Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Fostering a climate-smart intensification for oil palm


Oil palm production in Indonesia illustrates the intense pressure that exists worldwide to convert natural ecosystems to agricultural production. Oil palm production has increased because of expansion of cultivated area rather than due to average-yield increases. We used a data-rich modelling approach to investigate how intensification on existing plantations could help Indonesia meet palm oil demand while preserving fragile ecosystems. We found that average current yield represents 62% and 53% of the attainable yield in large and smallholder plantations, respectively. Narrowing yield gaps via improved agronomic management, together with a limited expansion that excludes fragile ecosystems, would save 2.6 million hectares of forests and peatlands and avoid 732 MtCO2e compared with following historical trends in yield and land use. Fine-tuning policy to promote intensification, along with investments in agricultural research and development, can help reconcile economic and environmental goals.


Indonesia hosts large tracts of pristine rainforests and tropical peatlands1. Of great concern is the conversion of these ecosystems for oil palm cultivation, which has contributed significantly to climate change and biodiversity loss2,3,4,5,6. Indeed, the sixfold palm oil production increase from 2000 to 2018 has been driven by a sharp expansion in plantation area (+10.2 Mha)7, with one-third of the converted land located in forests and peatlands8. By contrast, the average actual yield has remained unchanged during the same period7,9,10. While previous studies have highlighted the environmental impact associated with oil palm expansion11,12,13,14, there is a dearth of knowledge on how intensification, that is increasing average actual yield on existing planted area15,16,17,18, can help eliminate the expansion pattern of the past two decades of oil palm cultivation.

Oil palm is a perennial crop grown in tropical environments and it is the world’s largest source of vegetable oil7. The oil extracted from the palm’s fresh fruit bunches (FFB), typically referred to as crude palm oil (CPO), is used for cooking, and in processed food, cosmetics, cleaning products and biodiesel. Indonesia is the main palm oil producing country in the world (around 60% of global production)7 and the crop is a key driver of its economy7,19. Oil palm has contributed to rural development, with 42% of the oil palm area managed by smallholders and the remaining area managed by large plantations20. The oil palm sector, including government, research institutes, private companies and farmer associations, has projected that demand for Indonesian palm oil will reach 60 Mt of CPO by year 203521. Meeting that future demand would require a 46% increase in CPO production relative to the current level (41 Mt of CPO by 2018), which is consistent with the predicted CPO demand increase reported by other studies22,23. Indonesia has a critical decision to make about the most effective approach to follow in order to meet that production goal without further encroachment on fragile ecosystems.

Assessment of extra production potential for oil palm requires robust estimates of yield potential, that is, the maximum biological yield as determined by local weather, soil properties, and current planting material (Extended Data Fig. 1). Here we followed a data-rich spatial approach, consisting of crop modelling coupled with the best available sources of weather, soil, and production data, to determine the available room for increasing oil palm production on existing cultivated area located in mineral soils in Indonesia (Extended Data Figs. 2–5). We estimated potential environmental benefits, in terms of land savings and climate change mitigation (Extended Data Fig. 6), and discussed the required interventions and policies to promote intensification.

Available room for intensification on existing oil palm area

Reaching yield potential requires copious amounts of inputs and labour, together with a high degree of sophistication in crop management, in order to completely eliminate harvest losses, nutrient deficiencies and yield reduction due to pests and diseases24. Hence, reaching 70% of the yield potential is considered a reasonable yield goal for mature oil palm plantations where owners seek to maximize profit and return on input investments; this yield level is typically referred to as ‘attainable yield’ (Extended Data Fig. 1)25. Here we expressed attainable yields on an annual basis, based on long-term weather data and dominant soil types, and relative to the average plantation age for each farm type (Extended Data Fig. 5). We estimated an attainable FFB yield of 30.6 t ha−1 for Indonesia (Fig. 1, Supplementary Table 1). Data from high-yielding blocks provided to us by a number of large plantations corroborated our estimate of attainable yield derived from simulations (Extended Data Fig. 7). The attainable yield was slightly lower for smallholders compared with large plantations (29.1 versus 31.6 t ha−1) primarily because smallholders replant later than large plantations due to financial constraints (Extended Data Fig. 5)20,26.

Fig. 1: Yield gaps for oil palm in Indonesia.

a,b, Pie charts showing average farmer yield (yellow) and exploitable yield gap (red) expressed as a percentage of the attainable yield for large plantations (a) and smallholders (b) across 22 sites, representative of the oil palm producing area. Insets show national averages. Yatt, attainable yield; Ya, average actual farmer yield; and Yg, exploitable yield gap, expressed in tonnes of fresh fruit bunches per hectare per year. Means and temporal variability for each site are shown in Supplementary Table 1.

Source data

The exploitable yield gap was estimated as the difference between the attainable yield and average actual yield (Extended Data Figs. 1 and 5). The exploitable yield gap represents the available room for increasing productivity for an existing plantation via cost-effective agronomic management. We found a large exploitable yield gap for oil palm production in Indonesia, in both smallholder and large plantations (Fig. 1). At the national level, the exploitable yield gap represented 38% and 47% of the attainable yield in large and smallholder plantations, respectively (Fig. 1, Supplementary Table 1). The exploitable yield gap was larger in relatively new areas under cultivation, as it is in the case of Kalimantan, compared with traditional oil palm producing areas in Sumatra (Fig. 1). The large yield gap in these regions may be explained by the difficulties of adapting management practices and farm operations to these relatively new oil palm production areas and, in the case of smallholders, difficulties in acquiring healthy and certified seedlings, fertilizers and agrochemicals20.

Intensification and land use change scenarios and related global warming potential (GWP)

We then evaluated three scenarios that explored different levels of intensification and plantation area expansion as potential pathways to meet Indonesian CPO demand of 60 Mt by 2035 (Methods). In the ‘business-as-usual’ scenario (BAU) the projection of historical trends (2000–2018 period) in average actual yield and plantation area over the following 17 years results in 9.2 million hectares of new land brought into oil palm production without changes in average actual yield (Figs. 2 and 3). In the BAU scenario, 29% of land conversion occurs in peatlands and primary and secondary forests, with a tremendous loss in the biodiversity associated with these areas with high-conservation value (Fig. 3 and Extended Data Fig. 6). Despite carbon (C) gain due to oil palm expansion in land with low-C stocks, such as scrubland and grassland, the emissions from peat decomposition together with those derived from cultivation result into a net 767 MtCO2e released to the atmosphere (Fig. 3 and Extended Data Fig. 6).

