Pesticide report was Defra’s dodgy dossier

Monday 30th March 2015

A reappraisal of a Government scientific study has raised serious concerns about the scientific integrity of the pro-neonicotinoid UK Government – were civil servants distorting science so that it supported the policy position of Owen Paterson, the then Secretary of State for the Environment?

Earlier last week a neonicotinoid scandal broke in the media. Here are two examples of the coverage this scandal received, from the Guardian and Nature.

The story is that Government’s environment department, Defra, knew that there were correlations between neonicotinoids and bumblebee health in the data from a key study that they had commissioned from the Food and Environment Research Agency (Fera), but despite this they allowed Fera to ignore and massage the data to remove the correlations, then produced a Defra policy paper that supported the previously expressed position of the Owen Paterson’s department.

The Fera study was commissioned in early 2012 by Owen Paterson following the Whitehorn et al (2012) study that found that neonicotinoids could cause a 85% reduction in bumblebee queen production.  The pesticide companies were concerned that this and other evidence would result in the EC deciding to take regulatory action against neonicotinoids.

In November 2012, with EC action looming, Owen Paterson asked Fera to speed up its work on the bumblebee studies.

The report was not produced in time to influence the EU level scientific review, and on 16 January 2013 came the decision by the European Food Safety Agency (EFSA) that neonicotinoids posed an unacceptable risk to bees.

The UK Government’s Advisory Committee on Pesticides (ACP) met on 29 January 2013 Defra Chief Scientist Prof Ian Boyd and Fera report author Dr Helen Thompson were both present.  The minutes record that the committee flagged up the fact that the data from the Fera bumblebee study included at least three clear correlations between bumblebee health and neonicotinoids.

In March Fera and Helen Thompson produced their bumblebee study report, it concluded that “possible correlations with colony mass and the number of new queens produced were explored. No clear consistent relationships were observed”.

Instead of clearly detailing the correlations between bumblebee health and neonicotinoids the study features an overly complex statistical analysis, with results mixed together in a confusing table.  Some of the correlations are dispatched by removing highly polluted bumblebee colonies from the analysis, the reason for this is not explained.

One of the strongest statistical relationships is between the level of clothinaidin in nectar and the numbers of queens produced – where there is more clothianidin there are fewer queens produced, with very few queens being produced where the residue in nectar is more than 0.25ppb.  This relationship is not discussed in the report, and the statement that “Neither the non-parametric nor the parametric approaches showed evidence of a relationship between queen production and residues of …….clothianidin in nectar” is either a glaring mistake or deliberately misleading.

Concerns about the interpretation of the data were almost instant, the next day independent Cambridge scientist Dr Lynn Dicks pointed out that “looking at the data in the report it seems there could be some subtle impacts on bumblebee colony performance at these exposure levels.  I would really like access to the raw data.”

It was clear that the ACP had wanted to see the final version of the report before its release, the ACP wrote to ministers in March 2013 “Members commented that it was disappointing that the study was not available for the ACP meeting on 19 March”.

The current intention of Government appears to be to abolish the Advisory Committee on Pesticides and replace it with a new committee.

On the very same day that Defra released the Fera report it also revealed a new review of bee and neonicotinoid science.  This paper was based on just four studies; three published and peer reviewed lab/semi-field experiments that did show effects of neonicotinoids on bee health, and Fera’s artificial bumblebee nest field experiment – that the paper summarised as showing “no relationship between colony growth and neonicotinoid residues”.  Therefore the Government view was that the lab work was trumped by the field study and it concluded that “effects on bees do not occur under normal circumstances”.

It is this document that has been used as the basis for subsequent Defra policy and is still quoted on their website.

Owen Paterson touted the Fera study and new Defra policy position around Europe to convince other ministers to vote against the proposed partial ban – he failed and the ban went through.

Explaining why the UK had opposed the EC partial ban on neonicotinoids  Owen Paterson stated in April 2013 that “We voted against the ban because, backed my chief scientists and the Government’s scientific adviser, we all believe there has not been scientific evidence justifying the ban.”

In May 2013 EFSA published a scathing review of the Thompson 2013 study that concluded that “The objectives (problem formulation) were not clearly defined.”, “The lack of detailed reporting of the materials and methods” , and “Concerns were raised regarding the elaboration and interpretation of the results to reach the proposed conclusions”.

However, within weeks of producing the Fera report Helen Thompson was whisked off to a shiny new job working for neonicotinoid manufacturer Syngenta. No doubt they had been impressed with the 133 page report “Neonicotinoid Pesticides and Bees Report to Syngenta Ltd” (January 2013) that they had paid Fera to produce and had been written by Helen Thompson at the same time she was writing up the Fera/Defra bumblebee study.

EFSA’s dismissal of the results did not slow down Defra’s enthusiasm for neonicotinoids, responding to the HoC Environmental Audit Committee inquiry into Pollinators and Pesticides in September 2013, the department stated that “The Fera bumble bee study….found no relationship between colony growth and neonicotinoid residues within pollen or nectar in the colonies…. it is representative of an increasing number of field-realistic studies that have failed to find an effect of neonicotinoids on bees.”

Now, neatly two years later the truth has emerged.  A reanalysis of the Fera data by Prof Dave Goulson shows that not only was the data misrepresented, the conclusion to be taken from the Fera study should have been that despite small sample sizes and lots of variables, the study had in fact found strong evidence for a range of effects on colony health from neonicotinoids, and particularly clothianidin contamination below 0.3 parts per billion could cause significant harm to bumblebees, reducing both nest growth and the number of queens produced.

In June 2014, giving evidence on neonicotinoids in the House of Commons Prof Ian Boyd said “I would not allow publication of research that is not robust”

It would seem that on his watch this is precisely what happened, twice, the Fera report and the Defra science review were both misleading.  Flawed evidence was produced that supported the Government’s existing policy, and the only Government body that lifted a finger to stop its production is now due to be abolished.