Dear participants to the Open-ended Online Forum and AHTEG,

I am pleased to inform that the online discussion on the second draft of the guidance on "Risk Assessment of Living Modified Trees" is now open at http://bch.cbd.int/onlineconferences/discussiongroups_ra.shtml#topic2 .

Your feedback on ways to improve this draft guidance document is very welcome and may include your views on, for instance:

a) Whether all issues relevant to this topic have been included in this draft and, if not, which ones should be added;
b) Suggestions for improvements to the current text and/or structure;
c) Suggestions for background documents to be included (please provide the full reference and indicate to which section(s) of the guidance the background material is relevant).

As per the AHTEG Action Plan, this round of discussion should focus on concrete text proposals. It is appreciated if your suggestions could be made directly to the text of the draft guidance (attached) in "track changes" mode.

This discussion will close on 19 November 2011 at 1:00am GMT.

Please do not hesitate to contact me (manoela.miranda@cbd.int) if you have any question or encounter problems when posting comments.

Best regards,
Manoela

Dear all,
Moments ago I posted the message below on behalf of Dr. Ishii without realizing that he is not yet registered in the open-ended online forum. Please note that his message has been deleted and I am awaiting his registration in the online forum in order to re-post it.
My apologies for any inconvenient.
Best regards,
Manoela



The following message has been posted by Manoela Miranda, UNEP/SCBD/Biosafety on 2011-11-07 16:50.
RE: Opening of the discussion on the 2nd draft of the guidance on "RA of Living Modified Trees" [#2804]

POSTED ON BEHALF OF KATSUAKI ISHII

----

Dear Dr. Manoela

I would like to propose to add “other plants (allelopathy)” in P. 7, because I suppose checking effects of LM Trees to other plants is not negligible.

Sorry, I cannot create a new thread because of a technical problem.

Best regards,

Katsuaki Ishii (FFPRI)

Dear Beatrix:

Thank you for the revised draft of the Guidance on Risk Assessment for LM trees.  I note that there are quite a few changes to the text, including several of the text edits that I had suggested for the first draft, and I very much appreciate that these were incorporated.

There are a number of suggestions however that were made in comments by myself as well as others that do not yet appear to have been incorporated or addressed.  As an example, in the section titled “Uniqueness of trees” (now headed “Characteristics of trees”), one of the concerns I had commented on in the first draft was that this section suggested that the described characteristics were unique to trees and that all trees shared these same characteristics, whereas in fact, this is not the case.  I still have these same concerns about the current text.  There were also some characteristics that I felt were not mentioned or not discussed in sufficient detail to be helpful as guidance.   These included how experimental approaches with trees are set up to accommodate their physical size, and the existence of well established varieties where any form of breeding or backcrossing would be precluded.  

There still appears to be quite a lot of overlap with the Roadmap, particularly in several of the “Points to consider“ sections where the points listed are not unique or specific to trees and should instead be dealt with under the Roadmap itself.

I also provided (as a comment in the text) several web-links to published risk assessments on LM trees that could be used as examples for other risk assessors and strongly recommend that these be added.


As Manoela has requested that concrete edits to the text be submitted I will try to work on text edits that I hope will help to address more of the comments from the first on-line discussion. 

Best regards,
Les

Dear participants,

I appreciate the efforts of contributors (and thank them for this) in drafting this revised version of the document, which contains many improvements as compared with the first version.
Nevertheless, I still share the doubts expressed previously by some people as whether this document will be of practical use for people with limited experience in risk assessment of LM trees. In particular, it is not always clear why and how this document complements the general Guidance on Risk Assessment of LMOs. This relates to a major problem in the process, i.e. that this specific guidance should have been developed only AFTER the general Guidance on Risk Assessment of LMOs is tested and finalized. For instance, it would be very interesting to test the general Guidance as a tool in conducting and reviewing risk assessment of applications involving LM trees in order to determine what are the specific points that should be further considered in a specific guidance. That would make the whole process much more efficient and straightforward.
Keeping this in mind, I will nevertheless take this opportunity to provide a few comments/suggestions on this new version.