Fig. 2: Projected trends in mature oil palm area, average actual yield, and production in Indonesia.

ac, Projected trends in mature oil palm area (a), yield (b) and production (c) during the study period (2018–2035) for three scenarios: (i) business as usual (BAU), with historical trends in area and yield remaining unchanged in the future; (ii) intensification (INT), with complete closure of the exploitable yield gap in current plantation area in mineral soils and without physical expansion of oil palm area; and (iii) intensification plus target expansion (INT-TE), with partial closure of exploitable yield gap and oil palm area expansion into low-C land. Historical trends (2000–2018) are shown with triangles. Projected demand by 2035 (equivalent to 60 MtCPO) is shown as a yellow circle in c. Sources of uncertainty are discussed in Supplementary Section II.

Source data

Fig. 3: Cumulative land conversion and GWP associated with different scenarios of intensification and land use change.

a,b, Accumulated land conversion in high- and low-carbon stock areas (a) and associated GWP during the study period (2018–2035) for the three scenarios (b). The GWP is disaggregated by source: changes in carbon stock from aboveground biomass when land is converted to oil palm, peat decomposition and oil palm cultivation. Negative values indicate carbon gain, while positive values indicate net greenhouse gas emissions released to the atmosphere. Emissions derived from peatland converted into oil palm production prior to the baseline year were not included. Other sources of uncertainty are discussed in Supplementary Section II and Extended Data Fig. 8.

Source data

Our yield-gap analysis shows ample room to increase average actual yield on existing plantation area located in mineral soils (Fig. 1). Hence, we explored an ‘intensification’ scenario (INT), which assumes a large investment in agricultural research and development (AR&D) so that the exploitable yield gap for oil palm plantations located in mineral soils is closed (Fig. 2). Closure of the exploitable yield gap increases FFB production by 68% relative to current levels, allowing Indonesia to meet the CPO demand by year 2035, without any further increase in plantation area (Figs. 2 and 3). The accumulated GWP during the 2018–2035 period is 60% lower than the BAU scenario, mostly because there is no expansion in peatlands (Fig. 3, Supplementary Tables 2 and 3). However, the INT scenario assumes that an increase in average actual yield from 18 to 30.6 t ha−1 for the entire oil palm area located in mineral soils in Indonesia is feasible in only 17 years. That magnitude of yield increase would require an annual yield gain of 0.7 t ha−1, equivalent to a 3% annual compound rate (p.a.). While achieving that rate of yield gain may be possible for individual plantations, it would be difficult to scale that out to the entire country considering the timeline needed for diagnosing the causes of yield gaps, identification of technologies to overcome them, and farmer’s adoption rates20,26. We note that global rates of annual yield gain for major food crops are well below 3% p.a. (refs. 17,27).

The last scenario (INT-TE) explores a combination of moderate yield gap closure in mineral soil and a limited and targeted area expansion into low-C land, without encroaching on peatlands, forests or areas cultivated with food crops (Figs. 2 and 3). The INT-TE scenario assumes a more realistic annual yield gain of 0.27 t ha−1 (or 1.25% p.a.), which is similar to average actual yield gains reported for major food crops, and for oil palm in other producing countries such as Colombia7,17. In the INT-TE scenario, average actual yield increases from 18 currently to 22.5 t ha−1 by 2035, equivalent to a closure of the exploitable yield gap by one-third (Fig. 2). In comparison with the BAU scenario, the INT-TE pathway allows Indonesia to meet palm oil demand by 2035, avoiding conversion of 2.6 million hectares with high-C stocks and conservation value, and reducing GWP by 732 MtCO2e (Fig. 3 and Extended Data Fig. 6). The target expansion alone would not be sufficient by itself to meet future CPO demand unless it is complemented with a moderate yield gap closure and vice versa.


Our study provides estimates of yield gaps for the most important areas cultivated with oil palm across the Indonesian archipelago using a process-based model that accounts for the effect of water limitation on yield and based on long-term measured weather data in combination with detailed information on management, soil type and plantation age. None of the previous studies looking into the magnitude and/or causes for yield gaps in oil palm10,25,28,29,30 have aimed to generate spatially explicit estimates of yield gaps and to upscale them to national level to estimate extra production potential for the whole country. Instead, those previous studies have focused on single or few locations, sometimes relying on models that do not capture well the effect of water limitation on yield, and, in most cases, using coarse gridded weather and soil databases which, as documented in previous studies, can introduce substantial biases into the estimation of yield potential31. There is also a relatively large number of studies quantifying the environmental impact associated with oil palm expansion into fragile ecosystems14,32,33,34 but, again, none of those studies have aimed to determine the potential role of crop intensification in helping reconcile production and economic goals. Hence, our study makes an important contribution by using best available crop models and data sources, and a robust spatial framework to show that substantial room exists to increase palm oil production via crop intensification on existing plantation area in Indonesia, which, in turn, could potentially lead to land savings and a reduction in GWP compared to following historical trends in yield and land-use change.

From a global perspective, the case of oil palm in Indonesia serves to illustrate the intense pressure that exists to convert fragile natural ecosystems to agricultural production in many regions of the world, including hotspots for biodiversity such as the Amazon and Congo basins32,35. Our assessment for Indonesia revealed that different scenarios of land use change and crop intensification would allow the country to meet, and even exceed, the expected CPO demand by 2035 (Fig. 2). However, the ‘how’ matters, especially in relation to the environmental outcomes and implications to prioritize AR&D investments and inform policy. Indonesia hosts the globe’s largest portion of tropical peatlands as well as vast tracts of pristine rainforest, which, if converted into agriculture, would exacerbate climate change and cause a tremendous loss in biodiversity1,32,33,35. Despite the inherent uncertainties of the analysis (Supplementary Section II), our assessment indicates that it will be difficult to avoid these negative environmental outcomes without an explicit aim to intensify oil palm productivity on the existing plantation area. The potential climate change mitigation achievable via oil palm intensification, as estimated in our analysis (Fig. 3), is relevant, considering that agriculture and land use change accounts for half of the country’s GHG emissions6,36. Furthermore, the Indonesian government has committed via the Paris Climate Agreement to reduce by around one-third of the projected GHG emissions by year 20306. Although our analysis focused on Indonesia, which has accounted for 75% of the global oil palm area expansion during the past 10 years7, our study also gives an important lesson about the role of intensification at bridging the gap between environmental and production goals to other tropical regions of the world, such as those in South America and sub-Saharan Africa, where pressure exists to produce more palm oil and other agricultural products5,14,32,37.

