Government Response
GOVERNMENT RESPONSE TO THE ENVIRONMENTAL
AUDIT COMMITTEE REPORT: GM FOODS - EVALUATING THE FARM SCALE TRIALS.
SECOND REPORT OF SESSION 2003-04
The Government welcomes this opportunity to respond
to the Environmental Audit Committee's conclusions and recommendations
regarding the value and relevance of the Farm Scale Evaluation
GM crop trials (FSEs).
The Government continues to stand by the design and
operation of the FSEs. They are the biggest ecological study ever
undertaken on the effect of any farming practice, and are a credit
to British science. The work has been internationally recognised
and widely applauded. They produced robust and reliable data on
the impact on biodiversity of the herbicide management of the
GM crops in the trials. They have contributed to the consideration
of our policy on GM crops, and our position on the particular
crops in the trials. We await the results of the winter-sown
oilseed rape FSE with interest
The EAC report was published on 5 March, a few days
before the Government made its statement of policy on GM crops
to Parliament on 9 March. The statement was not delayed, as requested
by the EAC, because the report did not raise any new issues which
would have justified such a delay. The Government's statement
was the result of a long and painstaking assessment of GM crops.
In deciding our policy we gave careful consideration to all the
available evidence, including the reports from the GM public debate,
science review and costs and benefits study, the AEBC report on
co-existence and liability, as well as the results of the FSEs.
Indeed no other country has undertaken such a comprehensive and
rigorous assessment of the case for GM crops.
In our statement we have made it clear we would oppose
applications to grow GM herbicide tolerant (GMHT) beet and spring
sown oilseed rape commercially as managed under the regime in
the FSEs. We also indicated we were prepared to agree in principle
to the commercial growing of the GMHT maize subject to certain
conditions, but in the light of the recent announcement by Bayer
CropScience it is now clear that this particular variety of the
crop will not be grown commercially in the UK.
We were surprised that the EAC conducted an inquiry
evaluating the FSEs without talking to the research consortium
that carried out the trials. In our view this was a serious weakness.
By taking oral evidence from, and questioning, the researchers
the Committee could have perhaps clarified some of the issues
addressed in their report. This may have been particularly helpful
in relation to issues such as the design of the trials, the measurement
of yield, the follow up research on the withdrawal of atrazine
and the involvement of the industry in the trials. It is surprising
the Committee came to its conclusions without the benefit of expert
advice from those directly involved.
The FSEs also raised important questions about the
environmental impact of crop management practices as a whole and
the need to find a balance between a competitive and economically
viable farming industry and protecting our wildlife. Applying
the lessons learned from the FSEs to assess and improve the impact
of all forms of agriculture on the environment will be one of
the most important tasks for Government to take forward out of
this research.
We have asked our independent expert advisers, the
Advisory Committee on Releases to the Environment (ACRE) to consider
the range of issues that the EAC report has raised and it is expected
to issue advice in due course.
In the text below the Committee's conclusions and
recommendations are included in bold.
It is regrettable that the Government failed to
be transparent about the nature of any deal made with the industry
over the inclusion of beet. Given the public's concern and suspicion
on matters relating to the GM industry we would expect greater
openness. (Paragraph 25)
It is clear from the EAC's report that there was
some confusion in the evidence provided by different sources and
it is unfortunate that the EAC consider this reflects a lack of
transparency on the Government's part. The agreement with the
industry that led to the inclusion of the GMHT beet in the FSEs
was made in November 1999, and announced and published at that
time. This renewed the voluntary agreement between the Government
and industry and expanded it to include other provisions in light
of the experience of the pilot year for the FSEs.
When the FSEs were first proposed they were only
intended to evaluate GMHT maize and oilseed rape, as these two
crops were closest to commercialisation. An application for the
beet was in progress, but was not as advanced. In designing the
trials it was decided that a pilot year would be appropriate.
On the basis that the GMHT beet crop was likely to have to undergo
the same sort of process as the maize and the oilseed rape, the
industry was keen for the beet to be included in this pilot year.
It agreed to finance the trials in the pilot year for the beet
crop on the same basis as the other crops. However, it did not,
as suggested in the EAC report, offer to fund the entire FSE for
the beet crop. Therefore, it is not the case that "the industry
only accepted this toughened-up agreement on the basis that the
Government takeover the funding the beet trials". We are
not clear what evidence you have to make such a claim.
