Post-release monitoring and long-term effects of LMOs released into the environment

Dear Colleagues,
In respect to the ongoing discussion on post-release monitoring and long-term effects of LMOs released into the environment.
My opinion is that, it all depends on the kind of LMOs to be released into the environment which will determine how to monitor its release and its long-term effects.
In the process of doing post-release of LMOs monitoring, one question always come to mind as ‘Is LMOs present in the material of interest, and how much of it is present? This is the first step in assessing whether the presence of a given LMOs is correlated with specific effects either on the environment or on health. The ability to monitor LMOs in the environment is therefore an essential capacity required for biosafety assessment.

Hajara Yusuf Sadiq
National Biotechnology Development Agency
Abuja-Nigeria

There could be two categories for discussion


1) LMOs-FFP which have been released a long time ago with a decade of cmmercial experinece

2) LMOs under experimental trial under the confined environment or semi-closed condition such as isolated field or pond.

As for 1) There is more than ten years of experiences in many crop producing countries. Compilation of information as review shall be at first with the elements/modality listed in Annex III of RA.  WIth that there may be elements to be enforced to examine the long term effects.  There shall be a distinction of LMOs presence and the released environment and methods such as cultivation/agriculture system as that there was some confusion of the interpretation of the assessments in the past.

2) This may be yet within of the context of RA rather than monitoring, since there are many experience of trials, there could be some compilation of modalities to summarize the common features and specific features.


regards,

K. Watanabe, Japan

1. The interaction of LMOs with other organisms and environment is complex, and just monitoring the survival of LMOs in the environment is often meaningless. Therefore,when monitoring is considered,

"its need and utility should be considered, on a case-by-case basis, during risk assessment and its practicability should be considered during the risk management. The monitoring may be undertaken for the purpose of:

a) verifying conclusions about the absence or possible occurrence, impact and significance of potential environmental effects; and

b) monitoring changes in environmental parameters (that are deduced from the conclusions of the risk assessment) to determine their environmental impact."

The sentence in the quotation marks is modified (underlined part) from paragraph 20 of Principles for the Risk Analysis of Foods Derived from Modern Biotechnology.  

2. The objectives of monitoring and parameters to be monitored should be first identified through the process of risk assessment. "Time and spatial (e.g., how often and for how long) requirements for monitoring different types of LMOs, traits and receiving environments" should meet the objectives and should be practicable.

The type of baseline data is needed is the natural history of non-modified recipients or parental organisms.

In respect to the ongoing discussion on long-term effects of LMOs released into the environment my opinion is that it depends on the species, the potential lifetime (an-nual or long-living crops and trees), modified traits (herbicide-tolerance, inclusion of an insecticide, altered substances of content) and the intended use of the LMO (im-port only, processing only or cultivation). So this should be assessed by the authori-ties on a case-by-case basis.

Protection targets:
Monitoring should be primarily directed towards environmental protection targets, for instance conservation of the biological diversity in terrestrial and aquatic ecosystems, conservation of especially endangered or protected species, habitats or ecosystems, conservation of soil functions and soil biocenosis or human health. Terrestrial agro-ecosystems, aquatic ecosystems and air are strongly merged on the one hand and on the other hand they are also closely linked with other items such as human health, cultural and real assets, utilisation of the environment or the natural landscape. Con-servation of biodiversity and respective ecosystem functions, I think, are among the most essential protection targets.

Selection of  indicator organisms:
The essential step is to identify appropriate indicator organisms for long-term LMO effects. An important criterion for their selection is the potential to indicate LMO-induced changes. This potential may depend on
a) direct and/or indirect interrelationships with the LMO
b) a preferably widespread distribution of the indicator,
c) sufficiently high abundance
d) importance for ecological processes and ecosystem functions.
Another important aspect is the feasibility, economy and standardisation of detection methods.
To make selection criteria transparent they can be systematically assessed within a matrices (there are examples how this can be done)

Dear colleagues,

I agree with Hiroshi Yoshikura that “the objectives of monitoring and parameters to be monitored should be first identified through the process of risk assessment.”

With this I mean, that the need for and the ability of performing post release monitoring of both the spread of the LMO, its genes and/or traits as well as of its impacts on the environment, biodiversity or human health needs to be integral part of the risk assessment procedure.

The need for the ability to (1) monitor gene escape and the ability to (2) understand, anticipate and guard against long-term negative effects is evident for example in the case of genetically engineered trees.

(1) In order to be able to monitor gene escape, the extent, routes and means of gene flow must be understood. Whilst more, [though maybe not sufficient,] is known for some annual agricultural annual crop plants, the understanding of gene flow from trees is still very limited and does not allow predictions. The question thus arises of how to perform post release monitoring in cases where both time and special requirements are enormous and little detail of when and where to look is available.