Starting with the ‘Green Revolution’ in the mid-1960s, Indonesia was successful at increasing rice production without massive land conversion7. At that time, convincing smallholders in Asia to use newly developed rice varieties and associated fertilizer and pesticides inputs was easy because the results, in terms of yield increase, were large, fast and consistent across environments. As a result, annual rice production more than tripled between 1965 and 1990, with 70% of the production increase attributable to yield gain7. Identifying the means and methods to tailor a ‘Green Revolution’ for oil palm, with an explicit goal to close the exploitable yield gap in a sustainable way is a vital issue. From an agronomic perspective, there is a large body of research reporting on practices proven effective at increasing oil palm yield and profit, in both smallholder and large plantations, including proper harvest methods, nutrient management, pruning, field upkeep, and pest and disease control25,37. Among these practices, improving plant nutrition should be considered as a key element to support intensification efforts as current nutrient inputs are insufficient and imbalanced in relation to plant requirements, especially in the case of smallholder plantations38. Likewise, use of certified planting material with higher oil concentration, in concert with better agronomic management, plays an important role, but we note that the timeline for impact is longer given the typical replanting cycle of 25 years20,39.

What can explain the relatively large yield gap and the lack of yield gain for oil palm in Indonesia over the past 20 years? In contrast to rice, yield response to improved agronomic practices in oil palm has a time lag (from months to several years), which limits adoption, especially if the associated financial cost is high (for example, fertilizer use, certified planting material)26,40. Additionally, oil palm cultivation takes place in less densely populated areas where labour shortage may not allow fine-tuning plantation management to the level that is required to reach the attainable yield10. In some cases, expansion of oil palm area was considered as a long-term land investment, without an explicit goal to maximize yield and economic return in the short term41. In the case of smallholders, lack of access to inputs, markets and extension education, and lack of experience in growing oil palm limits adoption of improved management20,40. Additionally, some of the current support mechanisms for smallholders (for example, fertilizer subsidies) may not be effective to remove yield-limiting factors in oil palm plantations26,40. In terms of AR&D prioritization, our study can help pinpoint specific geographic regions, far from high-conservation value areas, where yield gaps of existing plantations are large, which could serve as a starting point to orient intensification programmes.

A solutions agenda that explicitly tackles how to narrow the existing yield gap in oil palm should provide productivity incentives and facilitate access to technological and knowledge inputs where these are not available. In the case of smallholders, it seems a priority to develop vigorous extension programmes and re-align current supporting mechanisms to overcome limiting factors and reduce financial risk. In contrast to other agricultural systems, where either smallholders or large plantations predominate, smallholders and large plantations co-exist across the entire area cultivated with oil palm in Indonesia20. Given the interplay between those two farm typologies (for example, smallholders selling FFB to the mills managed by large plantations and, in some cases, large plantation providing inputs and agronomic advice to farmers), other models to facilitate diagnosis of yield-limiting factors, knowledge sharing, and technology adoption can be explored26,42,43,44.

While our study focused on crop intensification and land-use planning to reconcile environmental and production goals for oil palm in Indonesia, we acknowledge that other approaches exist to reduce the negative aspects associated with oil palm production. These approaches include bans or limits on oil palm imports, and promotion of certification programmes. Considering that the global demand for vegetable oil will increase by 27% over the course of the next decade23, proposals to ban palm oil imports fall short in efforts to protect the environment as they may lead to indirect land use change in other countries connected to global trade15. In addition, full or limited bans on palm oil imports would produce negative impacts on the livelihoods and welfare of millions of smallholders who cultivate the crop as well as on the economy of the world’s fourth most populous nation26. In the case of certification programmes, these efforts have the potential to improve specific aspects associated with oil palm cultivation (for example, workers safety, preservation of high-conservation value land within current plantations) but they do not have crop intensification as an explicit goal45. While we acknowledge that intensification is only one piece of the challenge and must be complemented with policies and institutions to ensure land saving for nature3,14,35,46, our study shows that it has huge potential at helping preserve fragile ecosystems. Recent steps taken by the Indonesian government to prevent further expansion of oil palm production into primary forests and peatlands via land-use planning and moratoriums, coupled with foreign incentives to reduce conversion of C-rich natural ecosystems (for example, REDD+ programme), are promising47,48. These efforts would benefit from an explicit recognition of the need for intensification, in the strict sense of increasing average actual yields, and an associated blueprint for action, including re-setting priorities on AR&D in both public and private sectors. Such an approach would give Indonesia, as well as other developing countries with competing economic and environmental goals, a pathway to protect some of the last bastions of forests and biodiversity on the planet.


Conceptual framework

Yield potential is the maximum biological yield as determined by local weather and crop traits influencing interception and conversion of solar radiation into harvestable yield49. In the case of rainfed crops, water supply and soil properties influencing the crop water balance imposes another upper limit to yield potential, hereafter referred to as ‘water-limited yield potential’ (Yw)50. Finally, yield potential also varies with palm age28. Most plantations’ palms start to produce two to three years after planting (from that point onwards they are considered ‘mature plantations’), following a typical pattern in which there is a sharp yield increase during the first years until reaching a peak, which is followed by a gradual decline in productivity over time51. Commercial plantations are usually replanted 25 years after establishment although replanting tends to occur later on smallholder farms as a result of capital restrictions20,26.

Average plantation yield (Ya) is always below Yw. This ‘yield gap’ reflects the incidence of yield-limiting factors such as nutrient deficiencies and reducing factors such as the incidence of weeds, pathogens and insect pests10. Published data from cropping systems in the US Corn Belt, western Europe and Asia indicate that reaching around 70% of Yw is a realistic goal for farmers who have adequate access to inputs, markets and technical information24. Given the previous evidence from other crops, and consistent with previous studies on oil palm25, here we estimated the attainable yield (Yatt) as 70% of Yw (Extended Data Fig. 1). The exploitable yield gap was calculated as the difference between Yatt and Ya (Extended Data Figs. 1 and 5). The Yg provides an objective measure of the available room to increase production on existing cropland via improved agronomic management.

Description of the protocol used to estimate water-limited yield potential for oil palm in Indonesia

Given environmental concerns on conversion of peatland ecosystems for oil palm production and recent measures taken by the Indonesian government to prevent it47, we focused our analysis on assessing available room for intensification in existing plantations located in mineral soils across major producing areas in the archipelago (that is, Sumatra, Kalimantan and west Sulawesi). Similarly, the analysis excluded ‘frontier areas’ with high conservation value (and little oil palm cultivation) such as those in North Kalimantan and Papua.

