Pollinators and Pesticides - Environmental Audit Committee Contents

2  Pesticide approvals

EU approvals

19. The European Commission approves active substances for use in plant protection products in EU Member States. Defra described the European approvals procedure for active substances:

It is the job of the company which wishes to gain approval to put together the necessary scientific data to support its application … The applicant submits all of the information including study methodology and data generated, together with their own conclusions, in the form of a Dossier … The Dossier is scrutinised and assessed by a regulatory authority's experts in all of the various scientific disciplines involved. The regulatory authority's opinion—which may or may not coincide with that of the company—is set out in a Draft Assessment Report (DAR). The DAR produced by the regulatory authority of a Member State is then submitted to the European Food Safety Authority (EFSA), which organises a further scrutiny (known as peer review) by experts from all of the EU Member States. Following this peer review, EFSA sends its conclusions to the Commission. This is used as the basis for a proposal from the Commission for approval or not of the substance and any associated conditions. This proposal is adopted (or not) by qualified majority vote of Member States.[26]

20. The European Food Safety Authority (EFSA) was formed in 2002 as an independent source of scientific advice and communication on risks associated with the food chain. For pesticides work, EFSA conducts risk assessment and the European Commission is responsible for risk management. EFSA co-ordinates the peer review of active substances used in pesticides, provides scientific advice on broader issues that cannot be resolved within the peer review process and delivers scientific guidance on generic issues, commonly in the fields of toxicology, eco-toxicology and the fate and behaviour of pesticides. EU rules on the authorisation of pesticides allow the European Commission to seek EFSA's views on new evidence on the safety of a pesticide or active substance.

UK approvals

21. When an active substance has been approved by the European Commission, companies can apply to the regulatory authority in individual Member States—in the case of the UK, the Chemicals Regulation Directorate (CRD) of the Health and Safety Executive—for permission to place their product on the market. Most products include a range of substances in addition to the active substance—for example, the bait that attracts slugs to eat slug pellets. When a company seeks authorisation to market a product in the UK, the CRD prepares a scientific evaluation of factors such as the product's chemical properties, its potential toxicity to humans, dietary intake, exposure to operators and other workers, environmental fate and behaviour, efficacy and risk to crops. This draft evaluation is then considered by the Advisory Committee on Pesticides (ACP), which advises Ministers that an authorisation can be granted only if it is content that there are no "unacceptable risks".[27]

22. The ACP is a statutory independent advisory committee constituted under section 16(7) of the Food and Environment Protection Act 1985. It advises Ministers on matters relating to the control of pests and on the approval of pesticides in the UK. Appointments to the ACP are made by open competition and follow the requirements of the Office of the Commissioner for Public Appointments. All ACP Members are required to declare any interests they might have in the pesticides industry both on an annual basis and before the discussion of particular issues. The ACP has a current membership of 20, including two lay members, mostly drawn from academia. Defra has stated that "No [ACP] member has declared that they are in the employ of companies selling neonicotinoid pesticides."[28]


23. We heard that the data submitted by pesticide companies for regulatory purposes is not in the public domain, which makes it impossible for concerned stakeholders to examine the methods, assumptions and results underpinning risk assessment and risk management.[29] Professor Goulson commented:

I am very confused as to why they insist this information is confidential. We are talking about safety tests, so you have a new chemical that you want to bring to the market; you have to have it tested on a range of organisms to see at what level it kills them, what concentrations kill them or whatever. There would be tests on honeybees and worms and a range of other things. It is not clear to me why that information should not be made freely available to everybody or what commercial advantage a competitor would gain by finding out how many honeybees product X would kill at a certain concentration.[30]

24. Dr Cresswell described his experience of accessing studies submitted by pesticides companies to support applications to approve products:

I have just seen some of the studies and the way that I had to do it was I had to apply to the CRD, I had to go to York and then I had to sit in a room with a person looking to check—I don't know what he thought I might do. So I was allowed to look at the documents, make notes, but I could not have copies of them. So I did a pretty good transcription of all the data that I wanted and was able to take it away, but I am not sure that counts as transparent.[31]

