Forest fragmentation reduced carbon storage in a moist tropical forest in Bangladesh: Implications for policy development

Forest fragmentation is continued to be widespread in the tropics resulting in reduced ecosystem services including carbon storage. However, the effect of forest fragmentation is not considered in the current carbon policy. We investigated the effect of forest fragmentation on tree biomass carbon and soil organic carbon (SOC) storage in a moist tropical forest in Bangladesh. Above and below-ground tree biomass carbon were calculated by using widely accepted allometric equations and SOC was measured by sampling soils up to 10 cm depth and analyzing them in a soil laboratory. Results showed that carbon storage in tree biomass was significantly lower in fragmented forests (16.3 ± 1.37 t C ha⁻¹) than in contiguous forests (31.21 ± 2.75 t C ha⁻¹) (p < 0.001). Likewise, a significantly lower SOC was contained in the soils of fragmented forests (17.26 ± 0.83 t C ha⁻¹) than in contiguous forests (21.62 ± 0.78 t C ha⁻¹) (p < 0.001). Thus a total of 36% less carbon retained in tree biomass and soils in fragmented forests than in contiguous forests. Backward multiple linear regression analysis revealed tree density, tree height, tree DBH, height-diameter ratio (H/D) and tree species richness as influential factors of carbon variation in fragmented forests. All these structural parameters except tree species richness were significantly lower in fragmented forests, were positively associated with carbon storage and explained together 69% of the carbon storage variation. These findings suggest that the altered stand structure and tree allometry likely caused reduced carbon storage in fragmented forests and highlight the importance of landscape scale management intervention in the tropics. Here, we provided with the evidence of strong negative impact of forest fragmentation on carbon storage and argue that this effect should be in consideration which is currently overlooked in existing carbon accounting systems for tropical forests.     

Heterogeneity and assessment uncertainties in forest characteristics and biomass carbon stocks: Important considerations for climate mitigation policies

The management of forests to store carbon and mitigate climate change has received significant international attention during the last decade. Using in situ data from a 2008–2009 forest inventory field campaign in Sri Lanka, this study describes the structural characteristics and carbon stocks of six natural forest types. This paper has a dual scope: i) to highlight the variation in carbon stored in aboveground biomass within and between forest types and ii) to determine the implications of the allometric equations chosen to calculate biomass carbon stocks. This study concerns work related to climate change interventions, such as Reducing Emissions from Deforestation and Forest Degradation (REDD+) and other forest-related, performance-based initiatives that require proper monitoring, reporting, and verification of carbon stocks, sinks and emissions. The results revealed that forests are heterogeneous in terms of tree density and height–diameter relationships, both between and within the six forest types investigated. The mean aboveground carbon stock in the different forest types ranged from 22 to 181 Mg C ha⁻¹, and there were statistically significant differences in the carbon stocks of the six forest types in 7 of 15 cases. The estimated carbon stock depended heavily on the allometric equation used for the calculations, the variables, and its application to the specific life zone. Due to the diversity of forest structures, these results suggest that caution should be taken when applying default values to estimate forest carbon stocks and emission values in reporting and accounting schemes. The results also indicated the need for allometric equations that are context-specific for different forest types. Therefore, new field investigations and measurements are needed to determine these specific allometric equations, as well as the potential variation in forest carbon stocks in tropical natural forests.     

Interventions to better manage the carbon stocks in Australian Melaleuca forests

Forests and woodlands dominated by tree species of the genus Melaleuca cover around 7,556,000 ha in Australia and predominantly occur as wetland ecosystems. In this Viewpoint, we use published secondary data to estimate that there is likely to be between 158 tC/ha and 286 tC/ha stored in Melaleuca forests. These estimates are at least five times greater than the previous estimate made by the Australian Government (about 27.8 tC/ha). There are 2.1 million ha of protected Melaleuca forest which likely stock between 328 M tC and 601 M tC; equivalent to between 2.7% and 5.0% of total carbon storage of all Australian native forests. These estimates are significant because it appears that carbon stocks in Melaleuca forests are currently dramatically under-estimated in Australia's national greenhouse gas emissions inventory reported under the United Nations Framework Convention on Climate Change (UNFCCC). Whilst the precision of the estimates is limited by the availability of rigorous primary data, we also argue that the estimates are indicative and meaningful, and this synopsis highlights the fact that this forest type should be considered a significant carbon store nationally and globally.     