In the case of trees it is evident, that propagative plant material will travel and cross national borders. Thus risk assessment and post release monitoring, risk decisions and risk management need to be carried out in a “beyond national boundaries” process, also in the understanding that long-term effects will not be limited to the release site but potentially manifest across borders.

To recall (see also Steinbrecher & Lorch 2008),
most trees and their genes will spread not only through sexual reproduction (pollen and seed) but also by asexual (vegetative) reproduction, such as roots, shoots, twigs that can set root. These propagules can be dispersed by wind, water, pollinators (insects), animals and humans. To assess possible contamination a wide range of factors need to be taken into account, ranging from normal weather conditions in which pollen and seeds already travel long distances (depending on direction, speed and uplift of the wind), to extreme conditions like storms and floodings in which broken branches are swept along and can set root somewhere else. Animals and humans also attribute to the spread of seeds when they either take fruits, nuts, cones along (such as squirrels), or even when they consume fruits, thereby passing the seeds through their body and depositing them somewhere else.
“In any event, as we deploy vast plantations of transgene-bearing forest trees, we can expect the transgenes to escape into the wild population and to persist there for a long time. In conclusion, we can probably take the view that ‘propagules will travel’.” (Smouse et al. 2007)
The issue is not only contamination, but also invasiveness, especially where pioneer species such as GE poplar or birch are modified such that they gain an advantage over wild trees of the same or of other species. An example of a transgenic trait that can confer an advantage is cold tolerance (developed in eucalyptus), allowing trees to be cultivated in colder regions and thereby potentially enabling them to get established in ecosystems where this tree species previously did not grow or maybe where trees in general did not grow. Other examples are trees producing insecticidal protein (e.g. Bt toxins) and therefore possibly (more) resistant to specific pest insects, and trees with faster growth or bigger leaves who can out-compete other tree seedlings competing for light and space in forest settings.

“Transgenes which provide a large fitness advantage, perhaps by protecting from herbivores or disease, may enhance invasiveness.” “Transgenes which enhance fitness are most likely to increase invasiveness and frequency of recipient species outside cropping system.” (James 1998, see also Andow & Zwahlen 2006).

Pollen
Forest trees are largely wind-pollinated, with pollen highly adapted to be transported by wind, often over large distances. Whilst for white spruce (Picea glauca), the vast majority of pollen was found to cross-pollinate within a range of 250-3000m (O’Connell et al. 2007), travel distances of 1000 km have been documented for spruce pollen (Gregory 1973) and 100s of kilometres for birch pollen. For risk assessment purposes, pollen dispersal rates cannot be taken into account for individual years only, but have to be looked at cumulatively over time, e.g. a long distance dispersal (LDD) rate of 1% would amount to 9.6% over the period of a decade  (Smouse et al. 2007).

Seed dispersal
Seed dispersal needs to be taken into consideration when looking at gene flow. For trees we find, that they have developed a multitude of strategies to have their seeds dispersed either by abiotic means, such as wind or water, or by biotic means, mostly animals including humans.
Trees, especially forest trees, produce large quantities of seeds often well adapted to wind dispersal (abiotic seed dispersal). For examples as well as for vegetative propagules dispersal see attached paper Steinbrecher & Lorch (2008).


(2) In order to understand, anticipate and guard against long-term negative effects, it appears crucial to avoid assumption based prognosis. For this purpose it is crucial to have detailed and long term experience with the conventional parental plant of the LMO in question, thus any behavioural changes can be detected instantly and acted upon if necessary. Furthermore detailed and long-term experience is also required for the environmental and biodiversity settings and interactions in which the plant is commercially grown and/or naturally occurs. Whilst this strikes as common sense and might appear an easy task in particular cases, it is far from easy – if not currently impossible - in other cases, such as transgenic trees in the context of global forest biodiversity and global forest ecosystems.

In this context it is also important to remember, that a trait-confined risk assessment is insufficient for transgenic trees, where The ability to respond to biotic and abiotic stresses may be compromised by the performance of the transgene, its product(s) and the processes of genetic engineering. Vice versa, such stresses may also interfere with the performance of the transgene, e.g. induce gene silencing (Broer 1996, Meza 2001) 

Testing for any impacts on tree performance (internally as well as externally) will (or would) require a long time and additionally necessitate exposure to all different stresses across different developmental stages.


To summarise, the ability to carry out reliable post release monitoring, including assessing gene flow, and the ability to investigate and safeguard against long-term negative effects need to be assessed in the initial risk assessment itself. If data are insufficient or results are unsatisfactory, further research is required, especially to have all necessary base line data. However, to complicate matters further, field research must only be undertaken in a way that does not pose a risk to the environment in itself.



Literature cited:

Andow DA & Zwahlen C (2006). Assessing environmental risk of transgenic plants. Ecology Letter 9(2): 196-214.

Broer I (1996). Stress inactivation of foreign genes in transgenic plants. Field Crops Research 45: 19-25

Gregory PH (1973). The microbiology of the Atmosphere. 2nd edition. Leonard Hill, Aylesbury, UK. (In OECD consensus document vol 2, p.208).

James R, DiFazio SP, Brunner AM & Strauss SH (1998). Environmental effects of genetically engineered woody biomass crops. Biomass and Bioenergy 4(4): 403-414.

Meza TJ, Kamfjord D, Hakelien AM, Evans I, Godager LH, Mandal A, Jakobsen KS, and Aalen RB (2001). The frequency of silencing in Arabidopsis thaliana varies highly between progeny of siblings and can be influenced by environmental factors. Transgenic Research 10: 53-67

O'Connell LM, Mosseler A, and Rajora OP (2007). Extensive Long-Distance Pollen Dispersal in a Fragmented Landscape Maintains Genetic Diversity in White Spruce. Journal of Heredity 98(7): 640-645  (doi:10.1093/jhered/esm089)

OECD (2006). Safety assessment of transgenic organisms: Consensus documents on the biology of trees. OECD Consensus Documents Volume 2 (1996-2006), Chapter 4.

Smouse PE, Robledo-Arnuncio JJ & Gonzáles-Martines SC (2007). Implications of natural propagule flow for containment of genetically modified forest trees. Tree Genetics & Genomics 3(2): 141-152.

Literature attached:

Steinbrecher & Lorch: Genetically Engineered Trees and Risk Assessment – An overview of risk assessment and risk management issues. Federation of German Scientists, May 2008. Available at http://www.econexus.info/pdf/GE-Tree_FGS_2008.pdf     or     http://www.ifrik.org/en/gm-trees-risk-assessment

I am Tom Nickson, a scientist who has worked with Monsanto Company for over 27 years and an environmental risk assessment expert for biotech crops with over 16 years experience.  I have studied the overview and contributions to this subject of post-release monitoring and long-term effects with great interest.  It appears to me that a critical concept has not yet been mentioned, which is the role risk assessment plays in directing the monitoring and necessary post-release activities.  As such, this is not so much a specific reply as it is a comment on the entire topic.

Some important ideas have been put forward in the overview statement.  Firstly, a recent study “came to the conclusion that risk assessment focusing on long-term effects must be scientifically valid but also proportionate with respect to cost-effectiveness.”  Implicit in this statement is that knowledge has limits and science, even when it is “valid”, leads to additional questions.  Efforts to answer these questions, as well as the answers themselves will have associated costs.  These costs can be described in terms of regulatory and other resources, as well as lost opportunity to growers and the environment that could benefit from improved practices.   This cost-benefit must be balanced within a country where both the impacts of the costs and benefits are realized.  As such, post-release monitoring has broad cost implications that must be considered, and these should be guided by risk assessment.

A second important point made implicitly in the overview is that in any decision-making there will be some degree of uncertainty or unknown possibilities at the time a decision is made.  It states, “[t]he possibility of long-term persistence and accumulation of LMOs, parts of LMOs and transgene-products in the environment and the potential uncontrolled spread over long distances harbours a major potential for unforeseen environmental impacts, the temporal scale of which currently cannot be estimated.” This statement reflects a one-dimensional view of the uncertainty.  In fact, there would likely be just as much uncertainty around the long-term environmental impact associated with a “no” decision or delayed decision. 

At a high level, it is logical that some may see monitoring as a means of dealing with this uncertainty.  However, this should only be done with careful consideration and reflection on some basic assumptions.  When considering making a request for monitoring, a regulatory authority  must ask first ask the questions of what needs to be monitored, how it should be monitored, how long it should be monitored, where it should be monitored and what would an adverse outcome of the monitoring look like?  Importantly, each of these questions should be informed by a risk assessment, which itself would be guided by policy covering environmental protection.  These critical questions on monitoring may be more appropriately dealt with as questions about potential unforeseen and even unpredictable environmental impacts that are better addressed through basic research.  Failing to guide the risk assessment/risk management process using a policy-driven structure results in confusion, excessive costs and can even result in regulatory action based on wrong information.  Requirements on post-release monitoring should be an outcome of a process where risk assessment informs the various components of the decision,  resulting in any necessary and appropriate actions.

In summary, we must take care to not confuse the regulatory requirement of monitoring with an exercise in conducting basic environmental research. Regulators must avoid the temptation to pursue what might be nice to know with what they need to know in order to make a defensible and cost effective decision regarding regulatory approval, and use the risk assessment to guide them in decisions on monitoring.   In the context of the Risk Assessment and Risk Management under the protocol, monitoring should be seen as a regulatory activity guided by the risk assessment that is guided by policy-based environmental protection goals, and science is the tool to collect information.

I realize that the dealdine for postings is very short so this is a brief message.

With respect to GM trees it seems that the impression given is that ALL trees have long pollen or seed dispersal, can be propagated from vegetative materials and other attributes.  This is simply not the case.  Each species must be considered based on it's specific characteristics on a case-by-case basis.