Introduction - lines 21 to 40
As mentioned during the evaluation of the first version, I am of the opinion that these three paragraphs are not relevant in this guidance. Firstly, they focus on socio-economic considerations which fall outside the scope of the guidance. Secondly, such considerations were not included in the other risk assessment guidance documents developed so far and I think we should keep consistency between all guidance documents.

Unintentional transboundary movements
Issues related to unintentional transboundary movements, as currently mentioned in the document (e.g. dispersion though pollen or seed), are not specific to LM trees. If such issues are considered relevant to the risk assessment, they should be dealt with in the general Guidance.
Moreover, I do not see how the international trade of fruits from LM trees falls under the definition of unintentional transboundary movements.

Comparative approach
- Line 118: High potential for dispersal and establishment is not necessarily an inherent or specific characteristic of trees as compared to other plant species.
- Appropriate design of field trials to reveal differences that could appear on the long term in a LM tree as compared to the non-modified tree could sometimes be difficult to achieve. It is therefore important in this context to make a link with risk management strategies.
- Broadly speaking, it should emphasize that, as for any other plant species, the comparative approach will be challenging mainly in cases where little or no information is available concerning the behaviour/impact of the corresponding non-modified recipient organism in the likely potential receiving environment (adding a few concrete examples of such cases in this Guidance would be very useful). In such situations the use of closely related lines may provide a good alternative for the comparative risk assessment. In that respect, I suggest to add under points to consider "Whether one or more suitable comparators are available and the possibility of their use in the appropriate experimental design".

Transformation and propagation methods
- Line 140: Item (b) under points to consider is not per se specific to LM trees. However, it might be useful to give in the document a few words about grafted trees. In cases when the scion is not GM, fruits and seeds will most likely be non GM too, which is an important point to consider in the risk assessment.
- Line 141: Item (c) under points to consider is not specific to LM trees. I suggest to delete it.

Genotypic/phenotypic changes
Lines 147-169: The current wording gives the impression that any genotypic or phenotypic changes occurring over time should be considered. Genotypic and phenotypic changes in the LMO, either intended or unintended, should always be regarded in comparison with the non-modified recipient grown under the same conditions, and only those changes that occur specifically in the LMO and that could cause adverse effects should be further considered.
In addition, different information may be required or available in the case of releases for field trials and in the case of commercial releases. In the former case, it is likely that less information may be available but the objective will be to generate information for further risk assessments. Resulting uncertainty in such cases may be greater but could be addressed by risk management measures. In the later case, more information should be available leading to reduced uncertainty.

Dispersal mechanism
Once again, from the current wording of this section (lines 170-195), it is not clear what makes the points to consider specific of LM trees and why they are important to consider in the risk assessment. I suggest to rephrase the points to consider by focusing on the possible effect(s) of the genetic modification. Also, the "rationale" remains very general and of little practical use as guidance for non specialists. I would suggest to provide some concrete (eventually theoretical) examples of how changes in dispersal mechanisms could lead to an increased risk of the LM tree as compared to its non-GM counterpart.

Receiving environment
Line 204: "the potential for dispersal of propagative material..." has been considered in the previous section. I suggest to rephrase the sentence as follows "the potential for establishment of propagative material...".

Risk management
I am of the opinion that this section should be further elaborated given that LM tree cultivation could trigger the implementation of specific risk management strategies. Here are a few suggestions:
- clear distinction between pre-authorization strategies (such as bioconfinement) and post-authorization strategies (monitoring, detection...).
- clear distinction between experimental releases, where higher uncertainties could result from the risk assessment, and commercial releases.
- In the case of bioconfinement strategies, not only the risks should be considered but also the efficacy and effectiveness of these strategies.
A very important point that would probably merit some development in this guidance is the need to established clear criteria and thresholds for the acceptable/unacceptable levels of risk, as well as protection goals. From the previous round of discussion on this document, this seems particularly critical for forest trees.

Literature
I would like to draw the attention on the document "Report of the OECD Workshop on the Environmental Considerations of Genetically Modified Trees". This document (available at
http://www.oecd.org/LongAbstract/0,3425,en_33873108_33873681_1811440_119681_1_1_1,00.html) is a compilation of the presentations and plenary session reports made at a Workshop which was held in Norway in September 1999. It might contain relevant information for the current discussion.

