Safety assessment of the deliberate release of two model transgenic crop plants, oilseed rape and sugar beet

Background and objectives ⁽¹⁾

Biotechnological techniques will broaden the range of germplasm available to plant breeders by producing new gene combinations containing desirable DNA sequences that are not available from other sources or by making them accessible. The potential and limitations of the new techniques are not, as yet, fully understood. This study aimed to assess the entire range of safety issues associated with the deliberate release of genetically modified plants (GMP) into the environment. We studied two model crop species: oilseed rape (OSR) and sugar beet (Beta vulgaris).

(1) This project was a direct follow-on from EC projects:
BAP-0489/0490/0491/0492 and BAP-0371/0384/ 0408/0423

Approach and methodology

To assess the distribution of genes via pollen, field studies were carried out in the UK, Belgium and France. Studies were designed to consist of “source” area, containing the GMP and a “catcher” area. The source and catcher areas were either located adjacent to each other or were separated. The GMP carried an easily traceable marker such as a colour gene marker.

We aimed to assess the potential and likelihood of gene transfer from OSR to four of its most predominant wild Brassicaceae: B. campestris, B. adpressa, Sinapis arvensis and Raphanus raphanistrum. We studied the pollination ability, the production of F₁ hybrids under field and forced conditions and the reproductive potential of F₁ hybrids.

We used endogenous markers and genes introduced by genetic modification to produce interspecific hybrids with wild relatives of sugar beet: Beta maritima, B. atriplicifolia and B. macrocarpa. We evaluated the competitive ability of the hybrids in mixed populations in the greenhouse and in the field.

Computer modelling was vital for predicting the importance of each biological parameter and providing feedback on experimental design. Based on the general model incorporating the experimental parameters, simulations can be made on the impact of the particular traits.

Field of rape seed.

Main findings and outcome

Results of the pollen distribution experiments were consistent across years and locations but were dependent on the experimental design. The effective pollen dispersal is mainly limited to the immediate vicinity (25 to 35 metres) of each plant. However, pollen could be detected at low frequencies at 100 metres from the pollen source.

Outcrossing experiments confirmed the potential for gene transfer in B. campestris. B. adpressa, S. arvensis and R. raphanistrum could only transfer genes to OSR under specialised conditions.

Due to genetic barriers hybridisation of sugar beet with B. atriplicifolia and B. macrocarpa was less frequent than with B. maritima. The results of the fitness experiments confirmed that the traits investigated did not convey any competitive advantage to the transgenic sugar beet.

Conclusions

Methods were developed for the design and monitoring of greenhouse and field trials using GMP. We developed methods to quantify some aspects of the performance of GMP.

The agronomic performance, in the field, of GMP did not differ significantly from non-modified plants. For both modified and non-modified plants the performance in the field was not completely predictable from the performance in the greenhouse. However, greenhouse tests may be valuable in predicting major changes in growth behaviour.

The fitness of GMP evaluated in this study was not significantly different from the corresponding non-modified plants.

Pollen dispersal from plants allowed to flower openly in the field was largely confined to the immediate vicinity of the test field. Effective pollination at long distance was negligible.

The possibility for gene transfer to related species was dependent on the crop species.