DEFRA's Aim
Sustainable development, which means a better quality of life for everyone, now and for generations to come, including:
a better environment at home and internationally, and sustainable use of natural resources;
economic prosperity through sustainable farming, fishing, food, water and other industries that meet consumers' requirements;
thriving economies and communities in rural areas and a countryside for all to enjoy.
Carol Norris and Jeremy Sweet NIAB Huntingdon Road Cambridge CB3 0LE
Final report of monitoring studies of field scale releases of GM oilseed rape
crops in England from 1994 - 2000.
Enquiries should to Jeremy Sweet (jeremy.sweet@niab.com)
mindfully.org note: This file contains excerpts from the final report. Most of what is included are the conclusion and discussion sections of each chapter, with some chapters omitted entirely
The full report is available in PDF format in two sizes:
INDEX
1. SUMMARY
2. INTRODUCTION AND BACKGROUND
3. OBJECTIVES
4. MONITORING METHODS
5. GM CROP RELEASE SITE DETAILS
6. GENE FLOW BETWEEN ADJACENT OILSEED RAPE CROPS.
6.1. CROSS-POLLINATION AT THE FARM SCALE EVALUATION SITES.
6.2. CROSS-POLLINATION OF A COMMERCIAL CROP OF GEMINI FROM GM WINTER OILSEED
RAPE IN THE BRIGHT TRIAL AT NIAB CAMBRIDGE
6.3. CROSS POLLINATION OF A COMMERCIAL CROP OF APEX IN A FIELD ADJACENT TO THE
GM WINTER OILSEED RAPE TRIAL AT MELBOURN, CAMBRIDGE
7. FERAL AND VOLUNTEER RAPE: GENE FLOW, PERSISTENCE AND WEEDINESS
8. INTERSPECIFIC GENE FLOW: HYBRIDISATION WITH WILD RELATIVES AND OTHER CROP
SPECIES
9. HYBRIDISATION BETWEEN OILSEED RAPE AND TURNIP RAPE
10. GENE FLOW AND GENE INTROGRESSION INTO WEEDY B.RAPA
11. CONCLUSIONS
12. ACKNOWLEDGEMENTS
13. REFERENCES
14. APPENDIX 1. TRANSGENE DETECTION ASSAYS.
1. SUMMARY
Background
In 1994 the National Institute of Agricultural Botany (NIAB) and the Laboratory of the Government Chemist (LGC) were commissioned by the Department of the Environment, Transport and the Regions (DETR) to monitor the first agricultural releases of genetically modified (GM) oilseed rape (OSR) for a three year period. Subsequently NIAB received a second contract in 1997 to continue monitoring releases of GM OSR including all the previously studied sites and any new sites over 1ha. The contracts also required NIAB to conduct studies of monitoring methods and of the flow of transgenes to crops and wild relatives. The monitoring terminated at the end of 2000.
The first crops monitored were seed production crops sown in spring 1995 and 1996. Two 5ha areas of GM winter OSR sown in the autumns of 1995 and 1996, were also monitored. The seed production crops monitored were for the production of the Plant Genetics System (PGS) GM hybrid oilseed rape and consisted of a GM male-sterile female parent line, interplanted with a pollinator containing a male fertility restorer gene, and both lines containing the Bar marker transgene conferring tolerance to the herbicide glufosinate-ammonium. The winter rape areas were PGS trials containing a mixture of GM and non-GM parent lines and hybrids. The transformations were similar to those in the spring rape.
From 1997 several new sites containing trials or crops of glufosinate (Bar and Pat genes) and glyphosate tolerant transgenic varieties were monitored. In 1998, several sites growing a high laurate transgenic spring OSR variety were included in the monitoring study. In 1999 the monitoring included the two first Farm Scale Evaluation (FSE) trials of GM OSR and provided an opportunity to study gene flow between two large adjacent blocks of spring OSR at these sites. By the year 2000, a total of 11 sites that had grown GM OSR were being monitored. All sites continued to be monitored in the years following the GM OSR crop or trial until the end of 2000.
The monitoring programme studied the characteristics of herbicide-tolerant transgenic rape which were most likely to effect the crop, the cultivated and the non-cultivated environment. These characteristics were assumed to be the same as those of non-transgenic rape, namely dispersal into and colonisation of these environments and gene flow into other crops, feral populations and wild crucifers. The following factors were studied and comparisons made between the behaviour of transgenic OSR and conventional OSR where possible.
Intra-specific gene flow.
Gene flow was monitored from GM OSR crops to adjacent crops, OSR volunteers and feral rape populations.
No intra-specific gene flow was detected at any of the sites monitored between 1994 and 1997. During this time none of the GM release sites were near to other synchronously flowering oilseed rape crops. No gene flow was detected to OSR volunteers and feral OSR growing near the GM releases monitored at any of ites during this period.
In the period 1998 to 2000 gene flow was detected from GM trials into adjacent OSR crops. At one of the FSE sites gene flow decreased rapidly with distance from the pollen source. However at both FSE sites, levels of herbicide-tolerance in excess of 0.5% were found in some samples taken at 100m from the source while at one FSE site levels of herbicide-tolerance in excess of 0.5% were found in some samples taken at 200m, though the overall trend was for gene flow to decrease with distance. These could have resulted from several factors including, adventitious GM material in the original seed batch of Hyola 401, the possible presence of male-sterile individuals, weather conditions or a combination of these and other unknown factors.
Gene flow was also measured from 2 GM trials into adjacent fields of OSR in 2000. Gene flow levels were found to be substantially higher into a varietal association than a conventional variety, due to the male sterile component of these systems. Levels up to 3.2% herbicide tolerance were found at the edge of one field of the varietal association Gemini, at 105m distance from a small block of transgenic herbicide tolerant OSR. By contrast when a transgenic herbicide tolerant trial pollinated a neighbouring conventional crop of the variety Apex, at a different site, maximum levels of outcrossing at 100m were 0.2%. However at most sampling points less than 0.1% herbicide tolerance was found 70m from the pollen source.
Inter-specific gene flow.
Gene flow was monitored between GM OSR and related cruciferous species. In the first three years of the contract (1994 to 1997) a wider range of crucifers was monitored including Capsella bursa-pastoris (shepherd?s purse) and Sisymbrium officinale (hedge mustard). When the contract was renewed in 1997 it was considered that resources should be concentrated on species considered to be important candidates for hybridisation with OSR. The species that continued to be monitored were Brassica rapa (wild turnip), Raphanus raphanistrum (wild radish), Sinapis arvensis (charlock) and Sinapis alba (white mustard).
No gene flow was detected from OSR into the related species examined in this study during the period 1994 to 1997. Between 1997 and 2000 hybridisation was detected with B.rapa. One site was examined where weedy B.rapa occurred in an agricultural field. Hybridisation frequencies varied between plants and were between 0.0% and 48.5%. When seeds were germinated from hybrid mother plants, some evidence of backcrossing in the direction of both parents (B.napus and B.rapa) was also found. Backcrossing to B.napus plants was identified by their ploidy level, however back-crossing to B.rapa plants could not always be determined by their ploidy level as in many cases this was the same as or very similar to the ploidy level of B.rapa. The co-existence of the B rapa populations with B napus crops and the numbers of hybrids found, suggested that gene flow has been occurring for some time between these populations.
Seed dispersal.
Seed dispersal was usually associated with spillage and distribution by agricultural machinery, particularly combine harvesters. In the contract from 1994 to 1997, it was found that combine harvesters were often not thoroughly cleaned after the harvesting of the GM crop, and the crop harvested subsequently flushed out the GM rape seed onto the ground causing contamination of this field (often winter barley). OSR volunteers were frequently found in these fields in stubble, however they were generally controlled in the same way as conventional volunteers. Outside the cultivated area establishment and survival of seedlings was very poor, and few feral transgenic OSR plants survived to maturity
Persistence of transgenic OSR volunteers.
The persistence of transgenic OSR volunteers was compared to existing data and observations of non- transgenic volunteers. The numbers of GM OSR winter and spring volunteers were generally low in subsequent crops. The presence of a herbicide-tolerance transgene or high laurate transgene did not appear to increase the weediness or persistence of volunteer OSR in this study.
Feral Oilseed Rape
Only one feral OSR population was found to persist for more than one year at any of the sites being monitored. The herbicide-tolerance Bar gene was not detected in any of the feral OSR plants so that effects on weediness and persistence of these populations could not be assessed.
Development of optimal methodology for monitoring.
A practical, effective and economical combination of monitoring methodology was developed to cover all the above aspects of monitoring. This included familiarity with the species and sites involved, combined with phenotypic and genotypic testing for the presence of the transgene. The combination of methods used ensured that any major impacts of the GM plants on the agricultural and local environment occurring at each site were likely to be observed.
Conclusions
The high levels of isolation from other OSR crops flowering synchronously, and the relatively small GM pollen sources and low levels of cruciferous weeds present at the sites limited potential gene flow at the sites monitored in the first 3-year contract (1994 to 1997). Larger trials or crops released during the second 3-year contract (1998 to 2000) and the closer proximity of pollen receptive crops and related wild species allowed greater opportunities for gene flow to be studied. The results from these larger trials and crops indicate that commercial scale releases of GM OSR in the future could pollinate other crops and B.rapa, the levels of cross pollination depending on the environmental, varietal and agronomic factors prevailing at the time. There may be a need to review isolation requirements in keeping with current legislation on contamination thresholds in crops, in light of this research.
6 GENE FLOW BETWEEN ADJACENT OILSEED RAPE CROPS INTRODUCTION
Oilseed rape is predominantly self-pollinating, with about one third of seed produced by outcrossing. The outcrossing rate is greatly influenced by environmental factors and can vary between 12% and 47% (Becker et al., 1992). The environmental influences include geographical location, weather conditions at the time of flowering, and within-plant position of the flowers. Among flowers at different positions on the same plant, outcrossing varies from 11% at the top of the inflorescence to 39% at the bottom. Outcrossing in the experiments by Becker et. al., (1992) was estimated in plants grown in plots at a density normally found in an oilseed rape crop, in five geographical locations, by comparing isozyme patterns of mother plants with their progeny.
Oilseed rape is both wind and insect pollinated. Oilseed rape flowers contain nectaries, which make them attractive to honey bees (Free and Nutall, 1968), although good seed yields can still be achieved without insect pollination. Oilseed rape pollen grains are roughly spherical with a geometric diameter of about 25 µm (McCartney & Lacey, 1990). Fungal spores are of a similar size and are primarily dispersed by wind. A mathematical model from pollen trap data, produced by McCartney & Lacey (1990) predicted that more than 60% of pollen lost from an oilseed rape crop would still be airborne 100m downwind of the crop. The concentration at the ground at this distance, however, would only be 2 to 10% of the value at the edge of the crop. The conclusion from this work was that, although large amounts of pollen are released into the air from flowering oilseed rape crops, it seems unlikely that windborne pollen could play a significant role in c nation at distances greater than a few tens of metres from that crop.
