GENET archive


7-Misc: Precise Precaution Versus Sloppy Science

mailinglist genet-news

                          Hartmut Meyer
         German NGO Forum on Environment and Development

Bulletin of Science, Technology, and Society 19(2), April 1999,
p. 91-95
Copyright © 1999 Sage Publications, Inc.


THE Convention on Biological Diversity opens the possibility to
negotiate a legally binding Biosafety Protocol to assess and
minimize risks in the field of transboundary transfer, handling,
and use of organisms modified by genetic engineering. Two
principles - the Precautionary Principle and the Principle of
Familiarity - guiding the risk assessment as basis of import
decisions on such organisms are discussed. Developing and
European industrialized countries favor the Precautionary
Principle. The US, Australia, Japan, and some others call for the
Principle of Familiarity. These two principles exihibit opposite
effects on scientific progress in general and on scientific
methodology of risk assessment in particular. With the example of
risk assessment by the U.S. company Monsanto discussed below, it
could be illustrated that the Principle of Familiarity opens the
way for superficial evaluations based on citing arbitrary
references while the Precautionary Principle is an incentive for
developing and applying sound methodology in experimental risk


SINCE May 1997, negotiations on a legally binding Biosafety
Protocol to assess and minimize risks in the field of
transboundary transfer, handling and use of living modified
organisms (LMOs) are taking place in Montreal, Canada. LMOs are
organisms that are modified by means of genetic engineering. They
contain new combinations of genetic material which could not have
been gained by conventional breeding and that have been
transfered between sexually incompatible organisms, thus
overcoming natural barriers of reproduction. The Montreal
negotiations focus on the transboundary transfer of such LMOs,
for example, on the international exchange of and trade with
transgenic seeds, grains, animals or bacteria. The negotiations
are based on Articles 19.3 and 8(g) of the Convention on
Biological Diversity, elaborated during the Earth Summit in Rio
de Janeiro, Brasil, in 1992 and on Decision II/5 of the Second
Conference of the Parties to this Convention in Jakarta,
Indonesia, in 1995 (see the appendix; for more documents, see Until the Fourth Conference of the Parties in
1998, 172 states ratified this Convention. This Conference
decided to finish the Biosafety Protocol in February 1999. In
1998, transgenic crops were allowed to be planted in eight
countries of the world. As the country with the largest acreage
of transgenic crops (74% of total LMO-acreage) and as the main
trader of LMOs, the United States cannot become member of the
Protocol until they it has ratified the Convention on Biological


One of the main demands of civil society organizations during the
Montreal negotiations calls for the Precautionary Principle as
the basis for scientific risk assessment and political decision
making under the Biosafety Protocol. Civil society organizations
witnessed that some delegations try to establish contradiction or
even incompatibility between the Precautionary Principle and
sound science. It is the view of civil society organisations that
at their best, sound science as a methodological concept and the
Precautionary Principle as guideline for decision-making are
mutually supportive and can operate in perfect harmony.
What does this mean within the context of the Biosafety Protocol?
The Precautionary Principle enhances the search for scientific
knowledge by initiating a scientific risk assessment and thus
collecting data and providing scientific evaluations on
ecological, health and socio-economic risks of LMOs on biological
diversity and human health. Sound science serves precaution
because it contributes to an early warning system supporting
political efforts to minimize already obvious harm and to prevent
the emergence of new negative impacts.


However, not all delegates share this concept of mutual
supportiveness between science and precaution. The legal frame
work of the US, which is the leading exporter of LMOs and
products therof, stresses the Principle of Familiarity as basis
for decision-making in the context of modern biotechnology (FDA
1992). What task does this assign to science? Science has to
provide risk assessment on the intended novel traits of a given
LMO. However, as far as the unintended changes are concerned,
science has to provide arguments for avoiding further risk
assessment and has to support the judgement that the differences
in the composition of genetically engineered and non-engineered
organisms are insignificant. Thus, science is turned into an
instrument for stopping scientific curiosity and hindering the
quest for new scientific knowledge. The promotion of the
Precautionary Principle seems to civil society organisations the
appropriate and necessary safeguard to counteract the risks of
such science which is based on and reduced to serving the
interests of exporting companies. Precaution is the only chance
for all countries and populations to prevent harm. Moreover, it
is the only chance for poor for mitigating adverse effects of
modern biotechnology because they will not be able to finance
remedial actions.

