Introduction

In the last twenty years, large areas of Indonesian peatland have been converted, mainly into agricultural lands for estate crop production, and plantation forest areas for pulp production (Rehman et al. 2014; Gunarso et al. 2013; Miettinen et al. 2011; Koh et al. 2011; Murdiyarso et al. 2010). This conversion brought short-term economic gains, but poses major environmental and economic risks, resulting from health and economic damages due to peat fires, soil subsidence potentially leading to flooding of millions of hectares of coastal peat lands in the course of the next decades, the very large CO2 emissions from burning and oxidising peat, and from the loss of globally significant biodiversity contained in natural peat swamp forests (Wösten et al. 2008; Page et al. 2011; Joosten et al. 2009; Turetsky et al. 2015; Hooijer et al. 2012). For instance, drained peatland in Indonesia contributes 58% of global peatland CO2 emissions, with marked spike during El Nino years when emissions from fire are particularly high (Hooijer et al. 2006). In addition, a range of social issues have been related to peat conversion such as the loss of access to land of traditional forest users (Thorburn and Kull 2013; Sumarga et al. 2016).

A number of Indonesian national policies aim to enhance peatland management, for instance the Ministry of Agriculture Decree No. 14 year 2009 which prohibits oil palm establishment in peatlands with more than 3 m depth. The Indonesian government has also established a strict moratorium on peat conversion since 2011. Yet, this decree is often not effective because of a lack of enforcement at the level where many of the land-use decisions are taken i.e. at village, district and provincial level (Boer et al. 2012). The national policy on peatland management has not yet been widely translated into sub-national regulation, in part due to the lack of knowledge of local policy makers on short and long-term economic, social, and environmental consequences of different land use types. Nevertheless, in recent years, the emerging insights in the consequences of peat degradation including burning (e.g. World Bank. 2016; Turetsky et al. 2015), a better understanding of the various benefits provided by peatland ecosystems and their links to the stakeholders (e.g. Suwarno et al. 2016; Sumarga and Hein 2015) as well as new payment mechanisms (e.g. REDD+) have influenced the Indonesian playing field for peatland management. This may increase the chance for a transition to sustainability.

Although the potential effects of changes in Indonesian peatlands are now increasingly well understood, there is still no consensus on the economic benefits provided by peatlands under different uses at the scale of the country. This is important also in view of the major differences in peat development between the three major islands of the country that contain peat i.e. Sumatra, Kalimantan and Papua. In order to establish the effectiveness of proposed new policies on peatlands, based upon presidential guidance (PP 71 year 2014) and more recently articulated policy instructions (the direction of the President Republic Indonesia, on forest and peatland fires in a coordination meeting on 18 January 2016) a baseline assessment is needed of the current status of peatlands and the trends in their use.

The objective of this study is to analyse peatland uses and the ecosystem services supplied, the key sustainability issues, and the potential response options to move towards sustainability. We conduct a literature review and conduct spatial analysis to analyse peatland use in the period 2000–2014 in the three main islands (Sumatra, Kalimantan, and Papua). We specifically discuss the uncertainties in the current datasets on peat, which is one of the main barriers for effective policy implementation. The novelty in our paper is in bringing out economic benefits and sustainability issues in Indonesian peatlands in one paper, and in the review we conduct of the often contradicting datasets on this issue. We also propose a basic framework for identifying peatland management options.

Materials and methods

Study area

We specifically focused on peatland areas covering Sumatra, Kalimantan, and Papua covering 16 provinces of in total 33 provinces in Indonesia. This includes 10 provinces in Sumatra (Aceh, North-Sumatra, West-Sumatra, South-Sumatra, Riau, Kepulauan Riau, Jambi, Kepulauan Bangka Belitung, Bengkulu, and Lampung), 4 provinces in Kalimantan (West-Kalimantan, Central-Kalimantan, South-Kalimantan, and East-Kalimantan), and 2 provinces in Papua (Papua, and West-Papua). These three main islands together comprise the large majority of Indonesian peatlands. We show that these islands experience entirely different trends in the conversion of peatlands.