Carbon mitigation potential of the French forest sector under threat of combined physical and market impacts due to climate change

Objectives(1) To quantify the contribution of the French forest-wood product chain in terms of carbon sequestration and substitution when accounting for both the physical impacts (shifts in tree growth and mortality rates) and the market impacts (increased demand of harvested wood products (HWP)) of climate change (cc) and the subsequent forest managers adaptations; (2) To assess the uncertainty of the impacts on the above carbon balance and on forest allocation; and (3) To assess the role of managers’ expectations toward these future, uncertain but highly anticipated, impacts.MethodologyWe used a bio-economic model of the French Forest Sector (FFSM++) that is able to consider and integrate: (a) the effects of climate change over forest dynamics; (b) forest investment decisions (among groups of species) according to expected profitability; and (c) market effects in terms of regionalised supply, consumption and trade of HWP, depending on the forest resource stocks and international prices. By including both forest dynamics and forest products, we can evaluate the carbon balance taking the following elements into consideration: (a) carbon sequestered in live and dead biomass in the forest; (b) carbon sequestered in HWP; (c) carbon substituted when wood is used in place of fossil fuels or more energy-intensive materials; and (d) carbon released by forest operations.ResultsWhen the model is run at constant conditions for the next century, the average carbon potential of French forests is 66.2–125.3 Mt CO₂ y⁻¹, depending on whether we consider only inventoried wood resources, HWP pools and direct energy substitution, or if we also account for the carbon stored in tree branches and roots and if we consider the more indirect, but also largely more subjective, material substitution. These values correspond to 18.3% and 34.7%, respectively, of the French 2010 emissions (361 Mt CO₂). However, when we consider both the probable increment of coniferous mortality and changes in forest growth, plus the rise in HWP demand worldwide, the average sequestration rate of the French forest decreases by 6.6–5.8% to 61.8–118.0 Mt CO₂ y⁻¹. Running partial scenarios, we can assess the relative interplay of these two factors, where the price factor increases the HWP stock while decreasing the forest stocks (where the latter effect prevails), while the physical impact of climate change reduces both, but to a lesser extent. Considering short-sighted forest managers, whose behaviour is based uniquely on the observed conditions at the time decisions are made, we obtain a limited effect of the overall carbon balance but a relatively large impact on the area allocation of broadleaved vs. coniferous species.     

Public policies can reduce tropical deforestation: Lessons and challenges from Brazil

Reducing carbon emissions from deforestation and forest degradation now constitutes an important strategy for mitigating climate change, particularly in developing countries with large forests. Given growing concerns about global climate change, it is all the more important to identify cases in which economic growth has not sparked excessive forest clearance. We address the recent reduction of deforestation rates in the Brazilian Amazon by conducting a statistical analysis to ascertain if different levels of environmental enforcement between two groups of municipalities had any impact on this reduction. Our analysis shows that these targeted, heightened enforcement efforts avoided as much as 10,653 km² of deforestation, which translates into 1.44 × 10⁻¹ Pg C in avoided emissions for the 3 y period. Moreover, most of the carbon loss and land conversion would have occurred at the expense of closed moist forests. Although such results are encouraging, we caution that significant challenges remain for Brazil's continued success in this regard, given recent changes in the forestry code, ongoing massive investments in hydro power generation, reductions of established protected areas, and growing demand for agricultural products.     

