpoints were available, but eight of them were considered as non-reliable.

The biological variability of many of the reliable endpoints was considerably high and in many

cases

ﬂuctuated

both negatively and positively relative to the control

There is also variability between the experiments. The mean deviation from the control ranged

from negligible to medium negative

Exposure estimation and information on the exposure lengths was available for all reliable

experiments.

No dose–response pattern could be seen when all reliable endpoints are considered

Two experiments had more than one test concentrations. In one of them, a dose–response trend

could be observed, while this was not that clear in the other experiment (I.2017)

F.1.1.1.

Winter oilseed rape

Figure F.1:

Summary of the observed deviations for honeybee colony strength for seed treatment

use to winter oilseed rape

–

all available endpoints are indicated in the order of the

magnitude of mean deviation, scale for exposure level aligned to the available data

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Figure F.2:

Summary of the observed deviations for honeybee colony strength for seed treatment

use to winter oilseed rape

–

only the reliable endpoints are indicated in the order of the

magnitude of mean deviation, scale for exposure level aligned to the exposure

assessment goals (zoom in to the relevant part)

Figure F.3:

Summary of the observed deviations for honeybee colony strength for seed treatment

use to winter oilseed rape

–

only the reliable endpoints are indicated in the order of the

magnitude of the exposure estimation, scale for exposure level aligned to the available

data

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Interpretation for the treated crop scenario for the winter oilseed rape use

Indications for larger than negligible effect

The maximum negative deviation from the controls

was larger than negligible in eight experiments.

Although in four of these cases, the exposure was

too long or the estimated exposure was somewhat

higher than the exposure assessment goal, the

exposure regime of four experiments was not

unrealistically severe (I.2046, low dose of I.163,

I.2044, I.2045)

Considering the mean deviations from the control, it

was more than negligible in the negative direction in

case of three experiments (I.2044, I.2045, high dose

of I.163). Although in one of these cases, the

estimated exposure was somewhat higher than the

exposure assessment goal, the exposure regimes of

the other two experiments represented the lower

end of the realistic exposure situations (or even

below for I.2045)

Indications for lack of larger than negligible effect

The maximum negative deviation from the controls was

not larger than negligible in four experiments. Although

two of them represented a rather low exposure regime, in

the other two cases (I.1455-II, I.411), the exposure

regime, in general, was not too mild (and one of the

studies had an unrealistically long exposure)

Nine endpoints indicated negligible mean deviation from

the control. Although, two of them (I.2043, I.1498)

represented a rather low exposure regime, seven of them

had a realistic or severe exposure regime;

ﬁve

of these

seven were considered as reliable with minor deviations

The available evidence indicating not more than negligible deviation from the control has more weight than has

the evidence indicating more than negligible negative deviation from the control.This is because a relatively high

number of experiments with a realistic or severe exposure regime indicated an overall negligible deviation from

the control, while a fewer experiments with suitable exposure indicated a negative deviation from the control.

Also, there was a considerable biological variability and the endpoints indicating negligible deviation were, in

general, much more reliable.Overall, this indicates a

moderate evidence for negligible effect

F.1.1.2

Succeeding crop

Figure F.4:

Summary of the observed deviations for honeybee colony strength for seed treatment

use to the succeeding crop scenario

–

only the reliable endpoints are indicated in the

order of the magnitude of mean deviation, scale for exposure level aligned to the

exposure assessment goals (zoom in to the relevant part)

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Interpretation for the succeeding crop scenario

The exposure assessment goals for the succeeding crop scenario are only marginally different from the exposure

assessment goals for winter oilseed rape (see Section

5.1.1.2).

Also, when the exposure length of the

experiments is compared to the range of expected

ﬂowering

period, only marginal differences could be identiﬁed

between the

ﬁgures

for the succeeding crop scenario and for the winter oilseed rape. Therefore, the indications

for effects and for lack of effects with the overall balancing as presented for winter oilseed rape above is equally

applicable for the succeeding crop scenario

F.1.2.

Overwintering assessment (Class 1 endpoint)

Two

ﬁgures

are presented below for the treated crop scenario (winter oilseed rape) and one for the

succeeding crop scenario. All of them are based on the same data set (colony strengths of honeybee

colonies after overwintering), but with slightly different setups in order to help the interpretation of the

data. These differences are explained in the title of the

ﬁgures.

The general interpretation is relevant

for all

ﬁgures

for overwintering assessment.

It is noted that the colony strength after overwintering was calculated by considering the colony

strength of the dead colonies as 0. This was however done only for those studies where detailed data

on the individual colonies were available (i.e. it was possible to identify which and how many hives

were lost during the winter and what was the colony strength of the colonies that survived the

winter).

In addition, there were some experiments where the overwintering survival was studied, but not by

measuring the colony strength after overwintering. These experiments are brieﬂy summarized further

below and an integration of all the information on overwintering is provided.

General interpretation

•

Altogether nine endpoints were available, but four of them were considered as non-reliable

The variability between the reliable endpoints was very high; ranged from large positive to large

negative deviation from the controls

Exposure estimation and information on the exposure lengths was available for all the reliable

experiments

No dose–response pattern could be seen when all reliable experiments are considered

Within the experiments with more than one test concentration, a dose response trend could be

observed (considering the mean effects)

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

F.1.2.1.

Winter oilseed rape

Figure F.5:

Summary of the observed deviations for honeybee colony strength after overwintering for

seed treatment use to winter oilseed rape

–

all available endpoints are indicated in the

order of the magnitude of the deviation from the control, scale for exposure level aligned

to the available data

Figure F.6:

Summary of the observed deviations for honeybee colony strength after overwintering for

seed treatment use to winter oilseed rape

–

only the reliable endpoints are indicated in

the order of the magnitude of deviation from the control, scale for exposure level aligned

to the exposure assessment goals (zoom into the relevant part)

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Interpretation for the treated crop scenario for the winter oilseed rape use

Indications for larger than negligible effect

Two endpoints indicated more than negligible negative

deviation from the control. Although one of them had

high exposure level, the high dose of the experiment

I.411 had comparable but somewhat lower estimated

exposure than the exposure assessment goal and the

length of the exposure was in the realistic range of the

ﬂowering

period. This endpoint was considered as

reliable with major restrictions

Indications for lack of larger than negligible

effect

Three experiments indicated not more than negligible

negative deviation from the control. Although one of

them had too low exposure level, in the other two

cases the exposure regime, in general, was not mild.

These two endpoints were considered as reliable with

minor restrictions

The available evidence indicating not more than negligible deviation from the control has more weight than has

the evidence indicating more than negligible negative deviation from the control

This is because only one experiment with a realistic exposure regime indicated more than negligible negative

deviation from the control, while two experiments with suitable exposure indicated not more than negligible

negative deviation from the control. In addition, both of these two endpoints were more reliable than the one

with the more than negligible negative deviation

Overall, this indicates a moderate evidence for negligible effect

In addition, as reported above, there were three colony-feeder experiments (I.2017, I.843 and

I.362) where the overwintering success was studied, but not by measuring the colony strength after

overwintering. In these studies, the proportion of the successfully overwintered test colonies were

reported.