Trends in peatland use and ecosystem services

We first analysed peatland cover and subsequently we link these changes in peatland use to changes in ecosystem services provided by peatlands. We overlaid the Indonesia Land Cover Map for year 2000, 2003, 2006, 2009, 2012, 2014 produced by the Ministry of Forestry Republic Indonesia (MoFRI 2014) with the 2011 Indonesia Peatland Map Scale 1:250,000 produced by Balai Besar Sumber Daya Lahan Pertanian (BBSDLP) the Ministry of Agriculture Republic Indonesia (Ritung et al. 2011). The land cover map contains 23 land-cover classes; and for the purpose of this study the classes were reclassified into 10 land cover classes, namely undisturbed natural forest, disturbed natural forest, plantation forest, estate crop, degraded land, paddy field, dryland agriculture, urban, open water, and other uses. We considered all primary forest as ‘undisturbed natural forest’ class and all secondary forest as ‘disturbed natural forest’ class. In addition, dry shrub, wet shrub, savanna, grasses, and open swamps areas are presented as ‘degraded land’ class (based on Law et al. 2015). Agriculture areas for food crops are classified into ‘dryland agriculture’ class and ‘paddy field’ class, in which dryland agriculture class consist of pure and mixed dryland agriculture areas. The ‘other uses’ class is classified by aggregating fish pond/aquaculture areas, mining areas, port & harbour areas, and also cloud & no-data. In particular for analysing biodiversity habitat (protected areas), we overlaid this output with maps of protected areas produced by the Ministry of Forestry Republic of Indonesia. To analyse the trends in Indonesian peatland-use, we only considered peatland with the peat depth of at least 50 cm (Krisnawati et al. 2015) with an estimation of the total area around 14.9 million hectares (Ritung et al. 2011), although there is still uncertainty on the exact peat area and boundaries. We discussed uncertainty of the peatland data in the “Discussion” section of our paper. All spatial analyses were done with help of ArcGIS 10.2.

Next, we quantified seven ecosystem services i.e. timber production, oil palm production, biomass production for pulp, paddy production, carbon sequestration, biodiversity habitat, and ecotourism. These selected services are the most relevant ecosystem services in Indonesian peatland (Law et al. 2015; Sumarga and Hein 2014). The performance indicators, sources of data, and assessment methods for quantifying the flow of the six selected ecosystem services (excluding carbon sequestration) are described in Table 1. Note that for oil palm plantation and plantation forest areas in Indonesian peatlands, we used data for the three islands recorded by various sources (see Appendices 3 and 4 Tables 8, 9). Note also that we only considered forested areas inside protected areas in analysing biodiversity habitat given the difficulties and the high potential uncertainty in identifying habitat outside protected areas. This latter restriction is also based on the assumption that most of the forest outside (and to some extend also inside, in particular in Sumatra and Kalimantan) the national parks have been moderately to severely degraded due to in particular timber harvesting and slash and burn cultivation (Biancalani and Avagyan 2014; Posa et al. 2011). We acknowledge that there are many more ecosystem services provided by Indonesian peatlands (see e.g. Suwarno et al. 2015) but due to a lack of data we focus on the aforementioned services. We discuss the implications of our limited selection of services in the “Discussion” section.

Table 1 The physical units of selected ecosystem services
Full size table

The quantification method for carbon sequestration requires further explanation. We quantified carbon sequestration (a service) and carbon emissions (a disservice) based on the net carbon (CO2) flux of different types of peatland use, derived from several previous studies as listed in Appendix 1 (Table 6). The net carbon flux may be positive (sequestration higher than emissions) or negative (emissions higher than sequestration). We quantified the net carbon flux of eight peatland uses: undisturbed natural forest, disturbed natural forest, plantation forest (referred to acacia plantation), oil palm plantation, agricultural crops, shrubs (degraded lands), water, and other land uses (referred to degraded lands), with values ranging from −85 ton CO2/ha/year (in oil palm plantation, assuming a drainage depth of 90 cm) (Hooijer et al. 2010) to 19 ton CO2/ha/year (in undisturbed natural forest) (Suzuki et al. 1999). Except for undisturbed natural forest and water, we assumed that the areas are drained. As shown in Appendix 1 (Table 6), the net carbon fluxes are negative in most types of peatland use in Indonesia, indicating that what ecosystem provides in those land uses is a disservice. We multiplied the area of each peatland use with its net carbon flux data, and finally aggregated them all to derive the estimate of carbon sequestration at national level from 2000 to 2014.

