Oilseed rape

Cropping practice and use in Denmark

In Denmark two types of oilseed rape (Brassica napus) are grown, primarily winter oilseed rape (autumn-sown oilseed rape) and to a minor extent spring oilseed rape (spring-sown oilseed rape).

The crop is used for food, feed and fuel.

Crop area, Denmark, 2002 and 2006 Conventionally grown oilseed rape (seed) 600 ha 300 ha Conventionally grown oilseed rape in total 81,600 ha 124,800 ha Organically grown winter oilseed rape Organically grown oilseed rape (seed) 10 ha 0 ha Organically grown oilseed rape in total 890 ha 800 ha Oilseed rape in total 82,600 ha 125,600 ha Sources: Kristensen, 2007; The Danish Plant Directorate, 2006a.

Crop types and cultivation methods for oilseed rape are unchanged compared with 2002, but the area has increased by approx. 50%. At present, approx. 20% is grown specifically as oilseed rape for fuel, but after harvest approx. 70-80% of the total amount of harvested oilseed rape is sold for energy purposes due to better prices, and this development is expected to continue due to increased interest in using rape seed oil for biodiesel. The area is therefore expected to increase further in the years to come; the winter oilseed rape area sown in 2006 is thus 180,000-200,000 ha. However, due to diseases in the crop rotation the total oilseed rape area is not expected to increase to more than approx. 250,000 ha, which corresponds to approx. 10% of the cultivated area. The share of spring oilseed rape is falling, and this tendency will probably continue.

The seed used for establishing oilseed rape crops is mainly produced in Denmark and other EU countries. In 2002, seed production of winter oilseed rape was on 478 ha and of spring oilseed rape on 84 ha, whereas the corresponding figures for 2006 was 227 ha of winter oilseed rape seed production and 78 ha of spring oilseed rape seed. Propagation of hybrid varieties has had special attention in connection with co-existence due to a potential for major admixture of foreign pollen.

In 2002 propagation of seed of hybrid varieties of winter oilseed rape took place on less than 25 ha, while no hybrid varieties were propagated in 2006. During the intervening years there were very small areas propagating seed of hybrid varieties.

Organic fields constitute approx. 0.6% of the total oilseed rape area. The producers of organic oilseed rape also mainly grow winter oilseed rape.

Oilseed rape has several wild relatives in Denmark found in fields and has a range of related crops, with which it can hybridise such as swede, turnip, stubble turnip and brown mustard. The cultivated area with these crops still constitutes far less than 0.1% of the total Danish crop rotational area.

Also cabbage forms offspring by crossing with oilseed rape in rare cases in seed crops.

Experience with GM oilseed rape in Denmark

There have been no releases since 2003. Five herbicide tolerant GM oilseed rape lines have been approved for import and processing, and a similar application has been submitted for yet another herbicide tolerant GM line. There are no regulations established on cultivation of GM oilseed rape.

Experience with GM oilseed rape outside Denmark

At present there is no commercial propagation of GM oilseed rape for production purposes in the EU. There are large GM oilseed rape areas in the USA and Canada, mainly herbicide tolerant GM oilseed rape. Globally, GM oilseed rape constitutes approx. 5% of the total GM crop area (ISAAA, 2006).

Sources of dispersal

Oilseed rape plants are both self-pollinating and cross pollinating by wind and insects. Both pollen and seeds are therefore sources of dispersal of GM material as described in detail in the 2003 Report.

Pollen dispersal occurs mainly between oilseed rape fields, but pollen can also be dispersed from volunteer oilseed rape in fields or from naturalised populations along roads, on building sites, etc.

Pollen dispersal can also take place from oilseed rape to related weeds (especially wild turnip) in similar places.

Seeds are a major source of dispersal for GM material from oilseed rape. Volunteers from seeds that are spread adventitiously are also a source of pollen dispersal as mentioned above and a source of adventitious presence of unwanted seeds in later oilseed rape crops in the same field. Machinery is expected to have a major influence on the dispersal of seed within and between farms as seeds from oilseed rape are relatively small (1 kg of oilseed rape is equivalent of approx. 200,000 seeds).