During the pilot year the beet application advanced
through the regulatory process to the point where it became appropriate
to reconsider its inclusion in the FSEs and the Government agreed
to do this in the November 1999 agreement. Subsequently it was
confirmed that beet would be included in the FSEs on the same
terms as the maize and oilseed rape.
We consider it unfortunate that, as there was
no definite yield component to the results of the FSEs, rumours
and assertions have been allowed to proliferate concerning how
the crops performed. (Paragraph 41)
We expect future trials to incorporate robust
protocols for formal measurements of yield. (Paragraph 41)
Measurements of yield were not included as part of
the FSEs as they were not necessary for the purposes of the research.
The researchers, in consultation with Scientific Steering Committee
(SSC), chose instead to take a variety of measurements of the
growth stage of the crop. This approach was considered a more
reliable indicator of the performance of the crop under its respective
herbicide management throughout its growth. The detail included
in the FSE results (the paper by Champion et al)
of the manner in which the management of the crop was audited,
and the close similarity of the growth patterns of all three crops
provide convincing evidence that the herbicide management regimes
reflected 'cost effective weed control'.
Even if measurements of yield had been included it
does not follow that this would have provided the conclusive evidence
the Committee desires. Farmers in the FSEs were free to choose
the conventional varieties that they grew alongside the GM crop
according to their experience and local conditions. Yield comparisons
would not be useful in this context because of the wide variability
in yield between conventional varieties. For example, for maize
in 2002 the range across the national list was 12.9 to 17.1 t/ha
dry matter yield. A further complication for both maize and beet
was that the time of harvest was brought forward to a date earlier
than would be usual for the GMHT varieties, in order to meet the
conditions of the consent regarding the need to avoid material
entering the human food chain. Therefore, for maize and beet
at least, the GMHT yields could not have been compared fairly
with varieties in current commercial practice.
It was not the purpose of the FSE trials to assess
the yield performance of GMHT crops, and the requirements for
any future trials will be determined by what is appropriate to
assess the issue being researched. We will not insist that measurements
of yield are included regardless. As noted in the EAC report
yield is one of the factors assessed when considering seed varieties
value for cultivation and use as part of the National Seed Listing
process. Farmers make commercial decisions about which crop varieties
to grow, and the presumption must be that they will grow the variety
that is most beneficial for them. Yield may or may not be the
determining factor in their decision but that is a commercial
matter.
In the context of public concern about GM crops
and the North American experience with GM, we believe that in
order to determine the cumulative effect of rotational crops upon
biodiversity, the FSEs for those crops should have lasted longer
than three years. The trials on forage maize should also have
lasted longer. We believe that the Government must take account
of this in any future trials. (Paragraph 46)
The SSC overseeing the evaluations was satisfied
that the design of the trials was appropriate. The crops in the
evaluations are generally grown for one year as break crops in
arable rotations. The size and scale of the evaluations was chosen
to give sufficient power to test the hypothesis. On the basis
of the results, the researchers were confident that in the case
of rape and beet the effects seen after one year would have a
cumulative adverse effect over subsequent years. In the case of
maize the effects after one year indicated beneficial cumulative
effects. This view was endorsed by the SSC and by ACRE. Therefore
the hypothesis was rejected in each case and conducting additional
trials in subsequent years would not add anything.
If the crops were to be grown commercially they would
be subject to monitoring based on the environmental risk assessment
submitted with the application. If evidence were to come to light
from this monitoring that called into question the FSE results,
or any other aspect of the environmental risk assessment, then
the approval would be reassessed. It is also worth noting that
consents issued for GM crops are time-limited. The consent holder
must submit another application to renew this consent. This would
have to include the results of any monitoring, and any other new
evidence that has become available.
We are concerned that the GMHT forage maize trials
were based on an unsatisfactory, indeed invalid, comparison. It
is vital that the Government permit no commercial planting of
GMHT forage maize until that crop is thoroughly re-trialled against
a non-GM equivalent grown without the use of atrazine. (Paragraph
50)
It is clearly unsatisfactory that no definite
statement has yet been made as to what the results were from the
25% of conventional forage maize fields in which atrazine was
not used and whether or not this sample constitutes a large enough
base from which to extrapolate comparable results for non atrazine
conventional maize against GMHT maize. (Paragraph 51)
We do not accept that the maize trial was based on
an unsatisfactory or invalid comparison. Within the period of
the trials the use of atrazine was the predominant method of herbicide
management on conventional maize and it was entirely appropriate
that its use was permitted during the FSEs. The range of approved
herbicides is often subject to change. Over the course of any
long-term study there is a high probability of such changes, and
withdrawals occurred during the FSE for the chemicals bromoxynil
and avadex, used respectively on spring oilseed rape and beet.
