Overwhelming international attention has focused on land conversion and associated CO₂ emissions from oil palm (Elaeis guineensis), which expanded into humid tropical regions by approximately 160% from 1990 to 2010 . In 2010, two-thirds of the harvested oil palm area was in southeast Asia, primarily Indonesia and Malaysia . These two countries also contain 7% of remaining tropical forest and >50% of tropical peatlands . There has been a striking paucity of robust empirical data necessary to quantify carbon emissions from oil palm cultivation.

New research definitively demonstrates that oil palm is an extremely carbon-intense land use . While no single work provides a comprehensive estimate of oil palm emissions, two studies quantify carbon emissions from above- and below-ground sources, across peat and mineral soil types, over broad spatial scales and across the same period. From 2000 to 2010, Carlson et al. estimate that gross emissions from oil palm agriculture in Kalimantan ranged from 0.038 to 0.045 GtC yr^-1, while Lee et al. report gross Sumatran oil palm emissions at 0.021–0.028 GtC yr^-1. By comparison, recent assessments of gross carbon emissions from tropical land-use change in the 2000 era range from 0.81 to 2.8 GtC yr^-1, suggesting that oil palm in Indonesia alone contributed at least 2–9% of these emissions .

Governments and companies are eager to mitigate such carbon emissions embodied in oil palm products. For example, palm oil biodiesel does not meet the US EPA’s 20% life cycle greenhouse gas reduction threshold needed to qualify as a renewable fuel under the Renewable Fuel Standards (RFS) program . To meet consumer demand, companies committed to selling ‘sustainable’ oil palm products seek oil palm growers who have received certification by the Roundtable on Sustainable Palm Oil (RSPO) . Unfortunately, such sustainability initiatives fail to directly address the differential emissions from various forms of land cover conversion to oil palm – especially emissions from carbon-rich peatlands, as well as those indirectly linked to oil palm development.

Remote sensing analysis of land-use change across Peninsular Malaysia, Borneo and Sumatra – the global hub of palm oil production – suggest that from 1990 to 2010, approximately 50–85% of industrial oil palm plantations were established on intact or logged forested lands . Forest clearing contributes CO₂ to the atmosphere through combustion and decomposition of woody biomass. Over a 25-year typical oil palm plantation lifetime, intact forest conversion is estimated to contribute net emissions of approximately 9–20 tC ha^-1 yr^-1. Accordingly, the RSPO prohibits such intact deforestation for oil palm by certified companies after November 2005, although this restriction notably does not extend to logged forests or secondary forest regrowth .

Mineral soil carbon emissions compose only a fraction of those from peatlands. When oil palm is developed on peat soils, these anoxic organic substrates are drained, leading to ongoing oxidation and carbon dioxide emissions; peatland burning also releases carbon. In Peninsular Malaysia, Sumatra and Borneo, approximately 10–60% of oil palm plantations occupy peatlands, which are being cleared at increasing rates . If peatlands are forested and fire is employed during forest clearing, approximately 28–48 tC ha^-1 yr^-1 are emitted during a 25-year plantation lifetime, three-times the emissions from forested mineral soils . Unlike mineral soils, peatland emissions continue over time; oil palm developed from deforested peatlands could emit nine-times more net carbon than establishment on forested mineral soils after 100 years of oil palm cultivation . Yet, the RSPO may provide certification to oil palm developed on peatlands .

Carbon emissions disconnected from plantations – either in time or space – remain unaccounted for by current research. Logging before land clearing for oil palm may contribute 30–60% of emissions from plantation development ; establishing accountability for these emissions requires linking remote detection of logging to field interviews with timber contractors. Fires escaping from oil palm plantation edges into nearby forested lands, as well as accidental burns within peatland oil palm stands, remain largely unquantified . Especially during droughts, such wildfires could be a major cause of carbon loss . Draining and/or removing vegetation from peat domes, even if constrained to only a portion of the dome, may lead to peat subsidence and carbon emissions throughout the dome . Although advocates who promote oil palm ‘sustainability’ emphasize that palms sequester considerable carbon as they mature , such sequestration is ephemeral, emitted to the atmosphere when plantations are re-cleared after each approximately 25-year rotation. Assessments must incorporate net carbon flux from oil palm growth, clearing and re-planting cycles.

Land cover change also has a large biogeophysical impact (e.g., evaporation) on climate. In the Amazon Basin, large-scale forest loss may result in reduced rainfall through biogeophysical feedbacks . Incorporating projected future oil palm land-use change into a climate–vegetation model for Southeast Asia may reveal how plantation expansion alters climate through biophysical mechanisms.

Because oil palm evolved in Equatorial Africa, this species requires consistent rainfall and high temperatures, and is thereby narrowly restricted to humid tropical lowlands with correspondingly high biomass and associated potential carbon emissions . Yet across these tropical regions, previous human use and other disturbances (e.g., wildfires) have reduced carbon stocks. An appealing and widely touted solution to mitigating oil palm emissions promotes development primarily on such low-carbon lands . Under these prescriptions, pastures in Latin America and low-productivity swidden farm lands on mineral soils in southeast Asia are considered to be relatively low carbon . However, concentrating oil palm over existing agriculture rarely considers local interests and rights, especially when capitalized oil palm companies displace or replace smallholder farmers . Moreover, displaced land use could result in off-site carbon emissions .

Instead of considering the coincidence of low-carbon stocks and resident people as a problem to be avoided or – literally – bulldozed, we suggest that palm oil-producing countries reconsider their oil palm development strategy. What if solving the oil palm carbon emissions problem depended not on multinational capitalized companies, but on the widespread adoption of oil palm agriculture by tropical lowland resident peoples already managing ‘lower carbon’ lands?