Professor Goulson highlighted "the obvious inequity in that academic research that has shown evidence for harm of neonicotinoids on bees is picked apart and examined in minute detail by the agrochemical industry and yet in reverse we can't examine the evidence that they are safe."[32] Dr Dicks added that "it is the regulatory system itself, with its closed studies that you can't access, that is at fault".[33]

25. Defra appeared to be less concerned:

The studies commissioned in support of an approval application are sometimes described as secret, but that is not an accurate portrayal. These studies carry data protection rights under EU legislation, which means that they cannot be used by other companies to gain authorisation. However the data is accessible through access to information arrangements such as those under the Freedom of Information Act and Environmental Information Regulations. These access rights to the regulatory studies have been used in respect of neonicotinoids. The Government recognises the value of having the data more readily available for wider review and has suggested to the pesticide manufacturers that it would be a good idea to publish their studies.[34]

26. We agree with Defra that it would be "a good idea" if pesticide manufacturers were to publish the studies underpinning applications for pesticide approvals. The agrochemical industry has produced many studies on the environmental effect of neonicotinoids and other pesticides, but the data are allegedly confidential for commercial reasons. The lack of transparency in relation to trials and studies conducted by pesticide manufacturers has resulted in inequality between the pesticide industry on one side and academics and the public on the other. The agrochemical industry should place the results of its risk assessment trials in the public domain to inform academic research and increase transparency for the public. Defra should work with industry and academics to establish which, if any, genuinely commercially sensitive details should be redacted to make that possible.


27. On the tests that pesticide companies are obliged to conduct to support the approval of their products, the ACP told us that "the standard requirements do not include some of the specific sub-lethal effects suggested by recent academic studies."[35] The Pesticide Action Network highlighted that "the tests focus on short-term, acute toxicity to adult worker bees and mainly ignore chronic toxicity and sub-lethal effects on bee behaviour, on larvae and on hive overwintering."[36] In May 2012, EFSA published an Opinion on the risk assessment of plant protection products which addressed that point by proposing enhanced risk assessments including sub-lethal effects. Member States are currently consulting on that Opinion, which underpinned EFSA's recent revised risk assessments of neonicotinoids (paragraph 53).

28. The recent Gill laboratory study (paragraph 41) pointed to a need for the assessment regime to address the combined effect of multiple pesticides. Professor Goulson raised some of the practical problems:

It becomes very complicated very quickly because there are lots of chemicals that bees would be exposed to—fungicides as well as insecticides and herbicides and so on. If you were to demand that every new product had to be tested and all possible interactions had to be tested, in an ideal world that would be wonderful, but I think the costs would quickly become extraordinary.[37]

Dr Cresswell added:

There are two options: either you use your fundamental knowledge to predict what might happen, or you prescribe, in a regulatory framework, if those two things are going to be used together then we have to test those. I think there are ways forward but you have to be smart in what you test.[38]

29. The EU approvals process for active substances explicitly addresses only the risks to honeybees.[39] It does not explicitly refer to wild bees or other species of insect pollinator, although there is a general duty to ensure that plant protection products do not have "unacceptable effects on the environment".[40] The honeybee serves as a sentinel species for all insect pollinators in European risk assessment, but we heard that evidence derived from monitoring and testing honeybees cannot be used to draw reliable conclusions about outcomes for wild bees, let alone for other wild pollinator species. Dr Cresswell commented that "some kinds of bee are more sensitive than others. In fact, honeybees, in my view, are rather tough compared with, for example, bumblebees."[41] Dr Dicks pointed out why conclusions derived from monitoring honeybees do not apply to one important pollinator species:

In some parts of the country, hoverflies form a very substantial proportion of the flower-feeding insect community, providing an unknown amount of the pollination service. They have very different life cycles to bees. They feed on flowers exclusively as adults. Many species have different larval habits. Some of them are laying their eggs in a crop and the larvae are feeding in the crop, so their exposure routes are very, very different from bees in many ways.[42]

30. We recognise that it is impractical to conduct individual risk assessments for the thousands of species of bees, hoverflies, butterflies, carrion flies, beetles, midges, moths and other invertebrates that contribute to insect pollination, but we are not convinced that honeybees are an appropriate proxy for all such species. We urge Defra to introduce a representative range of sentinel pollinator species in UK pesticides risk assessments and work to agree a similar arrangement across the EU.