•

In the experiment I.843, honeybee colonies (n = 4) were fed by spiked sugar syrup at different

concentrations over a period of 9 weeks. For the next spring, all but one treated colonies died

while from the control, only one of the four colonies died. A slight overall tendency suggested

that the effect may have started earlier with the higher concentration. The endpoint was

assessed as

‘reliable

with minor restrictions’. The estimated exposure levels for the lowest test

concentration (40

lg/kg,

four of four colonies died) were forager

–

1.28–5.12 ng/bee per day,

nurse

–

1.36–2.0 ng/bee per day, larva

–

2.38 ng/larva per developmental period.

In the experiment I.362, over a 12-week period, spiked diet patties (pollen and sugary

solution) were offered to honeybee colonies (n = 10) with different concentrations. For

overwintering success, a decreasing trend with increasing concentration was found (although it

was statistically different only in case of the highest test concentration). All control hives

successfully overwintered. The endpoint was assessed as

‘reliable

with minor restrictions’.

The estimated exposure levels for the lowest test concentration (5

lg/kg,

2 of 10 colonies

died) were forager

–

0.32–1.28 ng/bee per day, nurse

–

0.37–0.53 ng/bee per day, larva

–

0.6 ng/larva per developmental period.

In the experiment I.2017 similar methodology was used as in the experiment I.362 (note:

I.2017 and I.362 are from the same publication). Three of the seven control colonies died and

four of the seven treated colonies (at each concentration) died for the next spring. The

endpoint was assessed as

‘reliable

with major restrictions’. The estimated exposure levels for

the lowest test concentration (5

lg/kg)

are the same as reported above for study I.362.

Taking the estimated exposure levels for the highest test concentration (100

lg/kg,

same

effect than at the lowest or at the middle concentration) were forager

–

6.4–25.6 ng/bee per

day, nurse

–

7.45–11.2 ng/bee per day, larva

–

12.0–12.1 ng/larva per developmental period.

An overall integration of the eight available and reliable experiments, which investigated the

overwintering, is presented below.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Reference

Exposure estimation

ng/bee per day

(for simpliﬁcation, it is

indicated only for foragers)

0.032–0.128

0.32–1.28

0.32–25.6

Exposure

length

Reliability

(day)

Description of effect/

deviation from control

I.411

I.2017

Reliable with major No more than negligible

restrictions

deviation from the control

Reliable with major Medium negative deviation

restrictions

from the control

Reliable with major No clear tendency (negative

restrictions

effect cannot be excluded nor

conﬁrmed)

Reliable with minor Clear tendency for a negative

restrictions

effect

Reliable with minor No more than negligible

restrictions

negative deviation from the

control

Reliable with minor Clear negative effect

restrictions

Reliable with minor Large positive deviation from

restrictions

the control

Reliable with minor Large negative deviation from

restrictions

the control

I.362

I.1455-II

0.32–1.28

0.39–1.56

I.843

I.163

1.28–5.12

1.6–6.4

3.2–12.8

Note: the exposure assessment goal for foragers is 2.1–1.6 ng/bee per day (acute and chronic, respectively) and the realistic

ﬂowering

duration of winter oilseed rape was considered as 10–42 days.

No dose–response pattern could be seen when all the endpoints are considered together. Estimated

exposures to honeybee castes of the lower dose of experiment I.411 were too low when compared to

the exposure assessment goal. No clear conclusion could be drawn from experiments I.2017.

Experiments I.362, I.843 and high dose of I.411 indicated a clear negative effect on honeybee

populations after overwintering. The estimated exposure of these experiments was at around or

somewhat below the exposure assessment goal. The high dose of the experiment I.163 also indicated

a clear negative effect; however, the exposure level of this endpoint was somewhat higher than the

exposure assessment goal. In the remaining two experiments, such negative effects were not present.

Experiment I.1455-II had only a negligible negative deviation from the control. The estimated

exposure of this experiment was comparable, but somewhat below the exposure assessment goal. The

low dose of the experiment I.163 had a realistic worst-case exposure regime and resulted in a large-

positive deviation from the control. Considering the deviations from the controls in combination with

the exposure estimations, it can be concluded that the available data set is contradictory. Nevertheless,

it may be considered that the available evidence indicating more than negligible negative effects has

more weight. However, two of the four experiments that indicated clear negative effects had

considerably longer exposure duration than the realistic

ﬂowering

period of oilseed rape (I.362 and

I.843). When the severity of these two endpoints is taken into consideration, the available evidences

become balanced and no clear trend is apparent. Therefore,

this line of evidence is inconclusive.

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

F.1.2.2.

Succeeding crop

Figure F.7:

Summary of the observed deviations for honeybee colony strength after overwintering for

the succeeding crop scenario

–

only the reliable endpoints are indicated in the order of

the magnitude of deviation from the control, scale for exposure level aligned to the

exposure assessment goals (zoom in to the relevant part)

Interpretation for the succeeding crop scenario

The exposure assessment goals for the succeeding crop scenario are only marginally different from the exposure

assessment goals for winter oilseed rape (see Section

5.1.1.2).

Also, when the exposure length of the

experiments is compared to the range of expected

ﬂowering

period, only marginal differences could be identiﬁed

between the succeeding crop scenario and for the winter oilseed rape. Therefore, the indications for effects and

for lack of effects with the overall balancing as presented for winter oilseed rape above based on colony strength

after overwintering is equally applicable for the succeeding crop scenario

The integration of the additional experiments and the

ﬁnal

conclusion on this line of evidence as presented for

winter oilseed rape above is also applicable for the succeeding crop scenario

F.1.3.

Mortality in front of the hive (Class 2 endpoint)

One

ﬁgure

is presented below for each of the treated crop scenario (winter oilseed rape) and for

the succeeding crop scenario. All of them are based on the same data set (honeybee mortality in front

of the hives). The general interpretation is relevant for both

ﬁgures.

General interpretation

•

Altogether 10 endpoints were available, but

ﬁve

of them were considered as non-reliable

All the

ﬁve

reliable endpoints were considered as reliable with major restrictions

The biological variability of all the reliable endpoints was very low; all the mean deviations from

the controls during the experiments were in the negligible range as well as the mean of the

extreme deviations

Exposure estimation and information on the exposure lengths was available for all the reliable

experiments

Dose–response analysis cannot be performed due to the very low variability between the reliable

endpoints

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

F.1.3.1.

