Memorandum from English Nature
THE FARM SCALE EVALUATIONS OF GMHT CROPS
IN BRITAIN: DEVELOPMENT, RESULTS AND IMPLICATIONS FOR BIODIVERSITY
IN THE UK
1. INTRODUCTION
1.1 English Nature is one of the four British
statutory nature conservation agencies (English Nature, Countryside
Council for Wales, Scottish Natural Heritage and the Joint Nature
Conservation Committee). The agencies are statutory consultees
in the process by which applications for release of genetically
modified organisms (GMOs) into the environment are considered,
and also have a general duty to advise government on conservation
policies. We are solely concerned with potential impacts of GMO
releases on the living environment and on sustainable use of our
natural resources, including protected sites and the wider countryside.
We have no locus on matters of public health and safety. The GMO
Lead Agency, based in English Nature, undertakes JNCC's work on
GMOs on behalf of the other statutory country agencies.
2. DEVELOPMENT
OF THE
FSES
2.1 The following narrative is partly derived
from a 2000 DETR (now Defra) paper "The History of the Farm-Scale
Evaluations".
2.2 The first experimental releases of GM
herbicide tolerant (GMHT) crops took place in the UK in the late
1980s. GMHT oilseed rape was first considered for placing on the
European market in 1994 closely followed by various maize varieties.
In advising the UK government on the application, the Advisory
Committee on Releases to the Environment (ACRE) was concerned
that while the risks of the GM crop by itself appeared to be low,
potential environmental effects associated with the use of broad-spectrum
herbicides on the crop should also be assessed.
2.3 In 1996 ACRE discussed specific
issues relating to GM herbicide tolerant crops with officials
from the Pesticides Safety Directorate (PSD) in attendance. ACRE's
main concern was to identify whether or not impacts of the use
of herbicides are adequately covered under the Pesticides legislation,
and whether there are any gaps in the legislation. Their discussions
and conclusions are published in Chapter 3 of the ACRE Annual
Report No 4 (1996-97). English Nature pressed ACRE to consider
the indirect effects of changing herbicide regimes when assessing
risks to the environment from GMHT crops, arguing that ACRE had
such a remit under the EU Directive on the Deliberate Release
of GMOs (EC 90/220).
2.4 In subsequent marketing notifications,
ACRE continued to raise issues with the Pesticides Safety Directorate
regarding herbicide use on GMHT crops.
2.5 In 1997 English Nature (EN) with
the other statutory nature conservation bodies called publicly
for a moratorium on the commercial use of genetically modified
herbicide tolerant (GMHT) and genetically modified insect resistant
(GMIR) crops until further ecological research was carried out.
They were especially concerned about the impact of farming practices
on farmland wildlife. GMHT crops could exacerbate wildlife declines
if they encouraged higher levels of weed control than necessary,
which in turn would reduce invertebrate and bird numbers. There
was some preliminary scientific information to support this contention
for GMHT maize, oilseed rape and beet. GMIR crops could have direct
effects on certain invertebrates such as butterflies, moths and
beetles.
2.6 By early 1998, DETR had started
in depth discussions with English Nature to lay the ground for
development of policy to address wider biodiversity issues related
to GM crops.
2.7 In June 1998 Michael Meacher
hosted a meeting to discuss the wider biodiversity issues with
experts from English Nature, RSPB, Green Alliance, National Institute
of Agricultural Botany, MAFF and ACRE members. Feedback from this
meeting was used to prepare a draft paper on the issue. In consultation
with ACRE, other Government Departments and English Nature the
text was developed along the lines of:
— Severe declines in farmland wildlife
had been detected, largely the result of increasingly intensive
agricultural practices.
— The UK has a statutory responsibility
under EU Directives (eg 79/409/EEC, relating to birds and habitats)
and other international agreements to maintain populations of
farm-dependent wildlife, especially birds).
— Biodiversity Action Plans are in
place to address this.
— Decisions on the marketing of GM
crops must not encourage further intensification, or Biodiversity
Action Plan targets will be prejudiced.
— Investigation of potential adverse
effects on farmland wildlife of GM crop management was needed
as well as statutory investigations already carried out on the
GM crop itself.
— A more strategic approach was needed,
in terms of what is wanted from agriculture in terms of food and
biodiversity, against which regulatory decisions can be made.