Case study: imidacloprid

31. We examined the neonicotinoid imidacloprid as a case study of how the trials, risk assessment and risk management of pesticides work in practice. Imidacloprid was first approved for use as an active substance in the EU in 1991, and individual products containing imidacloprid have been registered for use in the UK since 1993. Under Article 8(2) of Council Directive 91/414/EEC, which set out a rolling programme of reassessment for active substances, imidacloprid's status as an approved substance was subject to re-evaluation in 2006. Germany was the Rapporteur Member State, and therefore the German regulatory authority produced the Draft Assessment Report (DAR) for imidacloprid in 2006. Bayer CropScience, based in Germany, developed imidacloprid, and it markets several plant protection products in which imidacloprid is the active substance.

32. We heard concerns that neonicotinoids such as imidacloprid might accumulate in the environment to the detriment of insect pollinators.[43] Many invertebrates, such as some wild bees, nest in topsoil, which makes the extent to which a toxic active substance accumulates in soil an important environmental consideration (more widely, there is the further question whether such accumulations might make their way into groundwater). We therefore examined how the issue of environmental accumulation was addressed in the 2006 re-approval process.

33. The DAR described how imidacloprid's propensity to accumulate in soil was tested by two trials conducted in two separate locations (to minimise the possibility of an anomalous result) in the UK in the 1990s:

In order to demonstrate that imidacloprid does not persist in soil and that its use does not entail an accumulation in soil, long term dissipation studies with repetitive application of imidacloprid were conducted ... In a ... study performed in Great Britain, the long-term soil dissipation of imidacloprid following its use as seed dressing in winter barley was investigated in the course of six years. It was established that maximum concentrations in soil reach a plateau at rather low residue levels after four to six years.[44]

The DAR included the data on which that conclusion was based and expressed the results of the 1990s British soil accumulation trials in the form of two graphs (Figure 2).[45] The graphs shown in Figure 2 sit on top of the watermark, because they were added to the DAR at some point after 2006 to replace earlier graphs. The original graphs had erroneously added together the readings for different soil depths rather than averaging them. This basic arithmetic error was only partially corrected in the final version of the DAR, which still included incorrect figures which did not match the revised graphs. There was no acknowledgement in the DAR that this amendment had been made (accurate figures for the British trials appeared in an addendum in 2008).[46] Based on incorrectly calculated data, which would have exacerbated the apparent problem, the 2006 DAR concluded:

Long-term field dissipation trials of imidacloprid in soil with its repeated use as a seed treatment over six consecutive years have confirmed that the compound has no potential for accumulation in soil. Though the concentrations measured in the samples from the two test sites increase in the first three years the increase levels off and reaches a plateau.[47]

The conclusion in the DAR that "the compound has no potential for accumulation in soil" did not reflect the results of the trials. In reality, the trials did not show a plateau in accumulation in soil (see Figure 2, which graphs the correct data).