Although most wind-pollinated species are characterised by their smooth dry pollen grains, oilseed rape has sticky pollen which remains adherent to the anthers even under high wind velocities (Eisikowitch, 1980). Insects visit all stages of flowers, touching the anthers, which leads to pollination. It is thought that insects disturb pollen grains on dry days resulting in dispersal by wind. From Eisikowitch?s work it appears that only the pollen grains initially released by insects can be dispersed by wind. Pollination may also occur however by physical contact between neighbouring plants, and this may account for much of the local seed set occurring in the absence of insects.
Although these experiments have suggested that the majority of pollen from a source is likely to fall on the nearest plants, there are very rarely some long distance pollinations and therefore always a small risk of transgene movement outside isolation distances (Ellstrand & Hoffman, 1990).
The measured rates of hybridisation at distances from a pollen source in oilseed rape vary between different experiments. Part of this variation may be explained by the dual pollination modes of insects and wind. However, there are many other factors that can affect hybridisation rates, such as the shape and size of the source and sink, the density of the plants, flowering synchrony, weather conditions, environmental factors such as aspect and exposure, and the genotypes of the source and the sink. Lavigne et. al. (1997) measured the average pollen dispersal from single transgenic oilseed rape plants within a large plot. Results from this work suggest that approximately half the pollen from an individual plant falls within three metres, and that the probability of fertilisation decreases exponentially with increasing distance from that single plant. Lavigne et. al. (1997) also predicted that, as a result of insect movements among neighbouring plants and weak winds, a major proportion of pollen would fall around the source plant and the rest would be distributed over the field as a result of long-distance insect flights or strong winds and turbulence.
Both honeybees (Apis mellifera) and bumblebees (Bombus spp.) play an important role in cross- pollination of oilseed rape. Ramsay et. al. (1999) made observations of honeybee colonies and found that bees may forage up to 2km in all directions from a hive, and when returning to the hive they carry large numbers of loose grains. If these are then picked up by other bees, then this can mean that some pollen transfer and fertilisation could occur up to 4km away from the pollen source by bee transport.
Timmons et. al. (1995) measured airborne pollen levels along a linear transect up to 2.5km from fields of oilseed rape. Measurements were taken using volumetric spore traps. Although wide day-to-day fluctuations in airborne pollen densities were found, pollen levels generally declined with distance and pollen did not remain airborne for significant periods of time. Low pollen densities were consistently recorded at 1.5-2.5km from the source, representing background levels (Timmons et. al. 1995).
The use of male-sterile bait plants has also shown how far pollen moves either by wind or by insects. Timmons et. al. (1995), in a further experiment detecting pollen at long distances, used emasculated and de-petalled oilseed rape plants and placed them in fields at various distances from an oilseed rape field. Seed set on plants 2km away from the source showed that pollen is able to move considerable distances. The experiment could not determine the method of pollen movement, although the removal of petals may have reduced the attractiveness of the plants to pollinating insects in this case. Timmons et. al. (1995) concluded that insect pollination cannot be entirely ruled out, but wind is likely to be the major component of pollen movement in this case. Although precise data does not exist on the amount of pollination that occurs through insects or wind, it is clear that both modes of pollination are important, and that insects may have a key role in long distance pollination events.
The various experiments that have examined gene flow from fields of oilseed rape undertaken in the UK have mainly concentrated on small blocks of transgenic oilseed rape as the pollen source. The use of herbicide-tolerance, in particular glufosinate-ammonium as a marker system, means that these transgenes can easily be detected in conventional rape cultivars by phenotypic and molecular methods.
Although experiments such as the one by Lavigne et. al. (1997) have suggested that the majority of pollen from a source is likely to fall on the nearest plants, long distance pollinations have been reported so that there is always a small risk of transgene movement outside set isolation distances (Ellstrand & Hoffman, 1990).
Detection of herbicide tolerance in seed of male sterile oilseed rape plants at distances of up to 400m from a transgenic pollen source (Simpson et. al., 1999) show that there is potential for oilseed rape pollen to be dispersed by wind and remain viable over considerable distances. However over 1000 seeds from a population of feral rape plants flowering at the same time, and growing within 150m of the same transgenic pollen source were tested and no herbicide-tolerant plants were detected. This suggests that pollen competition is an important factor to consider when studying normally self-fertile oilseed rape.
In this study gene flow was monitored between GM OSR and adjacent crops, GM OSR and feral rape and between GM OSR and OSR volunteers in neighbouring fields. During the first 3 years of the contract (1994 to 1997) the monitoring of intra-specific gene flow was limited to seed production trials of spring GM OSR of between 1ha and just over 7ha. The trials contained between 1/3 and 1/5 male fertile pollinators and produced low seed yields. The fields were isolated from other rape crops by a minimum of 1km and the nearest crops were generally winter rapes with little synchrony of flowering. Harvested seed was bagged in the field and removed to limit seed dispersal. However all farms failed to clean their harvesters after harvest and some seed was dispersed in this way. The trials monitored during this period gave very limited data on gene flow and gene dispersal to other rape crops and fields.
However during the period 1997 to 2000 more fully fertile GM spring and winter GM OSR was grown and this provided some of the first real opportunities in the UK for measuring outcrossing from transgenic OSR into adjacent fields of commercial OSR.
The introduction of the BRIGHT trials and FSE trials in 1998/99 meant that for the first time cross- pollination between large blocks of OSR could be assessed directly by phenotypic methods.
6.3 DISCUSSION
The higher outcrossing levels at 5m distance recorded at the Lincolnshire site than the Oxfordshire site are most likely to have been caused by the gap between the GM crop and the conventional crop. Pollen would move either by wind or by insects directly over the gap and pollinated the first of the conventional plants at the 5m point. At the Oxfordshire site, there was no gap between the two varieties, and the plants located in the area between the GM crop and the 5m sampling point would have acted as a pollen trap. If the 5m and 10m results are excluded from the Lincolnshire totals then the rates of flow at these two sites become more similar so that at 15m - mean cross-pollination was 0.73% and 0.75% at Lincolnshire and Oxfordshire respectively.
The results from the Oxfordshire site indicate some long-distance outcrossing events, at 100m and 200m in Transect 2. There are several factors that may influenced these results:
i) The conventional variety Hyola 401 used in this experiment is a restored hybrid variety, which is the first generation of a cross between a female fertile, male sterile recipient line and a pollinator line. In the construction of this variety then, only the female rows carrying the F1 generation of hybrid seed are harvested. However there may be a low proportion of contaminating male-sterile plants (<10%) amongst the F1 generation sown as Hyola 401. These male-sterile plants in the conventional crop would be entirely outcrossing and thus more receptive to incoming pollen. If male sterile individual plants were sampled at 100 and 200m they were likely to have higher levels of outcrossed progeny.
ii) Advanta Seeds who supplied the Hyola 401 for these trials confirmed that some of the seed batches of Hyola from Canada were contaminated with GM glyphosate and possibly glufosinate tolerance in the 1999 and 2000 sowing years. If the conventional variety in trials contained a proportion of GM plants, this could have a major effect on the cross-pollination data obtained. If one of the sampling points had included a GM plant, or a GM plant was close to a sampling point and so likely to pollinate its neighbours, this might explain the anomalies seen at 100m and 200m in the Oxfordshire trial. Unfortunately the original seed batch of Hyola 401 used in these experiments was not available to test these hypotheses. However independent tests conducted at NIAB of Hyola seed accessions supplied by plant breeders since 1996 showed that several seed lots contained adventitious seed with both glufosinate and glyphosate tolerance up to levels of 0.5% (E. Simpson, pers. com.).
iii) An alternative possibility for the high cross-pollination levels observed at 100m and 200m from the GM pollen source may relate to the position of a small copse situated in the middle of the field (Plate 3.4). This copse may have disrupted the air currents over the field, thereby affecting pollen movement by wind. The flight paths of foraging bees may also have been disturbed by the position of the copse, leading to anomalous long-distance dispersal of transgene-containing pollen.
iv) A major cause of pollen movement at the Oxfordshire site may have been the invasion of the field by anti-GM demonstrators during the flowering period of the rape. Much of the GM crop was deliberately destroyed by this action. Human activity therefore may have carried some GM pollen into areas of the conventional crop where it might otherwise not have reached.
These results suggest that there are a number of unpredictable factors that may influence levels of outcrossing in agricultural systems. This study has shown some unusually high levels of gene flow at distances of 100m or greater. There is, therefore, some potential for significant levels of contamination between large agricultural fields. These data indicate, however, that effective gene flow in oilseed rape is limited in its extent. Only those plants within 5m of the boundary between varieties gave more than 5% cross-pollination. However, long-distance cross pollination at 100m or more from the boundary occurred in 5/6 transects, and attempts to prevent gene flow simply by distance seems an unlikely possibility. The total number of tolerant seeds amongst the seeds of the conventional variety found at Lincolnshire and Oxfordshire falls well below 1% (0.63% and 0.38% respectively). The frequency of contamination is thus low but reflects only 5/8th of the total outcrossing frquency. More data from large-scale plantings are needed to assess patterns of gene flow between varieties of oilseed rape under agricultural conditions. Only when such studies have been carried out, will it be possible to establish the most effective isolation distances between transgenic oilseed rape crops and conventional crops in order to keep contamination levels within defined limits.
The possibility of contamination of conventional seed sources by transgenic material discussed here also suggests that total isolation or exclusion from GM varieties is not practicable in current farming systems.
6.11 OVERALL CONCLUSIONS
Oilseed rape pollen is transferred between flowers by wind and insects, especially by honeybees (Apis mellifera) and solitary bees. Insects play an important role in pollination over long distances. Ramsay et. al.(1999) for example, found that honeybees in a colony in Scotland flew to an oilseed rape source 5km away from their hive. Theoretically then there is potential for pollen to be transferred within a radius of at least 10km by the mixing and contact within the hive of bees foraging in different directions. Thompson et. al. (1999) concluded however, that the relative importance of wind and insect pollination was difficult to determine under field conditions. In their study, male-sterile oilseed rape bait plants were placed at selected distances between zero and 4000m from the nearest non-GM oilseed rape crop. They found that pollination of the male-steriles occurred at all distances. Although airborne oilseed rape pollen was detected at all sites, its concentration declined rapidly with distance from the source. This perhaps indicates that bees may be important pollen vectors over long distances. However, Squire et. al.,(1999) suggested that airborne pollen shows a rapid decline in pollen concentration near the source, followed by a very slow decline with increasing distance. They concluded that the absolute level of airborne pollen then is dependent on the size of the pollen source.
The use of male-sterile bait plants provides some information on pollen movement, but is not an accurate guide to the likely levels of effective intercrossing between fully-fertile oilseed rape crops, because it does not take into consideration the effect of competition from self pollen. Levels of outcrossing decrease with increasing distance from the source, and very low levels of outcrossing (0.0038%) were reported at distances up to 400m from the pollen source with fully fertile recipient crops by Scheffler et. al., (1995). The FSE trials were sampled up to maximum distances of 200m (Oxfordshire) and 250m (Lincolnshire) and at these distances outcrossing was near to zero. Levels fell to below 1% by 25m distance into the receptor variety.