The potential and cost of increasing forest carbon sequestration in Sweden

This paper examines the potential and the cost of promoting forest carbon sequestration through a tax/subsidy to land owners for reducing/increasing carbon storage in their forests. We use a partial equilibrium model based on intertemporal optimization to estimate the impacts of carbon price (the tax/subsidy rate) on timber harvest volume and price in different time periods and on the change of forest carbon stock over time. The results show that a higher carbon price would lead to higher forest carbon stocks. The tax/subsidy induced annual net carbon sequestration is declining over time. The net carbon sequestration during 2015–2050 would increase by 30.2 to 218.3 million tonnes of CO₂, when carbon price increases from 170 SEK to 1428 SEK per tonne of CO₂. The associated cost, in terms of reduced total benefits of timber and other non-timber goods, ranges from 80 SEK to 105.8 SEK per tonne of CO₂. The change in carbon sequestration (as compared with the baseline case) beyond 2050 is small when carbon price is 680 SEK per tonne of CO₂ or lower. With a carbon price of 1428 SEK per tonne of CO₂, carbon sequestration will increase by 70 million tonnes of CO₂ from the baseline level during 2050-2070, and by 64 million tonnes during 2070–2170.     

Shifting patterns of oil palm driven deforestation in Indonesia and implications for zero-deforestation commitments

Oil palm plantations in Indonesia have been linked to substantial deforestation in the 1990s and 2000s, though recent studies suggest that new plantations are increasingly developed on non-forest land. Without nationwide data to establish recent baseline trends, the impact of commitments to eliminate deforestation from palm oil supply chains could therefore be overestimated. We examine the area and proportion of plantations replacing forests across Sumatra, Kalimantan, and Papua up to 2015, and map biophysically suitable areas for future deforestation-free expansion. We created new maps of oil palm plantations for the years 1995, 2000, 2005, 2010 and 2015, and examined land cover replaced in each period. Nationwide, oil palm plantation expansion occurred at an average rate of 450,000 ha yr⁻¹, and resulted in an average of 117,000 ha yr⁻¹ of deforestation, during 1995–2015. Our analysis of the most recent five-year period (2010–2015) shows that the rate of deforestation due to new plantations has remained relatively stable since 2005, despite large increases in the extent of plantations. As a result, the proportion of plantations replacing forests decreased from 54% during 1995–2000, to 18% during 2010–2015. In addition, we estimate there are 30.2 million hectares of non-forest land nationwide which meet biophysical suitability criteria for oil palm cultivation. Our findings suggest that recent zero-deforestation commitments may not have a large impact on deforestation in Sumatra, where plantations have increasingly expanded onto non-forest land over the past twenty years, and which hosts large potentially suitable areas for future deforestation-free expansion. On the other hand, these pledges could have more influence in Kalimantan, where oil palm driven deforestation increased over our study period, and in Papua, a new frontier of expansion with substantial remaining forest cover.     

Assessing the efficiency of switchgrass different cultural practices for pellet production

Switchgrass is a perennial crop producing high amounts of biomass for good quality pellet production. The objective of this study is to examine the efficiency of different cultural practices of switchgrass for pellet production under field conditions for four different N-fertilization (0, 80, 160 and 240 kg ha⁻¹) and two different irrigation levels (0 and 250 mm), in two soils in central Greece with rather different moisture status over the period 2009–2012. Moreover, comparison between three harvest methods (two different types of bales and silage) was made. The results derived from this study revealed that the bale at 22 kg is the harvesting practice with the highest costs while there was reduction of efficiency scores when nitrogen levels increased. At both environments the efficiency scores followed the same trend, confirming that low levels of nitrogen fertilization enhance the economic competitiveness of switchgrass production. Palamas site is the area where switchgrass for pellet production had positive income ranging from 400 to 1600 € ha⁻¹, while Velestino site had always negative. Therefore, places like Velestino with non-aquic soil should be avoided for switchgrass. These data suggest that growing switchgrass for solid biofuel production as energy crop is a worthwhile decision only in areas with a moderately shallow groundwater table (aquic soil) or maybe in high precipitation regions.     