Figure 2

34. The 2006 DAR was subject to peer review by EFSA, which provided a risk assessment to the European Commission. EFSA identified imidacloprid's apparent tendency to accumulate in soil in its risk assessment:

At the two UK study sites accumulation occurred over the full 6 year duration of the studies and the experts considered that a plateau was not reached.[48]

Plateau not reached at the end of study, data gap identified.[49]

The risk assessment to soil dwelling organisms cannot be finalised because the assessment of soil accumulation is not finalised.[50]

EFSA's comments on soil accumulation were simply included in the text of its risk assessment and were not highlighted as an action point or cause for concern. EFSA also extrapolated the half-life of imidacloprid in soil, which is a measure of persistence in the environment, from the results of the UK trials: it calculated a half-life of 1,333 days at one site and of 1,268 days at the other.[51] EFSA described its calculations as "conservative estimates."[52]

35. We asked Bayer CropScience, which developed imidacloprid, about soil accumulation. Bayer's spokesman, Dr Julian Little, told us:

It will depend on a huge number of different things, including soil type, climate, temperature, what has been grown in there, how many worms there are—everything will affect that figure. But if you are looking at something like imidacloprid or clothianidin you can be talking a half-life of anywhere between 16 and, say, 200 days.[53]

He later amended his estimate of the maximum half-life of imidacloprid: "in 'worse-case scenarios', the half-life of imidacloprid in normal soils would be variable but around 288 days, and would be expected to plateau upon repeated doses after three years."[54] That estimate is borne out by neither the 1990s UK field trials on which the re-approval of imidacloprid was based nor EFSA's 2008 risk assessment of those trials.

36. EFSA concluded its peer-reviewed risk assessment of imidacloprid in May 2008 and forwarded it to the European Commission. The EC Standing Committee on the Food Chain and Animal Health, which consists of representatives of EU Governments and public authorities and which manages risk in relation to pesticides on behalf of the European Commission, considered EFSA's risk assessment in September 2008: "The overall conclusion from the evaluation is that it may be expected that plant protection products containing imidacloprid will fulfil the safety requirements laid down in ... Directive 91/414/EEC … The review has also concluded that under the proposed and supported conditions of use, there are no unacceptable affects on the environment".[55] The Standing Committee's recommendation to approve imidacloprid did not mention the accumulation of imidacloprid in soil, and imidacloprid was re-approved in December 2008.[56]

37. We asked Defra and Bayer CropScience to comment on how soil accumulation was addressed in the re-approval process. Defra did not see the trials of imidacloprid as representative:

The UK field accumulation study considered by EFSA was a specific data requirement from the ACP and was designed to provide a worst case assessment, hence the incorporation of the entire straw rather than stubble only … EFSA noted that the incorporation of treated plant material was the one major difference between the experimental design of the UK and German studies and might be an explanation why very long half-life of 1,333 and 1,268 days were estimated at the two UK experimental sites and a plateau in soil residues had not occurred after 6 years of experimentation.[57]

Bayer CropScience told us that the UK trials were "a very specific study that is not designed to derive half-lives."[58] It also highlighted the effect of reincorporating straw: "Normally when we do these studies they are designed to reflect common agricultural practice. In this particular study, the barley was sown and we took the harvest of the grain, but then the straw remained on the soil and the straw was chopped and shallow-incorporated back into the soil bed."[59]

38. On the reincorporation of straw into the ground, UK Agriculture and Horticulture Development Board research indicates that 50% of stem and leaf material produced by oilseed rape, a crop commonly grown from neonicotinoid-coated seed, can be collected and baled, which suggests that the remaining 50% is reincorporated.[60] Similarly, some 60% of wheat and barley can be collected and baled, which means that the other 40% remains in the environment.[61] The extent to which it is possible to collect stem and leaf material might explain why the ACP specified that straw should be reincorporated when it designed the UK trials of imidacloprid in the early 1990s.