Analysis of sustainability issues in peatland

Based on a literature review and supported by our spatial analysis, we analyse the key sustainability issues related to the current use of Indonesian peatlands. In particular, we include the following issues in our study: fires and smoke, peat soil subsidence and flood risks, CO2 emissions (based on our analysis described above), loss of habitat, and social conflicts. We propose a general framework to order these sustainability issues. In this framework, we distinguish between four types of peatland condition: (1) forest use, drained; (2) forest use, no drainage; (3) agricultural use, drained; and (4) agricultural use, no drainage. With forest use is meant that the peatlands are not used for cropping systems including plantation crops or agroforestry and that they may be productive as forest systems with logging or supplying other ecosystem services (e.g. non-timber forest products, water regulation, carbon storage and sequestration), or that they may be degraded with little vegetation left. In the latter case the potential for rehabilitation to peat swamp forest exists, but rehabilitation may be hampered by recurrent fires that burn tree seedlings. The sustainability issues differ markedly for these categories as we will explore in our study. This has also repercussions for policy making, for example if peatlands are brought from the condition of productive use with drainage to non-productive use with drainage, for example because oil palm plantations are retired without subsequent peat rehabilitation including reducing drainage levels, this will not necessarily lead to sustainable peatland use. We also explore how these sustainability issues differ for the three islands that we consider.

Results

This section presents the results of our spatial analysis on peatland use and ecosystem services as well as our literature review on sustainability issues related to Indonesian peatlands. These sustainability issues are a consequence of the land use conversion to which the peatlands have been subject.

Indonesian peatland use

The distribution of the land cover in Sumatra, Kalimantan and Papua since 2000 reveals major changes in the use of Indonesian peatlands (see Table 2). Our study shows an ongoing, rapid conversion of natural forests to other land use in particular plantation crops (in particular but not only oil palm) and plantation forestry (in particular Acacia crassicarpa for pulp production). Indonesian tropical peatland occupied by disturbed and undisturbed natural forests decreased from about 9 million hectares in 2000 to about 6.4 million hectares in 2014. However, there is virtually no undisturbed peat swamp forest remaining in Sumatra and Kalimantan, i.e. all remaining undisturbed peat swamp forest is in Papua (where deforestation has been rapidly increasing in the last years). The fastest increase in land cover was related to expansion of oil palm plantations in Indonesian peatlands, which increased from about 700 thousands hectares in 2000 to almost 2 million hectares in 2014. Note that our figures are based on government data supplemented with data from industry for oil palm plantations. The figures are uncertain (see Appendices 2 and 3 Tables 7, 8 for more detailed assessment of uncertainties) and are likely to be conservative because new plantations are not immediately reflected in government statistics.

Table 2 Peatland-use area (in thousands of hectares) based on land cover type in Indonesia since 2000 according to government data and various sources
Full size table

Our analysis also shows major differences in land conversion between the three islands. The highest conversion of natural peat swamp forest took place in Sumatra (Fig. 1). Natural peat swamp forest has decreased from 51% (of which only 6% is undisturbed forests) of Sumatran peatlands in 2000 to only 17% in 2014 (of which 4% undisturbed forests, all located in protected areas). Recent years also show conversion of protected areas to plantation crops including in for example substantial encroachment in Berbak National Park in Jambi, Sumatra. Kalimantan takes an intermediate position with conversion of peatland to plantations still ongoing. In Kalimantan there are also large areas of degraded peatland, drained but not covered by plantations. These areas increased from 28% in 2000 to 35% of peatlands in 2014. As discussed in the next section, this has major repercussions for sustainability issues including peat fires. Most of the remaining peat swamp forests are in Papua. An issues is that data is particularly scarce and uncertain in Papua, for instance there are very few remote sensing based studies that we found with which we can compare government data. We compare our findings with other studies in the “Discussion” section, as well as in Appendices 2 and 3 (Tables 7, 8).