The non-adherance to sound standards of science and the strange
lack of inquisitiveness in the type of science that is not
exposed to the stimulus of the Precautionary Principle will be
demonstrated below in the attached evaluation of Monsanto's
attempt to prove the familiarity of consumers and environment
with Roundup Ready® Cotton.


Roundup Ready® Cotton was engineered by means of gene technology
to resist Monsanto¹s herbicide Roundup Ultra®. In addition to
the main product - that is, cotton fiber - residues from
processed cotton bolls are used as animal feed and for human
nutrition, notably, cotton protein and cotton oil. To register
herbicide resistant Roundup Ready® Cotton in the United States,
Monsanto had to prove its substantial equivalence to conventional
cotton. In 1996, Nida et al. published an article meant to
provide the supporting scientific evidence. One of the features
that were analysed by the authors was the content of gossypol, a
toxic terpenoid constitutent of various parts of the cotton

Nida et al. (1996) showed by their experimental results that
there is a statistically significant difference, and not an
equivalence, in the gossypol content between the seeds of two
tested LMOs (n=6, Ms=1.32 % and 1.01%) and the parental,
nonmodified line (n=6, M=1.19%). How did the authors then proceed
to prove that these significant differences are within the range
of familiarity? Did they continue the experimental approach or
did they just make use of available publications? Admittedly, it
is cheaper and faster to rely on preexisting publications.
However, choosing an experimental approach is the very way in
which science increases the range and depth of its knowledge.
Moreover, this is especially appropriate for research on
unintended effects. Nida et al. quoted, "previously reported
[gossypol] levels for cottonseed grown under various field
conditions, 0.39-1.70% (Berardi & Goldblatt, 1980; Abou-Donia,
1976)"(p. 1871).

The comparision of sets of data generated under different
conditions to prove the equivalence of the analysed biological
entities is one of the most demanding scientific challenges. The
experimental approach would imply a series of comparative
analyses, covering LMOs, with their parts and products, and
covering corresponding unmodified organisms, with their parts and
products. This should be preferably performed within the same
laboratory applying the same methods. The second approach
compares data taken from publications that were not designed to
be stricly comparable. The nonexperimental approach which
Monsanto has chosen (see Figure 1) has to adhere to the same
stringent standards of sound science as the experimental approach
(Nida et al., 1996). Does the article by Nida et al. (1996) do
so? One could think so because their choice of references passed
the peer review of the publishing scientific journal. Monsanto
made use of it as proof of substantial equivalence. The Principle
of Familiarity then freed Monsanto from conducting a risk
assessment. Are the data comparable and is their interpretation

Figure 1. 
References Cited and Visualization of Their Interrelationship
(not included in email version)



Abou-Donia (1976) wrote one sentence on the gossypol content in
cottonseed: "Cottonseed usually contains 0.4 to 1.7% gossypol.".
These numbers were cited correctly by Nida et al. (1996).
However, Abou-Donia (1976) neither stated the source of these
data, nor the cotton vareties, nor the country and year of the
field trials that provided these data. Therefore, this reference
is a very questionable proof of the gossypol content of

Berardi and Goldblatt (1980) give more details:
"Ranges of 0.68-2.36% gossypol in kernels of various cottonseed
varieties grown in India and of 0.33-2.40% gossypol in kernels of
seed grown in the USSR were reported (Markman & Rzhekhin, 1969;
Murti & Achaya, 1975). Carter et al. (1966) reported that
gossypol contents of seed from 11 (cultivated and noncultivated)
species of the genus Gossypium varied from 0% for the glandless
varieties of G. hirsutum L. to more than 9% for G. klotzschianum
var. davidsoni (Kellogg)." (pp. 192-193)

The reference of Nida et al. (1996) to this set of data lacks two
important features: (a) The numbers are not correctly quoted, and
(b) the data are not comparable, as the different authors are not
using the same parts of the cotton boll for their chemical

The most central component of the cotton seed is the kernel which
has the gossypol-producing glands. As such, the seed is composed
of the kernel and gossypol-free coating layers (hull and, in
North American species, the fibrous linter layer). Whilst Nida et
al. (1996) used delinted seed, Berardi and Goldblatt (1980)
refered to kernels. The concentration of gossypol in kernels is
approximately twice as high as in whole seed.


