link to page 2 link to page 2
ANNEX
NEONICOTINOIDS AND BEES
Scientific evidence from a study of bumble bee colonies under field conditions
1.
When the Council discussed the issue of neonicotinoids in March, the UK offered to share a
field study on effects on bumble bees, which was nearing completion.
2.
The study
1 was carried out by the Food and Environment Research Agency to consider the
effects of neonicotinoid seed treatments on bumble bee colonies under field conditions
and was
completed in late March. In the study, bumble bee colonies were placed within landscapes known
to contain oilseed rape treated with neonicotinoids. The colonies at all sites grew and produced new
queens. The researchers found no relationship between colony growth and neonicotinoid residues
within pollen or nectar in the colonies.
3.
To accompany the study, the UK Department for Environment, Food and Rural Affairs
prepared a short assessment of the key scientific evidence
2 on this subject. This paper concludes
that “
while this assessment cannot exclude rare effects of neonicotinoids on bees in the field, it
suggests that effects on bees do not occur under normal circumstances. This assessment also
suggests that laboratory based studies demonstrating sub-lethal effects on bees from neonicotinoids
did not replicate realistic conditions, but extreme scenarios. Consequently, it supports the view that
the risk to bee populations from neonicotinoids, as they are currently used, is low.”
4.
The documents have been circulated in full to the Commission and to Member States.
Executive summaries of each paper are appended to this document.
5.
We hope that this information will help Member States reflect on positions ahead of appeal
committee on 29th April. We also welcome the Commission request that EFSA provide a view on
this new information to assist the appeal committee deliberations.
1http://www.fera.defra.gov.uk/scienceResearch/scienceCapabilities/chemicalsEnvironment/documents/reportPS2371Ma
r13.pdf
2
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/181841/pb13937-neonicotinoid-bees-
20130326.pdf.pdf
8712/13
CC/mp
2
ANNEX
DG B 2
LIMITE EN
6.
The current evidence does not settle the question of the effects of neonicotinoids on bees
beyond doubt and further research is justified to strengthen the evidence. This should include
regulatory studies carried out by companies under the new EU risk assessment process, once this is
finalised. It might also include publicly-funded work to look at field effects across Europe.
Effects of neonicotinoid seed treatments on bumble bee colonies under field conditions
Food and Environment Research Agency March 2013
Executive Summary
1. The current study tested the hypothesis that exposure of bumble bee colonies placed in the
vicinity of crops treated with neonicotinoids had no major effect on the health of the colonies.
2. The study compared the development (mass and composition) of bumble bee Bombus terrestris
colonies placed in three landscapes (A-C) near oilseed rape (OSR) crops which had been grown
from (A) untreated seeds or from seeds treated with the neonicotinoid insecticides (B) clothianidin
or (C) imidacloprid). Twenty bumble bee colonies were established at 3 sites.
3. Systematic differences in the pesticide residues were found across the three sites, but these were
not related to the treatment applied to the adjacent crop suggesting bumble bees forage over large
distances.
4. Differences existed in flowering phenology across the three sites necessitating colonies to be
placed in the field at differing times. As a consequence of this there were systematic between-site
differences in the size and possibly other aspects of the colonies. These differences particularly
affected site C. In analysing the data attempts were made to control statistically for these baseline
differences.
5. Commercially sourced bumble bee colonies of known size which contained a mean of 16-24
worker bees were placed in each site for the duration of the flowering period (6-7 weeks) of the
OSR. At peak flowering times, samples of pollen and nectar were collected from each colony for
residue analysis (limit of detection 0.025 μg/Kg in nectar and 0.5 μg/Kg in pollen) and palynology
(pollen). Analysis was undertaken for the neonicotioinds and the two major honeybee-toxic
metabolites of imidacloprid (there is no published data on the chronic toxicity of these
neonicotinoids and metabolites to bumble bees).
8712/13
CC/mp
3
ANNEX
DG B 2
LIMITE EN
6. All colonies grew and survived to the end of the experiment. At sites A and B colonies grew to a
significantly greater terminal mass than at site C but all colonies grew to a greater mass than control
colonies in published laboratory experiments.
7. The number of gynes (new queens) produced in the colonies was not significantly different
across the three sites.
8. Despite the local abundance of flowering rape, pollen analysis showed bees foraged on a wide-
variety of flowers with the former contributing on average 35-37% of pollen collected. Possibly
because of the delay in placing colonies in the field, foraging patterns at site C differed somewhat
from the other two sites.