Kind regards,

Didier Breyer

POSTED ON BEHALF OF HIROSHI YOSHIKURA

----

Dear Manoela,
I have to ask you again to post this comment of mine to the CBD electronic discussion group. I am always very grateful to you.
This comment is on the guidelines on RA of LM trees, particularly to Dr. Didier Breyer’s comment posted on 2011-11-16.
1. I strongly concur with the his comment, which says “Broadly speaking, it should emphasized that, as for any other plant species, the comparative approach will be challenging mainly in cases where little or no information is available concerning the behaviour/impact of the corresponding non-modified recipient organism in the likely potential receiving environment. In such situations the use of closely related lines may provide a good alternative for the comparative risk assessment. In that respect, I suggest to add under points to consider "Whether one or more suitable comparators are available and the possibility of their use in the appropriate experimental design"”.
2. His statement is entirely in line with the codex’s approach that uses “conventional counterpart(s)” for the comparative assessment and practically limits the scope of guidelines to LMOs that have “conventional counterpart(s)”. It is implicit here that if LMOs without conventional counterpart(s) be used, we should accumulate sufficient knowledge of and be sufficiently familiar with the non-modified conventional counterpart(s).
3. Not only for LM trees but for other LMOs, so as not to impose unnecessary burden of environmental investigation of unmodified parental organisms with long history of use in the past thousand and thousand years, from which LMOs are derived, and so as to use the experience and common sense, I propose to introduce the idea of “conventional counterpart(s)” again. This approach can allow risk assessors to focus their assessment more on the modification and its consequence relevant to the environmental issues. The definition of conventional counterpart in codex is “'related organism/variety', its components and/or products 'for which there is experience of establishing safety based on common use as' food (Read only the parts between ' ').
4. We should note that most knowledge has been obtained not in the way as is laid out in the risk assessment guideline but just through long use and through many trials and errors. We may even not know entirely in detail why things that work in the environment so works.

Sincerely,
Hiroshi

First of all I want to thank Beatrix very much for preparing the revised version of the guidance document on the risk assessment of living modified trees.
In my view this document succeeds to cover the aspects that need to be considered when conducting a risk assessment of LM-trees in a very comprehensive and well-structured way.

I think that the document is now very balanced and also addresses in the introductory part the importance of specific aspects of trees regarding their biology and use, which is in my opinion crucial to set the scene for specific aspects of risk assemssement when compared to LM crop plants currently commercially used.

I believe that is very important in this context in order to build the basis for the risk assessment part of the document.

Just two specific remarks at this point:

Line 38-40: The sentence is a bit unclear- is it a theory that high production of biomass is combined with nutrient poor soils? If the ecological theory is only related to the second part of the sentence, it should be made clearer.

Line 58: The reference to the value for industry would fit better in the first para of the introduction. Since the topic of this section is the characteristic of trees it would seem better to write e.g. “Trees produce large biomass and contribute to ecological services and landscape architecture.”

Thank you Beatrix, and the AHTEG members,
for the preparation of the revised version of this guidance document on the risk assessment of LM trees.

It is a very clear document, and the introduction paragraphs add a much appreciated context. As an earlier contribution points out: other documents lack such an introduction but such lack is a weakness of those other documents, and no reason to delete it here.
Setting the context with reference to socio-economic factors does not interfere with the scope of the document since that is outlined clearly later, instead it positions the risk assessment of LM trees within the three goals of the Convention on Biodiversity.

I agree with an earlier contribution that grafted trees should be mentioned more strongly in this document because if either stock or scion are modified then the other part is likely to be affected by it. The counterpart of a grafted tree in which one or even both parts are LM trees, needs to be included in the risk assessment, but there doesn't seem to be an obvious place for it.
Unless there is a better place in this document, this could be considered following line 230 (Exposure of ecosystems).

Regarding the earlier proposal to change line 204, this could be reworded as "the potential of dispersal of propagative material and of its establishment into environments other than ..."