The varying results obtained in the experiments described here can be partly explained by the differences in the trial size and therefore the size of the pollen source. In the Melbourn study that estimated gene flow into the commercial variety Apex, the transgene source trial was a complex mixture of transgenic and non-transgenic plots so it was difficult to determine the exact size of the total transgenic pollen source or the density of transgenic-bearing pollen in the source output. The pollen source in the Gemini experiment at NIAB was smaller than at Melbourn. However, gene flow detected was higher, perhaps because of the male-sterile component of the Gemini receptor crop. This indicated that pollen competition may be an important factor in the amount of outcrossing that occurs. Simpson and Sweet (2001) studied cross pollination with another VA variety, Synergy, and found that, where there was little pollen competition from self produced pollen due to the high proportion of male-sterile plants in the varietal association, there was a higher frequency of outcrossing than with conventional oilseed rape. However in this study there was competition from pollen from neighbouring plots and barrier plots and it appears from comparisons of these three studies that fully fertile barrier (trap) crops surrounding the source have less effect on outcrossing than similar barriers around the receptor.
Although it is not possible to compare the individual studies reported here in a statistical fashion, the results presented show that different situations can give very different results under natural field conditions and that it is not always possible to extrapolate from small-scale experiments to field conditions. The results provide valuable data towards determining isolation distances for both VAs and standard fully-fertile oilseed rape crops to ensure cross-contamination in field scale crops is kept below allowable thresholds.
7. FERAL AND VOLUNTEER RAPE: GENE FLOW, PERSISTENCE AND WEEDINESS
Volunteer and feral rape arises from seed shed both before and during harvest of an oilseed rape crop, from seed shed from existing volunteer and feral populations and from seed spilled during transport on and off the farm . Mature pods opening during harvesting (pod shatter) results in seed dispersal before harvest of at least 10% of the yield each year, with up to 50% yield losses in some seasons (Child and Evans, 1989). Seeds lost at harvest (Plate 7.1) are generally left to germinate on the surface for 2 to 3 weeks after which the field is cultivated (Plate 7.2). Any ungerminated seeds are subsequently buried when the field is ploughed, and can germinate when brought to the surface again by cultivation in subsequent years. This results in unwanted oilseed rape plants as volunteers in the subsequent crops (Plates 7.3 and 7.4). and can reduce yield and affect the composition of following crops if allowed to persist.
In broad-leaved crops such as sugar beet, volunteer rape is generally not well controlled. However, in cereal crops, control of volunteer rape is relatively easy to achieve using selective broad-leaf herbicides. Various studies have provided evidence that rape seeds can persist in the soil for at least five years (Lutman, 1993; Schlink, 1995) and perhaps as long as ten years. A potential concern with regard to the introduction of transgenic oilseed rape is that the presence of a transgene may change the way in which oilseed rape persists in the agricultural environment over time. It is possible that genetic manipulation may cause changes in the agricultural and ecological behaviour of transgenic crop plants. If a transgene produces a trait that gives the plant a fitness advantage, then oilseed rape could become more persistent as a weed in other crops or more invasive of uncropped and marginal land.
Possession of a herbicide-tolerance transgene by an oilseed rape plant would not in itself offer any fitness advantage to the plant, unless the herbicide was applied. However there may be pleiotrophic effects which enhance fitness. Conversely there may be a physiological cost to carrying a transgene that would reduce the fitness of a transgenic rape plant compared to its non-transgenic counterpart. It may be however that the costs of carrying a herbicide-tolerance transgene are relatively low, in which case there may be no measurable differences between transgenic and non-transgenic lines. In the case of the high laurate transgene, for example, there is a small amount of evidence to suggest that the presence of high lauric acid may affect germination rates and seedling mortality ( Linder and Schmitt, 1995). This in turn could affect plant development and thus overall fitness of the adult plants.
Transgenic lines of glufosinate-tolerant, glyphosate tolerant and high laurate oilseed rape were studied for their persistence when occurring as volunteers in subsequent arable crops. The use of transgenic rape in this study also gave an opportunity to study the longevity of oilseed rape seed in the soil seed bank under arable conditions for the first time, since the presence of the transgene gives an absolute dating point. In addition some of the sites used had never grown oilseed rape on their land previously, and thus comparisons were made with non-GM rape also grown at these sites. Previous work on dormancy of oilseed rape and its persistence in the seed bank has concentrated mainly on germination and burial experiments and not on natural (agricultural) situations such as these (Pekrun and Lopez- Granados, 1995; Pekrun et. al., 1997).
In addition cross-pollination between volunteers, feral populations and rape crops was studied to determine the significance of this for gene flow of transgenes.
8 INTERSPECIFIC GENE FLOW: HYBRIDISATION WITH WILD RELATIVES AND OTHER CROP SPECIES
INTRODUCTION
Hybridisation between different species can only occur if the sexual compatibility of the two parental genomes is sufficient to allow the development of a mitotically and stable hybrid. Once seeds have been formed, the endosperm must be able to support normal hybrid development. If the endosperm does not function effectively, development of the hybrid embryo will be aborted (Dale, 1992).
The direction of the cross can also be important as the cytoplasm of one parent may support embryo and seed development more effectively than the other. Once viable seeds are formed, establishment is influenced by the adaptation and fitness of the hybrid to the habitat and the ecological conditions to which the plant is subject. Factors such as seed dormancy, the vigour of the hybrid plant in its habitat, competition with pests, other plants and diseases will determine the ability of the hybrid to survive.
Oilseed rape has several wild relatives with which it has been known to hybridise. In this research we have focussed on four of these species: wild turnip (Brassica rapa), wild radish (Raphanus raphanistrum), white mustard (Sinapis alba) and charlock (Sinapis arvensis).
Wild turnip (B. rapa ssp. sylvestris/ B. campestris)
Frequencies of hybridisation between oilseed rape and wild B.rapa have been reported in numerous studies. Frequencies vary between studies and appear to be much higher where B.rapa occurs as a weed in oilseed rape crops. Jørgensen et. al. (1996) measured frequencies of spontaneous hybridisation between oilseed rape and B.rapa and found that the frequency of interspecific hybridisation varied significantly with experimental design. Frequencies ranged from 9% in a 1:1 mix of B.rapa and B.napus with oilseed rape as the mother plant, to 93% where single isolated B.rapa mother plants were growing in an oilseed rape field. Hansen et. al. (2001) reported finding evidence of introgression in an organic field where B.napus and B.rapa coexisted.
Scott & Wilkinson (1998) also measured gene flow from oilseed rape into wild B.rapa populations growing outside field boundaries. Low hybridisation frequencies were found (between 0.4 ? 1.5%). In addition to this they found that less than 2% of hybrid seedlings survived, and suggested that establishment of GM B.napus x B.rapa populations would be poor and that introgression of the transgenes into the wild population would be unlikely or very slow.
Wild radish (R. raphanistrum)
This species is the second most likely of the four species tested to form hybrids with oilseed rape. In an experiment by Darmency et.. al. (1995) hybridisation was detected at a rate of 1:625, wild radish:oilseed rape. In this experiment 0.2% of the seeds derived from wild radish were intergeneric hybrids. In a similar experiment by Chevre et. al. (1999), wild radish plants were transplanted at different densities into a field of the partly male sterile oilseed rape variety Synergy. They found very low frequencies of hybrids were formed with R.raphanistrum as the female parent. Only one hybrid was found in the 189,420 seeds tested. Darmency et. al. (1995) obtained 45 intergeneric hybrids from each male sterile oilseed rape plant used as the female parent. Another study be Darmency et. al. (1998) found up to 3 intergeneric hybrids per 100 plants when R.raphanistrum was the female parent and isolated from other R.raphanistrum plants in a field of oilseed rape. Seed production of the F1 hybrids in this case was up to 0.4% of the wild species.
The use of a male sterile recipient and high pollen pressure from the wild species has enabled interspecific hybrids between oilseed rape and R.raphanistrum to be relatively easily produced. However under field conditions there would only be a very small chance of interspecific hybrids being produced with B.napus as the female parent because of the availability of pollen from rapeseed in large quantities. This pollen competition would only allow hybrids to be formed at very low frequencies. Studies to date have mainly used B.napus as the female as no male sterility is available in R.raphanistrum. However the presence of self-incompatibility in R.raphanistrum makes the reciprocal cross possible.
White mustard (S.alba)
Although hybrids between S.alba and B.napus have never been found under natural conditions, several methods have been used to overcome the interspecific barriers preventing hybrid production. Protoplast fusion, in vitro fertilisation of ovules, ovary culture and embryo rescue have all been used in the synthetic production of hybrids. 50 to 60 bud pollinations were made (Brown, et. al., 1997) by applying pollen from parent A to the stigmas of parent B, directly after emasculation. Developing ovaries were then excised from the plant 8 days after pollination and surface sterilised. The ovaries were then rinsed and transferred to a sterile culture medium. Embryos were then rescued from the ovaries and placed on fresh media.
All hybrid plants resulting from this procedure were found to be sterile. To overcome sterility cuttings were taken from the sterile plants and treated with rooting powder and then immersed in a colchicine solution in water to induce chromosome doubling. Only two plants were obtained from this, both of which were fertile. Pollination appeared to be more successful with S.alba as the female parent.
Charlock (S.arvensis)
S.arvensis like the other weeds mentioned in this study, is self-incompatible and pollinated by both insects and wind. In the UK B.napus and S.arvensis often grow alongside each other in agricultural fields and the two species often flower simultaneously though S arvensis individuals will flower over a much longer period from spring to late summer.
Artificial hybrids have been produced by emasculation and in vitro embryo rescue techniques (Mathias, 1991; Kerlan et. al., 1992). Kerlan et. al., (1992) succeeded only in producing hybrids where B.napus was the female parent and this was only possible by ovary culture, embryo rescue. No hybrids were produced using S.arvensis as the female parent. A study by Lefol et. al. (1996) revealed no hybrids found among 2.9 million seeds produced by S.arvensis growing amongst herbicide tolerant transgenic oilseed rape. In the same study, when male sterile B.napus plants were grown in the presence of S.arvensis , no more than 6 hybrids were obtained from 50,000 flowers from the oilseed rape plants (Lefol et. al., 1996). These results show that it is possible for a small number of hybrids to be produced under the most favourable conditions.
A more recent study (Chevre et. al., 1998) also using male sterile oilseed rape plants produced 0.18 interspecific hybrids from 100 seeds. The experiment involved growing several replicates of male sterile transgenic oilseed rape alongside S.arvensis under natural field conditions.
Results of similar trials by Moyes et. al. (1999) with seed samples from populations of S.arvensis in the UK and France are still awaiting publication. No interspecific hybrids between B.napus and S.arvensis have as yet been reported under natural field conditions and Downey (1999) concludes that from the data available so far there appears to be a general agreement that natural gene flow is unlikely to occur between B.napus and S.arvensis.
