Researchers went ahead with the spraying shortly after it was disclosed that the government committee which had approved the trials had not been fully informed of the risks.
The warning prompted the committee's chairman to make hurried checks on Friday afternoon with the head of the trial, David Bishop, who is on holiday in South Africa. Professor Bishop said the test is safe, and that the new information does not alter his risk assessment.
The new pesticide includes a virus that infects caterpillars. The extra genetic material from scorpions releases a nerve poison which also paralyses them and stops them feeding, allowing the virus to act more quickly.
The statistics which came to light on Friday are the results of experiments carried out between 1989 and 1993 at the Oxford Institute of Virology where Professor Bishop is director. These showed that of about 100 species of butterfly and moth 75 are susceptible to the virus in its natural form. This is far more species than implied in the material submitted to the Advisory Committee on Release to the Environment (Acre) which approved the test last month after protests from scientists.
Opponents say the virus is not native to Britain. Even if it were not given extra genes, releasing it would therefore be more risky than in an environment where it occurs naturally. The virus is native to California. They say the physical arrangements for keeping the virus from escaping into the environment are 'laughable'.
At 3pm yesterday, scientists from the institute prepared the site by laying absorbent papers around the young cabbages, planted inside 360 enclosures, to help trap some of the excess viral spray.