Managing tropical wetlands for advancing global rice production: Implications for land-use management

Being fertile lands, wetlands have been managed for traditional agriculture over millennia. However, the integrity and ecosystem services of wetlands are being jeopardized by intensive land-use comprising of drainage and excessive disturbance. An adhoc and intensive use of wetlands, without preserving ecological integrity is causing ecosystem disservices and threatening conversion of a large soil organic carbon (SOC) sink into a net source. Wetlands in the tropical parts of the world are distributed unevenly and represent ∼3% of the total world land area. Due to stagnancy and even reduction in rice (Oryza sativa) yield of many agricultural regions, there is a need for additional and alternative land-uses which can raise the global rice production to ∼1 billion Mg (megagram = 10⁶g = metric ton) by 2050 from ∼497 million Mg now. Wetlands can be a viable option to advance global food security because of high soil fertility and vast geographical distribution. A ‘3-tier rice production system’ is proposed herein based on specific hydrological niche to advance global food security without degrading the ecosystem services of wetlands. In addition to increasing agronomic yield, the proposed modus operandi can also improve the livelihood security of farmers through an additional income streams by: (i) trading of SOC credits generated through adoption of conservation agriculture in littoral zones, and (ii) promoting fish and duckery culture in conjunction with deepwater rice farming. Furthermore, the proposed strategies will also set in motion the process of restoration of wetlands while enhancing C sink capacity of the ecosystems.     

Opportunity cost of water allocation to afforestation rather than conservation of natural vegetation in China

Estimation of ecosystem service values is a hot area of research in ecological conservation and economics. However, the costs of these outputs are largely unknown. In this paper, we estimated the opportunity cost of water allocated to afforestation projects through mathematical modeling based on statistical data for all of China to provide support for restoration planning based on a fuller consideration of the true costs. To guide future ecological conservation and environmental policy development, we illustrate a neglected concept (ecosystem service costs) and use this concept to compare the ecological services provided by ecological restoration based on afforestation with those of restoration based on the conservation of natural vegetation using data obtained since 1949 in China. The results showed that afforestation and natural vegetation create annual costs related to use of the available water resources equal to 4800 and 3700 RMB ha⁻¹, respectively, representing a water opportunity cost of 1100 RMB ha⁻¹ for afforestation. This illustrates the rule that “there is no free lunch” for any service, including ecosystem services. Therefore, to support the development of more effective and sustainable environmental restoration policy, it will be necessary to evaluate the associated opportunity costs.     

Analysing foregone costs of communities and carbon benefits in small scale community based forestry practice in Nepal

Reducing emissions from deforestation and forest degradation, conservation and sustainable management of forests and enhancement of forest carbon (REDD+) are considered to be important cost effective approaches for global climate change mitigation; therefore, such practices are evolving as the REDD+ payment mechanism in developing countries. Using six years (2006–2012) data, this paper analyses trade-offs between carbon stock gains and the costs incurred by communities in generating additional carbon in 105 REDD+ pilot community forests in Nepal. It estimates foregone benefits for communities engaged in increasing carbon stocks in various dominant vegetation types. At recent carbon and commodity prices, communities receive on average US$ 0.47/ha/year of carbon benefits with the additional cost of US$ 67.30/ha/year. One dollar’s worth of community cost resulted 0.23 Mg of carbon sequestration. Therefore, carbon payment alone may not be an attractive incentive within small-scale community forestry and should link with payments for ecosystem services. Moreover, the study found highest community sacrificed benefits in Shorea mixed broadleaf forests and lowest in Schima-Castanopsis forests, while carbon benefits were highest in Pine forests followed by Schima-Castanopsis forests and lowest in Rhododendron-Quercus forests. This indicates that costs and benefits may vary by vegetation type. A policy should consider payment for other environmental services, carbon gains, co-benefits and trade off while designing the REDD+ mechanism in community based forest land use practice with equitable community outcomes. The learning from this study will help in the formulation of an appropriate REDD+ policy for community forestry.     