Here, we used results on Yw and yield gaps for Indonesia generated by the authors through the Global Yield Gap Atlas project following best available science and data sources31,52. Complete databases on weather, soil and detailed management information and productivity data from smallholder and large plantations were collected across 22 sites, located primarily in Sumatra and Kalimantan. An updated version of the oil palm crop model PALMSIM was used to estimate Yw (ref. 28,53). The model was calibrated using long-term yield data collected from well-managed, high-yielding plantations. PALMSIM was subsequently used to simulate Yw and estimate Yatt across the 22 sites based on local weather and soil, and average palm age. Resulting Yatt and average actual yield were used to calculate the yield gap for each of the 22 sites, separately for smallholder and large plantations. A scheme illustrating the steps followed to estimate yield gaps is shown in Extended Data Fig. 2. A detailed description of each step, associated data sources and uncertainty is provided in the Supplementary Section I.

Assessment of future scenarios of yield, production and land use change

We explored three future scenarios with different oil palm yield and area trajectories and assessed the production outcome and land use change associated with each of them. We used 2018 as the baseline year and we evaluated the degree to which each scenario would meet the CPO production target of 60 MtCPO (equivalent to 302 MtFFB) by 2035 set by the Indonesian oil palm sector, including government, research institutes, private companies and farmer associations21,22,23. Total oil palm area by the baseline year (2018) was 14.3 Mha, with 78% of total oil palm area corresponding to mature productive plantations and 18 t ha−1 of annual average actual yield9. For all scenarios, we considered mineral soils to account for 80% of oil palm area in the baseline year54,55.

Business-as-usual scenario

In this scenario, historical trends in oil palm yield and area over the 2000–2018 period remain unchanged over the next 17 years, that is, between the baseline year (2018) and 2035. In the BAU scenario, average actual yield remains stagnant and the total oil palm area increases at a constant rate of 0.54 Mha year−1; these values were derived from official statistics for the 2000–2018 period7,9. In this scenario, future oil palm area expansion follows the same pattern as during the 2000–2018 period in terms of the type of land cover that is converted for oil palm production, which roughly includes 1% and 7% of primary and secondary forests in mineral soils, respectively, and 21% of peatlands55. Predicted annual oil palm area expansion by land cover type is shown in Extended Data Fig. 6 and Supplementary Table 2.

Intensification scenario

To highlight here the available room for increasing production on existing plantation area located in mineral soils, we assumed a full closure of the exploitable yield gap so that average actual yield in mineral soils reaches the Yatt, that is, 70% of Yw. Average actual yield of oil palm in peatlands is assumed to remain unchanged over time and to be the same as in the baseline year. The intensification scenarios assumed stabilization of oil palm area (that is, a mature-to-total area ratio of 0.85) three years after the baseline year (2018) so that, while there is still a slight increase in mature area, physical expansion of oil palm area does not occur (Fig. 2).

Intensification plus target area expansion scenario

The level of yield gap closure that is required to achieve 70% of water-limited yield potential by year 2035, as investigated in the intensification scenario, would imply an annual yield gain in mineral soils of 3% compound rate per annum (p.a.) (equivalent to around 0.7 t ha−1 y−1). Such a high rate of yield gain is difficult to achieve at the national level and there is no evidence from the literature that rates of yield gain of this magnitude are possible for oil palm and other food crops over long periods of time18,27. We therefore propose a more reasonable 1.25% p.a. compound rate of yield gain for oil palm grown in mineral soils. This rate of yield gain is comparable to that for rice in Indonesia for the 2000–2018 period and to those observed in other oil-palm-producing countries such as Colombia7. An increase of 1.25% p.a. of average oil palm yield in mineral soils will close the current exploitable yield gap by 36% in 17 years. That degree of yield gap closure is, however, not sufficient by itself to meet CPO production goal by 2035 (53 versus 60 MtCPO). Hence, the INT-TE scenario considers further expansion of oil palm area into low-carbon land, that is, in areas where carbon stocks are lower than in oil palm plantations (Supplementary Table 3), following the same pattern as during the 2000–2018 period in terms of the type of land cover that is converted for oil palm production, avoiding primary and secondary forests as well as peatlands54,55. Similarly, the oil palm area is not allowed to expand into areas sown with annual food crops (for example, rice, maize) in order to avoid indirect land use change56. Overall, total oil palm area increases at a rate of 0.21 Mha year−1 in the INT-TE scenario, resulting into an increase in the total area by 3.6 Mha between years 2018 and 2035. Annual oil palm area expansion by land cover type is shown in Supplementary Table 2. That magnitude of oil palm area expansion into low-C land (that is, scrubland, grasslands, bare land) is realistic as it can be inferred from data reported by Austin et al.54, who estimated 30.4 Mha of low-C land suitable for oil palm production in Indonesia (excluding Papua). Similar to the INT scenario, our INT-TE scenario assumes no yield increase in peatlands, with associated average actual yield remaining at the same level as in the baseline year.

Estimation of global warming potential

We estimated the greenhouse gas emissions, including carbon dioxide (CO2), nitrous oxides (N2O) and methane (CH4) associated with land conversion (GHGluc) and with oil palm cultivation (GHGcul) for the three scenarios (BAU, INT, INT-TE) between the baseline year (2018) and 2035. The overall 100-y global warming potential (GWP) was estimated as the sum of GHGluc and GHGcul, both expressed as CO2 equivalents (CO2e) to account for the higher warming potential of CH4 and N2O, which are 25 and 298 times the intensity of CO2 on per mass basis, respectively.

GHGluc includes emissions associated with changes in carbon stocks from aboveground biomass when land is converted for oil palm production (GHGcon) and, in the case of peatlands, also for GHG emissions derived from peat decomposition that occurred following conversion of peatlands for oil palm production (GHGpeat)57.

For each land use type, the GHGcon was estimated for every year of the study period based on the change in carbon stocks between the land use type that was converted for oil palm production and the carbon stocks of a typical oil palm plantation (40 tC ha−1)58, and the amount of each land use type converted (Supplementary Tables 2 and 3)55.

$${\mathrm{GHG}}_{\mathrm{con}} = {\sum} {\left( {{\mathrm{ADM}}_i - {\mathrm{ADM}}_{\mathrm{op}}} \right) \times A_i}$$

where i is the land cover type, ADM is the aboveground dry matter (tC ha−1) in land cover type i and in oil palm (op) plantations, and Ai is the annual area converted from land use type i (Supplementary Tables 2 and 3). We assumed that GHGcon occurred during the first year after land conversion57. GHGcon was expressed as CO2 equivalents by multiplying changes in carbon stocks by 3.67. Following57, the additional emissions from peat decomposition for every year after conversion were calculated as follows:

$${\rm{GHG}_{peat}} = {\sum} {\left[ {\left( ({\mathrm{EF}_i + \mathrm{EF}_{\mathrm{op}})/2} \right) \times A_i + \mathrm{EF}_{\mathrm{op}} \times A_{i} \times \left( {t - 1} \right)} \right]}$$

where i is the land cover type, EF is the emission factor in land cover type i and in oil palm (op) plantations, Ai is the annual area converted from land use type i (Supplementary Tables 2 and 3), and t is the time (in years) after conversion. GHGpeat was expressed as CO2 equivalents by multiplying changes in carbon stocks by 3.67. We note that emissions from peatland only consider the peatland area converted into oil palm cultivation after the baseline year (2018).