2.8 By October 1998: DETR was exploring
the possibilities of a moratorium on the commercial planting of
GM crops. Discussions were held with Friends of the Earth (FoE),
English Nature (EN), RSPB and Genewatch, and separately with industry.
Calls from some NGOs for a blanket moratorium on GM crops (which
would be illegal under EU law) were rejected but instead the Government
proposed a programme of closely monitored commercial-scale plantings
of GM crops to assess the impact on wildlife of the new herbicide
regime. A voluntary agreement was made with industry to delay
commercial planting of GM crops for one year whilst research was
carried out. On 21 October, Michael Meacher and Jeff Rooker appeared
before the House of Lords European Communities Committee on Agriculture
and announced the concept of "managed development" and
the voluntary agreement with industry.
2.9 Following the announcement specifications
for the letting of contracts for the ecological research were
developed in consultation with English Nature and other ecologists.
The research was funded by DETR with contributions from MAFF and
the Scottish Executive. 15 major research organisations were invited
to tender for research contracts investigating effects of management
of herbicide tolerant oilseed rape (both spring and autumn sown),
and maize.
2.10 In autumn 1998 negotiations
started under the Austrian Presidency about amendment of Council
Directive 90/220/EEC on the release of GMOs into the environment.
The UK was instrumental in the development of technical annexes
on risk assessment and included text to ensure that indirect effects
such as the potential impact of the management of GM crops was
considered, in addition to the direct impact of the crops themselves.
The UK also played a key role in developing a technical annex
on post-market monitoring.
2.11 In December 1998, EU environment
ministers agreed that the technical annexes on risk assessment
and post-market monitoring should be implemented with immediate
effect whilst other provisions of the Directive continued to be
negotiated.
2.12 In February 1999 public concern
about GM crops reached fever pitch with daily headlines in the
media. ACRE published a discussion paper entitled: "The commercial
use of GM crops in the UK: the potential wider impact on farmland
wildlife". The following month an ACRE Sub-Group on Wider
Biodiversity Issues was set up and held its first meeting, adopting
the "Wider Issues" Paper as the basis for its deliberations.
Its agreed priority was to develop guidance for consent applicants
on how the wider biodiversity impacts of the management of GM
crops should be addressed in future applications for consent to
market.
2.13 In April 1999 the contract for
the farm scale evaluations was let to a Consortium of three research
organisations: Institute of Terrestrial Ecology (ITE) (now Centre
for Ecology and Hydrology, CEH), Institute of Arable Crops Research
(IACR) and Scottish Crop Research Institute (SCRI).
2.14 Secretary of State appointed an independent
Scientific Steering Committee (SSC), chaired by Professor Chris
Pollock to oversee the evaluations. The members of the committee
were Professor Mick Crawley, Imperial College, London, Dr David
Gibbons, RSPB, Dr Nick Sotherton, Game Conservancy Trust, Mr Jim
Orson, Morley Research Station, Dr Alastair Burn, English Nature
and Dr Nicholas Aebisher, GCT. The SSC first met in June 1999
and endorsed the FSE programme. The experiments started in spring
1999 with pilot plantings of two spring oilseed rape fields and
four fodder maize fields, followed by three autumn sown rape fields.
One field of sugar beet was also sown.
2.15 In early 2000, the results of
these pilot studies enabled the Steering Committee to make a final
decision on the design of the main experiments. They chose a "split
field" protocol and decided on the range of organisms that
could be feasibly measured and were biologically relevant to the
null hypothesis being tested. They also agreed the inclusion of
GMHT beet, both sugar and fodder, in the programme. At the same
time the statistical design was finalised on the basis of a "power
analysis" that enabled the size and number of fields to be
chosen. It became clear from this that around 20 fields per crop
per year would be needed and that these should both cover a wide
geographic range and include farms that differed in their management
intensity. The main experiments on spring-sown crops started in
March 2000, with winter-sown oilseed rape planted later
that year.
2.16 The FSEs were the largest manipulative
experiment ever carried out on farmland ecology anywhere in the
world, exceeding by more than threefold any of the comparable
experiments undertaken previously.