New knowledge

New knowledge in relation to adventitious GM dispersal in oilseed rape has appeared both in relation to pollen dispersal (dispersal distances, the influence of bees, hybridisation with wild turnip) and in relation to seed dispersal (presence of volunteers in the field, naturalised oilseed rape outside the field and impurity of certified seed). Part of the new knowledge is synthesised through research collaboration in connection with the EU project SIGMEA (see 4.3.2). For example, a database has been established with a considerable number of data sets that are relevant to the regulation of co-existence. Mathematical models at landscape level have also been developed further.

Pollen dispersal analyses have included new data for parametrising and further developing mathematical models (e.g. Cresswell & Hoyle, 2006; Deveaux et al., 2005; Klein et al., 2006;

NIAB, 2006; Weekes et al., 2005; Walklate et al., 2004; Østergård & Colbach, 2006). The new

studies do not change the general conclusion that the reproduction system of the receptor crop, the field size and the shape of donor and receptor fields as well as the isolation distance between fields are the most important factors for the extent of pollen dispersal.

Data from more than 100 field trials in the UK Farm Scale Evaluation (Weekes et al., 2005) including both spring and winter oilseed rape were used in a detailed simulation study to forecast adventitious presence in harvested seeds at field level as a function of isolation distance, field size and the quantity of GM copies per genome in the GM crop (NIAB, 2006). With the given preconditions the GM-DNA content per genome added from a GM (homozygote) neighbouring field at a distance 35 m will, with 98% certainty, stay below 0.3% if the non-GM field is 200 m deep, but below 0.6% (winter oilseed rape) or 0.8% (spring oilseed rape) if the field is only 100 m deep. If the GM field is hemizygote (i.e. only 1 GM allel), a content of less than 0.4% is expected for a field 100 m deep at this distance. The influence of field size (length and width), which is caused by the fact that incoming pollen only constitute a negligible part of the amount of pollen in the centre of a recipient field, was also pointed out in the 2003 Report (see also Damgaard &

Kjellson, 2005).

Data concerning pollination by insects have shown that the share of wind pollination between fields may be reduced at high bee activity; it is therefore important to consider this factor in future studies (Hayter & Cresswell, 2006). There can also be differences in both the amount of nectar and the quality of oilseed rape varieties. As bees are able to detect this (Pierre et al., 1999), this factor is also of importance in understanding dispersal mechanisms. Preliminary results of a Canadian study have shown a higher frequency of wild bees in a non-GM oilseed rape variety compared with a variety of GM oilseed rape with another genetic background (Morandini & Winston, 2005).

Finally, the consequences of hybridisation between oilseed rape and wild turnip for the dispersal of genes to related species were studied by means of a mathematical model (Hooftman et al., in prep.) The preliminary results show that the dispersal of GM material to non-GM oilseed rape in this way will be of minor importance. As a consequence, it may not be necessary to remove plants of related species near GM fields particularly as this may reduce biodiversity.

Volunteers in the field after oilseed rape crops are frequent and are of major importance to adventitious presence in subsequent oilseed rape crops. Lutman et al. (2003) forecast an adventitious presence of 6-15 volunteers/m² in a succeeding oilseed rape crop based on data for seed loss, seed survival and plant establishment. Provided that a volunteer produces the same quantity of seeds as plants from the crop, the 6-15 volunteers/m² will result in an adventitious presence of 8-20% (assuming a plant density in the field of 75 plants/m²).

In a recent analysis of plants from 3 Danish winter oilseed rape fields, adventitious presence was assessed based on DNA markers and statistical assignment methods (Jørgensen et al., 2007). The results showed that between 6 and 32% of the plants could be ascribed to other varieties than the one grown, and the highest presence occurred in a field where the farmer used his own machinery, certified seed and had not grown the assigned varieties for the last 7 years. In the field where 6%

came from another variety, this variety had been grown 17 years previously.

In another Danish trial with cultivation of organic winter oilseed rape it was shown that the number of volunteers can be reduced significantly by leaving shed seed and seedlings on the soil surface for as long as possible before soil surface tillage, and at least until there has been so much precipitation

that the seeds have had a chance of germinating (Rasmussen, 2004). However, there were still up to 800 seeds per m² in the seed bank two years after the oilseed rape crop, which is approx. 10 times the number of seeds sown in practice (I. A. Rasmussen, pers. com).