Fig. 1
figure 1

Trends of peatland use in Indonesia since 2000 (based on government data and various sources)

Full size image

Ecosystem services provided by Indonesian peatland

Table 3 shows estimates of the dynamics of ecosystem services provided by Indonesian peatland since 2000. The details of the ecosystem services data used for this analysis are presented in Appendix 1 (Table 6). The conversion of natural peat swamp forests to oil palm and plantation forest led to an estimated almost 50% decrease of timber production within 14 years (2000–2014), and a significant increase of CPO production (almost threefold) and biomass production for pulp (more than 20-fold), followed by a 3% decrease of paddy production during that period. Carbon emissions from peat nearly doubled in the period 2000–2014, to 105 million ton C per year or 385 million ton CO2 per year. This compares to emissions from other sources (e.g. households, industry) of around 595 million ton CO2 per year for Indonesia (DNPI 2010). Peatland deforestation also leads to loss of protected habitat with an average annual loss of about 8.6 thousands hectares. This reflects illegal encroachment in the protected forest areas. In 2014, around 28% of the total protected areas in peatlands in Indonesia were converted already. This protected peat swamp forest areas cover 17% of total peatland areas in Indonesia. For ecotourism, we calculated the number of visitors who visit national parks and recreation parks in peat. Our analysis shows a 21% increase of total number visitors from 97 thousands people in 2000 (of which 1% foreigners); to approximately 117 thousands people in 2014 (of which 33% foreigners). This reflects only 3% of total number visitors to all conservation parks in Indonesia during this period—given the specific biodiversity of peatlands this is relatively low but it may be relate to a lack of tourism infrastructure in peat areas where such infrastructure (e.g. boardwalks) is expensive to construct and maintain.

Table 3 Ecosystem services provided by Indonesian peatland since 2000
Full size table

Sustainability issues in Indonesian tropical peatland management

Table 4 summarizes the sustainability issues in Indonesian peatlands. Note that degradation may occur in under non-productive uses. Peatland areas with draining lead to abandoned areas, while peatland areas without draining remain as forest use areas. Shrubs, herbs, ferns or grasses are typically vegetation in abandoned areas which also categorized as degraded lands (Law et al. 2015) and having none of services and absent of Non Timber Forest Products (NTFPs). Peat swamp forest areas provide services like timber production and NTFPs, carbon stocks, biodiversity habitat, ecotourism, cultural services, etc. (Biancalani and Avagyan 2014).

Table 4 Sustainability issues in Indonesian peatlands
Full size table

Peatland areas under productive uses, with or without draining conditions, provide crop production services, including oil palm plantations, paddy fields, other horticultural lands (in drained areas); and paludiculture crops plantations (in non-drained areas) such as jelutung (Dyera spp.), sago palm (Metroxylon sagu), illipe nut (Shorea spp.), melaleuca, rattan, etc. Paludiculture is biomass cultivation in wet and or rewetted conditions (Biancalani and Avagyan 2014; Giesen 2013). Acacia plantations in peatlands are included as productive use with draining that provided biomass production for pulp service (Joosten et al. 2012).