9. At site A, colonies contained residues of thiamethoxam in all 19 available nectar samples
available (mean 0.885 μg/Kg) and in 9 of the 20 pollen samples (mean 0.730 μg/Kg). Residues of
clothianidin in nectar were detectable (but below the limit of quantitation; LOQ) in 17 nectar and 20
pollen samples. There were no detectable residues of imidacloprid, nor its major (toxic) metabolites
in pollen and nectar.
10. At site B, colonies also contained residues of thiamethoxam with a mean of 2.397 μg/Kg nectar
and 0.718 μg/Kg pollen. Residues of clothianidin, a metabolite of thiamethoxam, were above LOQ
in 14 nectar samples (mean 0.205 μg/Kg) and detectable, but below LOQ, in 6 nectar samples.
There were no detectable residues of clothianidin in pollen. Neither imidacloprid, nor its major
toxic metabolites, were detected in pollen and nectar.
11. At site C, of the 15 nectar samples, residues of imidacloprid were detected below LOQ in 6
samples, with both imidacloprid and clothiandin jointly detected (at or below LOQ) in a further 2
samples. In contrast, thiamethoxam was not determined in any of the nectar sampled. Residues of
thiamethoxam, clothianidin or imidacloprid (or metabolites), were not detected in any of the 20
pollen samples taken.
12. Using the observed variation in neonicotinoid residues across colonies within and between sites,
possible correlations with colony mass and the number of new queens produced were explored. No
clear consistent relationships were observed.
8712/13
CC/mp
4
ANNEX
DG B 2
LIMITE EN
13. This study was not a formal statistical test of the hypothesis that neonicotinoid insecticides
reduce the health of bumble bee colonies. Nevertheless, were neonicotinoids in pollen and nectar
from treated oilseed rape to be a major source of field mortality and morbidity to bumblebee
colonies, we would have expected to find a greater contribution of insecticide residues from nearby
treated crops and for there to have been a clear relationship between observed neonicotinoid levels
and measures of colony success. The absence of these effects is reassuring but not definitive. The
study underlines the importance of taking care in extrapolating laboratory toxicology studies to the
field, as well as the great need of further studies under natural conditions.
An assessment of key evidence about Neonicotinoids and bees
Department for Environment Food and Rural Affairs March 2013
Executive Summary
Three recent studies in which bees were dosed with neonicotinoids showed sub-lethal effects on
bees [1-3]. The results from these studies contrast with a growing body of evidence from field
studies that has failed to show an effect of neonicotinoids when bees are allowed to forage naturally
in the presence of crops treated with neonicotinoids [4-8]. The evidence suggests the reason for this
difference is over-dosing of bees in the dosing studies; in all cases there is evidence that the doses
of neonicotinoids presented to bees under laboratory or semi-field conditions were unrealistically
high. The dosing studies therefore represented the extreme case in a field situation. In the only study
in which dose was measured [1] the dose was much greater than would have ever been experienced
in a field situation.
A concentration of 1-5 μg/l of neonicotinoid in nectar appears to be the threshold below which an
effect tends not to be observed [9] and most residue measurements in the nectar and pollen of
treated crops are normally at or below this level [4,6,7,10,11]. In addition, examination of bee
foraging shows that they tend not to feed exclusively on treated crops [7], thus diluting any effects
of neonicotinoids. Consequently, the evidence of effects of neonicotinoids on bees come from
studies [1-3] in which doses were likely to have been at least 2-10 times above this threshold.
8712/13
CC/mp
5
ANNEX
DG B 2
LIMITE EN
There is a possibility that field studies did not have the statistical power to show effects but the
accumulated evidence across several independent studies suggests that this is unlikely and, any
effects that are present are likely to be small and not biologically significant. Moreover, oilseed rape
(OSR) requires insect pollinators to support its productivity [12-18]. The fact that OSR treated with
neonicotinoids has been a productive crop for over a decade in the UK is itself evidence that
pollinator populations, including bees, are not being reduced by the presence of neonicotinoids.
Conclusion: While this assessment cannot exclude rare effects of neonicotinoids on bees in the
field, it suggests that effects on bees do not occur under normal circumstances. This assessment also
suggests that laboratory based studies demonstrating sub-lethal effects on bees from neonicotinoids
did not replicate realistic conditions, but extreme scenarios. Consequently, it supports the view that
the risk to bee populations from neonicotinoids, as they are currently used, is low.
8712/13
CC/mp
6
ANNEX
DG B 2
LIMITE EN
Document Outline