8.4 DISCUSSION
These results suggest that if hybridisation occurs between oilseed rape and R.raphanistrum, S.arvensis or S.alba it is at a very low frequency and that there is poor survival of hybrids. The results this study supports previous work where hybrid formation only occurred under forced conditions or where male sterile oilseed rape was used as the female parent. Results of the work by Lefol et. al.(1996) where 6 hybrids were obtained from 50,000 flowers from male sterile oilseed rape growing amongst S,arvensis show that it is possible for a small number of hybrids to be produced under the most favourable conditions. In a natural situation with male fertile plants there would be pollen competition from the oilseed rape plants and S.arvensis pollen would be even less likely to be successful in fertilisation.
Despite the two species growing together in rapeseed fields for more than fifty years, no hybrids between B.napus and S.arvensis have yet been reported under natural conditions, so transgene movement from B.napus into S.arvensis seems highly unlikely. Likewise hybrids between B.napus and S.alba have never been found under natural conditions, and these results support the general agreement that natural gene flow is unlikely to occur between these two species.
No evidence of hybridisation between B.napus and R.raphanistrumn was found over six years at any of the sites monitored. However the numbers of populations of R.raphanistrum found were few, and therefore data on this species is limited. This weed was only found at two of the sites where glufosinate tolerant rape was grown, and at three sites where high laurate rape was grown. Often these R.raphanistrum populations were only found in the year the GM rape was grown, and were well controlled in following crops. Thus information on hybrids appearing in subsequent years is limited, and it would be necessary to carry out further monitoring at the sites where R.raphanistrum was found in subsequent crops of oilseed rape, in order to test for the presence of hybrids in these crops. As naturally occurring hybrids between B.napus and R.raphanistrum have been reported (Darmency et. al., 1995; Chevre et. al., 1999) this species is a more likely candidate for the possibility of hybridisation than either S.arvensis or S.alba.
There are strong reproductive barriers to inter-specific hybridisation between oilseed rape and most of its related cruciferous weeds. The probability of inter-specific gene flow is therefore very low, but nevertheless possible (Raybould & Gray, 1993). If these barriers are overcome and inter-specific hybridisation does occur, weedy characteristics may be transferred by introgression into the crop plant, making it fitter in natural or semi-natural habitats. Alternatively transgenes could be transferred to weeds making them more difficult to eradicate in agricultural environments or affecting their ecology in natural environments. The results of this research have shown that this is not likely to be the case with the species S.arvensis and S.alba. More data needs to be obtained on the spontaneous hybridisation between B.napus and R.raphanistrum under natural field conditions.
Rare hybridisation events occurring at very low frequencies would not necessarily be detected using these methods. It was not possible or desirable to test all plants in populations, since this would have artificially disrupted the monitoring sites. However the combination of tests and sites used in this would have allowed significant hybridisation events and their consequences to be detected.
Weedy B.rapa was found at only one of the sites monitored. Results of testing of this species are discussed in Chapter 10.
9. HYBRIDISATION BETWEEN OILSEED RAPE AND TURNIP RAPE
INTRODUCTION
In autumn 1998 a block of winter turnip rape (cv. Debut) was sown directly adjacent to a GM oilseed rape trial containing blocks of glufosinate and glyphosate tolerant varieties (Figure 9.1). The turnip rape was separated by approximately 5m on one side from a block of glufosinate tolerant rape and on the other side from a discard barrier plot of the conventional variety Apex. This study investigated gene flow from turnip rape into oilseed rape by screening seedlings grown from seed samples of oilseed rape by morphology. The reciprocal cross was also examined by measuring the proportion of GM seeds at set distances into the turnip rape plot..
9.4 DISCUSSION
The higher frequency of hybridisation detected in the Apex than the glufosinate tolerant variety may have occurred due to a slight difference in flowering times between the two varieties. The turnip rape was the first plot to start flowering, followed by the Apex and then the PGS W5 plot. A more likely explanation would be the establishment of the two varieties. The plot of Apex was badly damaged by pigeons and therefore had patchy and poor establishment with small plants. Pollen production from this plot therefore was not at its optimum, allowing turnip rape pollen to compete well. This could also explain the lower levels of cross-pollination occurring with B.rapa as the female parent. B.napus has been shown in previous studies to be the preferable parent for hybrid formation. However this study shows that hybridisation will occur between turnip rape and oilseed rape and thus there is a potential for gene flow and cross contamination between these crops. This study was not sufficient to indicate relative frequencies compared with outcrossing in oilseed rape varieties. However Downey (1999) suggested that inter-specific hybridisation between B.napus and B.rapa occurred at similar frequencies to within species outcrossing.
10 GENE FLOW AND INTROGRESSION BETWEEN BRASSICA NAPUS AND WEEDY BRASSICA RAPA
10.1 INTRODUCTION
Plate 10.7: Seeds from B.napus, B.rapa and B.napus x B.rapa hybrid. The hybrid seeds are larger than seeds from either parent and are misshapen and deformed.
B.rapa (wild turnip) is a self-incompatible annual or biennial plant which is reasonably common in weedy habitats throughout Britain and Ireland (Plates 6.1 and 6.2). It is locally abundant on roadsides, in arable fields, on waste ground and particularly along riverbanks. It has a chromosome number of 2n=20 and it is thought to be one of the progenitor species (AA) of oilseed rape (AACC) along with B.oleracea (CC). Perhaps because the genome of B.rapa is in common with part of the oilseed rape genome, the two species are sexually compatible and hybridise readily under certain conditions. The resulting hybrids are triploid with 2n=29, of genome constitution AAC. Oilseed rape can still be resynthesised by hybridisation of the two parental species (Prakash and Hinata, 1980).
B.rapa ssp. rapa (turnip) and B.rapa ssp. oleifera (turnip rape) are both grown as crop plants in the US and Europe. In addition, B.rapa ssp. sylvestris (wild turnip, Bastard turnip, Bargeman?s cabbage or navew) is found in both agricultural and non-agricultural environments as a weed and as feral populations. It is the most likely of the cruciferous weed species found in the UK to form hybrids with oilseed rape, and is the species on which the largest number of interspecific hybridisation studies have concentrated (Jørgensen and Andersen, 1994; Landbo et. al., 1996; Scott and Wilkinson, 1998).
Weedy B.rapa is smaller than B.napus and has open flowers overtopping the flower buds. In B.napus the open flowers are overtopped by the flower buds. B.rapa has brighter green leaves, which are hairier than those of B.napus. Interspecific hybrids are sometimes more vigorous than B.rapa and can be very close in appearance to B.napus but can be distinguished by their hairier leaves.
Interspecific hybrids with winter varieties of B.napus inherit a requirement for vernalisation. The B.rapa weedy parent however, does not need a cold treatment to initiate flowering (Jorgensen & Andersen, 1994).
Figure 10.7: DNA ratios of glufosinate-susceptible offspring of a B.napus x B.rapa hybrid (Plant 1) under field conditions from Patrington.
Hybridisation between oilseed rape and wild B.rapa has been reported on numerous occasions (Jørgensen et. al., 1996; Scott and Wilkinson, 1998). Hybrid frequencies in populations vary between studies and are much higher where B.rapa occurs as a weed in oilseed rape crops, rather than in its ?wild? habitat along riverbanks. Jørgensen et. al. (1996) measured frequencies of spontaneous hybridisation between oilseed rape and B.rapa in experimental fields in Denmark and found that the frequency of interspecific hybridisation varied significantly with experimental design. Frequencies of hybrids in progenies ranged from 9% in a 1:1 mix of B.rapa and B.napus with oilseed rape as the mother plant, to 93% where the mother plants were single isolated B.rapa plants surrounded by oilseed rape. B.rapa is highly self-incompatible a single plant will accept pollen from other plants and even pollen from other species more easily than its own.
Spontaneous hybridisation between B.napus and B.rapa has been observed in agricultural fields and also occasionally in nature (Jorgensen & Anderson, 1994; Jorgensen et al, 1996; Landbo et al, 1996). These observations show a large amount of variation in the frequency of hybrid seeds produced, depending on the spatial structure and flowering time of the two species. Interspecific pollination is more likely to occur where B.rapa is well spaced out in the field. Where B.rapa occurs in dense patches, hybrids are less likely to occur because of pollen competition.
Scott & Wilkinson (1998) measured gene flow from oilseed rape into natural populations of B.rapa growing outside field boundaries along the riverbanks of the Thames. A low level of hybridisation was found (between 0.4 ? 1.5%) in populations growing adjacent to oilseed rape fields. However they found that less than 2% of the hybrid seedlings survived. Scott and Wilkinson (1998) therefore suggested that establishment of GM B.napus x B.rapa plants would be at very low frequencies and that introgression of transgenes from B.napus into wild B.rapa populations would be very unlikely or at most would be very slow.
Previous hybridisation studies between oilseed rape and weedy B.rapa have focussed on the hybridisation and introgression of transgenes into weedy B.rapa under controlled conditions, but have not examined the extent of historical introgression with conventional oilseed rape varieties under field conditions. However some studies in Denmark have recently examined introgression from organically grown oilseed rape into weedy B.rapa using AFLP markers (R.B. Jørgensen, pers. com.). L.B.Hansen and R.B. Jørgensen (pers. com.) screened B.rapa growing in an organic rape crop with 24 species- specific AFLP markers. From 102 plants screened, they found 44 appeared to be introgressed beyond the F1 generation.
The review by Gray (1999) of the risk assessment on the PGS hybrid herbicide tolerant oilseed rape considered that the risk of gene transfer to B.rapa had been underestimated and that further research on the mechanisms, frequency and consequences of introgression of genes from B.napus to B.rapa should be undertaken. The second monitoring project was amended in order to gain more information on the extent or possibility of introgression from oilseed rape into weedy B.rapa, and to study crop and weed interactions under agricultural practices.
In 1998 it was brought to our attention by a consultant working in the North Humberside area that wild turnip or bargeman?s cabbage was a weed problem in oilseed rape in this area and he suspected that hybridisation was occurring between the two species. We examined sites near Patrington, North Humberside and observed plants of intermediate morphology between oilseed rape and wild turnip. The wild turnip appeared to be poorly controlled by the standard herbicides used on oilseed rape and farmers were using additional Fortrol (cyanazine) treatments in an attempt to control the weed. The wild turnip appeared in patches in fields that remained in much the same place from year to year. At some sites field margin populations were observed but plants were only found in cultivated ground. One of the sites had been growing conventional oilseed rape for many years and historical records had been kept of all varieties grown in the field for the last decade. AgrEvo (now Aventis) expressed interest in this site for a study of the use of HT oilseed rape to control wild turnip.
In the autumn of 1998 a trial area of 200m x 50m of glufosinate tolerant transgenic oilseed rape (BAR construct) was established in an area of the field known to have a recurring problem with B.rapa as a weed. The purpose of the trial was to study the control of weedy B.rapa with glufosinate but the opportunity was also taken to examine the interactions between weedy B.rapa and oilseed rape under agricultural conditions. Hybridisation between glufosinate tolerant oilseed rape and B.rapa was examined and the morphology and survival of interspecific hybrids and other crosses was also studied. In addition the possible historical introgression between the weedy B.rapa and the individual varieties grown in the field in the last ten years was investigated by using the amplified fragment length polymorphism (AFLP) technique. Individuals from the B.rapa population were screened for bands specific to oilseed rape to try and detect any evidence of past introgression having taken place.