The experience from Danish observations of volunteers is that large concentrations of oilseed rape can be found in set-aside fields (R.B. Jørgensen, pers. com.). These plants can be pollinated by GM crops resulting in adventitious presence of GM seeds in the seed bank of the set-aside field. This effect will to a large extent depend on the distance between the set-aside field and the GM field and on the shape and size of the set-aside field.

In a Swedish study, 4-5 GM oilseed rape volunteers per 100 m² were found in barley from a trial release of spring oilseed rape 10 years earlier, even though it had in any way been attempted to control volunteers by delayed ploughing, surface harrowing, no oilseed rape grown on the area, manual and chemical removal of volunteers, supervision and monitoring (D’Hertefelt et al., 2007, in prep.).

In a French study of a 10 x 10 km oilseed rape production area, 9% seeds from other varieties of oilseed rape than those grown were found in harvested seed lots (Deveaux et al., 2005). Here, the area was cultivated according to present practice without any special precautions.

The SIGMEA project is collating studies of volunteers in five European study areas and a report is being prepared which will be available in the middle of 2007.

New model-based forecasts of volunteer presence have also appeared. Thus, Begg et al. (2006) simulated the presence of volunteers in winter oilseed rape grown in a 3-year crop rotation with winter wheat according to common UK practice. The model is used for estimating both

requirements for time interval between the growing of GM oilseed rape and non-GM oilseed rape and the effect of reduced herbicide efficacy, tillage depth and seed shed on GM presence in subsequent non-GM crops. The modelling results conclude that volunteers are of major importance and that effective control is very important for achieving a low GM content. Since such control will be difficult to achieve, it was concluded that it will be difficult to change from GM oilseed rape to non-GM oilseed rape crops in the same field unless there are considerable cropping intervals.

New results on the presence of oilseed rape outside cultivated areas (naturalised oilseed rape) support the earlier assumption that naturalised oilseed rape is not going to be a major problem for co-existence in Denmark and Northern Europe. Results from five study areas of the SIGMEA project will be published during 2007. From the Danish study area near Bjerringbro (5 x 5 km), 10-12 flowering plants/km² were found in 2005 and 2006, corresponding to <0.002% of the oilseed rape grown in the area. Only few of the naturalised plants produced seeds since the verges were cut between flowering and seeding. These results agree with observations of naturalised oilseed rape in three other study areas (2 areas in Germany and 1 in the UK), but not with a study from France in which the presence of naturalised oilseed rape was considerably higher.

Seed dispersal is a major source of adventitious presence. In an analysis of the varietal purity in certified winter rape seed lots, it was shown by use of DNA markers that the content of other varieties was higher than the values permitted in 3 out of 15 lots tested (Jørgensen et al., 2007). As GM content is detected at DNA level, this shows the need of future DNA tests in connection with certification to avoid new cases of growing oilseed rape with an adventitious presence higher than

the threshold value. For example, the variety Hyola 401 was in 2000 grown in several places in the EU with an adventitious GM presence of up to 1% because this variety was not tested for GM content by the authorities (The Danish Plant Directorate, 2000). Now, all seed lots from non-EU countries, in which GM oilseed rape is grown, are tested for adventitious GM presence.

National proposals for co-existence of oilseed rape

Based on the conclusions in the 2003 Report, there is no legislation in Denmark on cropping distances for oilseed rape because the variation in and the uncertainty of pollen dispersal distances was considered too high to form the background for legislation, especially for seed production of hybrid varieties. At present, there is no legislation for co-existence of oilseed rape in other countries. In four countries apart from Denmark, proposals for regulations on co-existence have been submitted concerning cropping distances between fields with GM and non-GM oilseed rape (see Table 4.3).

In England, the proposals apply only to the growing of fully-fertile oilseed rape (DEFRA, 2006). A separation distance of 35 m is proposed based on forecasts from mathematical models and data from the UK Farm Scale Evaluation (see above), in which it is for example assumed that the non-GM field is 200 m deep and that pollen comes from one non-GM field only. The latter assumption is considered to be misleading in the long term, if the present development in the growing of oilseed rape continues. In addition, non-statutory measures in the form of “best practice” are proposed, for example minimising the presence of volunteers and cleaning machinery for both GM and non-GM growers. For partially-sterile oilseed rape, which at present is not grown in England, separation distances will be proposed if necessary.

