The much-anticipated results of Britain's field trials of genetically modified crops show that genetically modified, herbicide-tolerant (GMHT) oilseed rape (canola) and sugar beet cropping reduced biodiversity in and around fields relative to conventional crops, while GMHT maize increased it.

The trials were called “farm-scale evaluations of the impact on biodiversity of genetically modified, herbicide tolerant crops,” but the studies, published today (October 16) in the Philosophical Transactions of the Royal Society, are less about GM crops directly than about the herbicides used to manage weeds in GM and conventional varieties.

Chris Pollock, chairman of the trials' steering committee, speaking to The Scientist this morning, noted that herbicide-tolerant crops can be generated with out genetic manipulation. He stressed that this study was about the wider biodiversity implications of being able to use herbicide-tolerant crops.

“The novelty of this study is that for the first time, we've looked at this in advance [of a crop's introduction] and across an entire agricultural system… and not just the year they were cultivated, but the 2 following years also,” he said.

The management of GMHT maize had less impact on biodiversity than GMHT canola and GMHT sugar beet because the herbicides currently used in UK maize cultivation—atrazine and its derivatives, the triazines—are so powerful. But they are banned in several European countries and will also be banned in the United Kingdom, said Pollock, who is also chairman of ACRE, the UK Advisory Committee on Releases to the Environment.

“Maize farmers use atrazine to generate a clean landscape into which you put the maize,” said Pollock. When the herbicide is banned, “if they don't go for GMHT, maize growers will have to come up with a way of growing it that doesn';t use atrazine and yet suppresses early weeds.”

The scale of these trials was unprecedented in agronomic ecology. Some 66 beet fields, 68 maize fields, and 67 canola fields covering many soils types and environmental conditions were divided in two, one planted with a GM variety and one with a conventional crop. The farmers treated the GM variety according to the manufacturers' instructions, the other according to their own usual methods. Species of weeds, weed seeds, and invertebrates such as beetles and butterflies were counted throughout the year in both fields and compared across the country.

“Read the papers and be impressed,” said one of the scientists involved, Joe Perry of Rothamsted Research, at a press conference at the Royal Institution. “We counted half a million seeds, one and a half million invertebrates. We made 4000 visits to fields covering the length and breadth of England and Scotland.”

“Today is a really good news day for ecology,” said David Gibbons of the Royal Society for the Protection of Birds, “because we've learned more about farmland wildlife on a large spatial scale than we have for the last 50 years. “These trials were about four times bigger than any other agroeconomic study anywhere in the world.”

Because experiments of this size have not been attempted before, it wasn't known how consistent the effects would be, said Les Firbank of the Centre for Hydrology and Ecology. “The key surprise for me personally was that we tended to get the same kind of effect from one year to another and from one part of the country to another.”

Pollock said this consistency was crucial. “It means the results can be applied widely across Great Britain and can be built into mathematical models with some degree of confidence.”

UK Environment Secretary Margaret Beckett said in a statement the results of the trials would be forwarded to all other EU member states. “They will also, no doubt, want to consider them very carefully.”

For the follow-up, the scientific steering committee will be hosting two scientific meetings, one today (October 16) and one on October 28.

Then, ACRE will be calling for written comments on the significance of these results for the cultivation of GMHT crops,” said Pollock. “ACRE itself will host two open meetings in November and December on the implications of the evaluations, and then will advise ministers accordingly… But this is only part of the evidence that will be used. These results are significant both biologically and statistically, but they are not the whole story.”