Previous hybridisation studies between oilseed rape and B.rapa have focussed on the introgression of transgenes into weedy B.rapa but have not examined the extent of past introgression with conventional varieties under field conditions. Any evidence of historical introgression may give an indication of the likelihood of long-term persistence of oilseed rape genes in weedy B.rapa populations.
10.4 DISCUSSION
Extensive hybridisation between oilseed rape and B.rapa has occurred at the Patrington site. This finding supports previous work carried out on weedy B.rapa in oilseed rape fields in Denmark (Jørgensen and Andersen, 1994; Jørgensen and Andersen, 1996). No other research has been carried out on hybridisation between the two species in arable fields in the UK, although Scott and Wilkinson (1999) found hybrids in seeds collected from sympatric riverbank populations of B.rapa in the Thames valley
The spatial distribution and growth habit of B. rapa at the Patrington site offers an explanation for the high frequency of hybridisation. B. rapa is a mostly self-incompatible species, so very high proportions of inter-specific hybrid seed are likely to be set on B. rapa mother plants when individuals are isolated from sources of intra-specific pollen and surrounded by oilseed rape. It has also been shown that when B rapa and B. napus pollen are applied to B. rapa stigmas, they have equal fitness and so are equally likely to fertilise B. rapa (Hauser, et. al., 1997). The spatial distribution of B. rapa was patchy at Patrington, with many isolated plants surrounded by B. napus pollinators. Thus B. rapa pollen would have been under a great deal of competition from oilseed rape pollen in these fields.
Flowering time will have an influence on the probability of hybridisation between B. rapa and B. napus. Observations of B. rapa populations in Humberside showed that when growing amongst winter oilseed rape, B. rapa started to flower earlier than the crop. Some individuals finished flowering before the oilseed rape started. Flowering of a proportion of the B.rapa then continued throughout the period of the crop flowering, so there was some overlap. At Patrington the majority of the B.rapa population flowered synchronously with the oilseed rape. It is likely that the early-flowerers maintain a discrete B.rapa population despite a high degree of hybridisation in the population.
Differences in dormancy and germination patterns between B.rapa and B.napus may have an effect on emergence and growth of interspecific hybrids. B.rapa produces seeds with many and varying germination requirements. Seeds from the same mother plant can have different levels of dormancy, as can seeds from different plants in the same population. This contributes to making B.rapa a persistent weed in disturbed habitats by increasing the probability of germination at a favourable time in an unpredictable environment (Rees and Long 1992).
The ability of annual plants to survive in the environment is affected by certain environmental conditions and cues, and also by seed dormancy. Factors affecting the degree of dormancy include seed coat structure, light sensitivity, nutrient availability and cold stratification. Other factors, such as temperature and gaseous concentrations may also influence dormancy. Cold stratification in B.rapa has been found to induce light sensitivity and sensitise seeds to nutrient levels (Adler et. al., 1993). Adler et. al. (1993) suggested that B.rapa seeds that have dispersed in the autumn will be subjected to cold stratification over the winter and are likely to germinate in the spring. This is in contrast to the findings at Patrington where most B.rapa germinated in the autumn at the same time as the crop. Late germinating plants were at a disadvantage due to light competition from the oilseed rape canopy.
B.rapa seeds are more likely to germinate at the same time as the crop seeds when the soil is fertile and cultivated, and crop cover is sparse and there would be advantages to this. However there may also be selection for plants in the B.rapa population to germinate later than the crop if herbicide is applied to the crop soon after emergence. Although conventional selective herbicides such as Laser and Butisan are not fully effective against weeds such as B.rapa, treated plants will nevertheless be affected in their development and may well be stunted. In an agricultural environment, weedy B.rapa is subjected to frequent nutrient inputs. This may result in seeds becoming less responsive to changes in nutrient levels as a germination cue. In non-agricultural environments, however, B.rapa will be more likely to germinate after nutrient flushes because these do not occur as regularly as in agricultural fields. Far-red shade caused by the presence of a canopy of established plants can suppress germination of B.rapa and other species (Gorski 1975; Rees & Brown 1991). However this sensitivity to the presence of established plants may be over-ridden when nutrient levels become ideal for germination in non-agricultural environments.
B.napus seed at harvest does not show tendencies toward dormancy and its seeds will germinate under most conditions (see Chapter 6). At harvest, however, certain environmental conditions may induce secondary dormancy. B.napus seeds regularly survive burial in the soil for two to three years (Lutman and Lopez-Granados, 1998) and sometimes even up to ten years, depending on the agricultural practice and whether they remain undisturbed. B.napus seeds tend to germinate after being brought to the surface by cultivation, but will remain dormant and viable if left buried.
The results of seed germination from bait plants (Table 10.6) showed variable dormancy since the seed from some plants responded well to the addition of gibberellic acid. However, these seed progenies would have been a mixture of hybrids with some B.rapa so dormancy would be expected to be variable.
Hybrids between B.napus and B.rapa have been shown to inherit the dormancy characteristics of the maternal parent. This could be due to seed-coat influences that are maternally inherited, but organelle inheritance and endosperm formation are also unequally inherited. However, there may also be some zygotic influence on dormancy (Adler et. al., 1993) as the hybrid dormancy characteristics are closer to those of the maternal parent, but are not identical to it. Adler et. al., (1993) reported that the hybrid produced when B.napus is the maternal parent is slightly less dormant than the hybrid produced when B.napus is the pollen donor. The parentage of the hybrids found in the field at Patrington was unknown. The lower degree of dormancy of hybrid seeds compared to the weedy B.rapa parent could reduce the number of hybrid plants reaching the adult stage as they will predominantly be shed in disturbed habitats where the crop parent is found. A larger proportion of hybrid seeds will germinate immediately compared with the staggered germination pattern of B.rapa, which allows it to germinate in years when conditions are optimum for its growth. Some degree of dormancy in interspecific crop/ wild hybrids will allow alleles to persist over time, therefore affecting the genetic structure of the population. This could allow repeated backcrossing to wild relatives many years after the initial introduction of a transgene (Linder and Schmitt, 1994; Linder and Schmitt, 1995).
If the hybrid seeds germinate immediately after harvest, as do a large proportion of shed B.napus seeds, then the normal agricultural practice of leaving seeds to chit for a minimum of two weeks before cultivation will ensure that the majority of hybrid plants do not survive. The uniformity of their germination requirement could reduce their fitness under field conditions. Alternatively, interspecific hybrids growing in agricultural conditions may maintain similar growth stages to the crop plant, thereby flowering and producing seed at similar times. F1 interspecific hybrids vary considerably in their fitnesses, depending on parental genotypes (Hauser et al 1998). Some varieties of B.napus seem to produce hybrids more readily with particular B.rapa genotypes and the results of the bait plant experiment show that some B.rapa genotypes do not produce hybrids even when exposed to massive amounts of B.napus pollen.
For long-term survival, hybrid seeds must remain dormant until suitable conditions for B.rapa and its hybrids occur in the agricultural field when B.napus is grown. In normal agricultural rotations in the UK, this is likely to be every 4-5 years. Hybrid plants emerging in interim years will usually be growing amongst cereal crops and exposed to herbicides used to eradicate them. Thus few hybrids will reach maturity in years when cereals are grown. Because hybrids are thought to have dormancy characteristics somewhere between the two parental species, they are more likely to emerge when the soil is disturbed for cropping than are plants of B.rapa. Seeds germinated from the soil seed bank (Figure 10.6) show that B.napus tends to germinate in a flush after the input of seeds from a crop. However, numbers of seeds in the soil then reduce dramatically in the following years until the next crop of oilseed rape. Numbers of B.rapa and hybrids, however, remained fairly constant over the sampling period, and no rapid reduction of their seeds in the seedbank was observed. Hence the proportion of both hybrids and B.rapa in the total Brassica seed bank population increased with time.
The first backcross generation to B.rapa produces seeds with a more rapa-like germination pattern. Amongst the seed produced there will be differing germination requirements and some seeds will exhibit dormancy. Interspecific gene flow in the first backcross generation would not therefore be limited by seed germination characteristics to the same degree as that in F1 hybrids.
In a study where reciprocal crosses were made (Hauser et al 1998) those with B.napus as the female parent were found to produce more viable seeds than when B.rapa was the female parent. This is contrary to the results found when 7512 B.napus seeds from plants at Patrington were screened for hybridity. No hybrids were identified amongst these B.napus parents growing adjacent to B.rapa. It is a possibility that hybrids with morphology close to oilseed rape may not have been identified, although, from previous experience of hybrids, almost all are identifiable with a trained eye as having some intermediate traits.
The triploid (AAC) F1 hybrid with 2n = 29 often shows sterility or reduced fertility. Interestingly in the bait plant experiment, it was found that hybrids formed from unreduced gametes were a common occurrence. These had the genomic constitution AAAC (2n = 39). A selection of plants with DNA ratios between 0,36 and 0,39 were found which is higher than expected for a triploid hybrid. Many of these expressed tolerance to one or other herbicide, showing that they were pollinated by B.napus and were therefore a form of hybrid. Some plants seemed more likely to produce this type of hybrid than others did, implying a genotypic tendency. These plants were not grown to maturity therefore there is no information about the comparative fertility of the two types of hybrid. It is possible though that some AAAC hybrids exist in the field as they appear to be commonly present as seed. However they would be difficult to distinguish by either flow cytometry, molecular markers or morphology from backcrosses to B.napus. Chromosome analysis would be required to assess genomic constitution. Plants that were identified as backcrosses to B.napus by flow cytometry, and then analysed by AFLP, group together with the F1 hybrids, (Figures 10.4 and 10.5). This may indicate that these arose from F1 hybrids with unreduced gametes.
Although B.rapa has always been thought to be a highly self-incompatible species, results from the bait plant experiment (Table 10.7), suggest that this may not be entirely the case. Under conditions of isolation from other B.rapa plants, some B.rapa individuals produced a small number of progeny that were B.rapa. The most likely cause of this would be selfing, as there were no other plants of the same species within 15 m and the plants were subjected to a massive excess of oilseed rape pollen.
Some plants in the bait plant experiment produced no seed indicating a genetic incompatibility between some B.rapa plants and oilseed rape. All the plants flowered and set seed under the same conditions, so environmental influences seem unlikely. A high proportion of the hybrids produced from all B.rapa bait plants growing between glufosinate and glyphosate plots were not tolerant to either herbicide. This indicates that the transgene, if present, was not being expressed in these plants. It is known that expression can be lost under certain conditions of stress in oilseed rape (see Chapter 6). Perhaps this is more likely to occur in the genomic background of the hybrid due to its genetic constitution. A PCR test on these susceptible plants would have confirmed the presence or absence of the transgene. An alternative explanation for the susceptible hybrids is that they were fertilised by non-GM oilseed rape growing in a nearby plot. This is unlikely, given the pollen pressure from the two GM plots adjacent to these B.rapa plants.
