Large swathes of the Brazilian Amazon have come to resemble the midwestern United States in recent years, having been planted with soya as far as the eye can see. This development has unnerved conservation organizations, which fear that huge expanses of pristine rainforest are being felled to make way for the lucrative crop. A widespread hope is that large agribusiness, aware of both the need to protect this fragile environment and the importance of good public relations, can be induced to farm more sustainably. But a recent assessment of one major partnership between conservationists and the soya industry suggests the need for caution.

In 2006, it emerged that Cargill, the US agricultural giant based in Minneapolis, Minnesota, had expanded a port in Santarém, Pará state, to handle one million tonnes of locally produced soya — without having conducted a required environmental impact assessment. (The assessment was completed in 2010.) Environmentalists cried foul, fearing the impact of expanded production on the rainforest. With the world's attention on it, Cargill formed partnerships with several green groups, including The Nature Conservancy, headquartered in Arlington, Virginia. The aim was to ensure that Cargill was purchasing sustainably grown soya from local farmers and respecting the rights and opinions of other locals who oppose the expansion of soya production.

Well-intentioned schemes to protect the environment, respect local interests and boost the economy may fall short.

However, Cargill's apparently environmentally friendly operations in the Amazon may not be as green as they seem. Research by Brenda Baletti, a PhD candidate at the University of North Carolina at Chapel Hill, questions claims by green groups and Cargill that its soya operations avoid deforestation. In addition, the work raises concerns over how well the operations respect local opinion. The research will be presented this week at the International Conference on Global Land Grabbing at the University of Sussex in Brighton, UK.

Baletti contends that a satellite-imaging system capable of detecting deforestation on individual farms in Santarém was not available until two years ago. Before the monitoring system was up and running, there was simply no way to judge whether soya production by particular farmers in the area was sustainable, argues Baletti. This, she says, raises doubts about claims by green groups that their initiatives to protect the rainforest have succeeded.

Baletti also questions whether the Cargill–Nature Conservancy partnership took full account of the views and concerns of small farmers and others who were opposed to industrial-scale soya production in Santarém. Interviews conducted by Baletti indicate that many local people were unhappy with the developments and felt that their perspectives were ignored. The Nature Conservancy's partnership with Cargill may not have helped much: Baletti's research highlights, for example, that the organization reports those soya farmers who breach national forest-protection laws to Cargill, but not to the government.

The Nature Conservancy told Nature that it has been able to monitor broader patterns of deforestation for the past six years. It regards the joint programme with Cargill as successful because it has found clear reductions in forest loss over the past three years. The organization added that working with farmers would be difficult if it reported violations of forest-protection laws to the government, and said that its approach works: more farmers are complying with the laws and sparing the forest. Cargill, for its part, says that the initiative has discouraged the planting of soya in deforested areas and has helped to bring down the rate of deforestation in Santarém. The company also says that it regularly engages with locals through consultations and public meetings, and has taken on board some of their concerns about soya farming.

Yet the questions about Cargill's soya-production interests in the Brazilian Amazon echo wider concerns about whether market-driven approaches to conservation and sustainable development are always workable. Ensuring that all players, big and small, have a seat at the negotiating table is a worthwhile goal, and involving local communities is a pillar of the United Nations' programme to reduce greenhouse gases from deforestation. But it seems naive to assume that all voices and interests carry equal weight.

If not carefully designed, well-intentioned schemes that aim to protect the environment, respect local interests and boost the economy — all at once — may fall short. It should be borne in mind that if something sounds too good to be true, it often is.