In the case of mineral soils, the net change in soil carbon stock due to land conversion for oil palm production was assumed to be zero, as it has been reported in the literature based on field measurements59. In relation to the assumption of SOC neutrality in mineral soils, a recent study shows that SOC can decline up to 40% in the topsoil of mineral soils when forest is converted for oil palm cultivation60. Inclusion of this extra source of carbon would have increased our estimated GWP by 33 Mt (+4%) in the BAU scenario due to conversion of forests in mineral soils (Extended Data Fig. 8). For this calculation, we assumed that forests have, on average, SOC of 30 MgC ha−1 in the upper 0.1 m of the soil profile and, out of that, 40% is lost after conversion for oil palm cultivation60.

GHGluc was calculated as the sum of GHGcon and GHGpeat and expressed as CO2 equivalents. When the original land use type had higher carbon stocks compared with oil palm plantation (for example, forests and peatlands), there was a net loss in carbon stocks, and GHGluc emissions had a positive sign (that is, source of GHG emissions). When the original land use change had lower carbon stocks compared with oil palm plantation (for example, annual crops, shrubland, and bare land located in mineral soils), there was a net carbon gain, and GHGluc had a negative sign (that is, carbon sink). In the BAU scenario, oil palm area expanded at the same rate and on the same type of land use as over the 2000–2018 period, including carbon-rich ecosystems such as peatlands and forests, leading to net GHG emissions due to land use change (Extended Data Fig. 6). In the INT scenario, there was no land conversion; hence GHGluc was assumed to be nil. In the INT-TE scenario, the increase in oil palm area occurred in areas that had, on average, 21 tC ha−1 in the ADM in comparison with 40 t ha−1 for an oil palm plantation (Supplementary Table 3), leading to some carbon sequestration as a result of land use change (Fig. 3 and Extended Data Fig. 6). Similar results have been reported in the literature based on experimental data and modelling58,61.

Annual GHG emissions derived from oil palm cultivation (GHGcul) were calculated for each scenario and included those derived from manufacturing, packaging and transportation of agricultural inputs, fossil fuel use for field operations, and soil N2O emissions derived from application of nitrogen (N) fertilizer. In oil palm cultivation, N, phosphorus (P) and potassium (K) fertilizer accounts for 80% of GHG emissions62,63. Hence, we focused here on calculating the GHG emissions associated with NPK fertilizers and then we simply added an extra 25% to our calculation to account for other inputs (for example, pesticides, other nutrient fertilizer) and fossil fuel use for farm operations (for example, harvesting). In calculating GHG emissions associated with manufacturing, packaging and transportation of N, P and K fertilizers, we used specific updated emissions factors for Southeast Asia, selecting those fertilizer sources that are most commonly used for oil palm production64. For the BAU scenario, annual GHG from N, P and K fertilizer was calculated based on the current nutrient fertilizer rates used in Indonesia as derived from the management data for each farmer type across the 22 sites. In the case of large plantations, average fertilizer N, P and K application (expressed as elemental nutrients) averaged 170, 40 and 212 kg ha−1 y−1, respectively. Fertilizer applications were lower in the case of smallholder farmers, averaging 62, 17 and 51 kg N, P and K ha−1 y−1, respectively. Considering that projected yield level is higher in the INT and INT-TE scenarios compared with BAU, with a concomitant increase in nutrient requirements, we followed a balance approach to estimate appropriate fertilizer rates that could support the yield levels projected for the INT and INT-TE scenarios. That approach is typically followed in well-managed plantations to determine the amount of nutrient fertilizer that is needed to attain a given yield level65,66. Following this approach, we assumed that the amount of N, P and K fertilizer to be applied should basically replace the amount of nutrient that is removed from the field with the harvested FFB, after accounting for the nutrients that are stored in the trunk65,66. Soil N2O emissions were calculated assuming an N2O emission factor of 1.6% of the total N fertilizer applied based on the recommended emission factor for mineral soils in tropical regions67, which is consistent with soil N2O emissions measured in large and smallholders' oil palm plantations in Indonesia68,69,70. For each scenario, GHGcul was estimated at national level by multiplying the emissions per unit of area (hectare) by the amount of mature oil palm area in each year. We note that our calculation of GHGcul does not include emissions derived from FFB transportation, milling and processing as these would have been identical among the three scenarios given the similarity in national FFB production by year 2035.

The nutrient balance approach used for estimating nutrient fertilizer requirements in the INT and INT-TE scenarios, together with their lower land requirement, lead to an overall reduction in GHGcul compared with the BAU scenario (Fig. 3). Current nutrient fertilizer input is imbalanced leading to an excess of some nutrients (especially those contained in subsidized fertilizer)38. As a result, GHGcul derived from oil palm production in the BAU scenario are proportionally higher than those in the INT and INT-TE scenarios, despite the similarity in the national CPO output (Figs. 2 to 3 and Extended Data Fig. 6). This result shows the importance of coupling crop intensification with an explicit effort to reduce associated environmental footprint, for instance, by improving the synchrony between nutrient fertilizer inputs and crop nutrient requirements as shown here for the INT and INT-TE scenarios.

Data availability

The data on yield potential and yield gaps that support the findings of this study are publicly available via the Global Yield Gap Atlas website ( Source data are provided with this paper.


  1. 1.

    Pirker, J., Mosnier, A., Kraxner, F., Havlík, P. & Obersteiner, M. What are the limits to oil palm expansion? Glob. Environ. Change 40, 73–81 (2016).

    Article  Google Scholar 

  2. 2.

    Bonan, G. B. Forests and climate change: forcings, feedbacks, and the climate benefits of forests. Science 320, 1444–1449 (2008).

    CAS  Article  Google Scholar 

  3. 3.

    Colchester, M. et al. Justice in the Forest (CIFOR, 2006).

  4. 4.

    Johnson, C. N. et al. Biodiversity losses and conservation responses in the Anthropocene. Science 356, 270–275 (2017).

    CAS  Article  Google Scholar 

  5. 5.

    Seymour, F. & Harris, N. L. Reducing tropical deforestation. Science 365, 756–757 (2019).