Winter oilseed rape

Figure F.8:

Summary of the observed deviations for honeybee mortality in front of the hives for the

succeeding crop scenario

–

only the reliable endpoints are indicated, scale for exposure

level aligned to the available data

Interpretation for the treated crop scenario for the winter oilseed rape use

Indications for larger than negligible effect

The maximum negative deviation from the controls

was larger than negligible in two experiments and

the exposure regimes of these two experiments

represented the lower end of the realistic exposure

situations (or even below for I.2043)

Indications for lack of larger than negligible effect

The maximum negative deviation from the controls was

not larger than negligible in three experiments. One of

these experiments (I.2046) had exposure estimations

which are in the range of the exposure assessment goal.

The exposure lengths of this study fell in the realistic

range of the

ﬂowering

period, although tend to be at the

lower edge of this range

Considering the mean deviations from the control, all the

reliable studies fell in the negligible range. Only one of

these experiments (I.2046) had exposure estimations

which are in the range of the exposure assessment goal

The available evidence indicating not more than negligible deviation from the control has somewhat lower weight

than has the evidence indicating more than negligible negative deviation from the control. This is because only

one experiment with appropriate exposure indicated negligible effect, while in two experiments with mild

exposure regime, a temporal negative deviation was indicated. All the endpoints had low reliability, but the

biological variability was low

Overall, this indicates a

weak evidence for larger than negligible effect

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

F.1.3.2

Succeeding crop

Figure F.9:

Summary of the observed deviations for honeybee forager mortality in front of the hives

for the succeeding crop scenario

–

all endpoints are indicated, scale for exposure level

aligned to the available data

Interpretation for the succeeding crop scenario

The exposure assessment goals for the succeeding crop scenario are only marginally different from the exposure

assessment goals for winter oilseed rape (see Section

5.1.1.2).

Also, when the exposure length of the

experiments is compared to the range of expected

ﬂowering

period, only marginal differences could be identiﬁed

between the

ﬁgures

for the succeeding crop scenario and for the winter oilseed rape. Therefore, the indications

for effects and for lack of effects with the overall balancing as presented for winter oilseed rape is equally

applicable for the succeeding crop scenario

F.1.4.

Brood abundance (Class 2 endpoint)

Three

ﬁgures

are presented below for the treated crop scenario (winter oilseed rape) and one for

the succeeding crop scenario. All of them are based on the same data set (brood abundance of

honeybee colonies), but with slightly different setups in order to help the interpretation of the data.

These differences are explained in the title of the

ﬁgures.

The general interpretation is relevant for all

ﬁgures

for honeybee brood abundance.

F.1.4.1.

Winter oilseed rape

General interpretation

•

Altogether 22 endpoints were available, but eight of them were considered as non-reliable

All the 14 reliable endpoints were considered as reliable with major restrictions

The biological variability of several reliable experiments was considerable high, in some cases

ﬂuctuated

both negatively and positively relative to the control

There was also variability between the studies. The mean deviation from the control ranged from

medium negative to medium positive. The number of endpoints on both the negative and the

positive sides were balanced and all had the same reliability score. The mean of the mean

overall deviations during the duration of the studies is in the negligible range. This line is very

close to the zero line and the lines of the extreme deviations are also close to the edges of the

negligible deviation range

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

General interpretation

•

Exposure estimation was available for all of them and also, information on the exposure lengths

was available for all reliable studies

No dose–response pattern could be seen when all reliable experiments are considered

Within the experiments with more than one test concentration, a dose–response trend could be

observed (considering the mean effects)

Figure F.10:

Summary of the observed deviations for honeybee brood abundance for seed treatment

use to winter oilseed rape

–

all available endpoints are indicated in the order of the

magnitude of mean deviation, scale for exposure level aligned to the available data

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Figure F.11:

Summary of the observed deviations for honeybee brood abundance for seed treatment

use to winter oilseed rape

–

only the reliable endpoints are indicated in the order of the

magnitude of mean deviation, scale for exposure level aligned to the exposure

assessment goals (zoom in to the relevant part)

Figure F.12:

Summary of the observed deviations for honeybee brood abundance for seed treatment

use to winter oilseed rape

–

only the reliable endpoints are indicated in the order of the

magnitude of the exposure estimation, scale for exposure level aligned to the exposure

assessment goals (zoom into the relevant part)

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Interpretation for the treated crop scenario for the winter oilseed rape use

Indications for larger than negligible effect

The maximum negative deviation from the controls

was larger than negligible in eight experiments.

Three of these experiments had low estimated

exposure compared to the exposure assessment goal

in combination with exposure lengths which

represent the lower end of the realistic range of the

ﬂowering

period. An additional experiment from this

set (low dose of I.163) had a realistic worst-case

exposure regime

Considering the mean deviations from the control, it

was more than negligible in the negative direction in

case of four experiments

The exposure regimes in three of these four

experiments were not unrealistically severe and one

of these had a very low exposure estimation when

compared to the exposure assessment goal

Indications for lack of larger than negligible effect

The maximum negative deviation from the controls was

not larger than negligible in six experiments. One of these

experiments (high dose of I.2017) had a very severe

exposure regime and another one had a realistic worst-

case exposure estimation (I.2046). In addition, another

one (low dose of I.2017) had a relatively low estimated

exposure, but in combination with a very long exposure

duration

Ten experiments indicated negligible or positive mean

deviation from the control. Although, some of them

represented a rather low exposure regime, four of them

had higher or comparable exposure estimation than the

exposure assessment goal. Three of these four had an

exposure period considerable longer than the realistic

ﬂowering

period. An additional experiment (low dose of

I.2017) with lower exposure estimate than the exposure

assessment goal also had a long exposure period

The available evidence indicating not more than negligible deviation from the control has somewhat more weight

than has the evidence indicating more than negligible negative deviation from the control

This is because a slightly higher number of experiments with a realistic or severe exposure regime indicated an

overall negligible (or even positive) deviation from the control, while slightly fewer experiments with realistic or

mild exposure regime indicated more than negligible negative deviation from the control. There was a

considerable biological variability and all the endpoints had low reliability

Overall, this indicates a

weak evidence for negligible effect

F.1.4.2.

Succeeding crop

Figure F.13:

Summary of the observed deviations for honeybee brood abundance for the succeeding

crop scenario

–

only the reliable endpoints are indicated in the order of the magnitude

of mean deviation, scale for exposure level aligned to the exposure assessment goals

(zoom in to the relevant part)

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Interpretation for the succeeding crop scenario

The exposure assessment goals for the succeeding crop scenario are only marginally different from the exposure

assessment goals for winter oilseed rape (see Section

5.1.1.2).

Also, when the exposure length of the

experiments is compared to the range of expected

ﬂowering

period, only marginal differences could be identiﬁed

between the

ﬁgures

for the succeeding crop scenario and for the winter oilseed rape. Therefore, the indications

for effects and for lack of effects with the overall balancing as presented for winter oilseed rape above is equally

applicable for the succeeding crop scenario

F.1.5.

Homing success (Class 2 endpoint)

Two

ﬁgures

are presented below for the treated crop scenario (winter oilseed rape) and one for the

succeeding crop scenario. All of them are based on the same data set (homing success of honeybee

foragers), but with slightly different setups in order to help the interpretation of the data. These

differences are explained in the title of the

ﬁgures.

