Jeremy B Sweet,  Euan C Simpson, Carol E Norris, Jane E Thomas

The National Institute of Agricultural Botany, Huntingdon Road, Cambridge, CB3 OLE, UK



Cross pollination frequencies recorded between winter oilseed rape plots in cross pollination experiments varied according to variety, size of pollinator plots, spatial arrangement and environmental conditions and was measured at distances of up to 400m from a large scale release of herbicide tolerant (HT) rape. The incidence and persistence of  HT rape volunteers was not greater than that of non-HT varieties.


KEY WORDS:          Glufosinate, glyphosate, genetically modified.




The potential for gene transfer to related  Brassica species has been examined at a range of sites across the UK in addition to observations made of the incidence and persistence of volunteers and feral populations following GM releases (Sweet et al. 1997, Sweet & Shepperson 1996, Simpson et al. 1999, Norris et al. 1999).


Gene flow is an important part of evaluating the risks associated with growing GM oilseed rape, as the dispersal of transgenes via pollen could potentially contaminate neighbouring crops, feral or volunteer populations or hybridise with related Brassica species.  Seeds from oilseed rape also contribute to the dispersal and persistence of transgenes.  Sweet et al., (1997) reported that several comparative experiments have provided no conclusive evidence to suggest that transgenic rape is any more persistent or invasive than conventional types. Monitoring large scale releases of GM HT rape grown using normal farm practices is generating further data on the  dispersal, persistence and weediness of GM rape in agricultural environments. This paper discusses some of the  results of initial studies of cross pollination in herbicide tolerant oilseed rape, and on the incidence of oilseed rape volunteers at release sites.




1.                  Gene flow In National List Trials


National List (NL) trials of GM HT winter oilseed rape have contained two GM herbicide tolerant varieties tolerant to the broad spectrum herbicides glufosinate-ammonium and glyphosate as well as five non-tolerant conventional control varieties (Synergy, Express, Nickel, Falcon and Apex). Each GM trial was isolated from other rape crops and trials. Samples of harvested seed were taken from four NL sites in 1997. Samples were selected from plots nearest to the GM trial, 50m, and 100m or the furthest point from the GM trial. Seed samples were also tested from the varieties in GM trials to screen for single and double herbicide tolerance. The collected seed samples were sown in plots in a randomised block design replicated three times with negative controls (conventional winter rape variety - Express). Each trial was treated with either 200 g/l glufosinate-ammonium at 3l/ha or glyphosate 360 g/l at 4l/ha when plants were at the 3-5 leaf stage. The numbers of surviving plants were assessed approximately 7 DAT and 14 DAT for glufosinate and glyphosate treatments respectively. Surviving plants were sprayed on a second occasion with the appropriate herbicide to confirm tolerance.  Seed samples of both GM and conventional varieties from 3 NL GM winter rape trials were also grown in a glasshouse with positive and negative controls. Plants were grown to the two true leaf stage before being treated with either a 1% dilution of 200 g/l glufosinate-ammonium or a 1% dilution of 360 g/l glyphosate. The numbers of surviving plants were assessed approximately 7 DAT and 14 DAT after treatment for glufosinate and glyphosate respectively. Surviving plants were treated as appropriate with either glufosinate or glyphosate to detect double tolerance. Leaf tissue samples were taken from plants expressing double tolerance to confirm the presence of multiple transgenes by PCR using specific primers for the PAT and EPSPS GOX genes conferring glufosinate-ammonium and glyphosate tolerance respectively.


2.         Pollen dispersal from a large area of GM HT winter oilseed rape


The pollen source was a trial of approximately 9 ha of winter oilseed rape with about 65% of plants in this area containing the BAR gene conferring tolerance to the herbicide glufosinate ammonium. The male sterile portion of the spring oilseed rape composite hybrid variety ‘Concept’ was grown under glasshouse conditions so that flowering coincided with the onset of flowering of the pollen crop. Six male sterile plants were positioned (approximately 0.5m apart) in linear plots at a range of distances (100m, 200m, 400m) and directions north, south, east and west from the pollen source. Plants were left in position for the duration of the flowering of the crop and were removed after the main flowering period (approximately 4 weeks). Seeds were collected and tested under glasshouse conditions for glufosinate tolerance as described previously.


3.         Monitoring volunteer incidence


Post release monitoring of NL GM oilseed rape trial sites was conducted on two occasions, the first in winter (November) and a second in spring (April) to determine volunteer survival. Numbers of volunteers per m2 in a 50m radius around the GM trial site were assessed.


Trials of GM herbicide tolerant oilseed have been conducted on a farm in Cambridgshire since 1995. There are now four individual field sites where trials have been grown, with a fifth trial sown in Autumn 1998. The incidence of volunteers and feral rape plants in these fields and in the surrounding field margins and roadsides has been recorded since 1995. Volunteers were counted in the cereal crops following the various releases and where possible after the harvest of following cereal crops. Volunteer and feral plants were tested for herbicide tolerance using a non-destructive assay disk test (Sweet et al., 1997).




1. Gene flow between GM NL and NL trials


Herbicide tolerance was found in the seed samples of non-GM varieties in 3 of the trials tested. Examples of mean frequencies of glufosinate and glyphosate tolerance detected at a range of distances from the GM trials are shown in Table 1.  Frequencies are expressed as the percentage of herbicide tolerant plants per plot. Overall, frequencies of glyphosate and glufosinate tolerance tended to decrease with increasing distance from the GM trial and varied between varieties with the highest levels detected in the composite hybrid Synergy.