Line 230: Another issue to consider under "Exposure to ecosystems, (c) interactions" are parasitic plants, since they often live on specific tree species.

Two small points:
Line 59: for consistency, it should say "ecological functions and services"
Line 111: "to" is missing between "likely" and "change"

Again, thank you for the preparation of this document,
Antje Lorch

Manoela and All,

Initially I would like to thank everyone that has taken part of their precious time to contribute to the initial document and to the first draft revision. One thing that comes to my attention is that there were several comments to the first draft suggesting a complete re-write or extensive changes.  See: Strauss #2705; Gough #2708; Pearson #2712; Quemada #2715; van der Meer #2734; Nickson # 2737; and my own, Ulian # 2739.
While there have been edits made throughout the text, these do not address many of the comments raised by these commenters. Van der Meer, Nickson and myself suggested to engage experts in the re-writing of the document but it seems that this did not occur.
Some points of major concern that come out in the text of the "RA of Living Modified Trees" document are:
On lines 103 to 104: Where it discusses “whether and how a comparative approach can be carried out in the risk assessment of an LM tree”, it should not be a question about whether a comparative approach can be used – this is one of the main aspects under Annex III.
On Line 106: Where it says “unique adaptive and reproductive characteristics”.  These characteristics are NOT unique to trees.  This is the most notable example of the “unique” perspective, but there are many other examples in other parts of the document.
On lines 107 and 108: It says “may present a special challenge for an assessment based on a comparative approach developed for annual crops with high degree of domestication.”  There are several important points here.  First, this again appears to question the use of the comparative approach for LM trees.  Second, it suggests that the comparative approach was specifically developed for annual crops with a high degree of domestication, and that its use is only applicable to these crops. Third, it ignores the fact that many planted tree species already have degrees of domestication equivalent or higher than annual crops,
On line 133: It is written that “presence of mutations resulting from the transformation methods”.  If this is an issue to be addressed it should be addressed in the Roadmap, but this is not a part of the Roadmap and so should not be raised specifically for trees.
Other points to be considered are:
On lines 37 to 40: It discusses various characteristics of tropical forests but appears to confuse these and makes strong assumptions without good scientific references. As examples, native tropical forests typically are rich in biodiversity but this does not necessarily correlate with high biomass production on a yearly basis, as stated. 
Managed forests – that is, plantation grown forest trees (for example, Eucalyptus) in the tropics do have high yearly biomass production compared to other regions.
There should be a clear distinction between “forests” and plantation grown forest trees as noted by Strauss (see comment 2705 in the online forum).
On lines 52 and 53: It reads “the vegetative phase of a tree, where only vegetative propagation is possible,”  This suggests that before the onset of reproductive maturity all (or most) trees propagate naturally by vegetative means.  But this is not the case and can happen in only a few species (again, see Strauss comment).  We need to be scientifically clear and accurate in such statements.
On lines 65 and 66: There is a discussion on the view that domestication is low in trees.  Again, does not address the point raised by Strauss on interspecific hybrids (consider for example Eucalyptus urograndis and other crosses in Brazil) that could be considered highly domesticated.
On lines 113 to 114: There is a discussion on the special use of regional provenances in sustainable forestry.  All properly managed forestry systems should have a goal of sustainability.  This includes and is applied to the use of exotic species, such as Eucalyptus in Brazil. 
On lines 118 and 119: There is a discussion on the higher potential for trees to disperse and establish in natural ecosystems.  This is another example of a generalization and applying a characteristic to all trees.  This would not be the case for Eucalyptus in Brazil and there are recent studies showing that not to be the case (see reference in Ulian #2739).
In the points to consider sections on the dispersal of pollen or seed, and on exposure.  Again, none of these are unique or special to trees.

I would like to thank the organizers for the opportunity to input my comments to this important document,

Eugenio Ulian

Dear forum members.

I tried attaching a track changes version of the latest draft LM tree RA Guidance document but am having problems with the uplaod.  I will e-mail the file to Manoela to ask her to post it on my behalf.