The general interpretation is relevant for all

ﬁgures

for homing success.

General interpretation

•

Six endpoints were available from three experiments; all were considered as reliable with major

restrictions

These endpoints were derived from three studies; from each of them, results from two test

concentrations are presented. In all cases, the highest test concentration indicating no or the

smallest deviation from the control and the next higher test concentration was selected.

Exposure estimation was available for all the endpoints

When all the six endpoints are considered, some, but not entirely clear dose–response pattern is

indicated (note: for one study, range of doses are indicated; it is considered that the lower edge

of the ranges are more relevant since the homing

ﬂight

experiments do not lasts all day long)

Within each experiment, a dose–response trend was observed

F.1.5.1.

Winter oilseed rape

Figure F.14:

Summary of the observed deviations for homing success of honeybee foragers for seed

treatment use to winter oilseed rape

–

endpoints in the order of the magnitude of

deviation from control, scale for exposure level aligned to the available data

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Figure F.15:

Summary of the observed deviations for homing success of honeybee foragers for seed

treatment use to winter oilseed rape

–

endpoints in the order of the magnitude of

deviation from control, scale for exposure level aligned to the exposure assessment goal

(zoom in to the relevant part)

Interpretation for the treated crop scenario for the winter oilseed rape use

The estimated exposure levels considerably exceed the exposure assessment goal for winter oilseed rape in all,

but one case. These

ﬁve

endpoints included the lower test concentration from the experiment I.178, which

indicated a negligible deviation from the control. The only endpoint that had an estimated exposure level close

but slightly below the exposure assessment goal indicated no deviation from the control. The endpoints were

classiﬁed as reliable with major restrictions

Overall, this indicates a

weak evidence for negligible effect

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

F.1.5.2.

Succeeding crop

Figure F.16:

Summary of the observed deviations for homing success of honeybee foragers for the

succeeding crop scenario

–

endpoints in the order of the magnitude of deviation from

control, scale for exposure level aligned to the exposure assessment goal (zoom in to

the relevant part)

Interpretation for the succeeding crop scenario

The exposure assessment goal for the succeeding crop scenario is only marginally different from the exposure

assessment goal for winter oilseed rape (see Section

5.1.1.2).

Therefore, the interpretation of the available

evidences and the conclusion as presented for winter oilseed rape above are also applicable for the succeeding

crop scenario

F.2.

F.2.1.

Bumblebees

Queen production (Class 1 endpoint)

Two

ﬁgures

are presented below for the succeeding crop scenario. Both are based on the same

data set (queen production of queenright bumblebee colonies), but with slightly different setups in

order to help the interpretation of the data. These differences are explained in the title of the

ﬁgures.

The general interpretation is relevant for both

ﬁgures.

General interpretation

•

Six reliable endpoints were available, all were assessed as reliable with major restrictions

There was a high variability between the experiments. The deviation from the control ranged

from large positive to large negative

Exposure estimation was available for all, but one and information on the exposure lengths was

available for all experiments

No dose–response pattern could be seen when all endpoints are considered

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Figure F.17:

Summary of the observed deviations for queen production of bumblebees for the

succeeding crop scenario

–

endpoints in the order of the magnitude of deviation from

control, scale for exposure level aligned to the available data

Figure F.18:

Summary of the observed deviations for queen production of bumblebees for the

succeeding crop scenario

–

endpoints in the order of the magnitude of the exposure

estimation, scale for exposure level aligned to the available data

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Interpretation of the line of evidence

Indications for larger than negligible effect

Two endpoints indicated more than negligible negative

deviation from the control. Although one of them had

an unrealistically long exposure (when compared to

the realistic range of the

ﬂowering

period of any

succeeding crop), the other one had a realistic

exposure length falling to the lower end of the range

of the

ﬂowering

period of succeeding crops.

The estimated exposures of both of them were well

below the exposure assessment goals

Indications for lack of larger than negligible

effect

Three experiments with exposure estimation indicated

positive deviation from the control. Although two of

them had too low exposure levels, the exposure

estimations of one of them (C+I. 1248) was reasonable

close to the exposure assessment goals (but were

somewhat below) in combination with a very long

exposure length that was far beyond the realistic range

of the

ﬂowering

period of any succeeding crop

The available evidence indicating not more than negligible deviation from the control has less weight than has

the evidence indicating more than negligible negative deviation from the control

One experiment with appropriate exposure indicated negligible effect and also only one experiment with rather

mild exposure regime resulted in a large negative deviation from the control. However, there was another

endpoint with low exposure level indicating similar results. This experiment (I.937) was not considered to be

suitable to fully support the negative trend due to its very long exposure length. However, it was considered as a

weak indication for a likely negative effect at realistic exposure situation

Overall, this indicates a

weak evidence for larger than negligible effect

F.2.2.

Worker production (Class 1 endpoint)

Two

ﬁgures

are presented below for the succeeding crop scenario. Both are based on the same

data set (worker production of queenright bumblebee colonies), but with slightly different setups in

order to help the interpretation of the data. These differences are explained in the title of the

ﬁgures.

The general interpretation is relevant for both

ﬁgures.

General interpretation

•

Nine reliable endpoints were available, all, but one was assessed as reliable with major

restrictions

There was a high variability between the experiments. The deviation from the control ranged

from large positive to medium negative

Exposure estimation was available for all, but one and information on the exposure lengths was

available for all experiments

No dose–response pattern could be seen when all endpoints are considered

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Figure F.19:

Summary of the observed deviations for worker production of bumblebees for the

succeeding crop scenario

–

endpoints in the order of the magnitude of deviation from

control, scale for exposure level aligned to the available data

Figure F.20:

Summary of the observed deviations for worker production of bumblebees for the

succeeding crop scenario

–

endpoints in the order of the magnitude of the exposure

estimation, scale for exposure level aligned to the available data

www.efsa.europa.eu/efsajournal

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Interpretation of the line of evidence

Indications for larger than negligible effect

The maximum negative deviation from the controls

was larger than negligible in two experiments out of

the three, where the effect was studied at more than

one observation day. In both cases, the exposure

levels were below the exposure assessment goals,

although one of them had a very long exposure

period

The mean deviation from the control was negative

and larger than negligible in three experiments and

two of them had exposure estimations (lower dose of

I.1505, I.475). The exposure estimation for both of

them was lower than the exposure assessment goal

in combination with exposure lengths falling in the

realistic range of the

ﬂowering

period of any

succeeding crop. One of these experiments was

considered as reliable with minor restrictions

Indications for lack of larger than negligible effect

The maximum negative deviation from the controls was

not larger than negligible in one out of those three

experiments, where the effect was studied at more than

one observation day. When compared to the exposure

assessment goal, this experiment had a realistic exposure

level for the larvae in combination with a very long

exposure period

Six endpoints indicated negligible mean deviation from the

control or positive deviation from the control. Although

three of them had too low estimated exposure, in case of

the other three (C+I.1248, high dose of I.1505, I.2052) at

least the estimations for larvae were close to or

overlapped the respective exposure assessment goal.

