But how effective will REDD actually be at reducing emissions and protecting biodiversity? To answer this question a team from the UK, Indonesia and the US examined Indonesia's first REDD initiative, in Northern Sumatra.

"Our study suggests...[the] initiative will not significantly reduce deforestation in northern Sumatra and will have little impact on orangutan conservation," David Gaveau of the University of Kent, UK, and the Wildlife Conservation Society Indonesia Program told environmentalresearchweb, "because firstly a large amount of forest inside the proposed REDD project area is protected de facto by being inaccessible; and secondly much of northern Sumatra's lowland forests will remain outside of REDD and will be exposed to the combined expansion of high-revenue oil palm plantations and road networks."

The REDD project, financed by Bank of America and co-managed by carbon trading firm Carbon Conservation and non-governmental organisation Flora and Fauna International, covers 7500 sq km of forest, mostly in the uplands. For comparison, Northern Sumatra extends across 65,000 sq km and is home to 92% of remaining Sumatran orangutans.

The Sumatran orangutan (Pongo abelii) is critically endangered and acts as a symbol for southeast Asia's biological diversity. The team predicts that more than 1384 orangutans – 25% of the current population – will disappear by 2030 with or without REDD intervention, leaving just 5000 Sumatran orangutans in the wild.

"Those remaining would face increased threats from encroachment, conflicts with humans and extensive habitat fragmentation," said Gaveau. "These predicted losses of orangutans would largely arise from extensive losses (56%) of forest cover in lowland forests (<500>

To test the REDD project's effectiveness, Gaveau and colleagues from the University of Kent, UK, Wildlife Conservation Society Indonesia Program, Great Apes Trust of Iowa, US, Conservation International, US, and Center for International Forestry Research, Indonesia, examined satellite maps and modelled future deforestation for two scenarios – the current plan of implementing REDD within a new upland protected area, and an alternative arrangement that uses REDD across landscapes outside protected areas.

The team predicts that the current plan would save 1313 sq km of upland forest by 2030, but 7913 sq km of forest (around 22% of 2006 forest cover) would be lost across landscapes outside protected areas. That would cut orangutan habitat by 1137 sq km (16%), resulting in the loss of 1384 of the apes. Implementing REDD across all remaining forest landscapes outside protected areas, on the other hand, would save 7824 sq km of forest, with "maximum benefit for orangutan conservation".

"We argue that future REDD schemes will almost certainly be implemented within defined areas of land, essentially by establishing new protected areas, because this approach would leave some areas of forest open to plantation development, and road construction," said Gaveau. "In essence, REDD may simply end up resembling other types of conservation project that promote the establishment of protected areas, where conservation funds help finance monitoring and law enforcement within protected area boundaries."

Instead, the team believes that for REDD to be effective it must be implemented across large areas – such as a nation, state or province – rather than on a project-basis within small, well-defined areas of land.

"A national/provincial approach would include all remaining forest landscapes in a REDD scheme, including highly threatened lowland forests, prime habitats for primates and other species, and also huge carbon stores," said Gaveau. "This approach would of course leave no room for the expansion of oil palm – and other cash crops – and roads into the forest. Such plans for expanding roads and oil palm would either have to be cancelled, or relocated to vacant degraded non-forest lands (the idea of relocating oil palm plantations to degraded non-forest lands has been proposed by the World Resources Institute, under the name POTICO)."

The project-based approach is, however, favoured by private investors because it's easier to monitor a small area, is cheaper to set up, less of an investment risk, and more straightforward to identify who should receive compensation. But Gaveau and colleagues believe a national/state/provincial approach is the only chance to safeguard forest habitats, species and ultimately dramatically reduce greenhouse gases.

"The most important obstacle to establishing REDD across large regions is the opportunity costs of carbon concessions," said Gaveau. "REDD can only be implemented across very large spatial scales if REDD revenues are substantially more than profits from agricultural commodities. In other words, the compensation forest users would earn from REDD would have to be more than earnings they could make from, say, converting forest to oil palm or to another cash crop."

The team believes this may be feasible if post-2012 global climate agreements legitimize the trading of carbon credits from avoided deforestation on the compliance market.

Now the researchers hope to improve their modelling of future deforestation by incorporating socioeconomic variables, such as the receptivity of people to payments/compensation for forest and their willingness to participate in REDD programs; by evaluating economic incentives for a diverse suite of agents such as local and urban communities as well as the private sector; and explicitly considering the fluctuating price of carbon vis-à-vis competing commodity prices. "We will also incorporate empirical measurements of carbon stock change attributed to deforestation and degradation using a combination of remote sensing and field surveys," said Gaveau.

The team has made its maps of tropical deforestation available to the public using Google Earth technology. According to Gaveau, an additional factor in choosing Northern Sumatran forest for the case study was that it is the only place on Earth where all the animal characters of Walt Disney's animation feature film The Jungle Book co-exist.

The researchers reported their work in Environmental Research Letters.