Canadian Food Inspection Agency
Plant Products Directorate
Plant Biosafety Office

Workshop Proceedings

Canola Confined Research Field Trials: Revisiting Terms and Conditions
(September 10, 2003)

Executive Summary

The Plant Biosafety Office (PBO), Canadian Food Inspection Agency (CFIA), held a half day workshop to review the current terms and conditions for confined research field trials of canola (Brassica napus) outlined in its Directive 2000-07 (Dir2000-07), Conducting Confined Research Field Trials of Plants with Novel Traits in Canada. Invited to attend were representatives from federal government bodies, producer and trade associations as well as academia.

This report is meant to be an accurate reflection of the workshop proceedings. Please be aware that the views and personal opinions of participants expressed during the workshop have been included for accuracy but do not necessarily reflect the views of the CFIA. In addition, the objective of the plenary session was not to seek consensus among participants, but rather to provide a forum for participants to share their expertise, views, comments and recommendations.

The format of the workshop was a group discussion, which centered around focus questions developed by the PBO prior to the workshop. These questions were designed to give the PBO a better understanding of participants'views as well as concerns regarding the current terms and conditions for confined research field trials of canola. Specifically, the PBO wished to discuss reproductive isolation techniques, canola seed dormancy, sexual compatibility between Brassica juncea, Brassica napus, Brassica rapa and their wild relatives, as well as canola seed viability at various crop growth stages. In addition, the PBO wished to identify potential new concerns regarding the management of canola volunteers in post-harvest years as well as new management tools for their control.

Participants felt that increasing prescribed isolation distances on low risk confined research field trials would prove too costly and provide minimal gain in risk mitigation. For those confined research field trials considered high risk, such as those intended for plant molecular farming, participants agreed these trials may require more strict reproductive isolation measures, such as greater isolation distances or use of integrated reproductive isolation plans.

An increase in guard row width was also assessed to have minimal potential benefits as the difference in outcrossing frequency between B. napus in the trials and B. napus in the guard row is minimal past the 10m point. Participants felt more guidance with regards to the use of guard rows as a reproductive isolation technique should be provided in Dir2000-07. Please note that more detailed information on guard rows has been included in the revised version of Dir2000-07 (updated October 17, 2003).

The requirement for soil incorporation of all residual plant material is cause for concern among some workshop participants as they feel soil incorporation increases the risk of secondary seed dormancy of trial material. In an effort to reduce this risk, it was recommended that, where possible, propagable plant material be harvested from trial site. Following harvest, residual plant material on trial site should be lightly disked into the soil to prevent dispersal by wind and fauna as well as to discourage seed dormancy.

Available scientific information and practical field experience support the adequacy of the current terms and conditions for confined research field trials of canola set out in Dir2000-07. Participants agree that further research is needed on the biology of canola, more specifically, on seed dormancy, pollen flow and outcrossing potential. As outlined in the group discussion summary for question two (Q2), Sinapis arvensis (wild mustard) will be removed from the list of weedy species prohibited to grow within the isolation distance prescribed for such species in the terms and conditions for confined research field trials of canola. The PBO will carefully consider all new information and recommendations made by participants when it will review the terms and conditions for confined research field trials of canola.

Q1. Is there key information regarding canola biology relevant to confined research field trials that has been published recently or is soon to be published?

Robert H. Gulden from the University of Saskatchewan has just completed his Ph.D thesis on the "Secondary seed dormancy and the seedbank ecology of Brassica napus L. in Western Canada". His research showed "i) that on some farms, seedbank additions may be lowered by more diligent harvest practices, ii) seedbank persistence of B. napus may be reduced by growing low dormancy genotypes and avoiding seed burial for one year after seedbank establishment and iii) ABA abscisic acid + ABA-glucose ester (ABA-GE) and the ability of seeds to respond to ABA applications after seed dormancy induction may potentially be used to identify seed dormancy potential in this species." see Gulden (2003)

The current terms and conditions for B. napus recommend that "Seed or other propagable plant material from the confined research field trial must be harvested unless otherwise approved by the PBO. Plants must be harvested before full maturity to minimize silique shattering and seed dispersal. This is intended to decrease the seedbank in the soil and minimize the problems outlined above.

Beckie et al. (2001, 2003), Hall et al. (2002), and Warwick et al. (2004) published papers reviewing the "Impact of herbicide-resistant crops as weeds in Canada" . These papers reported that pollen flow for B. napus can extend up to 800 m. Beckie et al. 2003 report a table summarizing the results of all canola pollen flow studies. In a report on outcrossing studies entitled, "Genetically Modified Canola in Australia" , Phil Salisbury reported that outcrossing levels from small plot trials to a field are substantially lower than outcrossing levels from one large field to another. A small plot produces less pollen and, consequently, a less dense pollen cloud compared with the more dense pollen clouds produced by large fields.

The role of the PBO is to mitigate risk, where risk = hazard x exposure. A compromise between what is an acceptable level of pollen flow versus the economic viability of utilizing extreme isolation distances needs to be established. In order to maintain what has been established as an acceptable level of risk with respect to field trials in the environment, the environment's exposure to the novel trait must decrease as the potential and real hazards of that novel trait increase.

The PBO currently uses a 200 m isolation distance for B. napus, which is based on the Canadian Seed Growers' Association (CSGA) pedigreed seed guidelines. Workshop participants indicated that the difference in outcrossing frequencies at isolation distances of 200 m, 400 m and 800 m is minimal (Beckie et al. 2001, 2003). The PBO acknowledges the occurrence of a small outcrossing frequency at distances greater than 200 m, however, the environment's exposure to a novel trait from a small plot is minimal, therefore the outcrossing frequency at a distance of 200 m and greater from the trial is negligible. Where toxic products are involved, isolation distances may be increased in order to further mitigate the risk.

Christian Damgaard and Gösta Kjellsson from the Department of Territorial Ecology of Denmark's National Envrionmental Research Institute (Danmarks Miljøundersøgelser (DMU)), created a mega model from all published outcrossing data. In their study, it was observed that for crop species with a prescribed isolation distance of 100 m in their terms and conditions of confinement, every 5 m border strip (i.e., guard rows) adjacent to a pollen source exhibits the same level of pollen flow control as a 50 m isolation distance. For larger isolation distance requirements, every 100 m border strip adjacent to a pollen source would be equivalent to a 200 m isolation distance. The paper concluded that borders are indeed effective and may, in addition, be capable of limiting pollen movement (i.e., guard rows have demonstrated a certain level of control of bee pollination). Please note: This results of this study have not yet been published but are in the process of being peer reviewed for publication.

Q2. Is there new information about sexual compatibility between B. juncea, B. napus and B. rapa and their wild relatives that should impact on the terms and conditions for confined research field trials?

Ginette Séguin-Swartz, Hugh Beckie and Suzanne Warwick are currently conducting research in Saskatchewan on pollen flow between populations of B. napus, commercial B. juncea, commercial B. rapa and volunteer B. rapa. Studies from B. napus to B. juncea are still in progress, but preliminary data indicates that the highest outcrossing frequency occurs within the first 50 m of the pollen source. Gene flow from commercial B. napus fields to commercial B. rapa and to B. rapa volunteers in Western Canada was detected at low levels (0.01%) up to distances of 240 m (Warwick et al. Poster 2003b). In a recently published paper, Warwick et al. (2003) also reported the occurrence of herbicide tolerant B. napus x weedy B. rapa (bird's-rape) hybrids in Quebec in 2001. These F1 hybrid plants showed decreased fertility (ca. 60% fertility). Selfing, a B. napus trait, was also detected among the resulting F1 hybrids. Since 2002, Warwick et al. have been studying the persistence and introgression of the herbicide tolerance trait in two wild B. rapa hybrid swarm populations (Warwick et al. 2004, studies in progress). A paper, entitled "Hybridization between transgenic Brassica napus L. and its wild relatives: Brassica rapa L., Raphanus raphanistrum L., Sinapis arvensis L., and Erucastrum gallicum (Willd.)" , has also been written by Warwick et al. 2003. This paper indicates that no interspecific hybrids were detected in commercial fields of herbicide tolerant B. napus and Sinapis arvensis (wild mustard), Raphanus raphanistrum (wild radish) and Erucastrum gallicum (dog mustard).

A number of workshop participants questioned the inclusion of Sinapis arvensis (wild mustard) in the list of related weedy species which must not be present within 50 m of a B. napus field trial. Although no scientific data supports the occurrence of outcrossing between B. napus to S. arvensis in field experiments (reviewed in Warwick et al 2003b), S. arvensis had been included in the terms and conditions for B. napus trials to facilitate inspections. S. arvensis may be difficult to distinguish from other Brassica species, potentially rendering the inspection of a trial very difficult and tedious. For B. napus trials, plants within the trial site must be reproductively isolated from all Brassica species by a minimum 200 m, and isolated from weedy relatives by a minimum of 50 m, unless other methods of reproductive isolation has been successfully used. It was noted that if identification between S. arvensis and some Brassica species is problematic, it may be more logical to control S. arvensis within the 200 m reproductive isolation distance, rather than within the 50 m isolation distance, of a B. napus trial. The PBO has agreed to remove S. arvensis from the terms and conditions for B. napus trials based on 1) further discussion with inspection staff regarding the difficulty in distinguishing S. arvensis from other Brassica species, and 2) the data presented in the paper by Warwick et. al. (2003). The isolation distance of 50 m of S. arvensis plants from a B. juncea trial was also questioned. Based on research currently being conducted by Dr. Suzanne Warwick, it was suggested that a 200 m isolation distance may be more suitable as B. juncea and S. arvensis are sexually compatible.