Moreover, two of these three had very long exposure

period

Results from the two test concentrations of experiment

I.1505 are not consistent; the lower dose resulted in the

negative deviation from the control indicating some

uncertainties to consider this deviation as true effect

The available evidence indicating not more than negligible deviation from the control has not more, neither less

weight than has the evidence indicating more than negligible negative deviation from the control. Although the

available evidence indicating more than negligible negative deviations based on maximum deviations has slightly

more weight than has the evidence indicating not more than negligible deviation, overall this type of evidence

has a low weight (three experiments only). Considering the mean deviations, two experiments with appropriate

exposure indicated a clear negative deviation from the control. One of them had higher reliability than all the

other endpoints, while the other one bears some uncertainties. On the other hand, three experiments could be

considered as evidence indicating no negative deviation from the controls. However, two of them had somewhat

low exposure levels (but reasonable close to the exposure assessment goal at least for larvae) and the third one

had a high enough exposure level only for larvae. The lengths of the exposure of these experiments were either

realistic or considerable longer than the realistic range of the

ﬂowering

period of any succeeding crop. No clear

trend could be seen when all these positive and negative elements were balanced

Therefore,

this line of evidence is inconclusive

F.2.3.

Drone production (Class 1 endpoint)

Two

ﬁgures

are presented below for the succeeding crop scenario. Both are based on the same

data set (drone production of queenright bumblebee colonies), but with slightly different set-ups in

order to help the interpretation of the data. These differences are explained in the title of the

ﬁgures.

The general interpretation is relevant for both

ﬁgures.

General interpretation

•

Six reliable endpoints were available, all were assessed as reliable with major restrictions

There was variability between the experiments. The deviation from the control ranged from

medium positive to large negative

Exposure estimation and information on the exposure lengths was available for all experiments

No dose–response pattern could be seen when all endpoints are considered

www.efsa.europa.eu/efsajournal

100

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Figure F.21:

Summary of the observed deviations for drone production of bumblebees for the

succeeding crop scenario

–

endpoints in the order of the magnitude of deviation from

control, scale for exposure level aligned to the available data

Figure F.22:

Summary of the observed deviations for drone production of bumblebees for the

succeeding crop scenario

–

endpoints in the order of the magnitude of the exposure

estimation, scale for exposure level aligned to the available data

www.efsa.europa.eu/efsajournal

101

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Interpretation of the line of evidence

Indications for larger than negligible effect

The deviation from the control was large negative

for two experiments and the exposure estimation

was lower than the exposure assessment goal in

both cases. Although the exposure length was too

long for one of these experiments, the exposure

lengths of the other one (lower dose of I.1505) was

in the realistic range of the

ﬂowering

period of any

succeeding crop

Indications for lack of larger than negligible effect

Four endpoints indicated a positive mean deviation from

the control. Although all of them had too low estimated

exposure, in one case (high dose of I.1505) at least the

estimations for larvae were close to the respective

exposure assessment goal and another one had a very

long exposure period

Results from the two test concentrations of experiment

I.1505 are not consistent; the lower dose resulted in the

negative deviation from the control indicating some

uncertainties to consider this deviation as true effect

The available evidence indicating not more than negligible deviation from the control has not more neither less

weight than has the evidence indicating more than negligible negative deviation from the control.

Two experiments with low exposure level indicated a clear negative deviation from the control. One of them had

too long exposure length, while the other one bears some uncertainties. On the other hand, two experiments

could be considered bearing some evidences that indicate no negative deviations from the controls. However,

both of them had relatively mild exposure regimes. Overall, all the available evidences have very low weight and

no clear trend could be seen when they were balanced

Therefore,

this line of evidence is inconclusive

F.2.4.

Reproductive output of queenless microcolonies (Class 1

endpoint)

One

ﬁgure

is presented below for the succeeding crop scenario.

The endpoints in the available experiments with queenless microcolonies were reported in different

ways: (1) brood production (i.e. drone brood cells) or (2) drone production (number of adult drones)

or (3) both of these two reported separately or (4) both of the two, but reported as a combined

endpoint (brood cell

adults). Because of this and since all brood cells contain male larvae in this type

of study, when both endpoints were reported separately, they were summed up.

General interpretation

•

Altogether seven endpoints were available, but two of them were considered as non-reliable

All the

ﬁve

reliable endpoints were considered as reliable with major restrictions

The variability between the reliable endpoints was rather low; the deviations from the control

ranged from small to large negative

Exposure estimation and information on the exposure lengths was available for all the reliable

experiments

Meaningful dose–response analysis cannot be performed due to the low variability between the

reliable endpoints

www.efsa.europa.eu/efsajournal

102

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Figure F.23:

Summary of the observed deviations for drone production of queenless bumblebee

colonies for the succeeding crop scenario

Interpretation of the line of evidence

Indications for larger than negligible effect

All endpoints have larger than negligible negative

deviation from the control, three of the

ﬁve

have large

deviations, and there are no endpoints with negligible

or positive deviation from the control

Indications for lack of larger than negligible

effect

The experiment with the highest exposure estimation

(above the exposure assessment goal) and longest

exposure period indicated the smallest deviation from

the control, while all the other experiments with

milder exposure regime had larger deviations from the

control

One of the experiments (I.788) had a low exposure

level in combination with a realistic exposure length

compared to the range of the

ﬂowering

period of

succeeding crops (in fact the exposure length was at

the lower end of this range). Although the length of

exposure was too long for three other endpoints, the

exposure levels of these experiments were also lower

than the exposure assessment goal

The available evidence indicating not more than negligible deviation from the control has less weight than has

the evidence indicating more than negligible negative deviation from the control

This is because there were a number of endpoints with relatively low exposure levels and one of them with a

realistic exposure length, which endpoints indicated clearly negative deviations from the control. In addition, no

endpoint exists, which would indicate negligible or positive deviation from the control

Overall, this indicates a

moderate evidence for larger than negligible effect

F.2.5.

Brood production (Class 1 endpoint)

Three

ﬁgures

are presented below for the succeeding crop scenario. Both are based on the same

data set (brood production of queenright bumblebee colonies), but with slightly different set-ups in

order to help the interpretation of the data. These differences are explained in the title of the

ﬁgures.

