Published online 7 March 2011 | Nature | doi:10.1038/news.2011.139


Counting the carbon cost of peatland conversion

A study has quantified emissions from clearing peat-swamp forest in Southeast Asia for palm-oil plantations.

cleared Southeast Asian peatlandLarge amounts of carbon stored in both biomass and soils are released when peat-swamp forests are cleared.ROMEO GACAD/AFP/Getty Images

Up to 6% of carbon-rich peat-swamp forests had been cleared in Peninsular Malaysia and on the islands of Borneo and Sumatra to make way for oil-palm plantations by the early 2000s, according to a study published today. The clearances, a response to rising demand for food and biofuel, released as much carbon dioxide into the atmosphere as the entire UK transport sector does in a year.

Published in the Proceedings of the National Academy of Sciences1, the study is the first attempt to systematically assign a value to the carbon loss due to peatland destruction in Southeast Asia that can be attributed directly to conversion to oil-palm plantations.

Malaysia and Indonesia (which includes Sumatra and parts of Borneo) are the world's largest suppliers of palm oil, accounting for 87% of global production in 2008.

"Peatlands have been the focus of a lot of attention by developers who want to convert them to agricultural use, and also from conservation groups who want to protect the lands," says Lian Pin Koh, an ecologist with the Swiss Federal Institute of Technology Zurich and a member of the study team.

At least some of the expansion has been at the expense of cleared forests, including peat-swamp forests. Like most forests, peat-swamp forests store large amounts of carbon above ground as biomass, and this is lost when the forest is cleared. They also store large amounts of carbon in their soils, as dead organic matter decomposes slowly under marshy conditions. Draining peatlands to create agricultural land oxidizes the soil and releases large amounts of carbon dioxide into the atmosphere.

Peat carbon loss

For the first 25 years after an oil-palm plantation is established in a peat-swamp forest, about 60 tonnes of carbon dioxide are released per hectare every year, according to recent research2. More than half of those emissions come from the peat itself.

Koh and his colleagues note that by the early 2000s almost 90% of oil-palm development in Southeast Asia had taken place in non-peat areas; only 6% was on peat-forest soil. This might be because peatlands are low in nutrients and suitable for only some types of agriculture, making them an expensive option.

Looking at particular regions within their wider study areas, however, the team found significantly greater deforestation than the average. In the Indonesian provinces of North Sumatra and Bengkulu, and in Peninsular Malaysia, 38%, 35% and 27%, respectively, of peat-swamp forest had been converted to oil-palm plantations by the early 2000s.

Koh and his colleagues calculate that this conversion led to the release of about 140 million tonnes of carbon from biomass above ground, and 4.6 million tonnes of carbon from peat oxidation below ground. "Indonesia is among the largest contributors to carbon emissions," says Koh. And a quarter to a third of global greenhouse-gas emissions are the result of land-use change in forests, he says.

Peat-swamp forests are home to a number of endemic species, some of which have been affected by the changes, according to the new study. Using a "species extinction calculator" model3 that examines the effects of land-use change on wildlife, Koh and his colleagues found that the conversion to oil-palm plantations had put four species of bird at risk of extinction in Borneo, 16 species at risk in Sumatra, and 46 in Peninsular Malaysia.

Questions of scale and time

Koh's team also collaborated with remote-sensing experts to create a map of land-use change. The NASA satellite MODIS (moderate resolution imaging spectroradiometer) captured 490 images of the region early last year at a resolution of 250 metres. The images were examined and classified by type of land cover, including water, forest or plantation.

Given the rather coarse resolution of that satellite data, only large palm-oil plantations with closed canopies are easily identifiable, says Chandra Giri, a land-cover scientist and contractor to the US Geological Survey (USGS). Such plantations are at least ten years old, making all conclusions about land-cover change, even with the more up-to-date sensing data, relevant only to the early 2000s.

This is problematic, says Giri, because the planting of oil palms accelerated rapidly after 2002 together with forest clearance, although the two have not been correlated as in this study. And Koh and colleagues' ability only to map plantations larger than 200 hectares misses out smaller, open-canopy plantations, he says. Better imaging techniques using Landsat data at higher resolutions are available, he says, pointing to a recent study on loss of forest cover in Indonesia4.


A previous attempt at quantifying peat-forest conversion to oil-palm planting was made by the non-profit organization Wetlands International for the Malaysian state of Sarawak on Borneo. Their report, published in January, found that between 2005 and 2010, nearly 33% of the peat-swamp forests in the state had been cleared, more than half for oil palms. This study used Landsat satellite data from the USGS together with data from the Japanese satellite ALOS, which has a 50-metre resolution.

The project manager of this report, Niels Wielaard, at remote sensing institute SarVision in Wageningen, the Netherlands, says that although higher-resolution results are now possible and only older plantations are being considered, Koh's study is useful because of its large geographical focus.

"Such recent reliable information has not been available for the entire region until now," he says. 

  • References

    1. Koh, L. P., Miettinen, J., Liew, S. C. & Ghazoul, J. Proc. Natl Acad. Sci. USA doi:10.1073/pnas.1018776108 (2011).
    2. Murdiyarso, D. et al. Proc. Natl Acad. Sci. USA 107, 19655-19660 (2010).
    3. Koh, L. P. et al. Conserv. Biol. 24, 994-1001 (2010).
    4. Broich, M. et al. Environ. Res. Lett. 6 doi:10.1088/1748-9326/6/1/014010 (2011).
Commenting is now closed.