Fires and smoke

Fire and smoke occur through the burning of drained peat. Fire may involve burning of both above ground biomass and below ground peat. Often, Indonesian peat fires are the result of deliberate or accidental human interventions (Glover and Jessup 2006; Harrison et al. 2009). Plantation companies as well as smallholder farmers may deliberately use fire to clear land with the associated benefit that the ashes increases the pH of the otherwise acidic peat soils (Islam et al. 2016). In some cases, fire may be started accidentally or spread beyond the areas in which it was ignited (Harrison et al. 2009). Once started, fires in drained peat can spread easily (Miettinen et al. 2012; Turetsky et al. 2015). Peat swamp forests and other lands with wet conditions seldomly burn (Turetsky et al. 2015). Peat fires have been reported on drained unused land, on drained peat used for wood pulp and paper (in particular in Sumatra) and on drained land used for oil palm plantations (Marlier et al. 2015b). Peat fires contribute strongly to CO2 emissions and also cause smoke and haze (Marlier et al. 2015a; Heil et al. 2007). Because of often incomplete burning, the smoke contains a mixture of various gases including carbon monoxide, carbon dioxide, methane, ammonia, hydrogen cyanide, benzene, toluene, ethylbenzene, xylenes, formaldehydes, nitrous oxide, mono-nitrogen oxides, ethane, propone, butane, acrolein, acid gases, and particulate matter (PM or soot) (Stockwell et al. 2016; Gaveau et al. 2014; Heil et al. 2007). In the dry season, in particular during EL Nino years, smoke can cover major parts of Indonesia and even neighboring countries (Islam et al. 2016), with associated effects on human health. Reported impacts include negative health effects (acute and chronic), disruption on tourism, transport, and business, reduced enjoyment of life, contribution to the production of ozone, acid rain, and greenhouse gases, and reduced photosynthesis in plants by blocking some solar radiation (World Bank 2016). The cumulative impacts of (sequential) peatland fires, in combination with other disturbance factors such as forest conversion and peat subsidence, lead to the extinction and irreversible changes in forest species composition and vegetation structure and the disappearance of peat (Glover and Jessup 2006). Indonesia government data stated in World Bank (2016) indicated that during the fires from June to October 2015 about 2.6 million ha of land burned in Indonesia, of which 33% was peatlands. The total costs of the fires were estimated at IDR 221 trillion (USD 16.1 billion) (World Bank 2016). About 500 thousand people were hospitalized and other thousands people suffered including people in neighborhood countries Malaysia and Singapore.

Soil subsidence and flood risks

Soil subsidence occurs when peatlands are drained. Soil subsidence rates can be as high as 1.5 m in the first five years after the drainage and 3-5 cm in subsequent years as observed in drained peatland for acacia and oil palm plantation in Sumatera with a typical water table depth of about 70–90 cm (Hooijer et al. 2012). Subsidence is a consequence of both the physical drainage of the water (in particular in the first 5 years) as well as the chemical oxidation of dry peat. We assess (see Table 2) that there is about 4 million ha of drained peatland in Indonesia (in 2014), within the land cover types plantation forest, estate crops, dryland agriculture, paddy fields, and other uses. Other sources mention that about 7–12 million ha of peat is drained (Hooijer et al. 2010; Joosten et al. 2012; Miettinen et al. 2016). Consequently, soil subsidence leads to flood risks because many Indonesian peatlands are situated in coastal lowlands which will also be affected by sea-level rise because of climate change (Dommain et al. 2011; Hooijer and Vernimmen 2013). Soil subsidence progressively affects the possibility to use peat for crop production (Sumarga et al. 2016). Although water management involving 40–60 cm drainage levels has been promoted as best practice (Lim et al. 2012), this still involves considerable and irreversible peat subsidence (Sumarga et al. 2016). Peatland uses that do not require drainage (e.g. paludiculture crops) substantially lower the risk of subsidence (Joosten et al. 2012). Note that our assessment indicates that drainage of peatlands is still ongoing on all three islands, since new crop (including oil palm and Hevea rubber) and forestry (including Acacia) plantations require drainage.

CO2 emission

Carbon emission results from peat fires and peat oxidation (Hirano et al. 2007). Drained peat swamp forests for other peatland uses contribute to peat fires events and increasing peat oxidation that related to increase of CO2 emission (Hooijer et al. 2010), while the increased frequency and duration of flooding will slow down the processes of oxidation and subsidence (Biancalani and Avagyan 2014). Our calculation for CO2 sequestration in Table 3 shows that the historical emission from Indonesian peatland uses i.e. disturbed forests, plantation forests, oil palm plantations, agriculture crops (paddy fields and dryland agriculture areas), degraded lands, urban and other uses areas increased over time to almost 400 million ton CO2 per year in 2014. Sumatra is still the biggest emitter, contributing around 70% of the total carbon emission from Indonesian peat.

Loss of forest in protected areas

Forests are recognized as habitats with high biodiversity. Conversion of peat swamp forests to other land uses is associated with habitat loss and fragmentation affecting a range of endemic animal and plants species (Miettinen et al. 2012; Posa et al. 2011; Yule 2010). Given that many lowland forests on mineral soils have been converted to other land uses, in particular to oil palm plantations (e.g. Sumarga and Hein 2015; Sumarga et al. 2016), peat swamp forests are the last remaining refugium for a range of species including the Sumatran tigers and rhino, and including species that occur in peat but prefer forests on mineral land such as the orangutan. Logging and fire are additional pressures on biodiversity. In our study area, there are about 2.6 million hectare of protected peat swamp forests (equal to 17% of total Indonesian peatland areas). Based on our analysis, plantation forests and crop areas are also found inside these protected areas covering about 28% of land designated as protected area in 2014 (Fig. 2), which we interpret to be the result of illegal forest encroachment. This occurs in particular in Sumatra and Kalimantan, such as in Sembilang and Danau Sentarum National Parks.