3. THE RESULTS
3.1 The results were published in October
2003 as a series of papers in a theme issue of the Philosophical
Transactions of the Royal Society (Phil Trans, R Soc Lond
B 358). The research team and the steering committee have
produced a non-referred summary of the papers in a concise and
more accessible style and a commentary on the results that assesses
the implications of the GMHT crops for farmland biodiversity.
This assumes that the crops would be used on a large scale and
estimates what might be the effects on farmland biodiversity.
4. THE FSES
AND REGULATORY
ISSUES
Can this research design provide sufficient information
to enable the UK Government to discharge its responsibilities
under 2001/18/EC regarding assessment of indirect effects on the
environment of the three crops in question?
4.1 Annex II of Directive 2001/18/EC makes
it clear to regulatory authorities that they must:
— consider the indirect effects of
the GMO on the environment, including changes in agricultural
management associated with the GMO; and
— perform this assessment by comparing
the GMO and its associated management systems with equivalent
non-transgenic organisms.
4.2 Our view is that the design of the FSE
experiments conforms very well to these criteria. The experiments
were set up to investigate the impacts on farmland biodiversity
of changing herbicide systems from conventional weed control used
on non-modified crops with the broad spectrum weed control used
on the three GMHT crops. Importantly the methods used in our opinion
reflect very well the most likely outcome of general release of
these crops if they were to be used as set out in the application
dossiers. The use of herbicides on the conventional halves of
the fields was shown to be typical of the methods that were normally
used on the farms in question, and the broad spectrum herbicides
used over the growing GM crops were applied according to the recommendations
of the companies that submitted the dossiers. We are therefore
firmly of the view that the outcomes from the FSEs fairly represent
the most likely impacts on biodiversity from releasing these crops
commercially.
Does the evidence from the FSEs constitute sufficient
information to make robust regulatory decisions on the commercialisation
of GMHT spring oilseed rape, sugar beet, fodder beet and maize?
4.3 The FSE experiments proved to be statistically
over-powered in the sense of being able to detect smaller changes
in "between treatment" effects than was first envisaged.
This was somewhat surprising but partly due to remarkable consistency
of effects within crop type over a wide geographic range and a
variety of different farms. The data sets collected were very
large, covering a range of organisms sampled from different trophic
levels. This enabled robust analysis to be made of the impacts
of the different herbicide regimes on trophic levels and on factors
such as seed return that are vital in the maintenance of biodiversity
within cropped landscapes.
4.4 The results were, as the Steering Committee
and research team reported, clear and conclusive. The use of broad-spectrum
herbicides over GMHT beet and oilseed rape gave large reductions
in the primary trophic level, non-crop plants. This in turn had,
as ecological theory predicts, an adverse effect on abundance
of some invertebrates, especially those that feed on nectar. Seed
rain and seed banks in the following year were significantly reduced
in these GMHT crops. The later weed control in GMHT beet and oilseed
rape crops resulted in significantly higher densities of springtails,
small insects feeding on decaying weed residues following herbicide
applications, and lower densities of herbivores, predators and
parasitoids in beet. This is effectively a shift from a plant-eating
to a detritus-eating invertebrate community at a time of year
(early summer) that is unusual in arable systems. There is currently
little scientific evidence to be able to predict the impacts of
this change on higher trophic levels such as birds.
4.5 By contrast the results for GMHT maize
show that this herbicide system results in more biodiversity in
and around fields compared with the conventional residual herbicide
systems in general use on forage maize. However, the study showed
that both GMHT and conventional maize support much lower biomass
and densities of non-crop plants and invertebrates than the other
crops studied. This, as the authors suggest, may be attributed
partly to the persistent effects of atrazine and partly to strong
competition from the maize crop from July onwards.
4.6 We know that as break crops, oilseed
rape, sugar beet and fodder beet fields are relatively weedy and
allow seed banks to be replenished in between years of growing
cereal crops. This means that despite their comparatively small
area in arable areas of the UK they are disproportionately important
for farmland biodiversity, especially for birds such as skylarks,
reed buntings, corn buntings, linnets and stone curlews.
4.7 For example, recent BTO/Defra research
found that oilseed rape stubbles had higher densities of weed
and crop seeds on the soil surface than any other crop studied
and some fields supported high densities of birds, especially
early in winter (BTO 2003). Many sugar beet fields also had high
densities of seeds, although lower than oilseed rape. These figures
are backed up by the FSE results. Weed control is good in most
sugar beet fields, but is not very effective in around 10% of
fields, allowing weeds to flourish and set seed, providing an
important source of food for birds in comparison to other crops
(Lainsbury et al 1999).
4.8 A concerted programme of research over
the last 12 or so years has demonstrated the importance of seed
and invertebrate availability in the declines in farmland birds.
Robinson & Sutherland (1999) found that habitat preferences
of farmland birds in winter were related to seed density; skylarks,
grey partridge and red-leg partridge were never found at seed
densities lower than 150 m-2. In Scotland, overwintering oilseed
rape stubbles have higher densities of seed eating passerines
than any other crop (Hancock and Wilson 2003). Sugar beet is used
as a nesting habitat by some ground nesting species (eg skylark
and lapwing) that avoid dense cover, as it is sown in spring and
so is not too tall early in the breeding season. About 15% of
stone curlew nests in Breckland are in sugar beet (21 out of 145
nests in 2002), by far the most preferred crop type (EN/RSPB Stone
Curlew Project annual report). A lack of invertebrate food has
been linked to the decline of at least two species—grey
partridge and corn bunting (Potts 1997; Brickle et al 2000).
4.9 We have also received information from
plant recorders in southern England suggesting that oilseed rape
fields may be important refugia for rare arable plants. For example,
populations of Valerianella rimosa, Scandix pecten-veneris
and Galeopsis angustifolia have been found in oilseed rape
fields in Hampshire and Wiltshire (ref John Moon pers comm). This
is most likely attributable to the low and sporadic nature of
herbicide applications on oilseed rape.
4.10 We agree with the research team view
(Firbank et al 2003) that the implications for general
farmland biodiversity can be clearly seen from the results. If
GMHT beet and oilseed rape were released for general commercial
use, there would be a further reduction in farmland biodiversity.
If GMHT maize were released the opposite would be the case.
4.11 We cannot predict the magnitude of
adverse or beneficial effects on biodiversity of herbicide regimes
associated with the crops tested. Models developed by Sutherland
et al (2003) and Watkinson et al (2000) may eventually
give some insight into the immediate effects, but, as the FSE
research team pointed out, such estimates rely on assumptions
about uptake by farmers, distribution and areas of GMHT crops
cultivated and policy influences such as the impact of CAP reforms
on cropped areas. In addition any forward look would have to take
into account the impacts of UK and European measures to encourage
non-food uses of these crops. We argue that these factors are
almost impossible to predict with any degree of certainty.
4.12 What we can be sure about is that if
GMHT beet and spring oilseed rape were to be commercialised in
the near future by the unconditional general release that we believe
is implicit in the application dossiers, then there would be significant
reductions in farmland biodiversity associated with these crops.
Furthermore, there is a strong likelihood that these effects would
increase over time, as indicated by the significant declines in
dicot seedbank densities seen in fields used to grow GMHT beet
and oilseed rape. This would in our view take farmland biodiversity
in the opposite direction to that envisaged by the overall thrust
of UK government policies, which are to maintain and enhance biodiversity
on cropped land. This policy thrust is illustrated well by the
government's adoption in 1999 of wild bird populations as a headline
indicator of the sustainability of lifestyles and the general
"Quality of Life" in the UK, and Defra's Public Service
Agreement target 3(i) which aims to reverse the long-term decline
in the number of farmland birds by 2020.
4.13 Equally, we are firmly of the view
that using GMHT maize could have several environmental advantages
over current systems of cultivation, including higher in-crop
biodiversity, more biodiversity in the crop margins and verges,
and the possibility of having weeds or under-sown crops in maize
stubbles over winter. It is less clear that the benefits of GMHT
maize would be cumulative over time, since there was no significant
increase in seedbank density following GMHT maize, despite the
fact that the GM fields contained higher densities of weeds and
significantly higher dicot seed rain than conventional maize.
4.14 The most important elements of the
conventional herbicide systems against which the GMHT maize system
was compared in the FSEs (atrazine, simazine and cyanazine) have
already been banned in the EU or will be phased out during the
next 18 months, and it has been argued that this may invalidate
the results of the experiments as a method of predicting the net
impacts on biodiversity of growing GMHT maize. It is possible
that, following the phasing out of atrazine and related compounds,
growers of conventional maize will be forced to use other herbicides
that are less effective than either current methods, or the glufosinate-ammonium
GMHT system, and this could result in conventionally grown maize
supporting more biodiversity than GMHT maize. We do not have enough
information on this subject area to be able to predict either
the new herbicide regimes or their impacts on biodiversity. However,
we might reconsider our advice in the future should herbicides
carrying lower environmental risks replace the range of herbicides
currently used by growers of conventional maize.
4.15 Although not required by Directive
EEC/90/220 (under which the GMHT maize already has Part C consent
for commercial growing), we would expect a post market monitoring
plan to be drawn up and carried out to confirm the positive effects
on biodiversity that are predicted by the results of the FSEs.
This would have the additional advantage of enabling comparisons
with herbicide regimes used on conventional maize after atrazine
is withdrawn, with the possibility of reviewing the consent at
a later date if appropriate.
4.16 Our advice to government will be that
the decision to release these crops should be based on the above
factors and not on whether mitigation measures could be used to
alleviate the effects of the broad spectrum herbicides associated
with GMHT beet and oilseed rape. The application dossiers for
these crops contain no reference to mitigatory measures, and we
have not seen any specific proposals for crop management other
than unfettered general release. We also have serious doubts about
whether mitigation could be either delivered or "policed".
Neither should an attempt be made to estimate the magnitude of
adverse or beneficial impacts; decisions should be based on the
general impacts on biodiversity identified by the results of the
FSEs. We believe that this approach fits well with the precautionary
principle outlined in the Directive. Our farmland wildlife is
already at a very low ebb and we cannot risk any further reductions
caused by commercialisation of GMHT crops known to have significant
adverse effects on biodiversity. Equally the use of cropping systems
that have clear biodiversity benefits should be permitted, provided
that they can be demonstrated to have sufficiently low risks to
other aspects of the environment and to human and animal health.
November 2003
References
Brickle, N W, Harper, D G C, Aebischer N J,
and Cockayne S H (2000). Effects of agricultural intensification
on breeding success of corn buntings Miliaria calandra
in southern England. In press: J Appl Ecology Vol 37, no
5, pp 742-755. Oct 2000.
BTO (2003). The effects of different crop
stubbles and straw disposal methods on wintering birds and arable
plants. Project BD1610. Unpublished BTO report to Defra.
Burton et al (1996). The effects of
pre-harvesting operations on birds nesting in oilseed rape
Brassica napus. BTO Research Report 171.
Clarke R, Combridge P & Middleton N (2003).
Monitoring the diets of farmland winter seed-eaters through raptor
pellet analysis. British Birds 96, 360-375.
DETR 2000 The history of the Farm-scale Evaluations
http://www.defra.gov.uk/environment/gm/fse/background/history.htm.
Firbank L G, Perry J N, Squire G R, Bohan D
A, Brooks D R, Champion G T, Clark S J, Daniels R E, Dewar A M,
Haughton A J, Hawes C, Heard M S, Hill M O, May M J, Osborne J
L, Rothery P, Roy D B, Scott R J and Woiwod I P (2003). The implications
of spring-sown genetically modified herbicide-tolerant crops for
farmland biodiversity: A commentary on the Farm Scale Evaluations
of Spring Sown Crops.
Hancock, MH and Wilson, JD (2003). Winter habitat
associations of seed eating passerines on Scottish farmland. Bird
Study 50: 116-130.
Lainsbury, M A, Hilton, J G & Burn, A (1999).
Proc 1999 Brighton Conf Weeds 3, 817-822.
Potts, D (1997) Cereal farming, pesticides and
grey partridges. In: Farming and Birds in Europe.
Robinson, R A and Sutherland, W J (1999). The
winter distribution of seed-eating birds: habitat structure, seed
density and seasonal depletion. Ecography 22, 447-454
Watkinson A R, Freckleton R P, Robinson R A
& Sutherland W J (2000). Predictions of biodiversity response
to genetically modified herbicide-tolerant crops. Science,
289, 1554-1556.
Wilson, J D, Taylor, R and Muirhead, L B (1996)
Field use by farmland birds in winter: an analysis of field type
preferences using re-sampling methods. Bird Study 43: 320-332.
| 