F1 hybrids were found to be fertile under field conditions, although seed production was low compared to either parent species. Thus, F1 hybrids may provide an avenue for introgression to take place. Pollen production from hybrids was not measured but is probably low due to the small anthers observed. The growth of new flowering shoots on hybrid plants after pod formation also indicates low fertility. Sterile hybrids, or those with low fertility, often produce many flowers throughout the life cycle (Stace, 1974). Normally annual plants die after seed production, as energy is diverted from other parts of the plant to the seeds. In hybrids that are sterile or with low fertility, few seeds are produced so energy resources are directed to continuing flower production to compensate. Many of the seeds produced on F1 hybrid plants were inviable, had shrivelled or had germinated inside the pods. Hauser and Ostergard (1999) also reported germination of F1 and F2 seeds within pods, which may be an important loss of hybrid seeds. Hauser et. al. (1998) examined fitness of F1 hybrids between weedy B.rapa and B.napus and remarkably found that seed set in hybrids was intermediate between its two parents, with B.rapa producing fewer seeds per plant than F1 hybrids. This result is very different from that observed in the Patrington population where hybrids produced many fewer seeds per plant than B.rapa.
Backcrosses to F1 hybrids from oilseed rape plants were also found in the field but were less easy than the F1 hybrids to identify by morphology alone, as they appear very similar to oilseed rape. Backcrossing to B.rapa was likely to occur only in areas of the field where there were F1 hybrids and densely populated by B.rapa. Isolated hybrids from B.rapa were most likely to backcross to B.napus but the potential for self- pollination may exist.
Flow cytometry testing of twelve susceptible offspring from a hybrid parent showed a remarkable range of DNA ratios. Flowers on the hybrid parent could have been pollinated by B.napus, B.rapa, another hybrid or have been selfed. Offspring with DNA ratios between B.rapa and the hybrid were likely to have been pollinated by B.rapa - the first backcross generation. Seven plants fell into this category. One plant had a DNA ratio indistinguishable from B.rapa and in a field situation would be impossible to identify as a backcross plant due to its B.rapa morphology.
Two plants grown from the soil samples at the Patrington site, were shown by DNA ratio analysis to be backcrosses to B.rapa. These backcrosses appeared to be infrequent but were nevertheless possible, as shown by the putative backcrosses to B.rapa identified in the hybrid progeny discussed above. However, plants with intermediate DNA levels were not found growing in the field, which indicates that aneuploids do not generally survive to produce viable plants under natural conditions.
Backcrosses to B.rapa are likely to have the same, or very similar, chromosome numbers as B.rapa, so these may not have been identified by flow cytometry or by morphology. It is likely that individuals with a chromosome number close to B.rapa will be the most stable of the range of chromosome complements produced from backcrossing hybrids with B.rapa. They may occur at a higher frequency than other aneuploids due to selection taking place under field conditions. Hauser et. al. (1998) concluded that backcross and F2 plants are often aneuploid with unbalanced C chromosomes (backcrosses: AA + 0-9 C; F2 : AA + 0-18 C) and this may seriously reduce their fitnesses. Translocations between the A and C chromosomes of B.napus may also reduce fitness, because some backcross and F2 plants may then inherit incomplete sets of genes from the original B.napus A-genome.
The proposed reduced fitness of backcross plants is supported by the fact that no backcrosses to B.rapa individuals with DNA amounts between B.rapa and hybrid were found growing in the field, but when progeny from known hybrids were grown under glasshouse conditions several were identified. For this reason, numbers of backcrosses to B.rapa plants identified in the field may be underestimated, as those with very similar DNA amount to B.rapa were indistinguishable from B.rapa. It was difficult to distinguish between plants that were truly B.rapa and those that may have had some of the B.napus genome introgressed into them by using morphology and flow cytometry as means of identification.
This conclusion is supported by the AFLP results. Some plants were identified by morphology and flow cytometry as being B.rapa, but cluster together with F1 hybrids. This suggested that these plants were either first generation backcrosses to B.rapa, or later generation introgressed individuals. A larger sample size was needed to make firm conclusions about introgression taking place in this population. However the evidence here from AFLP, hybrid progeny testing and field-testing pointed towards the presence of an introgressing population.
A further problem in attempting to draw conclusions about introgression arises from the original parentage of B.napus itself. B.napus (2n = 38, AACC) is an amphidiploid derived from B.rapa (2n = 20, AA) and B.oleracea (2n = 18, CC) and therefore evidence of introgression from B.napus into B.rapa would have to involve the C genome for it to be readily detected.
The results of the field herbicide spot testing and the soil sampling showed that the population found growing each year was only a snapshot of the true population in the seed bank. Only a proportion of the F1 hybrids found were herbicide tolerant, suggesting that tolerant hybrids could still remain in the seed bank. Some of the non-tolerant hybrids may have arisen from hybridisation in previous oilseed rape crops several years before. This means that introgression may have occurred over a long period of time as oilseed rape crops are generally only grown on the same field every three or four years. Hybridisation and backcrossing between the two species can only take place in the years when oilseed rape is grown, as broad leaved weeds such as B.rapa are easily killed by the herbicides applied to cereal crops. Thus the long-term consequences of introducing a transgene into a weedy B.rapa population are as yet unknown, but these studies suggest that introgression could occur.
Familiarity with crop/ weed interactions has enabled predictions to be made about the fate of the herbicide-tolerance transgene under field conditions. Herbicide-tolerant hybrids have already been recorded post-harvest at the Patrington site, after a single GM crop. However, further data could be gathered when the next oilseed rape crop is grown and the opportunity will then arise to study the consequences of any hybridisation. Will the next stage of introgression (back crossing to the B.rapa parent) take place? The work discussed here took place mostly over one growing season and the results have given an insight into the complex relationships between the B.rapa weed population and the oilseed rape crop. The methods used have not been powerful enough to be able to distinguish with certainty between some backcross plants and F1 hybrids. At the Patrington site, all possible combinations of crosses probably occurred in the field, whether at the seed or the mature plant stage. It has been impossible to distinguish the genomic make up and parental origin of each individual without detailed molecular analyses and chromosome study.
The offspring from any of these F1 and backcross hybrids could potentially cross with any of the others: the number of combinations of possible crosses is thus enormous. The results of this study have categorised the plants into major classes only: the two parental species, F1 hybrids and backcrosses. What is more important than the exact identification of each cross, perhaps, is the discovery of extensive hybridisation here, and the evidence of backcrossing in both directions. Thus a transgene is likely to persist past the initial F1 hybridisation and will move into further generations in an unpredictable way. In addition, the way the transgene will move between generations may also be determined by the fitness it confers on the plant.
In an agricultural environment, the nature and extent of introgression will depend very much upon the type of cropping and the crop management. If weed management of cruciferous weeds is effective, there may be a higher potential for hybridisation between B.rapa and oilseed rape, since individual B.rapa plants may escape herbicide treatment and thus become isolated within the crop. If weed management is poor, more B.rapa plants will be left in the field, and perhaps there will be less hybridisation. For high levels of backcrossing to occur the opposite scenario will apply. Backcrossing to B.rapa is more likely when hybrids are present in the field and there is an abundance of B.rapa plants (when weed management is poor).
The results presented here have shown that introgression between B.napus and B.rapa has been occurring at the Patrington site for many years. The introduction of a transgenic oilseed rape to the site has allowed the extent of initial hybridisation to be determined, and will enable further studies to be carried out on backcrossing events in the future. However these data are from only one population of B.rapa and occupy only two years so cannot be deemed to be representative of every population. Further sites similar to this one at Patrington need to be studied over several years in order to gain more information about the nature and extent of introgression of oilseed rape genes into B.rapa.
11. CONCLUSIONS AND GENERAL DISCUSSION
11.1 CONCLUSIONS
The conclusions from this study can be summarised as follows:
• Cross-pollination between areas of oilseed rape
In fully fertile varieties outcrossing declined rapidly with distance from the source with the majority occurring within the first ten metres. However cross-pollination levels in excess of 0.5% were found at distances of 100 ? 200m in some samples so that further studies are needed of the factors influencing cross-pollination in order to accurately predict cross-pollination between neighbouring crops.
The levels of outcrossing found in samples from varietal associations (VAs) were considerably higher than those found in samples of fully fertile rape. This reinforces the requirement for greater isolation distances between GM releases and VAs. Any varieties of oilseed rape containing a male sterile component are more likely to outcross than fully fertile varieties. Pollen-mediated gene flow was also observed between oilseed rape and turnip rape at similar levels. • Cross-pollination of volunteer plants in adjacent fields
Pollination of volunteers in neighbouring fields by GM oilseed rape crops occurred at very low rates due to the volunteers being fully fertile and competition from selfing and other pollen. No volunteers with reduced male fertility were studied as outcrossing would be expected to be higher in these plants. (Simpson et al 1999).
• Feral populations of oilseed rape
Feral populations of oilseed rape were rarely observed near GM release sites in this study, and were transient, not usually persisting for more than one growing season. Observations made on individual plants growing in roadsides and verges showed that only low proportions survived to maturity and set seed. There was no indication that herbicide tolerant transgenic oilseed rape would survive any better outside of cultivation since the herbicides are unlikely to be used in these environments and thus the GM plants would have no additional competitive advantage. High laurate oilseed rape does not appear to be better adapted for survival as a weed or feral plant than conventional oilseed rape, although more data is needed for this to be conclusive.
• Gene flow into wild crucifers
Natural hybridisation under field conditions was not observed between GM oilseed rape and S.arvensis, S.alba or R.raphanistrum. If a rare hybridisation event were to take place, further introgression would be unlikely due to the reported lack of fitness in the hybrids and back crosses. Information on R.raphanistrum was scarce in this study and, as this is the most likely of the three species to hybridise with oilseed rape, more monitoring is necessary to make firm conclusions on the likelihood of hybridisation in other situations. Ongoing studies in France suggest that hybridisation and backcrossing occurs but that cytoplasmic incompatibility markedly reduces the vigour and viability of BC5 and BC6 and they become infertile ( Anne-Marie Chevre, pers comm, 2001).
Extensive hybridisation was observed when oilseed rape and B.rapa grew together in fields. Hybrids were fertile although seed production was low. Evidence of backcrossing in the direction of both parents was found, although backcross plants of an unstable chromosome complement appeared not to survive well in the field. Some evidence of possible past introgression was found from AFLP analyses. Therefore it is possible that introgression of a transgene from oilseed rape into weedy B.rapa could occur, a result which confirms the work of Hansen et. al. (2001), who found oilseed rape markers occurring at different frequencies in populations of B.rapa. Only one site was used to study hybridisation and introgression between oilseed rape and B.rapa in this project, therefore more information is needed from other sites and other B.rapa populations. The population of weedy B.rapa growing in oilseed rape fields at Patrington showed:
• Extensive hybridisation occurs between oilseed rape and weedy B.rapa in the field; • Hybrids seeds were found on B.rapa mother plants, but not on B.napus;. • Some evidence of an introgressing population between B.napus and B.rapa was detected from AFLP analyses.