Table 1.           Frequency of glufosinate and glyphosate tolerance detected at varying distances from the pollen source at Caxton National List trial site



Distance from source (m)

% glufosinate tolerant


Distance from source (m)

% glyphosate tolerant
















Var. A












Var. B
























Var. C










Gene flow in GM NL trials

Herbicide tolerance was detected in seed samples from conventional and herbicide tolerant varieties in samples from the three sites tested. A typical plot distribution is shown in Table 2. Overall, frequencies of herbicide tolerance decreased with increasing distance from GM plots. Levels of herbicide tolerance observed in the composite hybrid Synergy tended to be higher than in other varieties at the same distance from GM plots. Double tolerance to both glufosinate and glyphosate was detected in seed samples from herbicide tolerant varieties.


Table 2.           Plot configuration of a National List GM winter oilseed rape trial at Caxton  and the percentage  of glyphosate tolerance detected in seed samples from these plots.                                 


Synergy   4.33

GLU1    0.16

GLU1   0.83


Express   1.17

Falcon    0

Falcon     0.5

Nickel    0.5

Express  0.5

Falcon    7.6

Synergy  9.16

Nickel    0.16


Express  0.33

Apex     0

Nickel    0.83


Apex      0.16    

Apex       9.83


Synergy 1.0


GLU1    2.0



2. Pollen dispersal from herbicide tolerant winter oilseed rape


Male sterile bait plants were pollinated and set seed in all plots. Numbers of herbicide tolerant seed detected decreased rapidly with distance from the pollen source (Figure 2). Plots to the north of the GM pollen source produced the greatest proportion of herbicide tolerant seed. The incidence of pollinating insects particularly honey bees (Apis mellifera) and bumble bees (Bombus sp.) were notably low at flowering (mean high and low air temperatures were 18°C and 6°C respectively).  During the flowering period (01.05.98 - 29.05.98) wind direction was predominantly north and north west.


Figure 1.         Frequency of herbicide tolerant seed harvested from male sterile bait  plants at varying distances and directions from genetically modified herbicide tolerant oilseed rape1









1Total numbers of seed set per plot: north 100m=800, 200m=626, 400m=514; east 100m=1187, 200m=1061, 400m=1174; west 100m=88, 200m=967, 400m=317; south 100m=461


3. Monitoring studies


Table 3.           Volunteer incidence at Plant Genetic Systems Cambridgeshire field 1


Field No.1


GM Area

Following Crop 

No. Volunteers

No. Herbicide























WOR 5.2


winter wheat









winter barley









post harvest






Monitoring volunteer incidence indicates that oilseed rape seeds are persisting in the soil for up to three years post GM release (Table 3).. Herbicide tolerant plants were found both in following crops at all sites and in field margins at some sites. Consistently lower proportions of herbicide tolerant plants were detected compared to the proportion originally grown.




Gene flow through pollination decreased with increasing distance from the pollen source in both tests with NL trials and a large scale release using male sterile bait plants. Variations in gene flow levels in this study are likely to be due to different positions of GM plots within NL trials, differences in pollen source size, and various agronomic and climatic factors such as carry over of seed in harvesting machinery and prevailing wind direction. High frequencies of herbicide tolerant plants in seed samples of Synergy demonstrates the ‘susceptibility’ of composite hybrid material to alien pollen. Hybrids tolerant to both glufosinate and glyphosate were detected at all of the NL GM sites and the highest levels were observed in adjacent plots, 3.5% and 2% at the Caxton site which correspond approximately to rates of multiple tolerance reported by Messean (1997) of 2% at 1m from a much larger pollen source.                                                                                                                                                                                                                                                                                                                                             Detection of herbicide tolerance in seed of male sterile oilseed rape plants at distances of up to 400m show that there is potential for oilseed rape pollen to be dispersed by wind and remain viable over considerable distances.


Numbers of volunteers recorded at National List sites were low and it is evident from these results that GM herbicide tolerant oilseed rape does not appear to increase problems of volunteer management in following crops. A proportion of seeds sampled from GM plots were hybrids expressing tolerance to both glufosinate and glyphosate but there was no indication that these multiple tolerant hybrid plants were more difficult to control in following crops than conventional or single tolerant rape varieties.


The numbers of glufosinate tolerant compared to non-GM volunteer plants found both in following crops and in field margins were low at the Plant Genetic Systems Cambridgeshire site. Previous work looking at the survival and persistence of GM rape lines reflects the situation reported here (Sweet et al., 1997, Booth et al., 1996, Crawley et al., 1993). Incidence of GM herbicide tolerant rape plants in these volunteer populations suggest that weediness and invasiveness is not enhanced by this specific genetic modification.




Booth E J, et al. (1996)  Assessment of the ecological consequences of introducing transgenic rapeseed.  4th ESA Congress, Book of Abstracts (Persistence of oil- modified oilseed rape, Sinapis arvensis and Brassica nigra), pp 144-145.


Crawley M J, et al.  (1993)  Ecology of transgenic oilseed        rape in natural habitats. Nature,  363, 620-623.


Messean A (1997) Management of herbicide tolerant crops in Europe. Proceedings Brighton Crop Protection Conference - Weeds, 3, 947-954.


Norris C E. et al (1999) Monitoring weediness and persistance in genetically modified herbicide tolerant oilseed rape in the UK. Proceedings BCPC Conference on Gene Flow, Keele UK  - in press.


Simpson E. et al (1999)  Gene flow in genetically modified herbicide tolerant oilseed rape (Brassica napus) in the UK. Proceedings BCPC Conference on Gene Flow, Keele UK  - in press.


Sweet J B & Shepperson R (1996) Monitoring commercial releases of genetically modified       oilseed rape. Proceedings 10th International Weed Biology Conference Dijon, pp 217-            222.


Sweet J B; et al (1997) The impact of releases of genetically modified herbicide tolerant oilseed rape in the UK. Proceedings Brighton Crop Protection Conference - Weeds, 4, 291-302