The EU decision to phase out atrazine (and other triazines) was
not taken until a year after the spring sown FSE trials were complete
and will not take full effect until autumn 2005. It is also important
to emphasise that atrazine is being phased out not because of
its effectiveness for weed control, but because of the risk of
unacceptable contamination of groundwater. Future conventional
herbicide regimes that replace atrazine could have equivalent,
or even greater, impacts on weed populations.
We do not accept that there is a need for the GMHT
maize to be retrialled. It is the herbicide management of conventional
maize that will be changing as a result of the phasing out of
atrazine, not the GM maize. Further trials on the GM maize would
not tell us anything new about the impact of the herbicide management
of the crop on biodiversity, and thus would not be an appropriate
use of taxpayers' money. We made clear that if the consent holders
were to seek to renew the existing consent for the GM maize when
it expires in 2006 they would need to submit new evidence providing
a comparison against conventional practice current at that time.
A peer-reviewed analysis of the impact on the maize
FSE results of the phasing out of atrazine by members of the FSE
research consortium (Perry et al) was published by Nature
on 4 March. This concluded that the benefits to biodiversity of
growing GMHT maize compared to conventional maize would be reduced
but not eliminated by the withdrawal of atrazine. This suggests
that in the short-term the banning of atrazine will not invalidate
the conclusions of the FSE with respect to GM maize.
The problems in North America have not been taken
seriously enough. Defra should have advised the SSC to take account
of North American experiences with GM. (Paragraph 31)
We are unhappy that this work on north American
GM experiences has been left until after most of the FSEs have
reported. Consequently, the findings from that trans-Atlantic
research have not now been factored in to the decisions that are
already being reached on those GMHT crops in the UK nearest approval.
This is clearly unsatisfactory. No decision to proceed with the
commercial growing of GM crops should be made until thorough research
into the experience with GM crops in North America has been completed
and published. (Paragraph 31)
The North American experience with oilseed rape
and the devastation of organic rape production should serve as
an impetus to Government to bring in prudent guidelines for separation
distances as quickly as possible. (Paragraph 37)
We are very concerned about possible contamination
by gene-flow and pollen spread of non-GM crops and insist that
the issue of liability be settled before any GM crops are allowed
to be commercially grown in the UK. The Government should ensure,
through primary legislation, if necessary, that it puts in place,
before any GM crops may be grown commercially in this country,
a clear and comprehensive liability regime to underpin any future
regulations dealing with co-existence issues. Moreover, liability
should lie with the industry and not with farmers. It would be
wrong for the Government to allow farmers to be used as a firewall
for the industry. (Paragraph 38)
Any evidence from North America that is relevant
to the safety assessment of GM crops in the UK is always considered
very carefully. However, the FSEs were focused on a specific
issue. The general experience of commercial GM cropping in North
America is of more relevance to a wider discussion about GM crops
rather the conduct of a specific research project. In addition
it is a valid argument that agronomic conditions differ significantly
between the UK and the US. For example different climates, weed
populations, and the fact that maize in the US is typically cultivated
for grain production whereas in the UK it is typically a fodder
crop.
There has been contradictory information coming out
of North America on the performance of GM crops. In our view the
EAC's assessment of the experience of GM crops in North America
is unduly negative. In particular we are surprised at the prominence
given to the views of the Canadian National Farmers Union given
that we understand that they represent less than 3% of farmers
in Canada. A cross section of views would have helped come to
a balanced conclusion. If it turns out that a crop ceases to
offer farmers any tangible benefit then the presumption must be
that they will stop growing it. However, the use of GM crops
in Canada and the USA continues to increase, and increased again
last year. For example GM varieties accounted for 40% and 58%
of US and Canadian maize, and 84% and 68% of the respective oilseed
rape crops, in 2003. This would suggest that the majority of farmers
growing GM crops in North America see commercial benefits.
We accept that there are issues that have been raised
in the context of the experience in North America that have to
be addressed, but these are separate from the FSE research. For
example, to the extent that there have been contamination problems
in North America, this may reflect the fact that few, if any,
steps have been taken there to manage the co-existence of GM and
non-GM crops. As we made clear in our policy statement we intend
that co-existence measures will be in place here before any commercial
GM crops are sown. Our intention is that GM producers should
observe a code of practice on co-existence with statutory backing,
based on the 0.9% EU labelling threshold for adventitious GM presence.
We will consult on this, on whether a lower GM threshold might
apply for organic production, and on options for compensating
non-GM farmers who suffer financially because a GM presence exceeds
statutory thresholds.
We are concerned that the industry was responsible
for a number of key inputs into the operation of the trials which
appear to have been assessed only against very broad or vague
criteria, or which were taken on trust. Even if these inputs had
no cumulative effect upon the results, they were significantly
integral to raise significant concerns as to the extent to which
the industry was in practice capable of influencing the trials.
(Paragraph 62)
It was the industry body SCIMAC which provided the
inputs to the evaluations. SCIMAC is an umbrella body comprising
UK Agricultural Supply Trade Association, National Farmers Union,
Crop Protection Association, British Society of Plant Breeders
and British Sugar Beet Seed Producers Association. As the EAC
report itself acknowledges there were various inputs into the
operation of the FSEs that it was entirely appropriate for the
industry to provide, and indeed the trials could not have taken
place without this degree of industry involvement. There is no
substantiated evidence to support the concern that the industry
may have unduly influenced the trial results. The research consortium
audited the management of the trials and concluded that there
was no evidence of bias in the inputs into the trials (See Champion
et al).
Damage to the trial sites should be seen by those
responsible for that damage as counterproductive, since it undermines
the scientific validity of evidence that could well support their
claims. We support the lawful right to protest but feel that future
trials should be better protected in order to safeguard scientific
evidence that may prove very valuable in domestic and international
discussions as to whether the commercial growing of GM crops should
proceed. Defra must consult with appropriate security bodies about
achieving more secure trials in future. (Paragraph 57)
We share the Committee's concern about damage to
trial sites. The Government and the industry put significant
effort into protecting the FSE sites. The Government is committed
to the principle of transparency and as such we made all locations
of sites public, and made information explaining the trials available
to people and farmers close to the crop trials. We also liased
with the police, and ensured they were fully informed. We understand
that in some cases the industry provided private security for
farmers, their families and their property at certain times in
response to threats and intimidation. However, it was simply not
feasible to provide this at all the FSE sites for the duration
of the trials.
It is unfortunate that a small number of saboteurs
have abused our policy of publishing detailed site locations.
But the actions of these extremists have not lessened our commitment
to transparency, particularly the need to provide adequate information
to people and farmers close to GM crop trials. We appreciate
the concern that a legitimate activity is being hindered and we
are keeping our policy on site disclosure under close review.
However, it is difficult to see, without devoting disproportionate
resources to the task, what can practically be done to increase
protection of the trial sites.
It is vital the Government makes clear in its
decision exactly what will be required of applicants in future,
and how it will assess whether there is evidence of biodiversity
harm from the use of the GM crop and herbicide regime for which
the particular application is made. (Paragraph 65)
We agree that the industry should pay for any
future trials including the future trials we think necessary for
forage maize. (Paragraph 65)
We recommend that future GM crop assessments of
biodiversity impact should be no shorter than four years. (Paragraph
67)
We expect to see thorough multi-year and multi-site
trials for any new applications. We likewise expect comparative
assessment of biodiversity harm to be undertaken on a crop-by-crop
basis. (Paragraph 68)
ACRE has carefully considered what will be required
of future applicants seeking to market GM crops in the UK in terms
of wider biodiversity risk assessment. ACRE established a sub-group
in 1999 specifically to address this question, full details of
this 'ACRE sub-group on wider biodiversity issues' are available
on its website[2], including
minutes of its meetings.
The sub-group published a guidance document in July
2001[3]. This guidance
reinforces the need for case-by-case assessment of crop management
effects. It is conceivable that some potential crop management
systems might differ little from existing practice while others
may present novel systems of which there is little previous experience.
One may require little or no field testing while another might
require years of testing and/or monitoring. It would not be appropriate
to be prescriptive about the level and duration of testing required.
One of the most notable factors in the FSE results was the consistency
of the treatment effects over many independent variables, including
years. This suggests that trials of this scale and length are
unlikely to be necessary in the future.
It is clear from Directive 2001/18/EC that the onus
is on the applicant to undertake such trials as are considered
necessary to provide the level of evidence required by the regulatory
authorities. This was not the case under the preceding Directive
90/220/EEC. The detailed design of any future trials will need
to be decided on a case-by-case basis.