• Genetic incompatibility exists between some B.rapa individuals and oilseed rape. • Hybrids were fertile under field conditions, despite their triploid nature, although seed production was low compared to the parents.
• Interspecific hybrids backcrossed with both B.napus and B.rapa; so that the progeny from a hybrid mother plant had a range of DNA amounts.
• Individuals backcrossed to B.rapa had DNA amounts similar to B.rapa and could not be distinguished by a combination of morphology and flow cytometry. Chromosome analyses and DNA studies are required to monitor the backcrossing.
• Aneuploids were found in soil seedbank samples but were not found growing in the field. • Hybrids formed from unreduced gametes were relatively common. • Hybridisation occurred between oilseed rape and turnip rape.
In the recently commissioned BBSRC/NERC Gene Flow Programme, NIAB, together with Reading University, CEH, HRI and other partners, is conducting a detailed study of gene flow from B napus to B rapa and B oleracea. This study is addressing some of the issues raised by this report and is hoping to gain more information on historical gene introgression from oilseed rape to these species using variety specific markers.
Weediness and persistence
In studies of volunteers of GM rape, weediness and persistence of did not appear to be enhanced by the presence of herbicide-tolerance transgenes or a high laurate transgene. This confirms the findings of Simpson and Sweet (2001) and Booth et al (1995).
11.2 GENERAL DISCUSSION Crop-to-crop gene flow: Pollen
Although this study has given data on various components of gene flow in oilseed rape, it has also raised new questions about the extent of gene flow that potentially may occur in the agricultural environment. Many of the studies reported here gave the first opportunity in the UK to carry out gene flow studies in ?real? agricultural situations using transgenic markers to estimate the extent of gene flow. However if transgenic oilseed rape is grown on a large scale in the UK, then gene flow will occur between fields, farms and across landscapes.
In the large scale gene flow studies between blocks of oilseed rape, such as those carried out on the FSE trials in Lincolnshire and Oxfordshire and the VA variety Gemini at NIAB, it was possible to demonstrate certain features of gene flow between varieties. However, quantification of the extent of gene flow remains problematic for a number of reasons, as described in Chapter 7 . The current FSE trials were established to investigate the effects on diversity of wildlife of the herbicides used in the herbicide-tolerant system of oilseed rape cultivation. They were not designed to study pollen related gene flow though they are contributing some useful information and will give an indication of gene flow on a large scale. Gene flow at this level should be investigated on a landscape scale using larger numbers of transgenic pollen sources, and examining different genotypes, (both of the transgenic plants and the conventional varieties), the extent of pollen flow at further distances from sources, a range of environmental conditions, geographical locations and patterns of cropping of GM and non-GM crops. It is only when these studies have been conducted under a range of UK conditions that farmers and seed producers will be able to accurately predict outcrossing levels and develop appropriate strategies for managing it.
Crop-to-crop gene flow: Seed
The persistence of GM oilseed rape volunteers reported here and observed in other studies (Simpson and Sweet, 2001, BRIGHT, Genesys etc.. ) indicate that they are likely to be able to persist for as long as conventional oilseed rape. However, to give more definitive information on specific GM types, further long term experiments are needed to study survival of seed banks of GM volunteers over years in a range of crops and rotational circumstances, and in different soil types and conditions. In addition, this study, Simpson and Sweet (2001) and the Genesys study in France, have shown that seed from GM rape is likely to become dispersed by normal agricultural practices so that it will become widely distributed on farms growing GM rape, on farms sharing equipment with GM rape growers, contractors machinery, bulk transporters etc.
It is likely that the EU will set standards of 99.7% purity for non-GM certified seed of oilseed rape. Tests of certified seed of a particular variety imported from N America since 1996, conducted by NIAB, detected GM contamination in c 40% of samples ranging from 0.05% ? 0.5%. It is anticipated that contamination of this type will become more frequent as GM rape is commercialised globally.
Simpson and Sweet (2001) showed that one volunteer per square metre (which is approximately 1 plant in 100 depending on sowing rate), whether arising from seed or soil contamination, will result in crop contamination levels of between 0.6% in conventional varieties to 1.5 % in varietal associations. Thus even low levels of seed or seed bed contamination in some varieties will result in the establishment and persistance of contamination in fields and subsequent crops. However this study showed that there was a decline in glufosinate tolerance in subsequent volunteer populations, probably due to the segregation of the Bar gene in subsequent populations. Therefore it is not clear how transgenes will persist in subsequent generations of volunteers and feral plants.
As the commercialisation of GM oilseed rape proceeds, farmers growing non-GM rape will face contamination from several sources and will need to be able to accurately predict likely contamination rates and implement a range of practices in order keep below thresholds for GM contamination. Seed and seed bed contamination is likely to be just as important as the more widely discussed and publicised contamination from outcrossing. Interspecific Gene Flow:
Further investigations are needed to determine the extent of spontaneous hybridisation between oilseed rape and certain wild crucifers and the production of backcrossed and introgressed populations. This study, plus those in Denmark ( Hansen et al 2001) have shown that hybridisation with B.rapa and survival of hybrids occurs, though the extent in the UK is not yet known. Hybridisation with R.raphanistrum has been reported in France (Darmency et. al., 1998) though recent research suggests that introgression is restricted by genetic/cytoplasmic incompatability (Chevre , pers comm). In this study R raphanistrum only occurred at a few sites and thus there were limited opportunities for hybridisation to occur. Further studies of hybridisation between R.raphanistrum populations and B.napus under UK environmental conditions are needed to determine the likelyhood of gene flow into this species.
This study is significant in that it revealed for the first time that there are weedy populations of B rapa co- existing and hybridising with oilseed rape in England, in situations similar to those reported in Denmark. It is likely that these populations will readily acquire genes from oilseed rape particularly if they enhance the survival or fitness characteristics of the B rapa growing as a weed in oilseed rape crops. These studies have revealed the evolutionary interaction between tetraploid B.napus and one of its diploid parents, B.rapa. This could serve as a model system in clarifying our understanding of the evolution and maintenance of polyploidy.
Further studies are needed to determine the extent of gene introgression from rape into weedy B.rapa. Molecular investigations of the historical introgression of B.rapa and oilseed rape are required and AFLP analyses on a further range of populations of B.rapa in other parts of the country are required to assess the extent of gene exchange between these species. The development and use of micro-satellites (simple sequence repeats) will allow gene flow and paternity to be studied by the use of co-dominant markers for clearly definable alleles. Chromosome analysis is also required to confirm the status and meiotic behaviour of backcross and hybrid individuals identified by flow cytometry. The fate of the herbicide-tolerance transgene in the B.rapa population at Patrington is intriguing. This marker will allow the extent of backcrossing to either parental species in following generations to be assessed.
BBSRC and NERC research projects started in 2001 are investigating the rate, extent and consequences of gene flow into B.rapa, by studying historical introgression with markers from conventional rape varieites and using herbicide tolerance, insect resistance and disease resistance as model systems.
12. ACKNOWLEDGEMENTS
The authors acknowledge the assistance and support of Rachel Shepperson, Euan Simpson, John Law, Adam Cockley, Emma Singer, Katie Eastham and many other staff at NIAB. We acknowledge the materials, technical information and access to trials granted by PGS/Aventis, Monsanto and John K King and Sons Ltd. We also thank the many farmers who hosted GM OSR trials between 1995 and 2000 for their considerable support and co-operation in allowing us access to their sites. We particularly thank
13. BIBLIOGRAPHY
Brian Beeney for help with the studies of gene flow in wild turnip (Brassica rapa). Finally we thank DETR (now DEFRA) for their financial support and the inputs from several of their staff involved in this project . 13. BIBLIOGRAPHY
Adler L.S., Wikler K., Wyndham F.S., Linder C.R., & Schmitt J. (1993) Potential for persistence of genes escaped from canola: germination cues in crop, wild, and crop-wild hybrid Brassica rapa. Functional Ecology 7: 736-745
Al-Kaff, N.S., Covey, S.N., Kreike, M.M., Page, A.M., Pinder, R. and Dale, P.J. (1998) Transcriptional and Post-transcriptional plant gene Silencing in Response to a Pathogen. Science, Vol 279: 2113-2115
Becker, H.C., Damgaard, C. and Karlsson, B. (1992) Environmental variation for outcrossing rate in rapeseed (Brassica napus). Theor. Appl. Genet. 84: 303-306
Booth E. J., Walker K. C., Whytock G.P., Sovero M. (1996) Assessment of the ecological consequences of introducing transgenic rapeseed. In: 4th ESA Congress, Abstracts pp 144-145.
Brown J. Brown A.P Davis J. B. Erickson D. (1997) Intergeneric hybridisation between Sinapis alba and Brassica napus. Euphytica 93: 163-168.
Champolivier, J., Gasquez, J. and Messean, A. (1999) Crop management of transgenic rapeseed: risk assessment of gene flow. GCIRC 10th International Rapeseed Congress, Canberra, Australia: 6
Charters, Y.M., Robertson, A., O?Brien, E.T. & Squire, G.R. (1996) Investigation of Feral Oilseed Rape Populations. Department of the Environment Contract PECD 7/8/237 Final report
Chevre, A.M., Eber, F., Baranger, A. & Renard, M. (1998) Risk assessment on crucifer species. Proc. Int. Symp. on Brassicas. Acta Hort. 459: 219-224
Chevre, A.M., Eber, F., Renard, M. and Darmency, H. (1999) Gene flow from oilseed rape to weeds. In: Gene Flow and Agriculture, Relevance for Transgenic Crops. Proceedings of BCPC Symposium, University of Keele. 125-130
Child, R.D. and Evans, D.E. (1989) Improvement of recoverable yields in oilseed rape (Brassica napus) with growth retardants Aspects of Biology 23:135-143
Crawley, M.J., & Brown, S.L., (1995) Seed limitation and the dynamics of feral oilseed rape on the M25 motorway. Proc. R. Soc. Lond. B. 49-54
Crawley, M. J., Brown, S. L., Hails, R. S., Kohn, D. D. and Rees, M. (2001) Transgenic crops in natural habitats. Nature 409: 682 - 683
Dale, P.J,, (1992) Spread of engineered genes to wild relatives. Plant Physiol 100: 13-15
Darmency H. Fleury A. Lefol E. (1995) Effect of transgenic release on weed biodiversity: oilseed rape and wild radish. Brighton Crop Protection Conference 1995 (5A-2):433-438
Darmency, H., Lefol, E. and Fleury, A. (1998) Spontaneous hybridisations between oilseed rape and wild radish. Molecular Ecology 7: 1467-1473
De Block, M., Botterman, J., Vandewiele M., Dockx T., Gossele V., Movva N.R., Thompson C., Van Montagu M., Leemans J. (1987) Engineering herbicide resistance in plants by expression of a detoxifying enzyme. EMBO Journal 6: 2513-2518
Delourme, R., Eber, F and Renard, M. (1995) Breeding double low restorer lines in radish cytoplasmic male sterility of rapeseed (Brassica napus L.). Proceedings of 9th International Rapeseed Congress. Cambridge, UK 1: 6-8
Delourme, R., Foisset, N., Horvais, R., Barret, P., Champagne, G., Cheung, W.Y., Landry, B.S. and Renard, M. (1998) Characterisation of the radish introgression carrying the Rfo restorer gene for the Ogu-INRA cytoplasmic male sterility in rapeseed (Brassica napus L.) Theoretical Applied Genetics 97: 129-134
Eisikowitch, D. (1980) Some aspects of pollination of oilseed rape (Brassica napus L.) J. Agric. Sci. Camb. 96: 321-326
Ellstrand, N., and Hoffman, C. (1990) Hybridization as an avenue of escape for engineered genes. Bioscience 40: 438-42
Free, J.B. and Nutall, P.M. (1968) The pollination of oilseed rape (Brassica napus) and the behaviour of bees on the crop. J. Agric. Sci. Camb. 71: 91-94
Gorski T., (1975) Germination of seeds in the shadow of other plants. Physiologia Plantarum 34: 342- 346
Hansen, L.B., Siegismund, H.R., R.B. Jørgensen (2001). Introgression between oilseed rape (Brassica napus L.) and its weedy relative B. rapa L. in a natural population. Genetic Resources and crop Evolution 48 (6): 621-627.