The general interpretation is relevant for both

ﬁgures.

www.efsa.europa.eu/efsajournal

103

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

General interpretation

•

Twelve reliable endpoints were available, all but one were assessed as reliable with major

restrictions

There was some variability between the experiments. The deviation from the control ranged

from medium positive to large negative

Exposure estimation and information on the exposure lengths was available for all experiments

No dose–response pattern could be seen when all endpoints are considered

Figure F.24:

Summary of the observed deviations for brood production of bumblebees for the

succeeding crop scenario

–

endpoints are indicated in the order of the magnitude of

mean deviation, scale for exposure level aligned to the available data

www.efsa.europa.eu/efsajournal

104

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Figure F.25:

Summary of the observed deviations for brood production of bumblebees for the

succeeding crop scenario

–

endpoints are indicated in the order of the magnitude of

mean deviation, scale for exposure level aligned to the exposure assessment goals

(zoom in to the relevant part)

Figure F.26:

Summary of the observed deviations for brood production of bumblebees for the

succeeding crop scenario

–

endpoints in the order of the magnitude of the exposure

estimation, scale for exposure level aligned to the exposure assessment goals (zoom in

to the relevant part)

www.efsa.europa.eu/efsajournal

105

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Interpretation of the line of evidence

Indications for larger than negligible effect

The maximum negative deviation from the controls

was larger than negligible in one experiment out of

the three, where the effect was studied at more than

one observation day. The exposure level was below

the exposure assessment goals, although it had a

very long exposure period

Indications for lack of larger than negligible effect

The maximum negative deviation from the controls was

not larger than negligible in two of those three

experiments, where the effect was studied at more than

one observation day; these two endpoints are originating

from the same experiment, they are two test

concentrations

Although the results from the two test concentrations are

not consistent in terms of dose–response trend, they

provide an indication for not more than negligible

deviation from the control at realistic exposure level for

the larvae (high dose) compared to the exposure

assessment goal in combination with a very long exposure

period

The mean deviation from the control was larger than Only two endpoints (from the same experiment) indicated

negligible in 10 experiments and the exposure

no negative mean deviation from the control. Although

estimation for eight of them was lower than the

one of them had too low estimated exposure, for the

exposure assessment goal. For

ﬁve

of these

others the exposure estimations for larvae overlapped the

experiments, the exposure lengths fell in the realistic respective exposure assessment goal. Moreover, both of

range of the

ﬂowering

period of any succeeding crop them had a very long exposure period

The available evidence indicating not more than negligible deviation from the control has less weight than has

the evidence indicating more than negligible negative deviation from the control

This is because only two endpoints (from the same experiment) indicated no negative deviation from the

controls and only one of them had a high enough exposure level and this was only for the larvae. Nevertheless,

the lengths of the exposure of these experiments were considerable longer than the realistic range of the

ﬂowering

period of any succeeding crop. On the other hand, a number of experiments indicated clearly negative

deviations from the control. Five of these experiments had a mild exposure regime

Overall, this indicates a

moderate evidence for larger than negligible effect

F.2.6.

Colony strength (Class 1 endpoint)

Only two endpoints were available for colony strength. The deviation from the control in the

experiment I.2004 ranged from negligible to large negative deviation and the median effect (over the

study duration) was assessed as a medium negative effect. The endpoint was considered as fully

reliable. The estimated exposure level was 1.83–3.73 ng/bee per day for workers and 5.95 ng/larva

per developmental period for larva (the exposure assessment goal for workers is 2.69–3.12 ng/bee per

day and for larvae is 6.5 ng/larva per developmental period). The length of the exposure was at the

higher end, but within the realistic range of the

ﬂowering

period of any succeeding crop.

The deviation from the control in the experiment C+I.2017 was a large-negative deviation. The

endpoint was considered reliable with major restrictions. The estimated exposure level was

0.31–0.63 ng/bee per day for workers and 0.99 ng/larva per developmental period for larva (the

exposure assessment goal for workers is 2.69–3.12 ng/bee per day and for larvae is 6.5 ng/larva per

developmental period). The length of the exposure was higher than the realistic range of the

ﬂowering

period of any succeeding crop, but only with 1 day.

Overall, this indicated a

strong evidence for larger than negligible effect.

F.2.7.

Emergence rate (Class 2 endpoint)

Only one endpoint was available for emergence rate (I.2004). The deviation from the control

bridged from large positive to large negative and the median effect (over the study duration) was

assessed as a medium-negative effect. The estimated exposure level was 1.83–3.73 ng/bee per day

for workers and 5.95 ng/larva per developmental period for larva (the exposure assessment goal for

workers is 2.69–3.12 ng/bee per day and for larvae is 6.5 ng/larva per developmental period). The

endpoint was considered as fully reliable. The length of the exposure was at the higher end, but within

the realistic range of the

ﬂowering

period of any succeeding crop.

This indicated a

moderate evidence for larger than negligible effect.

www.efsa.europa.eu/efsajournal

106

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

F.2.8.

Mating ability of the new queens (Class 2 endpoint)

Only one endpoint was available for mating ability (I.1388). This experiment, which was a

ﬁeld

study on sunﬂower, indicated a negligible (negative) deviation from the control. The endpoint was

considered as reliable with major restrictions. No exposure estimation was possible. The length of the

exposure was 9 days.

Due to the lack of information on the exposure of the single endpoint that was available, this

line

of evidence is considered to be inconclusive.

www.efsa.europa.eu/efsajournal

107

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Appendix G

–

List of study references

Exp. ID

110403

CAN/US

All+.2003

Doc.ID

Author

distiller

N/A

Scott-Dupree, CD; Spivak, MS;

Bruns, G; Blenskinsop, C;

Nelson, S

Pohorecka K, Skubida P,

Miszczak A, Semkiw P, Sikorski

P, Zagibajlo K, et al.

Pohorecka K, Skubida P,

Miszczak A, Semkiw P, Sikorski

P, Zagibajlo K, et al.

Study title

The impact of Gaucho and TI-435 seed-treated Canola on

honeybees,

Apis mellifera

Year

2001

Type

Appendix

to EFSA

(2018a)

File name

EFSA

(2013a)

All

ﬁeld

Residues Study

evaluation

notes

Field

1080

Residues of Neonicotinoid Insecticides in Bee Collected Plant

2012

Materials from Oilseed Rape Crops and their Effect on Bee Colonies

Residues of Neonicotinoid Insecticides in Bee Collected Plant

2012

Materials from Oilseed Rape Crops and their Effect on Bee Colonies

2015

2013

All+.1080

1080

Residues M

All

Residues

All

Lab

C+I_Field

All+.1084

C+I.1081

1084

1081

Poquet Y, Kairo G, Tchamitchian Wings as a new route of exposure to pesticides in the honeybee

S, Brunet J-L, Belzunces LP

Pohorecka K, Skubida P,

Effects of exposure of honeybee colonies to neonicotinoid seed

Semkiw P, Miszczak A, Teper D,

Sikorski P, et al.

Schneider CW, Tautz J,

Gruenewald B, Fuchs S

Scholer J, Krischik V

RFID Tracking of Sublethal Effects of Two Neonicotinoid

Insecticides on the Foraging Behavior of

Apis mellifera

Chronic Exposure of Imidacloprid and Clothianidin Reduce Queen

Survival, Foraging, and Nectar Storing in Colonies of

Bombus

impatiens

Lab

Field

C+I.1244

C+I.1248

1244

1248

2012

2014

Colony

feeders

Colony

feeders

Colony

feeders

Colony

feeders

Colony

feeders

C+I_Colony

feeders

C+I_Colony

feeders

C+I_Colony

feeders

C+I_Colony

feeders

C+I_Colony

feeders

C+I.2017

926

C+I.423

423

Moffat C, Pacheco JG, Sharp S,

Samson AJ, Bollan KA, Huang J,

et al.

Fischer J, Mueller T, Spatz A-K,

Greggers U, Gruenewald B,

Menzel R

Lu C, Warchol KM, Callahan RA

Chronic exposure to neonicotinoids increases neuronal vulnerability 2015

to mitochondrial dysfunction in the bumblebee (Bombus

terrestris)

Neonicotinoids Interfere with Speciﬁc Components of Navigation in 2014

Honeybees

Sub-lethal exposure to neonicotinoids impaired honeybees

winterization before proceeding to colony collapse disorder

2014

C+I.844

844

www.efsa.europa.eu/efsajournal

108

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Exp. ID

C*I.1324

Doc.ID

Author

distiller

1324

€

Staffel J; Luckmann J

Study title

Assessment of Potential Impacts on Honeybee Colony

Development, their Hibernation Performance and Concurrent

Monitoring of Aerial Dust Drift During the Sowing Operation of

Poncho Beta Plus - Treated Sugar Beet Pills with Typical

Commercial Vacuum-Pneumatic Sowing Technology, Directly

Adjacent to Full-Flowering Phacelia tanacetifolia in Germany

Residue levels of imidacloprid and imidacloprid metabolites in

nectar, blossoms and pollen of summer rape cultivated on soils

with different imidacloprid residue levels and effects of these

residues on foraging honeybees. Laacher Hof 1999

Residue levels of imidacloprid and imidacloprid metabolites in

nectar, blossoms and pollen of summer rape cultivated on soils

with different imidacloprid residue levels and effect of these

residues on foraging honeybees

Determination of Side-effects of SeedOprid 600 FS coloured

treated Seeds of Winter Oil-Seed Rape on Honeybees (Apis

mellifera

L.) in the

ﬁeld

in Germany 2011/2012

Determination of Side-effects of SeedOprid 600 treated Seeds of

Spring Oil-Seed Rape on Honeybees (Apis

mellifera

L.) in the Field

in Germany 2011

Effects of sublethal doses of crop protection agents on honeybee

(Apis

mellifera)

global colony vitality and its potential link with

aberrant foraging activity

Sub-lethal effects of imidacloprid on bumblebees,

Bombus

terrestris

(Hymenoptera: Apidae), during a laboratory feeding test

Year

2014

Type

Field

Appendix

to EFSA

(2018a)

File name

C*I Field

E 370 1548-8 N/A

Schmuck, R; Schoening, R;

Schramel, O

1999

Residues Study

evaluation

notes

Residues N/A

EFSA

(2013a)

E 370 1553-4 N/A

Schmuck, R; Schoening, R;

Schramel, O

2007

EFSA

(2013a)

I.1094

1094

Probsting A

2012

Residues E

I residues

I.1095

1095

Probsting A

2011

Residues E

I residues

I.137

137

I.1386

I.1388

I.1455 - Part

I and II

1386

1388

1455

Belien T, Kellers J, Heylen K,

Keulemans W, Billen J, Arckens

L, et al.

Tasei JN, Lerin J, Ripault G

Tasei JN, Ripault G, Rivault E

2009

Colony

feeders

Colony

feeders

ﬁeld

Colony

feeders

I colony

feeders

I colony

feeders

ﬁeld

I colony

feeders

2000

2001

2014

Hazards of imidacloprid seed coating to

Bombus terrestris

(Hymenoptera: Apidae) when applied to sunﬂower

van der Steen JC, Hok-Ahin and De invloed van imidacloprid en de interactie met gereduceerd

Cornelissen B

stuifmeelaanvoer op de vitaliteit en de overwintering van

bijenvolken. Translation from Dutch: The inﬂuence of imidacloprid

and the interaction with reduced pollen supply on the vitality and

overwintering of honeybee colonies

www.efsa.europa.eu/efsajournal

109

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Exp. ID

I.1498

Doc.ID

Author

distiller

1498

Wehner A

Study title

Year

Type

Field

Appendix

to EFSA

(2018a)

File name

2.I

ﬁeld

I.1505

I.163

I.178

1505

163

178

Whitehorn PR, O’Connor S,

Wackers FL, Goulson D

Bocksch S

Bortolotti L, Montanari R,

Marcelino J, Medrzycki P, Maini

S, Porrini C

Bryden J, Gill RJ, Mitton RAA,

Raine NE, Jansen VAA

Hecht-Rost

2013

Determination of Side-effects of SeedOprid 600 FS coloured

treated Seeds of Sunﬂower on Honeybees (Apis

mellifera

L.) in the

Field in Italy 2012

Neonicotinoid Pesticide Reduces Bumblebee Colony Growth and

2012

Queen Production

Honeybee brood and colony level effects following Imidacloprid

intake via treated artiﬁcial diet in a

ﬁeld

study in North Carolina

Effects of sub-lethal imidacloprid doses on the homing rate and

foraging activity of honeybees

Chronic sublethal stress causes bee colony failure

A residue study with Nuprid 600 FS treated maize seed,

investigating residues in crops, soil and honeybee products in

Germany 2007

Assessment of Chronic Sublethal Effects of Imidacloprid on

Honeybee Colony Health

Effect of reduced risk pesticides for use in greenhouse vegetable

production on

Bombus impatiens

(Hymenoptera: Apidae)

Learning performances of honeybees (Apis

mellifera

L) are

differentially affected by imidacloprid according to the season

Risk posed to honeybees (Apis

mellifera

L. Hymenoptera) by an

imidacloprid seed dressing of sunﬂowers

2014

2003

Colony

feeders

Colony

feeders

Colony

feeders

I colony

feeders

I colony

feeders

I colony

feeders

I colony

feeders

I semiﬁeld

I.2004

I.2016

201

553

2013

2009

Colony

feeders

Semiﬁeld E

I.2017

I.2019

I.2020

I.2025

I.2043

362

501

340

1243

1092

Dively GP, Embrey MS, Kamel A,

Hawthorne DJ, Pettis JS

Gradish AE, Scott-Dupree CD,

Shipp L, Harris CR, Ferguson G

Decourtye A, Lacassie E,

Pham-Delegue MH

Schmuck R, Schoning R,

Stork A, Schramel O

Probsting A

2015

2010

2003

2001

Colony

feeders

Colony

feeders

Lab

Colony

feeders

Field

I colony

feeders

I colony

feeders

I Lab

I colony

feeders

ﬁeld

I.2044

1093

Probsting A

Determination of Side-effects of SeedOprid 600 FS coloured

2011

treated Seeds of Maize on Honeybees (Apis

mellifera

L.) in the

Field in Germany 2011

2011

Determination of Side-effects of SeedOprid 600 FS coloured

treated Seeds of Sunﬂower on Honeybees (Apis

mellifera

L.) in the

Field in Germany 2011

Determination of Side-effects of SeedOprid 600 FS coloured

treated Seeds of Winter Oil-Seed Rape on Honeybees

(Apis

mellifera

L.) in the

ﬁeld

in Germany 2011/2012

2012

Field

ﬁeld

I.2045

1094

Probsting A

Field

ﬁeld

www.efsa.europa.eu/efsajournal

110

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Exp. ID

I.2046

Doc.ID

Author

distiller

1095

Probsting A

Study title

Determination of Side-effects of SeedOprid 600 treated Seeds of

Spring Oil-Seed Rape on Honeybees (Apis

mellifera

L.) in the Field

in Germany 2011

Repression and Recuperation of Brood Production in

Bombus

terrestris

Bumblebees Exposed to a Pulse of the Neonicotinoid

Pesticide Imidacloprid

Effects of novel pesticides on bumblebee (Hymenoptera: Apidae)

colony health and foraging ability

Year

2011

Type

Field

Appendix

to EFSA

(2018a)

File name

ﬁeld

I.2051

787

Laycock I, Cresswell JE

2013

Colony

feeders

Colony

feeders

Colony

feeders

Colony

feeders

I colony

feeders

I colony

feeders

I colony

feeders

I colony

feeders

I semiﬁeld

I colony

feeders

I Lab

I.2052

I.362

I.411

937

362

411

Morandin LA, Winston ML

Dively GP, Embrey MS, Kamel A,

Hawthorne DJ, Pettis JS

Faucon JP, Aurieres C,

Drajnudel P, Mathieu L, Ribiere

M, Martel AC, et al.

Gels JA, Held DW, Potter DA

Gill RJ, Ramos-Rodriguez O,

Raine NE

2003

Assessment of Chronic Sublethal Effects of Imidacloprid on

2015

Honeybee Colony Health

Experimental study on the toxicity of imidacloprid given in syrup to 2005

honeybee (Apis

mellifera)

colonies

Hazards of insecticides to the bumblebees

Bombus impatiens

(Hymenoptera: Apidae) foraging on

ﬂowering

white clover in turf

Combined pesticide exposure severely affects individual- and

colony-level traits in bees

2002

2012

I.462

I.475

I.545

462

475

545

Semiﬁeld E

Colony

feeders

Lab

I.724:

724

Monceren BB

(1)

I.725:

Monceren

(BB2)

I.787

725

Sublethal doses of imidacloprid decreased size of hypopharyngeal 2013

Hatjina F, Papaefthimiou C,

Charistos L, Dogaroglu T, Bouga glands and respiratory rhythm of honeybees

in vivo

M, Emmanouil C, et al.

Klein O

A Field Study to Evaluate Effects of Monceren G on the Bumblebee 2014

(Bombus

terrestris

L; Hymenoptera, Apidae) in Potato in Southern

Germany in 2014

Klein O

Field Study to Evaluate Effects of Monceren G on the Bumblebee

(Bombus

terrestris

L; Hymenoptera, Apidae) in Potato in Southern

Germany in 2014

Repression and Recuperation of Brood Production in

Bombus

terrestris

Bumblebees Exposed to a Pulse of the Neonicotinoid

Pesticide Imidacloprid

Effects of imidacloprid, a neonicotinoid pesticide, on reproduction

in worker bumblebees (Bombus

terrestris)

In situ replication of honeybee colony collapse disorder

2014

Field

ﬁeld

Field

ﬁeld

787

Laycock I, Cresswell JE

2013

Colony

feeders

Colony

feeders

Colony

feeders

I colony

feeders

I colony

feeders

I colony

feeders

I.788

I.843

788

843

Laycock I, Lenthall KM, Barratt

AT, Cresswell JE

Lu C, Warchol KM, Callahan RA

2012

www.efsa.europa.eu/efsajournal

111

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Exp. ID

I.848

Doc.ID

Author

distiller

848

€

Luckmann J, Staffel J

(L mentioned in Appendix to

EFSA, 2018a)

Study title

Assessment of Potential Impacts on Honeybee Colony

Development, their Hibernation Performance and Concurrent

Monitoring of Aerial Dust Drift During the Sowing Operation of

Imidacloprid FS 350A G - Treated Winter Barley with Typical

Commercial Pneumatic So

Effects of novel pesticides on bumblebee (Hymenoptera: Apidae)

colony health and foraging ability

Risk assessment for side effects of neonicotinoids against

bumblebees with and without impairing foraging behaviour

Sensitivity of non-target arthropods and beneﬁcial fungal species

to chemical and biological plant protection products: results of

laboratory and semiﬁeld trials

Risk assessment for side effects of neonicotinoids against

bumblebees with and without impairing foraging behaviour

Reconciling laboratory and

ﬁeld

assessments of neonicotinoid

toxicity to honeybees

Year

2014

Type

Field

Appendix

to EFSA

(2018a)

File name

ﬁeld

I.937

I+T.2015

937

931

Morandin LA, Winston ML

Mommaerts V, Reynders S,

Boulet J, Besard L, Sterk G,

Smagghe G

Sterk G, Heuts F, Merck N,

Bock J

2003

2010

Colony

feeders

Colony

feeders

Colony

feeders

Colony

feeders

Colony

feeders

Field

I colony

feeders

I+T. Colony

feeder

I+T. Colony

feeder

I+T. Colony

feeder

I+T. Colony

feeder

T*I

ﬁeld

I+T.2029

I+T.2031

I+T.931

1341

2003

931

Mommaerts V, Reynders S,

Boulet J, Besard L, Sterk G,

Smagghe G

Henry M, Cerrutti N, Aupinel P,

Decourtye A, Gayrard M,

Odoux J-F, et al.

2010

I*T.583

583

2015

www.efsa.europa.eu/efsajournal

112

EFSA Journal 2018;16(2):5178

Peer review of the pesticide risk assessment for bees of the active substance imidacloprid

Appendix H

–

Used compound codes

Code/trivial

name

5-OH-

imidacloprid

Chemical name/SMILES notation

(2E)-1-[(6-chloropyridin-3-yl)methyl]-5-hydroxy-N-

nitroimidazolidin-2-imine

[O-][N+](=O)/N=C2\NCC(O)N2Cc1cnc(Cl)cc1

Structural formula

–

Imidacloprid

oleﬁn

(2E)-1-[(6-chloropyridin-3-yl)methyl]-N-nitro-1,3-dihydro-

2H-imidazol-2-imine

[O-][N+](=O)/N=C2\NC=CN2Cc1cnc(Cl)cc1

–

www.efsa.europa.eu/efsajournal

113

EFSA Journal 2018;16(2):5178