I have made suggested changes to the text throughout but in particular offer a restructuring of the section on the characteristics of trees that I hope will give more specific background to risk assessors who may not be familiar with tree species.  I feel the current draft still has considerable overlap with the Roadmap and suggest edits to address that also.  Having just read the comments posted by Eugenio Ulian I hope that my text edits can address at least some of the concerns he raised.

I reiterate from my comment on the first draft my belief that we are still some way from being close to a useful, practical document.  A key point in the risk assessment I believe must be to consider the existing experience and understanding within the broader scientific community that is available to the risk assessor.  Not captured in the current document is the understanding that there is a wealth of scientific information on trees that exists today, despite their perceived low degrees of domestication.  As an example I provide in my edits a link to the recent IUFRO (International Union of Forest Research Organizations) conference on tree biotechnology as a reference to examples of the state of the art in LM trees and forest tree research.  This group in IUFRO – Unit 2.04.06 Molecular Biology of Forest Trees, is just one of dozens of working groups that address all areas of forestry research ranging from silviculture to forest policy and economics.  While I think we are very fortunate to have the participation in the forum by some experts familiar with LM Trees, we are just a few, and none of us would presume to represent the full extent of available knowledge and expertise in LM trees, much less the vast diversity of trees and their values as described in the document’s Introduction.  In this regard I again strongly recommend that the AHTEG seek ways to engage a broader set of tree experts in the development of this document as a way to ensure that it is concise, yet insightful and practical as guidance to others, and especially as an adjunct to the Roadmap itself. 

Best regards to all,
Les

Dear all,

As a general comments, on behalf of the Bolivian competent authority on
biosafety, we agree and support the approach taken in the current draft of
the RA guidelines for LM trees. Particularly, we appreciate underlying that
(LM) trees differ significantly from annual (LM) crops in terms of
characteristics, life cycle and ecosystem dynamic.

Attached our specific comments and suggestions.

Kind regards,

Georgina Catacora-Vargas

Dear Georgina, dear forum moderators,

has this attachment gone lost?

Dear Members,

I would like to congratulate AHTEG and forum members for the sound work and advancements that resulted in the present version of the guidance on "Risk Assessments of Living Modified Trees". I nevertheless share with doctors Pearson, Breyer and Ulian the preoccupation with the presence of exceedingly overlapping information with the “Roadmap for Risk Assessment of Living Modified Organisms”, and the absence of important contributions made previously. Being a complementary guidance to the “RA of LMO”, it is clear that RA of LM trees has to be guided by the more general document. Therefore, repetition is unnecessary, as clearly pointed out by the previously mentioned members. I would like to support the suggestion of Dr. Pearson of considering IUFRO (and others) as excellent sources of bibliographic material about LM trees, biosafety and forest tree research. Last IUFRO conferences have shown, for instance, that trees like poplars, eucalypts and grapevine have sequenced genomes and gene annotations with much higher qualities, with much more DNA marker-saturated maps and associations of phenotype-genotype than important transgenic crops approved for commercialization (or close to it) like common beans, sugarcane and others. Considering this, I agree with Dr. Ulian that the degree of domestication of some trees are equivalent to agriculture crops, based on the knowledge accumulated and the dependence of such trees on forestry techniques. Finally, I definitely disagree with the text of the guidance concerning “transformation and propagation methods” as a special concern of LM trees, and I suggest the removal of this session. In summary, I would like to strongly recommend the careful consideration of the suggestions made by the above mentioned forum members.
Thanks to the Organizers for the opportunity and, again, congratulations on the work accomplished so far.

G. Pasquali - UFRGS, Brazil

POSTED ON BEHALF OF STEVE STRAUSS

----

Dear Manoela:

Can you please post this comment and proposed revisions to the Guidance Document on GM Trees?  Please confirm, thanks. 

The first two pages of the attached include the text below my signature in this email.   

The comments represent the views of the public sector scientists listed below. 