Fig. 2
figure 2

Map of habitat inside protected areas in Indonesian peatlands in 2014 (insert area: Danau Sentarum National Park, West Kalimantan)

Full size image

Social conflicts

In Indonesia, social conflicts related to land use are often triggered by overlapping land ownership or land use rights. This is the result of a lack of consistent national base map integrating cadaster information, land use, concessions applied for or granted, etc., in combination with sometimes opaque procedures involving a range of government agencies (Goldstein 2015; Galudra et al. 2011, 2014; Marlier et al. 2015b). Indonesia has about 8 sectoral maps of government agencies that have the authority to make their own sectoral maps for their own purposes (e.g. Ministry of Forestry with forestry maps for determining forestry areas, Ministry of Agriculture with maps of standard competence of agriculture human resources in order to support allocating land for agriculture purposes, etc.). We analysed maps from several government agencies and noted that they were indeed different, even though they covered the same subject matter such as forestry, conservation, mining areas, etc. The different outputs of these maps lead to conflicts between different companies but more often between companies and local residents whose traditional land use rights are often set aside by new permits and concessions. However, there are differences between the islands. For instance in Sumatra, there is increasing competition between companies (acacia and oil palm plantation) and local people (both transmigrants and indigenous) who also want to start or expand oil palm plantations (including on peat). This is related to the increasing scarcity of mineral land available for new plantations. On a specific occasion, local people protested outside the Regency Forest Agency until they were granted a concession to plant oil palm inside a protected area (Galudra et al. 2014). In Kalimantan, for instance, there are reports on conflicts between local communities who started to reclaim peatlands based on customary/tribal right, whereas the central and local governments used a different interpretation of the legality of different management regimes (Galudra et al. 2011; Suwarno et al. 2015). In Papua, conflicts on forestland utilization and concessions occurred due to overlapping regulations issued at the national level, provincial level, and district level leading to protests and human right violations against the local indigenous people (Hidayat et al. 2014). Hence, the pressure of land and the culture differ between the islands, but the lack of transparency in allocating land is a common factor.

Discussion

Uncertainties in baseline data

There is much uncertainty related to the occurrence of Indonesian peatland. The absence of common definitions, measurement techniques and other peatland-related information (forest status or intensive converted peatlands) leads to major differences in the various estimates of the area covered by Indonesian peatland. In this study, we considered peatland with at least 50 cm peat depth, however the lack of data on peat depth in many parts of the country means that this boundary is often highly uncertain. Studies reporting on the area covered by Indonesian peatlands, provide a considerable range from 12 to 26.4 million ha (see Appendix 2 Table 7). There are also substantial differences in the maps of peatland distribution in Indonesia, including the maps published by BBSDLP Ministry of Agriculture (Ritung et al. 2011), Wetlands International (Wahyunto and Suryadiputra 2008), and the Ministry of Environment (MoEFRI 2015). These different maps reflect the potential uncertainty related to estimation of both Indonesian peatland area and its spatial distribution (see Appendices 2–4 Tables 7, 8, 9), and the uncertainty propagates when it is combined with other sources of data, for example to estimate ecosystem services provided by multiple uses of peatland as analysed in this study.

We estimate ecosystem services supply based on data on land use in peatlands from a range of sources but in particular from Indonesian government data. Estimates of visitors to national parks, forest production, paddy production, acacia production are from the Indonesian government, and are generally based on survey and census data. The area covered by oil palm was analysed using remote sensing (Gunarso et al. 2013) in a study for the RSPO and we believed this to be more up-to-date than Indonesian government data. We were not able to map the spatial diversity of the supply of these services, for example forest timber production will not be equally spread over the different peat swamp forests but depend upon forest quality and species composition. Given the status of Sumatran and Kalimantan lowland forests (MoFRI 2014) it is likely that currently the majority of timber production takes places in Papua. We may also underreport the supply of specific services. For instance, oil palm productivity in Indonesia ranges from 4 to 8.6 ton Crude Palm Oil (CPO)/ha/year according to World Growth (2011) whereas census data from BPS (2000–2014) indicates an average yield of between 3 and 4 ton CPO/ha/year (depending upon the year).