Hauser T.P. and Ostergard H. (1999) Precocious germination of Brassica rapa x Brassica napus seeds within pods. Hereditas, 130:89-93
Hauser, T.P., Jørgensen, R.B. & Ostergard H. (1998) Fitness of F1 hybrids between weedy Brassica rapa and oilseed rape (B.napus). Heredity 81: 429-435
Ingram, J. (2000) The separation distances required to ensure cross-pollination is below specified limits in non-seed crops of sugar beet, maize and oilseed rape. Plant varieties and seeds. 13: 181-199
Jorgensen R.B., & Andersen B. (1994) Spontaneous hybridisation between oilseed rape (Brassica napus) and weedy campestris (Brassicaceae): a risk of growing genetically modified oilseed rape. Amer. J. Bot. 81 (12): 1620-1626
Jorgensen R.B., Andersen B., Landbo L., Mikkelson T.R. (1996) Spontaneous hybridisation between oilseed rape (Brassica napus) and weedy relatives. Acta Hort 407: 193-199
Krishna, P. J. (2001) Gene Silencing. Environmental Influences on Gene Silencing. [Online] 09/01/02 <http://www.biotech-info.net/influences.html>
Landbo, L., Andersen, B., Jorgensen, R.B. (1996) Natural hybridisation between oilseed rape and a wild relative: hybrids among seeds from weedy B.campestris. Hereditas 125: 89-91
Lavigne, C., Klein, E.K., Vallee, P., Pierre, J., Godelle, B., Renard, M. (1997) A pollen-dispersal experiment with transgenic oilseed rape. Estimation of the average pollen dispersal of an individual plant within a field. Theor. Appl. Genet. 96: 886-896
Lefol, E., Danielou, H., Darmency, H. (1996) Predicting hybridization between transgenic oilseed rape and wild mustard. Field Crops Research 45: 153-161
Linder, C.R., and Schmitt, J. (1994) Assessing the risks of transgene escape through time and crop- wild hybrid persistence. Molecular Ecology 3: 23-30
Linder, C.R., and Schmitt, J. (1995) Potential persistence of escaped transgenes: Performance of transgenic, oil-modified Brassica seeds and seedlings. Ecological Applications. 5(4): 1056-1068
Lutman, P.J.W., (1993) The occurrence and persistence of volunteer oilseed rape (Brassica napus). Aspects of Applied Biology 35: Volunteer Crops as Weeds: 29-36
Lutman, P.J.W and Lopez Granados, F. (1998) The persistence of seeds of oilseed rape (Brassica napus). Apsects of Applied Biology 51, 147-152
Manasse, R. and Karieva, P. (1991) Quantifying the spread of recombinant genes and organisms. In: Assessing Ecological Risks of Biotechnology. Ed L. Ginzburgh. 215-231
Mathias, R. (1991) Improved embryo rescue technique for intergeneric hybridization between Sinapis species and Brassica napus. Cruciferae Newsletter, 14-15: 90-91
McCartney, H.A., and Lacey, M.E. (1990) Wind dispersal of pollen from crops of oilseed rape (Brassica napus L.) Journal of Aerosol Science 22: 467-477
McNeilly, T. (1968) Evolution in closely adjacent plant populations. III. Agrostis tenius on a small copper mine. Heredity 23: 99-108
Metz, P.L.J., Jacobsen, E., Nap, J.P., Pereira, A., Stiekema, W.J. (1997) The impact on biosafety of the phosphinothricin-tolerance transgene in inter-specific B.rapa x B.napus hybrids and their successive backcrosses. Theor Appl Genet 95: 442-450
Morris, W. F. Kareiva, P. M. Raymer, P. L. (1994) Do barren zones and pollen traps reduce gene escape from transgenic crops? Ecological Applications 4 (1): 157-165
Moyes,C., Lilley , J., Casais C., Dale, P.J. (1999) Gene flow from oilseed rape to Sinapis arvensis: variation at the population level. In: Gene Flow and Agriculture, Relevance for Transgenic Crops. Proceedings of BCPC Symposium, University of Keele. 143-148
NIAB. (2000) Oilseeds variety handbook. Supported by HGCA Research & Development: 22
Oilseed Rape Manual (1989) Oilseed rape growth stages. Agricultural Division Advisory Service. 1-13
Pekrun, C. and Lopez-Granados, F. (1995) The effect of water stress and light conditions on the induction of secondary dormancy in seeds of Brassica napus L. In: 9th International Rapeseed Congress. Crop Protection: Forecasting and Chemical Control, pp. 1052-1054.
Pekrun, C. and Lutman, P.J.W. (1996) Population dynamics of volunteer rape and possible means of control. Second International Weed Control Congress, Copenhagen: 1223-1227
Pekrun, C., Lopez-Granados, F. and Lutman, P. J. W. (1997)a Studies on the persistence of rape seeds (Brassica napus L.), emphasizing their response to light. In: Basic and Applied Aspects of Seed Biology. Eds: R.H. Ellis, M. Black, A.J. Murdoch and T.D. Hong pp. 339-347.
Pekrun, C., Lutman, P.J.W. and Baemer K. (1997)b Germination behaviour of persisting oilseed rape seeds in relation to temperature. Weed Research 37: 419-431
Pekrun, C., Potter, T. C. and Lutman, P. J. W. (1997)c Genotypic variation in the development of secondary dormancy in oilseed rape and its impact on the persistence of volunteer rape. BCPC ? Weeds, 4A-2, 243-248.
Pekrun C., Hewitt J.D.J., Lutman P.J.W. (1998) Cultural control of volunteer oilseed rape (Brassica napus). Journal of Agricultural Science Cambridge 130:155-163.
Pekrun, C. and Lutman, P.J.W., (1998) The influence of post-harvest cultivation on the persistence of volunteer oilseed rape. Aspects of Applied Biology 51: 113-118
Prakash, S. and Hinata, K. (1980) Taxonomy, cytogenetics and origin of crop Brassicas, a review. Opera. Bot. 55: 1-157
Ramsay, G., Thomson, C.E., Neilson, S. and Mackay, G.R. (1999) Honeybees as vectors of GM oilseed rape pollen. In: Gene Flow and Agriculture, Relevance for Transgenic Crops. Proceedings of BCPC Symposium, University of Keele. 209-214
Ratcliff, F., Harrison, B.D., and Baulcombe, D.C. (1997) A similarity between viral defence and gene silencing in plants. Science, Vol 276: 1558-1560
Raybould, A.F., and Gray, A.J., (1993) Genetically modified crops and hybridisation with wild relatives: a UK perspective. Journal of Applied Ecology 30: 199-219
Rees M. and Brown, V.K. (1991) The effects of established plants on recruitment in the annual forb, Sinapis arvensis. Oecologia 87:58-62
Scheffler, J.A., Parkinson, R. and Dale (1993) Frequency and distance of pollen dispersal from transgenic oilseed rape (Brassica napus). Transgenic Research 2: 356-364
Scheffler, J.A., Parkinson, R. and Dale, P.J. (1995) Evaluating the effectiveness of isolation distances for field plots of oilseed rape (Brassica napus) using a herbicide-resistance transgene as a selectable marker. Plant Breeding 114: 317-321
Schlink, S. (1995) Uberdauerungsvermogen und Dormanz von Rapssamen (Brassica napus) im Boden. Proceedings of 9th EWRS Symposium, Budapest. 65-72.
Scott, S.E. and Wilkinson, M.J. (1998) Transgene risk is low. Nature 393:320
Simpson, E., Norris C.E., Law, J.R., Thomas, J.E. and Sweet, J.B. (1999) Gene flow in genetically modified herbicide tolerant oilseed rape (Brassica napus) in the UK In: Gene Flow and Agriculture, Relevance for Transgenic Crops. Proceedings of BCPC Symposium, University of Keele. 72: 75-81
Simpson, E.C. and Sweet, J.B. (2001) Consequence analysis of the impact on agriculture and the environment of the release of herbicide tolerant oilseed rape. MAFF report RG0217
Squire, G.R., Crawford, J.W., Ramsay, G., Thompson, C. and Bown, J. (1999) Gene flow at the landscape level. In: Lutman, P.J.W. Gene flow and Agriculture: Relevance for Transgenic Crops. BCPC Symposium Proceedings no. 72: 57-64
Stace, C.A. (1975) Hybridisation and the Flora of the British Isles. Academic Press London. New York. San Francisco 4; 38
Staniland, B.K., McVetty, P.B.E., Friesen, L.F., Yarrow, S., Freyssinet, G. and Freyssinet, M. (2000) Effectiveness of border areas in confining the spread of transgenic B.napus pollen. Canadian Journal of Plant Science 80: 521-526
Stringham, G.R. and Downey (1982) Effectiveness of isolation distances in seed production of rapeseed (Brassica napus). Agronomical Abstracts. Madison. 136-137
Timmons, A.M., O?Brien, E.T., Charters, Y.M., Dubbels, S.J., & Wilkinson, M.J. (1995) Assessing the risks of wind pollination from fields of genetically modified Brassica napus ssp. Oleifera. Euphytica 85: 417-423
Thompson, C.E., Squire, G., Mackay, Bradshaw, J.E. Crawford, J. and Ramsay, G. (1999) Regional patterns of gene flow and its consequences for GM oilseed rape. In: Gene Flow and Agriculture, Relevance for Transgenic Crops. Proceedings of BCPC Symposium, University of Keele. 72: 95-100.
|
If you have come to this page from an outside location click here to get back to mindfully.org |