    CAS  Article  Google Scholar 

  6. 6.

    First Nationally Determined Contribution Submitted to UNFCCC (Republic of Indonesia, 2016).

  7. 7.

    FAOSTAT (FAO, accessed 1 March 2020);

  8. 8.

    Austin, K. G., Schwantes, A., Gu, Y. & Kasibhatla, P. S. What causes deforestation in Indonesia? Environ. Res. Lett. 14, 024007 (2019).

  9. 9.

    Tree Crop Estate Statistics of Indonesia 2017–2019 (Directorate General of Estate Crops, 2019).

  10. 10.

    Woittiez, L. S., van Wijk, M. T., Slingerland, M., van Noordwijk, M. & Giller, K. E. Yield gaps in oil palm: a quantitative review of contributing factors. Eur. J. Agron. 83, 57–77 (2017).

    Article  Google Scholar 

  11. 11.

    Wilcove, D. S., Giam, X., Edwards, D. P., Fisher, B. & Koh, L. P. Navjot’s nightmare revisited: logging, agriculture, and biodiversity in Southeast Asia. Trends Ecol. Evol. 28, 531–540 (2013).

    Article  Google Scholar 

  12. 12.

    Curtis, P. G., Slay, C. M., Harris, N. L., Tyukavina, A. & Hansen, M. C. Classifying drivers of global forest loss. Science 361, 1108–1111 (2018).

    CAS  Article  Google Scholar 

  13. 13.

    Gaveau, D. L. A. et al. Rapid conversions and avoided deforestation: Examining four decades of industrial plantation expansion in Borneo. Sci. Rep. 6, 32017 (2016).

    CAS  Article  Google Scholar 

  14. 14.

    Srinivas, A. & Koh, L. P. Oil palm expansion drives avifaunal decline in the Pucallpa region of Peruvian Amazonia. Glob. Ecol. Conserv. 7, 183–200 (2016).

    Article  Google Scholar 

  15. 15.

    Byerlee, D., Stevenson, J. & Villoria, N. Does intensification slow crop land expansion or encourage deforestation? Glob. Food Sec. 3, 92–98 (2014).

    Article  Google Scholar 

  16. 16.

    Cassman, K. G. Ecological intensification of cereal production systems: Yield potential, soil quality, and precision agriculture. Proc. Natl Acad. Sci. USA 96, 5952–5959 (1999).

    CAS  Article  Google Scholar 

  17. 17.

    Cassman, K. G. & Grassini, P. A global perspective on sustainable intensification research. Nat. Sustain. 3, 262–268 (2020).

    Article  Google Scholar 

  18. 18.

    Tilman, D., Cassman, K. G., Matson, P. A., Naylor, R. & Polasky, S. Agricultural sustainability and intensive production practices. Nature 418, 671–677 (2002).

    CAS  Article  Google Scholar 

  19. 19.

    Statistics Indonesia (BPS, accessed 1 March 2020);

  20. 20.

    Jelsma, I., Schoneveld, G. C., Zoomers, A. & van Westen, A. C. M. Unpacking Indonesia’s independent oil palm smallholders: an actor-disaggregated approach to identifying environmental and social performance challenges. Land Use Policy 69, 281–297 (2017).

    Article  Google Scholar 

  21. 21.

    Roadmap for the National Oil Palm Industry Towards 2045 (Indonesian cross-ministry team and oil palm institutions and local associations, 2019).

  22. 22.

    The Palm Oil Dilemma: Policy Tensions Among Higher Productivity, Rising Demand, and Deforestation (IFPRI, 2019).

  23. 23.

    OECD-FAO Agricultural Outlook 2020–2029 (OECD, 2020).

  24. 24.

    Lobell, D. B., Cassman, K. G. & Field, C. B. Crop yield gaps: their importance, magnitudes, and causes. Annu. Rev. Environ. Resour. 34, 179–204 (2009).

    Article  Google Scholar 

  25. 25.

    Hoffmann, M. P. et al. Yield gap analysis in oil palm: framework development and application in commercial operations in Southeast Asia. Agric. Syst. 151, 12–19 (2017).

    Article  Google Scholar 

  26. 26.

    Molenaar, J. W., Persch-Orth, M., Taylor, C. & Harms, J. Diagnostic Study on Indonesia Oil Palm Smallholders: Developing a Better Understanding of their Performance and Potential (IFC, 2013).

  27. 27.

    The Future of Food and Agriculture: Trends and Challenges (FAO, 2017).

  28. 28.

    Hoffmann, M. P. et al. Simulating potential growth and yield of oil palm (Elaeis guineensis) with PALMSIM: model description, evaluation and application. Agric. Syst. 131, 1–10 (2014).

    Article  Google Scholar 

  29. 29.

    Euler, M., Hoffmann, M. P., Fathoni, Z. & Schwarze, S. Exploring yield gaps in smallholder oil palm production systems in eastern Sumatra, Indonesia. Agric. Syst. 146, 111–119 (2016).

    Article  Google Scholar 

  30. 30.

    Soliman, T., Lim, F. K. S. S., Lee, J. S. H. H. & Carrasco, L. R. Closing oil palm yield gaps among Indonesian smallholders through industry schemes, pruning, weeding and improved seeds. R. Soc. Open Sci. 3, 160292 (2016).

    CAS  Article  Google Scholar 

  31. 31.

    Grassini, P. et al. How good is good enough? Data requirements for reliable crop yield simulations and yield-gap analysis. Field Crop. Res. 177, 49–63 (2015).

    Article  Google Scholar 

  32. 32.

    Tilman, D. et al. Future threats to biodiversity and pathways to their prevention. Nature 546, 73–81 (2017).

    CAS  Article  Google Scholar 

  33. 33.

    Mitchard, E. T. A. The tropical forest carbon cycle and climate change. Nature 559, 527–534 (2018).

    CAS  Article  Google Scholar 

  34. 34.

    Zabel, F. et al. Global impacts of future cropland expansion and intensification on agricultural markets and biodiversity. Nat. Commun. 10, 2844 (2019).

    Article  CAS  Google Scholar 

  35. 35.

    Barlow, J. et al. The future of hyperdiverse tropical ecosystems. Nature 559, 517–526 (2018).

    CAS  Article  Google Scholar 

  36. 36.

    Indonesia. Second Biennial Update Report. Under the United Nations Framework Convention on Climate Change (Directorate General of Climate Change, Ministry of Environment and Forestry, 2018).

  37. 37.

    Rhebergen, T. et al. Closing yield gaps in oil palm production systems in Ghana through best management practices. Eur. J. Agron. 115, 126011 (2020).

    Article  Google Scholar 

  38. 38.