Q3. Of the approved methods for maintaining reproductive isolation in field trials, which are currently preferred and why? Is there need for changes to the approved methods? Is it possible to improve the effectiveness of guard rows? Does the use of guard rows require further study?

Most applicants for confined research field trials appear to be using isolation distances or guard rows (or, in some cases, both) to achieve reproductive isolation. Many factors must be taken into consideration when guard rows are used as a means for reproductive isolation. The guard rows must contain early, mid and late flowering varieties in order to have concurrent flowering with the trial plot. In the event of guard row failure, applicants must be prepared to default to the required reproductive isolation distance, or, alternatively, remove/destroy their trial plots. It was also suggested that the use of two seeding dates may enable guard rows to flower throughout the entire flowering phase of the trial plot.

There was some question as to whether a 10 m guard row for B. napus was sufficient. Studies reported in Beckie et al. (2001, 2003) indicate effective pollen flow, i.e hybridization rates of 0.19% at 50 m compared with 1.25% at a common border between commercial B. napus fields. From personal experience, other workshop participants believed that 15-20 m guard rows should be required for B. napus trials. A paper by Cuthbert and McVetty (2001) reported that the outcrossing frequency beyond the10 m point is approximately the same ( i.e., the outcrossing frequency at 15 m from trial plot is approximately the same as the frequency observed at 30 m). It was pointed out that guard rows show merit from the perspective of limiting bee pollination, but are not necessarily useful in the prevention of wind pollination. However, wind pollination does not play a major role in the outcrossing of B. napus as this species'pollen is very heavy (CFIA, 2001).

Applicants have requested more guidance on a number of issues surrounding guard rows:

These issues have been addressed in the updated Directive 2000-07, "Conducting Confined Research Field Trials of Plants with Novel Traits in Canada".

Studies are currently being conducted on soybean fields to provide more information to those using pedigree seed systems on the effectiveness of guard rows as a reproductive isolation tool. These trials have indicated that guard rows are more effective in controlling gene flow than isolation distances. More specifically, guard rows of the same species have demonstrated to be more effective than guard rows of different plant species. As soybean is generally considered to be a mostly self-pollinating crop and where Brassica species are considered to be mostly cross-pollinating crops, the PBO will require studies to be conducted on the various Brassica species in order to make informed decisions with regards to the efficacy of guard rows for these crops.

The merits of tenting and bagging as methods of reproductive isolation techniques were also discussed. Tents appear to be effective for controlling bee and fly pollination but not for controlling pollination by smaller insects. Tents also limit the occurrence of wind pollination as the probability of pollen being blown out of the tent and subsequently cross pollinating another plant is very low. However, standards need to be set regarding the acceptable mesh size for tents. Bags seem to be more difficult to use accurately and are more fragile, making them less effective than other means of reproductive isolation. The revised version of Directive 2000-07 requires that applicants provide scientific rationale to justify the effectiveness of isolation measures such as tenting or bagging.

Industry/applicants may be more comfortable with two standards for reproductive isolation, depending on whether the trait is a high risk or low risk (i.e., molecular farming etc.). For example, field trials containing high risk traits, such as some molecular farming products, would have more stringent standards for reproductive isolation than common low risk traits. Currently the PBO imposes higher levels of reproductive isolation for high risk trials on a case by case basis, but has not set a standard for high risk traits.

Q4. Are there new concerns regarding canola volunteers in post-harvest years of confined research field trials, or are there new management options for control of volunteers in post-harvest years to bring trial sites back into compliance?

In recent studies by Beckie et al. (2004), greenhouse experiments were conducted to investigate the dose response of three single herbicide tolerant (HT) cultivars (glyphosate, glufosinate, imidazolinone), one non-HT cultivar, and seven multiple HT experimental lines (double or triple) treated at various growth stages with 2,4-D amine, 2,4-D ester, MCPA ester, and metribuzin. These studies suggest that volunteer canola with multiple herbicide tolerance traits does not differ from cultivars that are non-HT or single HT in its sensitivity to herbicides commonly used to control volunteers. All volunteers, whether non-HT, single HT, or multiple HT should be treated when plants are most sensitive to herbicides (two- to four-leaf stage) to reduce their interference against crops and their perpetuation of gene flow.

Q5. Is there any new information about canola seed dormancy that might impact on the terms and conditions of confined research field trials?

Gulden (2003) reports a high level of secondary seed dormancy for B. napus in the field. Therefore, for the purposes of confined research field trials, it would be beneficial if breeders selected for the low secondary seed dormancy phenotype as part of their breeding strategy in order to help control B. napus volunteers.

In the current terms and conditions for confined research field trials of B. napus, the PBO requires that residual plant material be soil incorporated after harvest in order to prevent dispersal of plant material by wind or fauna. Soil incorporation has been observed to increase secondary seed dormancy and, subsequently, to increase the number of volunteers the following spring. One mitigation option adopted by the PBO is requiring confined research field trial material to be harvested prior to full maturity in order to reduce seed shatter. However, harvesting before maturity is not always an appropriate course of action for a developer collecting the required information for an unconfined release submission. In addition, harvesting green material may obstruct combines, resulting in even more dispersal of plant material on the soil surface. As such, alternative mitigation options for secondary seed dormancy need to be developed. Workshop participants felt that guidelines should be interpreted as "seed is required to be harvested from trial site with as little seed shatter as possible (e.g., prior to the seeds drying out or seed shattering). After harvest, residual plant material on trial site should be lightly disked into the soil to prevent dispersal by wind and fauna as well as to discourage seed dormancy.

An increasing amount of research has been indicating that seed dormancy rates for B. napus may be up to five years, recent papers by Légère et al. (2001) and Simard et al. (2002) have reported that volunteers can persist up to 4 yrs and 5 yrs in western and eastern Canada, respectively. However, the workshop participants have observed that very few B. napus volunteers emerge during and after the third year of the post-harvest restriction period.

It was also postulated that a five year post-harvest land use restriction for B. juncea trials may be too long. Field researchers have noted that B. juncea volunteers do not seem to occur. The PBO is waiting for published data before considering amending its terms and conditions for B. juncea confined field trials .

Q6. Is a policy needed to address seed viability at different growth stages (i.e., at what growth stages can viable seeds be recovered from a tilled plant?)?

Currently, if related species are found within the isolation distance of a current year trial or, if volunteers with developed pods are found on a site still subject to post-harvest restriction requirements, the trial site will be placed under additional post-harvest requirements. For example, the isolation distance may be included in the post-harvest monitoring area or additional years may be added to the original post-harvest restriction period.

It has been recommended that germination tests on the different stages of seed development should be conducted to determine at what stage seed becomes viable. The resulting generated data could potentially be used to develop a growth stage scale for both field managers and CFIA inspection staff to use in the determination of presence of viable seed. Generally, a seed is considered to be viable if its seed coat has developed. However, the in-field determination of the seed coat's presence is very difficult as a magnifying glass is required to identify the seed coat.

Until other identification methods are available, the best practice remains to assume that all seed, whether mature or immature, is viable and must be treated accordingly (i.e., trial site under post-harvest restriction will be subject to additional post-harvest requirements).

Q7. What other issues regarding canola confined research field trials need to be discussed?

Over the years the PBO has dealt with fewer and fewer compliance problems and has seen a better understanding of the regulatory system by its confined research field trial applicants. Compliance problems are inevitable as confined research field trials deal with the biology of plants, an area where the unexpected can always occur, even under the strictest terms and conditions. The best option available remains risk mitigation to the best of our ability.

CropLife Canada has developed an interactive training program for conducting confined research field trials and hopes to implement a certification program. This program would require the successful completion of an exam for certification. This program could also be potentially included with other licensing and registration programs from Health Canada's Pest Management Regulatory Agency (PMRA), the Canadian Seed Institute (CSI) , as well as other programs for licensed operators, accredited graders, etc. However, it is still unclear whether CropLife Canada's certification program will be recognized by the PBO.

Workshop participants felt that, should this confined research field trial training program become recognized by the PBO, it should be taken not only by the individual who is legally responsible for the field trial(s) [the applicant and/or Canadian Agent] but also by any individual who will be working on the trial plot(s).

Some industry representatives expressed a desire for the PBO's confined research field trial program to include a written notification by the PBO to the applicant at the conclusion of the post-harvest monitoring period of a trial, indicating that all requirements for the confined field trial site are now complete.

Workshop participants agreed that it is more beneficial to hold small expert meetings, such as this workshop, in order to address changes to the terms and conditions of confined research field trials as oppose to large multi-stakeholder meetings. The PBO is looking into conducting similar expert workshops for the terms and conditions of confined field trials of wheat, alfalfa and other perennials in general.

List of Participants

References

Beckie, H J; Hall L M; Warwick S I (2001). Impacts of herbicide-resistant crops as weeds in Canada. Proceedings Brighton Crop Protection Conference - Weeds. British Crop Protection Council, Farnham, Surrey, UK, Vol. 1 pp.135-142

Beckie, H.J., Warwick, S.I., Nair, H. and Séguin-Swartz, G. (2003). Gene flow in commercial fields of herbicide-resistant canola (Brassica napus). Ecological Applications 13: 1276-1294. ***[Provides a summary table of all pollen flow studies in canola].

Beckie, H.J., Johnson, E., Séguin-Swartz, G., Nair, H., Warwick, S.I. (2004). Multiple herbicide resistant canola (Brassica napus) can be controlled by alternative herbicides. Weed Science 52: "In Press" .

Canadian Food Inspection Agency, Plant Biosafety Office (2001). Directive Dir94-09: The Biology of Brassica napus L. (Canola/Rapeseed).

Cuthbert, J.L., McVetty, P.B.E., Freyssinet, G. and Freyssinet, M. (2001) Comparison of the performance of bromoxynil-resistant and susceptible near iso-genic populations of oilseed rape. Canadian Journal of Plant Science 81: 367-372.

Damgaard C; Kjellsson G. Department of Territorial Ecology DMU, Vejlsovej 25, 8600 Silkeborg, Denmark (not yet published)

Gulden, R. (2003) Secondary seed dormancy and the seedbank ecology of Brassica napus L. in western Canada. Ph.D. thesis, Department of Plant Sciences, University of Saskatchewan.

Gulden, R.H., Shirtliffe, S.J. and Thomas, A.G. (2003) Harvest losses of canola (Brassica napus) cause large seedbank inputs. Weed Science 51: 83-86.

Hall, L.M., Good, A., Beckie, H.J. and Warwick, S.I. (2002). Gene flow in herbicide-resistant canola (Brassica napus): the Canadian experience. pp. 57-66. In Ecological Impact of GMO Dissemination in Agro-Ecosystems, Edited by T. Lelley, E. Balázs, and M.Tepfer. Facultas Verlags- und Buchhandels AG, Berggasse, Austria. This 2003 publication is based on presentations at an International OECD Workshop held in Grossrussbach, Austria, 27-28 September, 2002.

Légère, A., Simard, M.-J., Thomas, A.G., Pageau, D., Lajeunesse. J., Warwick. S.I. and Derksen, D.A. (2001). Presence and persistence of volunteer canola in Canadian cropping systems. Proceedings Brighton Crop Protection Conference B Weeds. British Crop Protection Council, Farnham, Surrey, UK. pp. 143-148.

Salisbury, P. (2002) Genetically Modified Canola in Australia. Agronomic and Environmental Considerations. Edited by K. Downey. Published by Australian Oilseeds Federation. 69 page report.

Simard, M.J., Légère, A., Pageau, D., Lajeunnesse, J. and Warwick, S.I. (2002). The frequency and persistence of canola (Brassica napus) volunteers in Québec cropping systems. Weed Technology 16: 433-439.

Warwick, S.I, H.J. Beckie, G. Séguin-Swartz, E. Seidle, C. Voloaca and T. James. (2003a). Gene flow between transgenic herbicde resistant (HR) Argentine canola (Brassica napus L.) and cultivated Polish canola (Brassica rapa L.) Annual Meeting of the Canadian Weed Science Society, Halifax, 30 Nov. - 4 Dec. 2003, Abstract, Poster.

Warwick S I; Simard M J; Legere A; Beckie H J; Braun L; Zhu B; Mason P; Seguin-Swartz G; Stewart C N Jr. (2003b) Hybridization between transgenic Brassica napus L. and its wild relatives: Brassica rapa L., Raphanus raphanistrum L., Sinapis arvensis L., and Erucastrum gallicum (Willd.). Theoretical and Applied Genetics 107: 528-539

Warwick, S.I, Beckie, H.J. Simard, M.J., Légère, A., Nair, H. and Séguin-Swartz, G. (2004). Environmental and agronomic consequences of herbicide-resistant (HR) canola in Canada. BOOK CHAPTER No. 24- CABI Publishers; Papers from the European Science Foundation Conference 21-24 January 2003 titled "Introgression from Genetically Modified plants into Wild relatives". "In Press" .

Document in PDF format: http://www.inspection.gc.ca/english/plaveg/bio/consult/canolae.pdf