The benchmark against which GMHT crops were measured
was not ambitious, since biodiversity in conventional crops has
suffered greatly over the last half-century. (Paragraph 6)
Biodiversity levels have slipped intolerably over
the last fifty years and Government has a duty to attempt to regain
some of that lost ground. Indeed, the Government, in the light
of the Curry report, should establish a benchmark for biodiversity
in conventional crops, at the less intensive end of the spectrum.
It is against this benchmark that future trials should assess
innovatory practices and regimes in conventional agriculture.
This ought to make the benchmark used in the FSEs irrelevant.
(Paragraph 72)
While we applaud the steps the Government has
taken to assess biodiversity in a rational way before permitting
an agricultural innovation in the form of GM, we believe that
even if some GM crops with some associated herbicide regimes are
eventually shown to be less harmful to biodiversity than their
conventional counterparts, the Government and its advisory bodies
are still guilty of setting too low the level of harm. (Paragraph
73)
We therefore recommend that in future trials the
biodiversity benchmark against which GM crops should be assessed
should be that associated with the less intensive and more biodiversity-friendly
end of the spectrum found in UK agriculture, such as organic crops.
(Paragraph 73)
We share the EAC's concern that biodiversity levels
have suffered over previous decades. However, in the specific
case of the FSEs we do not accept that the 'benchmark' should
have been set at a higher level. The FSEs must be placed in their
regulatory context. The indirect environmental impact of the management
techniques associated with GM crops compared to the management
of the non-GM counterpart is one aspect of the risk assessment
of a GM crop required by Directive 2001/18/EC. The key comparison,
in order to make regulatory decisions, was thus an assessment
of the herbicide management of the GMHT crop against the current
herbicide management of its conventional counterpart. Some conventional
crops performed poorly in relation to biodiversity in relation
to the chemical inputs. It is not clear why even when the Committee
accepts some GM crops may be less harmful to biodiversity they
alone must pass a threshold not applied to conventional ones.
The FSEs also raised important questions about the
environmental impact of crop management practices as a whole and
our strategy for sustainable farming. The trial crop with the
'best' results for the environment was a conventional crop and
the one which was 'worst' was also a conventional crop.
We are looking carefully at how the experience of
the evaluations can help us to strike the appropriate balance
between the needs of an economically healthy and competitive farming
sector and the need to protect our wildlife. This will include
a consideration of what is the most appropriate 'benchmark' against
which crops and management systems should be assessed. The FSEs
are invaluable in informing our consideration of such a benchmark
as the studies are the most thorough assessment of any crop management
systems ever undertaken. The FSE results in particular will contribute
to the development of the Sustainable Farming and Food Strategy
which incorporates our objectives to increase biodiversity in
farmland and also will inform the development of the England Biodiversity
Strategy and future agri-environment and cross-compliance schemes.
In addition we intend to see how the methodology
used in the FSEs can be streamlined and improved to be less resource
intensive, and will consider how a similar approach may be more
widely applied to assess the risks of other (non-GM) crop or farm
management situations, such as pesticides or the introduction
of new bio-fuel crops or to assess the true value of organic systems.
In our agricultural research programme we will apply the lessons
learned about how different farming practices interact and how
these in turn affect the environment at a landscape scale.
It is inconceivable that beet or spring sown oilseed
rape will be given consents to be grown if managed under the same
regime as applied in the FSEs. (Paragraph 63)
The scope of the trials was very narrow and the
results cannot be regarded as adequate grounds for a decision
to be taken in favour of commercialisation (Paragraph 74)
It would be irresponsible for the Government to
permit the commercialisation of GM crops on the basis of one narrow
component of the entire evaluation of GM technology. This would
be the case even were there no significant doubts about as to
the robustness, validity and relevance of the FSE results. (Paragraph
75)
The FSE trials were designed to address the specific
question of the impact on biodiversity of the herbicide management
regimes associated with GM crops. They addressed only one aspect
of the required risk assessment and on their own they were never
intended to be the sole basis for a decision to commercialise
GM crops. When considering applications the full range of safety
issues required for the risk assessment will be considered, as
specified in Directive 2001/18/EC.
The Government's policy statement made it clear we
would oppose applications to grow GMHT beet and spring sown oilseed
rape commercially as managed under the regime in the FSEs.
Department for Environment, Food and Rural Affairs
April 2004
2 www.defra.gov.uk/environment/acre/biodiversity/index.htm Back
3
www.defra.gov.uk/environment/acre/biodiversity/guidance/index.htm. Back
| 