    Woittiez, L. S., Slingerland, M., Rafik, R. & Giller, K. E. Nutritional imbalance in smallholder oil palm plantations in Indonesia. Nutr. Cycl. Agroecosyst. 111, 73–86 (2018).

    Article  Google Scholar 

  39. 39.

    Corley, R. H. V. & Lee, C. H. The physiological basis for genetic improvement of oil palm in Malaysia. Euphytica 60, 179–184 (1992).

    Google Scholar 

  40. 40.

    Jelsma, I., Woittiez, L. S., Ollivier, J. & Dharmawan, A. H. Do wealthy farmers implement better agricultural practices? An assessment of implementation of Good Agricultural Practices among different types of independent oil palm smallholders in Riau, Indonesia. Agric. Syst. 170, 63–76 (2019).

    Article  Google Scholar 

  41. 41.

    Deininger, K. Challenges posed by the new wave of farmland investment. J. Peasant Stud. 38, 217–247 (2011).

    Article  Google Scholar 

  42. 42.

    Agricultural Innovation Systems: An Investment Sourcebook (The World Bank, 2012).

  43. 43.

    Cock, J. et al. Learning from commercial crop performance: oil palm yield response to management under well-defined growing conditions. Agric. Syst. 149, 99–111 (2016).

    Article  Google Scholar 

  44. 44.

    Jelsma, I., Slingerland, M., Giller, K. E. & Bijman, J. Collective action in a smallholder oil palm production system in Indonesia: the key to sustainable and inclusive smallholder palm oil? J. Rural Stud. 54, 198–210 (2017).

    Article  Google Scholar 

  45. 45.

    Carlson, K. M. et al. Effect of oil palm sustainability certification on deforestation and fire in Indonesia. Proc. Natl Acad. Sci. USA 115, 121–126 (2018).

    CAS  Article  Google Scholar 

  46. 46.

    Sahide, M. A. K. & Giessen, L. The fragmented land use administration in Indonesia: analysing bureaucratic responsibilities influencing tropical rainforest transformation systems. Land Use Policy 43, 96–110 (2015).

    Article  Google Scholar 

  47. 47.

    Presidential Instruction no. 5 (President of the Replublic of Indonesia, 2019).

  48. 48.

    REDD+ (UNFCCC, accessed 1 March 2020);

  49. 49.

    Evans, L. T. Crop Evolution, Adaptation and Yield (Cambridge Univ. Press, 1993).

  50. 50.

    van Ittersum, M. K. et al. Yield gap analysis with local to global relevance—A review. F. Crop. Res. 143, 4–17 (2013).

    Article  Google Scholar 

  51. 51.

    Fairhurst, T. H. & Griffiths, W. Oil Palm: Best Management Practices for Yield Intensification (International Plant Nutrition Institute, Southeast Asia Program, 2015).

  52. 52.

    Global Yield Gap Atlas (University of Nebraska, Wageningen University, accessed 1 March 2020);

  53. 53.

    Hekman, W., Slingerland, M. A., van den Beuken, R., Gerrie, V. & Grassini, P. Estimating yield gaps in oil palm in Indonesia using PALMSIM to inform policy on the scope of intensification. In International Oil Palm Conference (IOPC) (2018).

  54. 54.

    Austin, K. G. et al. Shifting patterns of oil palm driven deforestation in Indonesia and implications for zero-deforestation commitments. Land Use Policy 69, 41–48 (2017).

    Article  Google Scholar 

  55. 55.

    Land Cover Data (Ministry of Environment and Forestry, Indonesia, accessed 1 March 2020).

  56. 56.

    Searchinger, T. et al. Use of U.S. croplands for biofuels increases greenhouse gases through emissions from land-use change. Science 319, 1238–1240 (2008).

    CAS  Article  Google Scholar 

  57. 57.

    National Forest References Emission Level for Deforestations and Forest Degradation (Ministry of Environment and Forestry, Indonesia, 2016).

  58. 58.

    Khasanah, N., van Noordwijk, M., Ningsih, H. & Wich, S. Aboveground carbon stocks in oil palm plantations and the threshold for carbon-neutral vegetation conversion on mineral soils. Cogent Environ. Sci. 1, 1119964 (2015).

    Article  CAS  Google Scholar 

  59. 59.

    Khasanah, N., van Noordwijk, M., Ningsih, H. & Rahayu, S. Carbon neutral? No change in mineral soil carbon stock under oil palm plantations derived from forest or non-forest in Indonesia. Agric. Ecosyst. Environ. 211, 195–206 (2015).

    Article  Google Scholar 

  60. 60.

    van Straaten, O. et al. Conversion of lowland tropical forests to tree cash crop plantations loses up to one-half of stored soil organic carbon. Proc. Natl Acad. Sci. USA 112, 9956–9960 (2015).

    Article  CAS  Google Scholar 

  61. 61.

    Quezada, J. C., Etter, A., Ghazoul, J., Buttler, A. & Guillaume, T. Carbon neutral expansion of oil palm plantations in the Neotropics. Sci. Adv. 5, eaaw4418 (2019).

    CAS  Article  Google Scholar 

  62. 62.

    Harsono, S. S., Prochnow, A., Grundmann, P., Hansen, A. & Hallmann, C. Energy balances and greenhouse gas emissions of palm oil biodiesel in Indonesia. GCB Bioenergy 4, 213–228 (2012).

    CAS  Article  Google Scholar 

  63. 63.

    Archer, S. A., Murphy, R. J. & Steinberger-Wilckens, R. Methodological analysis of palm oil biodiesel life cycle studies. Renew. Sustain. Energy Rev. 94, 694–704 (2018).

    Article  Google Scholar 

  64. 64.

    Brentrup, F., Lammel, J., Stephani, T. & Christensen, B. Updated carbon footprint values for mineral fertilizer from different world regions. In 11th International Conference on Life Cycle Assessment of Food 2018 (LCA Food) (2018).

  65. 65.

    Lim, Y. L. et al. An update on oil palm nutrient budgets. In International Oil Palm Conference (IOPC) (2018).

  66. 66.

    Tiemann, T. T. et al. Feeding the palm: a review of oil palm nutrition. Adv. Agron. 152, 149–243 (2018).

    Article  Google Scholar 

  67. 67.

    Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (eds Calvo Buendia, E. et al.) (IPCC, 2019).

  68. 68.

    Caliman, J. P. N-Fertiliser losses quantification in term of N2O emission and NH3 volatilisation. In Oil Palm Best Practices Workshop (MOSTA, 2019).

  69. 69.