The uncertainty in peat cover, and in particular in peat depth and the current land use on peat makes the implementation of policies at the local level very difficult. The various government agencies involved in evaluating applications for concessions sometimes lack accurate and up-to-date information on peat location, peat depth, existing concessions and pending concessions applications. By preparing an updated national peat map, the current One-Map policy by Indonesia government may improve the basic data as a basis for decision making (Wibowo and Giessen 2015).

Policy recommendations

The Indonesian government has voluntary pledged in 2009 to reduce GHG emissions nationally 26% by its own efforts, and up to 41% with international assistance in 2020. A more ambitious target was unveiled in 2015, specifically GHG emissions reduction up to 29% by 2030 (INDC 2015). To support these targets, the Indonesian government published government regulation PP number 71 year 2014 on the protection and management of peat ecosystems. This regulation mandated a maximum water drainage in peat of 0.4 m where appropriate. This has the potential to reduce emissions by around 60 tonnes of CO2/ha/year if applied, however the challenge is that in practice it is extremely difficult to maintain the water level in large areas, year round, at this level. The level is also very close to when crops will start experiencing flood damages, and hence it may be very difficult for plantations in peat to implement this water level. In addition, even a drainage of 0.4 m still leads to soil subsidence. Hence, we believe that whereas this is a welcome initiative, it will not be sufficient to safeguard peat from fires and soil subsidence. Our analysis of Indonesian peatland conditions points to four main potential approaches for Indonesian peatland use depending upon their condition (Table 5).

Table 5 Policy priorities for sustainable peatland uses in Indonesia
Full size table

Paludiculture crops (e.g. jelutung, sago palm, etc.) are crops that do not require drainage and therefore pose much lower fire risks, CO2 emissions and enable cropping over the long-term given that there is no soil subsidence. However, currently they are less financially attractive compared to oil palm and rubber productions (Giesen 2013; Joosten et al. 2012; Sumarga et al. 2016) and therefore their cropping will depend upon policies and regulations that limit growing the crops that require drainage in peat. We also note that the ‘traditional’ crops such as oil palm have benefitted from a long period of breeding and value chain development, which is still in its infancy for the paludiculture crops. From an economic perspective, i.e. when the costs of externalities such as CO2 emissions, health effects, soil subsidence and loss of productive land in the longer term are considered (e.g. World Bank 2016), paludiculture crops such as jelutung already are more profitable than oil palm and Hevea rubber on peat (Sumarga et al. 2016).

Conclusion

Indonesian peatlands have increasingly been converted for agricultural and plantation forest purposes in particular for oil palm, acacia and rubber. In the process, ecosystem services provided by peat swamp forest (e.g. carbon sequestration, biodiversity conservation) have been replaced by the production of agricultural commodities. The highest conversion of natural peat swamp forest took place in Sumatra. In Kalimantan conversion started later, and some peat swamp forest is still remaining—but the island is undergoing rapid land use change at the moment. Most of the remaining peat swamp forests are in Papua, where unfortunately there is also the largest lack of reliable information on forests and peatlands. On the positive side, this has led to major increases in palm oil production (nearly a factor 3 increase in production on peatlands between 2000 and 2014) and biomass production for pulp (a factor 20 increase in the same period). On the negative side, these production levels are not sustainable since progressive soil subsidence will lead to seasonal flooding of the drained plantations in the coming decades ensuring that they will need to be taken out of production (e.g. Sumarga et al. 2016). In addition there are significant externalities related to peat fires and health problems, CO2 emissions and loss of habitat. To move towards sustainability, alternative peat development scenarios should be developed, which should involve a gradual phasing out of oil palm and other drained crops on peat and replacing them by crops that do not require drainage in combination with forestry including timber and non-timber forest production.