For Poland the proposed separation distance is about twice as large as the 150 m proposed in the Danish 2003 Report, while Luxemburg and Latvia have considerably higher figures. The proposals for Poland and Latvia show different distances between conventional and organic fields, as was the case in the Danish 2003 Report.

As regards cropping intervals, in England this is considered to be outside legislation, as it concerns issues within the farm. For Poland and Latvia intervals of 3 and 6 years respectively between the growing of GM and non-GM oilseed rape are proposed on conventional and organic fields respectively, which is less than the proposed Danish values of 8 and 12 years respectively (Table 4.4).

Need for further knowledge

There is a general wish for more studies under practical Danish field conditions and for more focus on experiments with winter oilseed rape, as most experiments till now have been in spring oilseed rape.

¾ Data at DNA level in connection with seed dispersal, including

o presence and assessment of viability and survival of volunteers and naturalised populations

o extent of seed dispersal by machinery within the property and regionally o seed purity

¾ Data and modelling with the purpose of studying the importance of o separate harvesting of the field margin

o dilution of pollen in the field and in connection with field distances and sizes o pollen dispersal by honeybees

o pollen dispersal to hybrid oilseed rape

¾ Monitoring, based on DNA markers, of dispersal from future GM oilseed rape fields in order to make a continuous adjustment of the measures.

Conclusion

In 2006 oilseed rape was grown on approx. 5% of the cultivated area, mainly as winter oilseed rape.

The area has increased considerably (approx. 50%) since 2002 and is expected to increase to approx. 10% of the total cultivated area in the years to come. The organically grown area is only 0.6% of the total oilseed rape area.

At present, there is no commercial production of GM oilseed rape in the EU, but a few herbicide tolerant lines have been approved for import and processing. Research is going on into developing GM oilseed rape for bioenergy purposes.

Some new knowledge about oilseed rape is available, which is of relevance to co-existence. The general conclusion is that the contribution of pollen dispersal to adventitious GM presence can be kept low with suitable distance criteria but that limitation of seed dispersal and presence of GM volunteers will be difficult and require strict measures. As results of different trials and models vary depending on the actual conditions and assumptions, future adjustments of measures must be expected concurrently with potential future growing of GM oilseed rape.

No changes are proposed in the proposed separation distances between production fields compared with the 2003 Report. For fully-fertile winter as well as spring oilseed rape, this means that the proposed separation distance of 150 m is retained for oilseed rape for production. These distances are based on experience from propagation of certified seed, for which a minimum distance of 100 m is required for self-fertile varieties to morphologically keep an adventitious presence of other varieties below 1%. As a new compulsory measure, a distance is proposed of 50 m to other large populations of oilseed rape, for example in some set-aside fields; an alternative to this could be an agreement on cutting the area before a certain date. Concerning the separation distance from GM oilseed rape fields to non-GM oilseed rape fields for seed production, we refer to 2003 Report as these distances have not been re-assessed.

The new knowledge has confirmed that volunteers from the seed bank are very important to adventitious presence in the field. There is much uncertainty of the expected presence of volunteers depending of soil treatment and control efficiency. As regards the cropping interval, no change is proposed from the intervals in the 2003 Report of 8 and 12 years before again growing conventional and organic crops respectively. The fact that a longer interval is proposed before growing organic oilseed rape is due to fewer options for controlling volunteers in organic farming.

As obligatory measures for limiting seed dispersal and controlling volunteers the Working Group proposes minimum tillage after harvest, transport of GM oilseed rape in seed-tight containers, extra cleaning of machinery and control of volunteers in the field, in adjacent areas and on transport routes within the property.

GM admixture in seed is also an important factor, and there will be a need for more thorough testing of certified seed. This will especially become important in future growing of GM oilseed rape for production within the EU.

More research, especially under practical Danish cropping conditions, is required on seed dispersal and varietal purity, pollen dispersal in connection with the growing of hybrids, the importance of honeybees for pollen dispersal and the importance of harvesting methods and field sizes.

More research, especially under practical Danish cropping conditions, is required on seed dispersal and varietal purity, pollen dispersal in connection with the growing of hybrids, the importance of honeybees for pollen dispersal and the importance of harvesting methods and field sizes.