The International Symposium on Biosafety of Genetically Modified Organisms (ISBGMO)
is a biennial international meeting organised by the International Society for Biosafety
Research (ISBR; www.isbr.info/). ISBR promotes research and application of science
in the fields of agricultural and environmental biotechnology and risk analysis. In
particular, ISBR encourages research that supports the safe and effective use of biotechnology
in agriculture, food production, and public health, and assists the development of
the relevant policy and regulation.
The first ISBGMO was held in Kiawah Island, North Carolina, USA in November 1990.
Since then, many countries have hosted the meeting: Germany (twice), USA (California),
Japan, Canada, China, France, South Korea, New Zealand and Argentina. The 12th ISBGMO
was held in St Louis, Missouri, USA from the 16th to the 20th of September 2012, and
was attended by about 500 delegates from 47 countries.
This special section of Transgenic Research features thirteen papers developed from
lectures and workshops at the meeting. Taken together, the papers indicate that, for
first generation genetically modified (GM) crops at least, biosafety research is increasingly
addressing questions about the sustainable deployment of the crops in agricultural
systems, and is focussing less on the basic characteristics of GM crops as a class.
Making regulatory risk assessment efficient and effective, and realising the economic,
environmental and social opportunities presented by commercialised GM crops, now seems
more important than, say, further basic research on unintended effects of transformation
or gene flow from GM crops to wild species. This is a hugely significant development.
While ISBR and ISBGMOs focus on science, the role of policy in framing research questions
and ensuring effective application of new knowledge is increasingly recognised. In
his keynote address, Raven (2013) placed GM crops in the context of fulfilling agricultural
policy. If our policy is to increase food production to meet increasing need, then
GM crops ought not to be singled out for “burdensome” regulatory treatment: “it is
no longer acceptable to delay the use of any strategy that is safe and will help us
achieve the ability to feed the world’s people.”
A source of onerous regulatory treatment was concern that that GM crops might “escape”.
At the time of the first ISBGMO in 1990, scientists were thinking about the ecological
consequences of transgenes not being contained by agricultural management. One concern
was that genetically modification of crops would create more serious weeds of agriculture
or plants more invasive of non-agricultural habitats. These problems may occur if
the introduced trait provides tolerance or resistance to a factor that was controlling
the persistence or spread of the crop (Keeler 1989)—a process called ecological release
(Schmitt and Linder 1994). A more speculative mechanism for ecological release was
that transformation would lead to unintended effects, perhaps through disruption of
native genes, pleiotropy or epistasis (Regal 1994).
Discussions about ecological release of GM crops and unintended changes caused by
transformation focussed on technical questions, such as the mechanisms by which unintended
effects may arise, and the potential changes in the population dynamics of feral GM
crops or hybrids. Often such discussions missed a crucial element of risk assessment:
what changes ought to be regarded as harmful? Without agreed definitions of harm from
using a GM crop, biosafety research often confuses rather than clarifies risk assessments.
Without harm as a context for evaluating the significance of changes, regulatory decision-making
becomes “burdensome” as it tries to catalogue all possible changes that might arise
from using a GM crop, rather than the likelihood and seriousness of harmful effects.
This question of how to define harm was central to several papers presented at ISBGMO12.
Devos et al. (2013) point out that disputes about the risks—and opportunities—from
using GM crops may result from differences in values. In such circumstances, clarification
of policy objectives is vital. Without such clarity, additional scientific research
may not help people to reach agreement, and indeed may make disagreements worse. Gray
(2013) reinforces this point with examples from applied ecology. He notes that misunderstandings
between scientists and policymakers are not unique to GM crops, although the debate
about the use of GM crops may be uniquely intense.
The benefits of clarifying policy objectives are demonstrated in the papers by Andrade
et al. (2014) and Nang’ayo et al. (2014), writing about Brazil and Africa respectively.
In Brazil, there is clarity: post-market monitoring concentrates on detecting harmful
effects from cultivating a GM crop, not on comparing the agro-ecosystems where the
GM crop is cultivated with a baseline of conventional (i.e., non-GM) crop cultivation.
The advantages of the system are flexibility, cost-effectiveness and proportionality.
In Africa, on the other hand, there is ambiguity. GM crops provide opportunities to
fulfil policy objectives of improving the lives of smallholders through higher yields
and better quality of crops, while simultaneously reducing unsuitable use of pesticides
and fertilisers. Nevertheless, many countries in Africa also have policies that result
in regulations that constrain or prevent research and development of GM crops tailored
to local conditions and needs.
A weight of evidence from field studies of GM crops in uncultivated land (e.g., Crawley
et al. 1993; 2001), phenotypic comparisons of GM crops with non-GM near isolines (e.g.,
Horak et al. 2007), and molecular analysis of genetic changes induced by transformation
and other breeding methods (e.g., Ricroch et al. 2011) suggests that GM crops are
no more likely to be serious weeds than are other new crop varieties. Nevertheless,
new data are often required to complete regulatory assessments of the weediness and
invasiveness of GM crops (e.g., Raybould et al. 2012). Three papers from ISBGMO12
may help to reduce or eliminate these requirements.
First, Garcia-Alonso and Raybould (2013) describe how general concerns about the effects
of weedy or invasive GM crops may be translated into operational definitions of the
environmental entities that are to be protected. Secondly, Keese et al. (2013) describe
risk assessments for importing plants new to Australia, and how these established
methods may be applied to the risks posed by cultivating GM crops. Finally, Roberts
et al. (2013) point out that assessment of the risks of weedy or invasive crops too
often focuses on exhaustive characterization of potential hazards, when an estimate
of exposure—the amount of gene flow from the GM crop through pollen or seed—would
be sufficient to conclude with high confidence that the risks are minimal. By explicitly
stating objects of concern, and putting the risks from GM crops in context with those
from other plants, these papers will help risk assessors to judge whether data requirements
for weediness assessments are proportionate.
In addition to gene flow, weediness and invasiveness, a recurrent theme of ISBGMOs
has been the potential effects of Bt crops on non-target organisms (NTOs). Between
2006 and 2012 sound theoretical and practical frameworks for risk assessment and testing
methods covering Bt crops and NTOs were developed and published (Garcia-Alonso et
al. 2006; Romeis et al. 2008, 2011, 2012). As Burns and Raybould (2013) show, these
methods are becoming routine for regulatory risk assessments for GM insect-resistance
traits in the USA.
NTO risk assessments often make conservative assumptions about which organisms will
be exposed to Bt proteins. Hence in many cases, studies of the hazard of proteins
to groups of NTOs may be unnecessary because those groups are unlikely to be exposed.
Romeis et al. (2014) present a database that contains bio-ecological information on
arthropods found in relevant agro-ecosystems in Europe. This information could help
focus ecological risk assessments by identifying NTOs that are ecologically important
and that are likely to be exposed. Better knowledge of the ecology the non-target
fauna in crops may reduce the amount of hazard testing needed for ERAs of insect-resistant
GM crops.
We often hear that risk assessment for GM crops must be case-by-case. With a huge
variety of potential traits, environments and policies, a risk assessment for a particular
GM crop in a particular country cannot cover all uses of all GM crops. Nevertheless,
case-by-case does not mean that each risk assessment must start from scratch; each
risk assessment may contain data and analysis that are useful for subsequent risk
assessments. Two papers presented at the conference describe different aspects of
this topic.
Kearns et al. (2013) describe the work of the Organisation for Economic Co-operation
and Development (OECD) in producing consensus documents for risk assessment of GM
crops. These documents compile information on the biology of crops and traits that
are agreed to be relevant to comparative risk assessment. The OECD documents mean
that regulatory authorities need not always start from scratch when faced with a crop
or trait new to their country.
Garcia-Alonso et al. (2014) tackle the question of whether field trial data must be
produced in the country to which the risk assessment applies. They propose a conceptual
framework, based on agro-climatic zones, for determining whether data produced in
one country are relevant to other countries. If accepted, the framework will allow
researchers to design trials that produce data for use in many countries. The framework
would reduce the time and expense of producing regulatory data without compromising
the rigour of risk assessments produced from those data.
Finally, technology development and application is continuous. While biosafety research
on herbicide-tolerant and insect-resistant GM crops may have moved from fundamental
questions about their properties to applied questions about effective risk assessment
and regulation, we may need to return to basic research for crops developed from new
technology. Furthermore, as Hokanson et al. (2013) show, not all future products of
agricultural biotechnology will be crops. Nevertheless, experience from first-generation
GM crops teaches us that regulation ought to be designed to deliver clear policy objectives
about real products. This experience will also help us to identify data that are essential
for assessing the risks from future products.