Carter et al. (1966) analysed the gossypol content in kernels of
different species of the genus Gossypium, including wild species.
This dated publication provides genuine experimental data. A
comparision of a distinct transgenic cotton line with any other
cotton species might be useful for a general understanding of
different levels of gossypol production; yet, to use such a
comparision to establish substantial equivalence through
familiarity is seriously doubtful.

Markman and Rzhekhin (1969) reviewed Soviet literature on the
gossypol content of various parts in different cotton varieties
and species. Berardi and Goldblatt (1980) chose one table on
kernel data published in that review, originally compiled by
Ismailov in 1959.

Murti and Achaya (1975) reviewed Indian publications on the
composition of various parts in different cotton varieties and
species. Data on gossypol contents in kernels are taken from
Carter et al. (1966), Narayana Rao and Krishnamurthy (1953),
Raghavendar Rao (personal communication), and Anonymous (no
date). From this review, Berardi and Goldblatt (1980) have chosen
data, originally compiled by Narayana Rao and Krishnamurthy
(1953) and Raghavendar Rao (personal communication). The
publication of Narayana Rao and Krishnamurthy (1953) is not
publically available to the international scientific community;
personal communications are in general not appropriate in a sound
risk assessment. In addition to the above citations, Murti and
Achaya (1975) also present data on the gossypol content of whole
seeds of four cotton species by citing the results of Murti and
Appu Rao (1963) and Smirnova (1936), respectively. These data are
the only data which may lend themselves to comparison with the
LMO data of Nida et al. (1996), yet they were not used.


Ismailov (1959) compiled data on the gossypol content of cotton
kernels from varieties within six cotton species. This
publication is not publically available to the international
scientific community. Again as argued above, kernels and whole
seeds are not comparable entities.

Murti and Appu Rao (1963) present data (0.27%-0.82%, as cited in
Murti & Achaya, 1975) that contradict the assumption of
substantial equivalence of Monsanto¹s Roundup Ready®-Cotton. The
whole seed they used contained approximately half the amount of
the toxin gossypol than did the seeds of Monsanto¹s cotton
plants. Again, this publication is not publically available to
the international scientific community.

Smirnova (1936) uses whole seeds (0.15%-1.59%, as cited in Murti
& Achaya, 1975) and has some aspirations to be comparable with
Nida et al.'s (1996) results. Again, this publication is not
publically available to the international scientific community.
This data was produced 60 years before Nida et al., and methods
of analysis have changed. Strict comparability could be better
served by more recent research data.


All documents negotiated in Rio de Janeiro should introduce the
Precautionary Principle as formulated in § 15 of the Rio
Declaration into international legislation. In the field of
assessing risks of organisms and their products altered by
genetic engineering, existing European legislation adheres to the
Precautionary Principle while the national legislation in the
United States is based on the Principle of Familiarity. In the
Montreal negotiations, the European countries together with the
G77, representing the developing countries, call for the
inclusion of the Precautionary Principle in the Biosafety
Protocol. Apart from the advantages that this principle gives
parties to ensure effective health and environment protection,
the implementation of the Precautionary Principle in the
Biosafety Protocol will have direct effects on the quality of
science and progress of knowlegde. While the application of the
Principle of Familiarity causes a lack of incentives for new
scientific investigations, the application of the Precautionary
Principle will stimulate scientific accuracy and progress in risk
assessment. The above case study on the gossypol content of
Roundup Ready® Cotton is one of several possible examples to
illustrates the inediquacy or the Principle of Familiarity as
basis for sound risk assessment. The final round of the Biosafety
Protocol negotiations in Cartagena, Columbia, in February 1999 is
the last possibility for the governments to create a Protocol
with innovative and demanding legal and scientific instruments to
ensure optimal protection of human health and biological



1992 - Convention on Biological Diversity
§ 19.3
The Parties shall consider the need for and the modalities of a
protocol setting out appropriate procedures, including, in
particular, advance informed agreement, in the field of the safe
transfer, handling and use of any living modified organism
resulting from modern biotechnology that may have adverse effect
on the conservation and sustainable use of biological diversity.

§ 8 (g)
Each Contracting Party shall, as far as possible and as
. . .
(g) Establish or maintain means to regulate, manage or control
risks associated with the use and release of living modified
organisms resulting from biotechnology which are likely to have
adverse environmental impacts that could affect the conservation
and sustainable use of biological diversity, taking also into
account the risk to human health;
. . .

1995 - Jakarta-Mandate Decision II/5:

The Conference of the Parties,
. . .
Affirming that international action on biosafety should offer an
efficient and effective framework for the development of
international cooperation aimed at ensuring safety in
biotechnology through effective risk assessment and risk
management for the transfer, handling, and use of any LMO
resulting from modern biotechnology that may have environmental
impacts that could affect the conservation and sustainable use of
biological diversity, taking also into account the risks to human
health, and taking also into account Articles 8(g) and 19,
paragraph 4, of the Convention,
. . .
1. Decides to seek solution to the above-mentioned concerns
through a negotiation process to develop, in the field of the
safe transfer, handling and use of living modified organisms, a
protocol on biosafety, specifically focusing on transboundary
movement, of any living modified organism resulting from modern
biotechnology that may have adverse effect on the conservation
and sustainable use of bioogical diversity, setting out for
consideration, in particular, appropriate procedure for advance
informed agreement;
. . .


I would like to thank the Günter-Altner-Foundation and the
Hatzfeldt-Foundation for supporting the work for this
contribution. My participation in the Working Group on Biosafety
was made possible by financial support by the BUND (Friends of
the Earth Germany), the Forum Environment and Development, and
the Günter-Altner-Foundation.

Abou-Donia, M. B. (1976). Physiological effects and metabolism of
  gossypol. In F. A. Gunther & J. D. Gunther (Eds), Residue
  Reviews - Residues of Pesticides and other Contaminants in the
  Totel Environment (Vol. 61, pp. 126-158). New York: Springer

Anonymus. (no date). Characteristics of important commercial
  varieties of Indian cottonseed. Technology Series 10.
  Hyderabad, IN: Indian Central Oilseed Committee.

Berardi, L. C., & L. A. Goldblatt. (1980). Gossypol. In I. E.
  Liener (Ed), Toxic Constituents of Plant Foodstuffs (pp. 183
  237). New York: Academic Press.

Carter, F. L., A. E. Castillo, V. L. Frampton, & T. Kerr. (1966).
  Phytochemistry, 5, 1103-1112.

Food and Drug Administration (FDA). (1992). Statement of policy:
  Foods derived from new plant varieties (Docket No. 92N-0138).
  Federal Register, 57, 22984-23005.

Ismailov, A. (1959). Chemical Investigation of Gossypol, the
  Specific Pigment of Cottonseed, in Russian. Tashkent,
  Uzbekistan: Author¹s Summary.

Markman, A. L., & V. P. Rzhekhin. (1969). Gossypol in cotton
  plant. Content and localization. In A. L. Markman & V. P.
  Rzhekhin (Eds), Gossypol and its derivates (pp. 1-9).
  Jerusalem: Israel Program for Scientific Translation.

Murti, K. S., & K. T. Achaya. (1975). Cottonseed Chemistry and
  Technology in its Setting in India. New Delhi: CSIR.

Murti, K. S., & B. Appu Rao. (1963). Studies on the compositional
  and varietal characteristics of Indian cottonseed during the
  decenium 1951-52 to 1960-61. Hyderabad, IN: Indian Central
  Oilseed Committee, 1963.

Narayana Rao, & M.; K. Krishnamurthy. (1953). Bull. cent. Fd.
  tech. Res. Inst., 3,: 103.

Nida, D. L., S. Patzer, P. Harvey, R. Stipanovic, R. Wood, & R.
  L. Fuchs. (1996). Glyphosate-tolerant cotton: The composition
  of the cottonseed is equivalent to that of conventional
  cottonseed. Journal of Agricultural and Food Chemistry, 44,

Smirnowa, M. I. (1936). Bulleten Prikladnoi Botaniki, Series III,
  15, 227.

Hartmut Meyer
Working Group on Biological Diversity
German NGO Forum on Environment and Development
Reinhäuser Landstr. 51
D-37083 Göttingen, Germany

phone:  +49-551-7700027
fax:    +49-551-7701672


-| Hartmut Meyer
-| Co-ordinator
-| The European NGO Network on Genetic Engineering
-| Reinhaeuser Landstr. 51
-| D - 37083 Goettingen
-| Germany
-| phone: #49-551-7700027
-| fax  : #49-551-7701672
-| email:

Genet News