By Richard Black
BBC environment correspondent
A new UK study of a number of specific GM crops has found no evidence that they are more harmful to the environment than conventional varieties.
The Bright Link project studied sugar beet and winter oil-seed rape which had been engineered to make them tolerant of specific herbicides.
This modification allowed them to be sprayed and still prosper while all the weeds around them died.
The novel crops were grown in rotation with non-GM cereals, and compared with similar rotations involving non-GM beet and rape.
The project concluded that the GM varieties used in this way did not deplete the soil of weed seeds needed by many birds and other wildlife.
The findings of the Botanical and Rotational Implications of Genetically Modified Herbicide Tolerance (Bright) Link project have been released eight months after another major GM investigation, called the Farm-Scale Evaluations or FSEs.
The FSEs found that two GM varieties, a sugar beet and a spring rape, were more damaging to biodiversity than conventional crops. There were fewer insect groups, such as bees and butterflies recorded among the plants.
A GM maize, on the other hand, appeared to do better than its conventional cousin.
There were more weeds in and around the biotech maize crops, more butterflies and bees around at certain times of the year, and more weed seeds.
The FSEs were considerably larger in scale than Bright, involving 60-70 fields across the UK, and follow-up measurements were made in years following the planting.
On the face of it, the Bright conclusions appear to contradict the FSE results. Bright found fewer adverse impacts, whereas the FSEs found more.
"They asked different sets of questions," Chris Pollock, who chaired the FSEs' scientific steering committee but was not involved in Bright, explained to BBC News.
The FSEs were a straight comparison of GM versus non-GM in a single growing season, whereas Bright aimed to reflect normal farming practices in each location, indicating how GM varieties might perform if they were integrated into UK agriculture.
"The FSEs looked at single-management options," said Professor Pollock, "and we suggested there were important questions of how management practices would need to be modified to include herbicide-tolerant varieties - Bright is now giving us some answers."
Professor Pollock, research director of the Institute of Grassland and Environmental Research in Aberystwyth, was speaking in a personal capacity.
Dr Jeremy Sweet, Bright's scientific co-ordinator, believes that in both studies, the impact on seeds - and so on wildlife - is down to the herbicides used, rather than the GM crops themselves.
"There have been studies in the UK now for many years which have shown the detrimental effect of herbicides, and how they've depleted farmland birds, and all the rest of it," he told BBC News.
THE BRIGHT LINK PROJECT
Four-year study on relatively large plots (0.25-0.5 hectares)
Studied GM winter rape and beet grown in rotation patterns
Outcomes compared with plots of conventional plants
GM crops tolerant to glyphosate or glufosinate herbicides
One rape had cross-bred resistance to imidazolinone
Studied numbers and diversity of weed seeds left in soil
Looked for emergence of multi-herbicide-resistant plants
"The critical thing is how the herbicides are used on these crops; so what we need to do is to ask whether these herbicides have the potential to do more harm than current ones or, if managed properly, they can be less harmful than the current ones."
Dr Sweet concludes that the herbicides used with the GM varieties can be less harmful than those used on the conventional crops.
"One of the interesting things about the herbicide-tolerant systems is that you can apply the herbicides later, when you've got a much better idea of what spectrum of weeds is in the field, and therefore you can target your weed control more effectively.
"This means that you have some scope for manipulating populations of weeds so that if you do want to retain a reasonable weed flora in the field, you can do that.
"You can also control the weeds which are competing with the crop, particularly when the crop is being established."
Professor Pollock added: "The most important element of Bright is the fact that it confirms comments made in the FSEs about the importance of management.
"You can use management to generate rotations that give you the best balance between weed control and preservation of the seed supply."
Bright did show some potential problems with cross-breeding between herbicide-tolerant varieties of rape, producing seeds immune to more than one herbicide.
THE BRIGHT FINDINGS
GM crops showed increase weed seeds over 4 years
Weed seed diversity in soil similar in GM and non-GM
Advantages seen in GM in reduced weed control costs
GM crops offered greater flexibility in spray timing
Emerging multi-herbicide-resistant crops controllable
"We did create a stock of oil-seed rape seeds in the soil following the growing of GM crops," said Dr Peter Lutman, from Rothamsted Research, one of the agricultural centres that took part in Bright.
"And that seed bank, although it declines quite rapidly, does stay in the field for a number of years; so there is a question about how soon you could grow a non-GM crop in the same field and not have a problem arising from the GM plants from the previous crop."
'Lack of trust'
Dr Lutman believes there could be further problems if, in the future, GM beet and rape were grown in rotation with cereals which were also genetically modified to be tolerant to the same herbicide.
"My experience of managing weeds over many years is that if you use the same herbicide year on year on year, then you will build up problems.
"Indeed there are problems arising in North America where people are growing Roundup-Ready soya and Roundup-Ready corn in the same rotations.
"And I would think one would need to look very hard about how one managed a rotation of GM crops."
Bright has not impressed the anti-GM lobby. "The study is really quite limited," Clare Oxborrow, from Friends of the Earth, told BBC News.
"It looked at a very limited aspect of biodiversity, and it doesn't really give a definitive result in terms of whether the GM crops were any better or worse for the environment in terms of the seeds left in the weed bank.
"It won't really have a very significant impact, I don't think, on the big picture. It doesn't really look at big issues like liability - who will pay if things go wrong.
"We've got an enormous amount of political pressure being exerted from the US though the World Trade Organisation process, where they are complaining about the European precautionary approach with respect to GMOs.
"They are trying to force through GM approvals and imports into Europe despite the fact that overwhelming consumers don't trust the technology."
For English Nature, the UK government's independent wildlife advisor - it sees no contradiction between the FSEs and Bright but neither does it see anything to cheer.
Having studied the latest findings, it has concluded that Bright supports the evidence from the FSEs that the farming methods used with modified herbicide-tolerant sugar beet and oilseed rape crops can be harmful to wildlife.
It believes that because weed control in these crops is more effective and reliable than for conventional intensive agriculture, this in itself poses significant danger for wildlife.
"This new study adds little to what we already know about the impacts on wildlife of these cropping systems," said Dr Brian Johnson, English Nature's biotechnology advisor.
"We know from the government's larger study that using these systems with GM oilseed rape and beet crops would reduce densities of wild plants and insects in our already impoverished countryside. They should not therefore be grown commercially."
Bright was sponsored by the Department for Environment, Food and Rural Affairs (Defra) and the Scottish Executive Environment and Rural Affairs Department (Seerad).
It was carried out by Broom's Barn, Niab, Rothamsted Research, The Arable Group; and the Scottish Agricultural College, Aberdeen. It also included industrial partners.