    Meijide, A. et al. Measured greenhouse gas budgets challenge emission savings from palm-oil biodiesel. Nat. Commun. 11, 1089 (2020).

    CAS  Article  Google Scholar 

  70. 70.

    Hassler, E., Corre, M. D., Kurniawan, S. & Veldkamp, E. Soil nitrogen oxide fluxes from lowland forests converted to smallholder rubber and oil palm plantations in Sumatra, Indonesia. Biogeosciences 14, 2781–2798 (2017).

    CAS  Article  Google Scholar 

Download references


We are thankful to the personnel at the former International Plant Nutrition Institute (South East Asia Office) and a number of private plantation companies (SMART PT, Asian Agri, Lonsum and Wilmar) for providing productivity and management data from a number of well-managed blocks across the Indonesian archipelago for model calibration and testing. We are also grateful to large plantations and smallholder farmers’ associations, non-governmental organizations and the Research Center for Climate Change at Universitas Indonesia (RCCC-UI) for their support to the project and useful discussions. We also thank J. Matthews (formerly R&D Head at PT Bumitama Gunajava Agro) and T. Fairhurst (Tropical Crops Consultants Limited) for useful feedback at many stages of this project. This project was funded by the Norwegian Ministry of Foreign Affairs (grant INS-19/0007 to P.G.), with some additional funding from the Global Engagement Office at the Institute of Agriculture and Natural Resources, University of Nebraska-Lincoln.

Author information




P.G., M.A.S., S.R., F.A., T.O. and T.F. conceived the project. F.H., I.P., D.K.G.P. and Y.L.L. collected the data. W.H. and R.v.d.B. ran the model simulations. P.G., J.P.M., F.A., J.F.A., A.C., J.I.R.E., C.R.D. and H.S. analysed the data. P.G. and J.P.M. wrote the manuscript with input from all authors.

Corresponding author

Correspondence to Patricio Grassini.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature Sustainability thanks Wan Yee Lam, Ana Meijide and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Water-limited yield potential (Yw), attainable yield (Yatt), and average actual oil palm yield (Ya).

The Yatt is calculated as 70% of Yw. The exploitable yield gap (arrow) is calculated as the difference between Yatt and Ya.

Extended Data Fig. 2 Scheme illustrating the methodology used to build the yield gap atlas for oil palm in mineral soils in Indonesia.

Yields are expressed as fresh fruit bunches (FFB) per hectare per year. Yatt: attainable yield.

Extended Data Fig. 3 Selected 22 sites (yellow circles) and associated buffer zones (polygons with red borders).

Note that buffers’ borders are irregular as they were clipped by the borders of the climate zone where each buffer is located. Mature oil palm area located in mineral soils is shown in green. Lines show administrative boundaries. Details on each specific buffer are shown in Extended Data Table 1. Inset shows the location of the study area within Indonesia.

Extended Data Fig. 4 Scheme illustrating estimation of water-limited yield potential using different sets of weather data to account for variation in water-limited yield potential at a given plant age due to weather variation.

Separate sets of simulations were started at different years, filling the missing years at the end with the early years of the weather file if needed.

Extended Data Fig. 5 Scheme showing estimation of the exploitable yield gap for large plantations and smallholders.

Water-limited yield potential (Yw; blue solid line), attainable yield (Yatt; green solid line), average actual yields (solid triangles), and exploitable yield gaps (red arrows) are shown. The Yatt was estimated as 70% of Yw at a given age.

Extended Data Fig. 6 Projected trends in accumulated land use change and associated global warming potential (GWP).

Accumulated high- and low-carbon land converted for oil palm production (a, b) and associated accumulated GWP (c) during the study period (2018–2035) for three scenarios: (i) business as usual (BAU), with historical trends in area and yield remaining unchanged in the future; (ii) intensification (INT), with complete closure of the exploitable yield gap in current plantation area in mineral soils and without physical expansion of oil palm area; and (iii) intensification plus target expansion (INT-TE), with partial closure of exploitable yield gap and oil palm area expansion into low-carbon land.

Source data

Extended Data Fig. 7 Comparison of simulated and measured water-limited yield potential and attainable yield.

In the case of water-limited yield potential (Yw), simulated values correspond to those derived from crop modelling in this study (see Extended Data Table 1) while published data (PD) correspond to highest recorded yields in plantations located in Southeast Asia as reported in the oil palm literature (see Supplementary Information, Section IIa). In the case of attainable yield (Yatt), values were estimated as 70% of simulated Yw in this study (see Extended Data Table 1) while large plantation (LP) values were derived from long-term yield records from 14 well-managed commercial blocks in Indonesia as provided by a number of private large plantations companies, including the four blocks used for calibration. Boxes indicate the 25th and 75th percentiles, whiskers represent the 10th and 90th percentiles, and means are shown with crosses. Also shown is the sample size (n), statistical significance for the differences and degrees of freedom (d.f.) for the comparison between the values reported in this study versus those reported in the literature (left) or provided by large plantations (right) using unpaired two-tailed Student’s t-test. All variables were normally distributed (D’Agostino’s test; p > 0.30).

Source data

Extended Data Fig. 8 Global warming potential (GWP) associated with different scenarios of intensification and land use change and different assumptions in relation to changes in soil organic carbon (SOC) in mineral soils.

The GWP during the study period (2018–2035) was estimated for three scenarios and two different assumptions in relation to SOC changes in mineral soils after land conversion for oil palm cultivation: (i) no change in SOC following Khasanah et al.59 and (ii) 40% decline in SOC in the topsoil after conversion of primary or secondary forest for oil palm cultivation following van Straaten et al.60. For the latter, we assumed no change in SOC in mineral soils when non-forest land is converted for oil palm cultivation; hence, GWP in the other two scenarios (INT and INT-TE) remained unchanged. Dashed portion of the bar indicates the increase in GWP as a result of including changes in SOC in mineral soils when forest is converted for oil palm cultivation. We note that GHG emissions derived from peat decomposition are accounted for in the calculation of GWP, regardless the assumption on SOC changes in mineral soils.

Source data

Supplementary information

Supplementary Information

Supplementary text and Tables 1–3.

Source data

Source Data Fig. 1

Raw data used to create figure.

Source Data Fig. 2

Raw data used to create figure.

Source Data Fig. 3

Raw data used to create figure.

Source Data Extended Data Fig. 6

Raw data used to create figure.

Source Data Extended Data Fig. 7

Raw data used to create figure.

Source Data Extended Data Fig. 8

Raw data used to create figure.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Monzon, J.P., Slingerland, M.A., Rahutomo, S. et al. Fostering a climate-smart intensification for oil palm. Nat Sustain 4, 595–601 (2021).

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing