Examination of Witnesses (Questions 280-299)
3 DECEMBER 2003
PROFESSOR CHRISTOPHER
POLLOCK, DR
NICK BRICKLE,
DR BRIAN
JOHNSON AND
DR MARK
AVERY
Q280 Chairman: This is a matter of some
confusion. Your Steering Committee first met on 14 June, according
to the memorandum. I have a memorandum here from Defra, which
says that the decision to include beet was taken at a meeting
in November, and yet you have just told us that beet was already
in this when you came on the scene?
Dr Brickle: Is that the decision
for Defra to adopt the funding for beet? I can read the first
minute, that separate farmers are evaluating, looking at the effect
on wildlife, and are being funded by SCIMAC. "SCIMAC have
asked if the Steering Committee could advise on the progress of
the sugar beet evaluations." Then it would be in December
that Defra, for separate reasons, took over the funding.
Q281 Chairman: Defra is quite clear that
the meeting took place in November 1999 and it says: "At
the same time, Ministers agreed to include GMHT sugar beet in
the evaluation on the same terms as rape and maize". There
is, I am afraid, some contradiction over the precise timing here.
We may have to pursue that.[5]
Dr Brickle: I would suggest that
refers to funding conditions. Their agreement to oversee it as
a scientific study is clearly minuted in that 14 June SSC[6]document.
Q282 Joan Walley: Yes, but presumably
that begs the question that at the very outset Defra had not approved
the funding for beet, so therefore there was a change to what
was going to be included in the trials?
Professor Pollock: From a scientific
standpoint, not at all, no, because what we were doing was administering
four sets of studies under four identical scientific conditions,
which is what the Scientific Steering Committee was required to
do.
Q283 Joan Walley: Why was not beet included?
Where did the pressure come from for beet to be included?
Dr Johnson: I might be able to
shed some light on that. Before the field scale trials were set
up, English Nature presented evidence to the Government that there
was indeed a real problem with the impact of the crop management
system on biodiversity. That evidence came from three sets of
experiments that were carried out at a research station under
research station conditions on maize, oilseed rape and beet. English
Nature was as concerned about the impact on beet fields as we
were about the impact on oilseed rape and on maize fields because
we know that they are important for some farmland wildlife. We
continued to press the Department to include beet in these trials.
If my memory is right, one of the logistical problems with including
beet at that time was the shortage of seed. The industry had told
me that it would be very difficult to provide sufficient seed
for a large-scale trial; they did not really know whether they
would or would not be able to do that. There had been some problems
with GMHT sugar beet seed production at that time. I think that
is one of the reasons why the decision to include beet was delayed.
There are no doubt others which Dr Brickle has alluded to. Certainly
that was one that I recall.
Q284 Joan Walley: I still do not understand
why the Department did not provide the funding so that the funding
would be in place, subject to the availability of seeds. It seems
to me that it was added as an added-on extra?
Dr Johnson: We were pressing for
it all the time. We thought that beet should be included in these
experiments.
Q285 Mr Savidge: My initial question,
for a general comment from any of you, is: to what extent do you
have any concerns about the SCIMAC guidelines? Were they sufficiently
stringent; were they properly applied; are you content with the
independent audit that ADAS did for SCIMAC; and were there any
instances where the guidelines were not properly followed through?
Professor Pollock: The guidelines
represented SCIMAC's assessment of how the herbicide would be
used to obtain effective weed control. The use of the herbicides
was monitored throughout the trials and all the information about
the usage is in the eight refereed papers and the summary. We
can validate the effectiveness of this in terms of effective weed
control. Again, the data are in the refereed papers. I am content
that the data show that that the guidelines were appropriate to
obtain levels of weed control that were commercially sensible,
that there was no indication, for example, as has been suggested
by others, that the timing and intensity of spray was altered
in a manner that would get less effective weed control and therefore
perhaps improve the level of biodiversity within the crop. I think
you have a paper by Champion et al in the Royal Society Philosophical
Transactions that presents all the information that would allow
people to conclude that that is what was attempted and what was
obtained in that paper.
Q286 Mr Savidge: Do you want to comment
on the matter of whether there were any cases where the guidelines
were not properly followed?
Professor Pollock: In terms of
whether it affected the quality of the output from the study,
that is, across the broad range of the numbers of fields that
were sampled, I am content that cost-effective weed control in
the herbicide-tolerant crop was generated. Obviously there is
going to be some variance within this, dependent upon field, environment
and year, as there was in the level of weed control that was obtained
in the conventional crops, but we could not distinguish a difference
in terms of effective weed control between the two treatments.
Q287 Mr Savidge: Does it concern you
that the Head of the Scottish Executive GM Co-ordination team
under oath in a Scottish court appeared to indicate that the guidelines
were never properly applied at all in Scotland? Could that apply
elsewhere?
Professor Pollock: Again, I can
go by the data that there are in the paper of Champion et al,
and that does not support the contention that the way in which
the herbicides were applied significantly affected the outputs
of the trials in terms of one part of the trial having less effective
weed control than the other.
Q288 Mr Thomas: You referred to the paper
by Champion in your site transaction. However, we are also aware
as a Committee of the actual measurement, the yield from the field
scale evaluations to evaluate that commercial sense. Also, in
your evidence when you outlined what the Scientific Steering Committee
assessed in advance of doing these trials, you do not include
yield. Are you confident that the commercial aspect was really
measured within these field scale evaluations?
Professor Pollock: I will need
to go back over information that I have given in a number of cases.
The trials were designed to look at the relative effects on biodiversity
of herbicide application to obtain effective weed control, the
"null hypotheses" being that there was not a difference,
and that hypothesis was there for falsification or for support.
In order to prove that you have got effective weed control and
that you are not, if you like, distorting one part of the comparison
against the other, a large number of measurements were made upon
what is known as the crop phenology; that is the growth stage,
the change in growth stage and the time to maturity, et cetera.
This is actually a much more reliable indicator of crop performance
than absolute yield because, do not forget, the varieties were
different; yield measurements of maize are notoriously difficult
anyway because yield measurements are rarely taken in the field
as it is a forage crop, and different management conditions with
beet were specifically imposed in order to minimise the occurrence
of volunteers. The Scientific Steering Committee spend a very
long period of time talking about how we would determine whether
the weed control was conforming with agronomic practice in a way
that did not introduce bias. The phenological measurements
that are described in Champion et al we recognise as being the
best way of doing this, and that was accepted by the peer review
process within the journal.
Q289 Mr Thomas: That is extremely useful
because that is the first time as a Committee I think we have
heard that analysis of how that was done. You will be aware, however,
that the allegation, if you like, has been made that these particular
crops in other circumstances, in North America in particular,
have a different herbicide regime, have more applications during
the growing season than were made in these field scale evaluations.
Does that concern you at all? Are you quite confident in relying
on that?
Professor Pollock: I am confident
that the experimental design that we used was robust and appropriate
for UK conditions. Do not forget that there is very little forage
maize grown in the US; it is nearly all grain maize and the agronomy
is different. I would stand by the published refereed data.
Chairman: We may want to come to the
question of the US slightly later on.
Q290 Mr Savidge: Professor Pollock, were
you and the research consortium provided with enough choice by
the NFU and SCIMAC as far as choosing adequately farmers and sites?
Do you feel that there was sufficient geographic spread and sufficient
numbers to be satisfactory? Were there any particular problems
in relation to particular crops?
Professor Pollock: Across the
three years of the trial, we were fully satisfied. Obviously there
were fluctuations from year to year and we did express concerns
in our minutes as a sort of encouragement to SCIMAC to address
the totality of ranges of intensity across the three years of
the trials, but in general terms they met those challenges.
Q291 Mr Savidge: There have been some
suggestions that you chose less intensively managed sites. Would
you say that was valid and, if so, was there a reason for that?
Professor Pollock: There is a
reason for choosing a range of intensity and for ensuring that
those are proportionate, if you like, because less intensive sites
are obviously likely to carry more biodiversity and therefore,
if there are differences, those would be likely to have more impact
than there would be on more intensively-managed sites. By analogy,
with the benefit of 20-20 hindsight, obviously changes in maize,
which is relatively poor in biodiversity and poor crop, are going
to be less significant than changes for other crops.
Q292 Mr Savidge: Since the trials were
not supposed to be about measurements of yield, why did you require
from farmers information about yields in previous years of conventional
crops from the chosen sites?
Professor Pollock: Because that
gives us a very good proxy measure for intensity.
Q293 David Wright: I would like to continue
questioning on this whole issue of yield but also more broadly
on the trials themselves in terms of their design and operation.
You have mentioned that the trials were drawn up not to take account
of yield. Clearly, when farmers are assessing the practicality
of growing these crops, they are going to want to know whether
they will be commercially viable or not. Indeed, we took some
evidence from Bob Fiddaman of the NFU and SCIMAC who said in evidence:
"As a farmer, if you are looking to a new technology, you
want to see what the potential profitability is".[7]Do
you collectively have any view on the commercial viability of
crops grown in the farm scale evaluations?
Professor Pollock: No. I think
the Scientific Steering Committee set out to produce a reputable,
scientific experiment to test the "null hypothesis".
Dr Avery: I think the whole business
of SCIMAC guidelines is important in that the farmers needed some
guidelines as to how to grow these new crops and the industry
is the best place to get those guidelines, but if those guidelines
were rather kind to the environment or were rather harsh to the
environment compared with what farmers might do if they were allowed
to grow those crops however they wanted, then Defra, if they ever
came to give commercial approval for these crops, could impose
conditions on how they were managed in line with the guidelines
that were imposed on the farm scale evaluations. If farmers chose
or attempted to grow the corps in a different way, then farm scale
evaluations would not tell you so much. It would be within the
remit of the regulatory process to say, "This is how you
have to grow them". It is quite important to get those guidelines
specified so that they could be attached as a condition to any
commercial approval.
Q294 David Wright: Does this not leave
us in a position, though, where we have claim and counter-claim
from elements of the industry, interested parties and different
agencies promoting GM crops that yield is variable and, if only
we had farmed them in a different way, it would be incredibly
productive, or not? Does not this whole approach leave the context
of the trials open to challenge and that will not enable us to
come to any detailed conclusion?
Dr Johnson: No, because the trials
from our point of view were most certainly set up to test a simple,
"null hypothesis". They were never set up to test the
agronomic performance of these crops. Had we been asked to do
that and had the Steering Committee been asked to do that, they
would have designed a very different kind of experiment. The agronomic
performance is really not the issue. That is an issue for national
listing, for example, where the agronomic performance of crops
is compared but not for this experiment.
Q295 David Wright: We are going to get
into a situation, are we not, and maybe we are promoting that
through the inquiry and through some of the questions I have been
asking. I have been asking farmers for their impression and if
they feel that the process of growing GM crops is more productive
and what kind of impression they had of yield. That kind of debate
is going to go on within the farming community outside of the
trials. Is that not the problem that we have?
Dr Avery: In a way, that is a
problem but the results of the trials were so clear as to the
environmental damage that any commercial approval of these crops
would cause that we would hope we would never get to that stage.
That was what these trials were actually designed to test. I have
to say, the worries that English Nature and the RSPB had at the
beginning of them that caused these trials to come into place
were confirmed by a very extensive and thorough piece of scientific
research, which I think should lead to the management regimes
of these crops, the two that have been shown to cause damage,
being banned, but, if they were approved, it is up to farmers.
If there is no environmental damage, then it is up to farmers
to discover whether they can make any money out of these crops
because of their yields or not. I am not sure that it would have
been appropriate for public money to go into these trials to measure
yield in any big way; it is up to the industry to convince farmers
whether or not they have a good product. Five years ago, the biotech
industry was telling us all that not only would these crops lead
to higher yields, but they would be great for wildlife as well.
That last claim has been shown to be palpably false but they made
the claim. Now we need to decide whether we grow these crops or
not, and I just hope we do not.
Q296 Chairman: Would it be unfair to
ask you, Professor Pollock, whether you agree with that or not?
Professor Pollock: Yes, it would
be unfair!
Q297 David Wright: I want to return to
the issue about the quantity and timing of herbicide applications
regarding both conventional and GM crops. Did you have concerns
about the structure and the terms of application during the trials?
Professor Pollock: No, because,
as I say, the monitoring regimes that were being carried out would
have alerted us to any distortion in the comparison between the
two halves of the field.
Q298 David Wright: I want to ask you
a brief question on the Broom's Barn study. How do you regard
the validity in scientific and in commercial farming terms of
the Broom's Barn study?
Dr Johnson: Perhaps I could make
a comment about that. The study was set up to demonstrate that
if you delayed spraying of broad spectrum herbicides of GMHT crops,
you would end up with more biodiversity in the field. The question
that I would ask always of experiments like that is: yes, you
may have more biodiversity at some point in the growing cycle,
but is that going to have a significant impact on farm and biodiversity
across the arable landscape? I think the field scale evaluations
show us quite clearly that that would not have any significant
benefit, simply because what happens is that you do get more weeds,
for example in the field early in the season, but then you have
to hit those weeds before they start to affect the yield and when
you hit the weeds, you take away at a stroke the foundation stone
of biodiversity in that field, including at the margins, and the
weeds will never set seeds because you are hitting them very early.
The field scale evaluations demonstrated quite clearly that if
you use that system, your seed return is very much diminished.
There was another study in Denmark that demonstrated exactly the
same phenomenon. It really looks as though a system that at first
sight looked attractive, and it does look attractive when you
see it in the field, does not actually benefit the kind of biodiversity
that we are concerned about.
Dr Avery: I think the ideas behind
that study are quite interesting. The idea is: can you fiddle
about with the GM system and generate some benefits for wildlife?
I think that is interesting, but I do not think the study is very
compelling in suggesting that that study showed those benefits
to wildlife. Both Brian Johnson and I are members of the Government's
GM Science Review Panel, which published its first report on GM
science in the summer in July, and in that report we say that
the study at Broom's Barn was interesting but it does not tell
us very much and we await the result of farm scale evaluations,
which are much bigger and look at a much wider ranger of important
issues. Now, having got the farm scale evaluations, those are
very clear. I think much more work would need to be done along
the lines of the Broom's Barn study to approach remotely the certainty
that the results of the farm scale evaluations have reached.
Q299 Mr Thomas: Professor Pollock first,
looking again at the evidence you gave to the Committee, you outline
the main discussions in the first year, which were all about the
design of the field scale evaluations, the sample of the fields,
the selection of the fields, the selection of species to be studied,
and so forth. One of the things that is not mentioned there is
if the evaluations should have looked at whether the fields were
used for more than one growing season after the other. My understanding
is that this was one growing season in each case. Did you look
at that as a Scientific Steering Committee?
Professor Pollock: As I say, these
are actually four separate experiments. Rape, both winter and
autumn, and beet are grown in a rotation, so it is very unusual
to have identical follow-up crops and that was never an issue.
Maize is cultivated in a range of different ways and follow-on
is actually quite common, and so a proportion of the fields were
follow-on fields. That has been quite interesting, and of course,
having collected this data, the contractors and others, now that
it is moving into the public domain, will be looking at subsets
of this information to see whether they can peel apart some of
these secondary effects. When we started these trials, we were
very concerned to make sure that we were going to be able to measure
the major effects reliably and that was done by the Statistical
Power Study that is mentioned in my evidence. As it turned out,
the effects are remarkably uniform across time and space. I think
this is probably the most surprising element of the study. That
does mean that a certain amount of secondary delamination of the
data will be possible to some considerable advantage, and this
will be going on now. There are secondary and tertiary analyses
which will throw further light on the system in due course, once
the data has all entered the public domain.
5 Please see an explanation from Defra in Ev 136. Back
6
Scientific Steering Committee. Back
7
Please see Ev 65. Back
| 