Thanks very much / Steve

Steve Strauss, Professor, Oregon State University
Cell: 541 760 7357 -- http://www.cof.orst.edu/coops/tbgrc/Staff/strauss/


These proposed revisions submitted by

Dr. Steven H. Strauss, Distinguished Professor of Forest Biotechnology, College of Forestry, Oregon State University, Corvallis, OR 97331 USA

Dr. Wout Boerjan, Professor, VIB Department of Plant Biotechnology and Genetics, University of Gent, 9052 Gent, Belgium

Dr. Richard Meilan, Associate Professor, Department of Forestry and Natural Resources, Purdue University, West Lafayette, IN  47907 USA

Dr. Scott Merkle, Professor, Warnell School of Forestry and Natural Resources, University of Georgia, Athens, GA 30602 USA

Dr. Gilles Pilate, PhD, Directeur de recherche, Unité Amélioration, Génétique et Physiologie Forestières, Centre d'Orléans, INRA, France

Dr. William A. Powell, Professor and Director, Council on Biotechnology in Forestry, SUNY College of Environmental Science and Forestry, Syracuse, NY 13210 USA

Summary of proposed changes

We have submitted a number of detailed revisions on the following/attached pages. We are all forest scientists and biotechnologists with direct experience in conducting laboratory and field research with GM trees.  We therefore understand the importance of field research and also how excessive regulatory requirements can make essential research difficult and in many cases impossible.  It is in the spirit of trying to balance the risk that regulations are so strong that they effectively remove the possibility of research and ultimate benefit with GM trees, with the need for reasonable precaution and of course for compliance with required Conventional on Biological Safety processes and national laws, that we submit these comments. 

We make the following general recommendations for framing risk assessments; all relate to issues raised within this document.

1. Do not establish stringent blanket regulations for risk assessments for all GM trees.  Some have modifications that are well within those routine in conventional breeding, for example those induced by gene suppression or directed mutagenesis, while others are more novel as they might encode novel and ecologically active proteins.  In depth studies should only be for truly novel kinds of traits that go well beyond those possible with conventional breeding for those genera. 

2. Consider the benefits as well as the risks.  Projects that seeks to address a serious economic and ecological problem, such as by producing disease or climate-stress resistant trees, should receive lesser stringency in risk assessment due to the urgent problems they seek to address. 

3. The focus should be on the phenotype and its effect, not on detailed studies of mutagenesis or gene expression, as there can be extensive genetic change associated with conventional breeding methods, such as species hybridization. 

4. Many studies have shown that beyond the first year or two of study, transgenic traits are highly stable.  Long term studies are generally not recommended or needed. 

5. All studies must be comparative and phenotype based.  Detailed studies of ecological processes such as of mycorrhizal associations and microbial populations, are not needed when normal plant growth and stand structure can be observed via normal breeding-silviculture types of field trials. 

6. Absolute genetic containment is essentially never obtained in conventional breeding, and will rarely be needed with GM trees grown in large research trials or commercially, unless they have been show to have radical ecological impacts.  Provisions for confinement (quantitative not qualitative isolation) during research and commercial use have been developed and used for conventional breeding and can also be applied to GM trees.

POSTED ON BEHALF OF DAVID HERON

----

Dear Colleagues,

I appreciate the efforts of contributors in drafting this document and thank you for the opportunity to give the following suggestions.  I have several overall comments about the guidance document.  First, The Risk Assessment Roadmap is still in development, which makes reviewing the development of this guidance difficult.   In general, the guidance in this document on conducting the risk assessment is applicable to more than LM trees and is covered by the roadmap.  For this reason the comments in the attached document has removed text that appears to be redundant with the Roadmap. 

In addition, the suggested text changes have removed guidance that is not unique to LM tree risk assessments.  It is important that the document more effectively identify the different types of trees and associated unique characteristics that may impact risk assessment for confined and unconfined environmental releases. 

The suggested text revisions in the introduction are offered to make clear the wide range of uses for trees in agriculture, forestry, and ornamental use.  As one example, fruit tree cultivation, breeding etc. is fundamentally different from that of forest trees.  There is also very little information in this document that is relevant to fruit trees or trees used for ornamental purposes. 

The attached file contains suggestions for clarifying these and other aspects of the document.  I have attached a version of the document with track changes and a clean version that may be easier for others to review.  We have retained some comments in the margin on the clean copy to remind the reader of overarching reasons for the suggested changes.  I look forward to reviewing input from others as this document progresses. 

Thank you,
David Heron

Dear all.

thank you, Beatrix, for this draft, and thank you all for the current debate.

Much has been said, and so I would like to add just a few points:

As compared to annual crops, it is important to highlight the time, spacial and ecosystem difference between trees, especially forest trees, and annual crops.

In general, it is also important to emphasise ecosystem function and resilience, which are underlying - sustainable ecosystem services amongst others, rather than referring to the more limited ecosystem (or ecological) ‘services’ on their own.  (e.g. line 59)

Line 59 after architecture. Add:  Forest trees, as part of the forest ecosystems, both require water for growth and contribute to the maintenance of watersheds.

Line 73: delete “in an attempt to improve”  and replace with: “with the aim to achieve” to have a more neutral approach.

Compared to the first draft, there is no particular mention of the different approach taken with regards to heritage trees, which are clearly intended to spread. If this guidance is to cover them (they were excluded in the scope of the previous draft), then there might be a need for specific wording in the various steps to address the approach of intended spread of trees – as compared to preventing spread and outcrossing for other LM trees.
I also find it problematic that bamboo, palm trees etc might be included, as they are clearly different from trees, though they might give rise to similarly severe risks to forest ecosystems if they should spread.  The previous draft had included text that we should reconsider, i.e. : “The family of Palmae or Aracaceae (palm trees) and the herbaceous flowering banana plants as not being trees, and bamboos belonging to the family of true grasses (Poaceae) are not covered by this guidance.”

As mentioned by others, a comparative approach might be challenging and we need look at other assessment approaches and reemphasize the precautionary approach when faced with situations where there is no way to predict or safely assess the behaviour and impact of LM trees.

In line with comments already made by others, I want to emphasize that trees with all their characteristics as well as their role for ecosystems and ecosystem function and resilience – and therefore the assessment of LM trees – are distinctively different to annual, highly domesticated crops. Any risk assessment will have to take on board and address the integral position and role of trees within interactive and interdependent networks,  especially in forests.

Line 106: the word “capabilities” of the first draft has been replaced with “characteristics”. These terms are clearly different and should both be used here linked by ‘and’.

Line 111: the word ‘effects’ might be better here than ‘impacts’

Line 113: the term “regional provenances” needs to be detailed in use of terms (or further explained here with regards to eg using local seed from trees etc.

Line 142: affect or effect?

Line 151: add after ‘grafting’: “as epigenetic changes are inheritable”.
It is important in this paragraph and this section to highlight the role of epigenetics in adaptation as well as a response mechanism of organisms to external and internal changes, and that epigenetic alterations are inheritable but can also be undone over time.

Line 190 (f): include “across snow and ice cover” in the list.

Line 196 (i): add “building materials, transport of wood” to the list

Line 240/1: add new (d): Exudes from LM trees, such as bt toxins, into soil, and into ground water.
Of course their might be other compounds - than those directly linked to a newly introduced trait - that are newly, increased or less produced by the tree and passed to the soil via the roots. This should also be reflected in this point, as the interactive network in soil is crucial to the growth and heatlh of the trees (amongs others).

250: add (in line with comment above): “Different again are heritage trees, intended to spread.”

I will suggest further references at a later stage.

with kind regards,

Ricarda

Dear Participants,

Two recurring themes in the postings to date have motivated my posting to this forum:

1.  Lack of benefit from experts familiar with LM trees.

2.  Lack of clarity regarding why the currently proposed points to consider are unique to LM trees and therefore require further repetition in a separate document.

Regarding point 1, it would certainly be an easy matter to incorporate the suggestions posted on this forum by recognized experts in the field.  Therefore, I would expect to see those suggestions reflected in the next draft.  However, I believe that with regard to point 2, the lack of clarity could be addressed by relying on examples of risk assessments already conducted with LM trees and available online, as Les Pearson points out.  I believe these examples would be of more benefit in assisting inexperienced regulators in determining relevant points to consider than the current points.  Therefore, the changes that this would involve in the text are relatively simple, as I show in the attached document.

Regards,

Hector