39. When imidacloprid was re-approved for use as an active substance in 2008, Directive 91/414/EEC did not set a limit on the half-life of a substance in soil, but stipulated that an approved substance should have "no unacceptable influence on the environment … having particular regard to its fate and distribution".[62] EU Regulation 1107/2009, introduced in 2009, set criteria for 'persistence', one of which is a half-life in soil of more than 120 days (and 'very persistent' where the half-life in soil is higher than 180 days).[63] Defra told us that "active substances which are deemed to be persistent are not excluded from approval unless they are also bioaccumulative and toxic (so-called PBT substances). Imidacloprid does not meet the bioaccumulative criteria and so is not a PBT substance".[64] [65] However, Regulation 1107/2009 includes a catch-all stipulation that an active substance "shall have no unacceptable effects on the environment, having particular regard to ... its fate and distribution in the environment, particularly contamination of surface waters, including estuarine and coastal waters, groundwater, air and soil".[66] In short, Defra acknowledged that imidacloprid is persistent (Regulation 1107/2009 indicates that it is "very persistent") and toxic, but justified its approval on the grounds that it is not bioaccumulative. Defra did not engage with the question whether imidacloprid's apparent half-life in soil might constitute an "unacceptable influence on the environment".[67]

40. For Governments, scientists and the public to have confidence in the EU-wide pesticide approvals regime, data and analysis should be rigorously scrutinised and quality checked to form a credible evidence base. The 2006 re-approval of imidacloprid for use in the EU shows two flaws in the system. First, EFSA identified the issue of soil accumulation in its peer review, but the European Commission proceeded to sign off imidacloprid as an approved active substance for use in Member States without explicitly addressing that risk. There seems little point in EFSA's assessing risk if the Commission ignores environmental threats identified in that process. We recommend that the Government exercises its influence in Europe to empower EFSA to include action points in future peer reviews which the European Commission must explicitly address before approving active substances. Secondly, the choice of Germany as the Rapporteur Member State in the case of a substance developed and manufactured in Germany raised a potential conflict of interest. The Government should seek a common understanding in Europe that active substances should be assessed by the regulatory authority of a Member State other than the one in which the applicant company is based.

26   Ev 196 Back

27   Ev 223 Back

28   HC Deb, 4 March 2013, col 806W Back

29   Ev 125 Back

30   Q 109 Back

31   Q 100 Back

32   Q 101 Back

33   Q 103 Back

34   Ev 196 Back

35   Ev 216 Back

36   Ev 125 Back

37   Q 130 Back

38   Ibid. Back

39   Ev 199 Back

40   Council Regulation (EC) No. 1107/2009, Article 4 Back

41   Q 98 Back

42   Q 99 Back

43   Ev 141, w2 Back

44   Draft Assessment Report, "Imidacloprid", 2006, vol 1, p 42 Back

45   Ibid, Annex B.8, p 640 Back

46   Draft Assessment Report, "Imidacloprid", 2008, Final Addendum, p 8 Back

47   Draft Assessment Report, "Imidacloprid", 2006, Annex B.8, p 640 Back

48   EFSA Scientific Report, "Conclusion on the peer review of imidacloprid", vol 148 (2008), p 26 Back

49   Ibid, p 79 Back

50   ibid, p 58 Back

51   ibid, p 25 Back

52   ibid, p 26 Back

53   Q 187 Back

54   Ev 236 Back

55   European Commission, Review report for the active substance imidacloprid, SANCO 108/08, September 2008 Back

56   Commission Directive 2008/116/EC Back

57   BEE 38, Defra, para 8 Back

58   Q 434 Back

59   Ibid. Back

60   Harley Stoddart and Jack Watts, "Biomass feedstock, residues and by-products", Agriculture and Horticulture Development Board, 2012 Back

61   Ibid. Back

62   Council Directive 91/414/EEC, Article 4(1)(v) Back

63   Council Regulation (EC) No. 1107/2009, Annex II. Back

64   Ev 237 Back

65   Bioaccumulation refers to the accumulation of substances, such as pesticides, or other organic chemicals in an organism. Bioaccumulation occurs when an organism absorbs a toxic substance at a rate greater than that at which the substance is lost. Back

66   Council Regulation (EC) No. 1107/2009, Article 3(e) Back

67   Council Directive 91/414/EEC, Article 4(1)(v) Back

previous page contents next page

© Parliamentary copyright 2013
Prepared 5 April 2013