Wood biomass use for energy in Europe under different assumptions of coal,gas and CO2 emission prices and market conditions

This study examines the effects of different coal, natural gas and carbon emission prices and market situations on the use of wood for electricity and heat production in the European Union. The analysis is carried out using the global forest sector model EFI-GTM expanded to cover electricity and heat production from wood, coal, natural gas, wind and solar energy. Analysis shows that with low coal and gas prices, use of wood for energy will be limited to low cost logging residues. With high coal, and especially natural gas prices, industrial wood also comes to be used for energy. At a carbon price of 100 €/tCO₂, some 32 Mm³ of industrial wood, in addition to 224 Mm³ of logging residues, are projected to be used for electricity and heat in the EU region (including Norway and Switzerland) in 2030. The relatively low quantity of industrial wood used by the energy sector despite the collapse of the use of coal is explained by the fact that under high CO₂ prices, other energy forms like natural gas, solar and wind energy become more and more competitive. However, the amount of industrial wood used for energy may substantially increase with subsidies for using wood for electricity and heat, even with relatively low carbon prices. With a high coal and gas price and a carbon price of 100 €/t, a subsidy of 30 €/MWh to the wood based and coal with wood co-firing electricity production will have a significant impact on the European wood based sector. Depending on the development of the market demand for forest industry products, such a subsidy may cause a 10–12.5% reduction in forest products production, a 6–9% increase in harvest level, about 30–60% increase in the pulpwood prices, and a 6–9 fold increase of wood imports in the EU, compared to the respective case without a subsidy in 2030.     

Trends in land ownership fragmentation during the last 230 years in Czechia,and a projection of future developments

A high level of fragmentation of farmland ownership is an important underlying cause of land degradation and, at the same time, an obstacle to sustainable land management. This study makes the first-ever analysis of long-term trends in the rate of fragmentation. Our study covers the period from the earliest stages of the current form of ownership patterns at the end of the 18th century until the present day. On the basis of significant predictors that have been identified (initial fragmentation, population growth, historical development of inheritance laws and of the land market, natural soil fertility and landscape type), we go on to project probable developments for the period from 2016 to 2045. A total of 102,984 land parcels in 56 cadastral units in the territory of Czechia have been analysed on the basis of data from four years (1785, 1840, 1950, 2015). Our study considers the development of two basic indicators of fragmentation – Mean Parcel Size and Number of Owners per 100 ha. The Mean Parcel Size has decreased over a period of 230 years from 1.08 ha to 0.64 ha, at a mean rate of −0.26% year⁻¹. During the same period, the Number of Owners per 100 ha has risen from 17.50 to 79.66, at a mean rate of 0.61% year⁻¹. A detailed analysis of the development trends confirms significant spatial variability and also time variability of the rates of the two indicators. The analysis also shows their mutual complementarity: growth in the rate of one of the indicators is usually accompanied by a drop in the other. The general trend that we project for the territory of Czechia in the upcoming 30 years is that there will be further diminution of the physical size of land parcels (continuing fragmentation of land parcels) accompanied by a reduction in the Number of Owners (defragmentation of land ownership).     

A DEA approach to assess the efficiency of radiata pine logs to produce New Zealand structural grades

An efficiency analysis revealed the relative magnitude of wood traits that distinguishes efficient radiata pine logs to produce New Zealand structural grades. Technical and cost efficiencies were obtained by using data envelopment analysis (DEA). Wood trait prices used to perform the cost efficiency corresponded to economic weights derived from a partial regression. These values were 1.1, 29.7, 0.3 and ?0.4 NZ $/m³ for small end diameter (cm), stiffness (GPa), basic density (kg/m³) and largest branch (mm) respectively. The most efficient logs were those with the highest difference between recovery value and price. There were positive and significant correlations between technical efficiency and wood stiffness (0.46, p < 0.05) and between cost efficiency and log recovery value (0.85, p < 0.05). The most efficient logs had a ratio of 1:4 between stiffness and small end diameter whereas logs that did not generate structural lumber presented ratios close to 1:8. This information will inform the development of breeding objectives, and help segregating and pricing logs by using traits patterns that result in efficient logs for the production of structural wood.     

Risk analysis of water scarcity in artificial woodlands of semi-arid and arid China

All activities are inherently risky, including seemingly beneficial activities such as ecological restoration. However, small risks are easy to ignore, even if they may accumulate to create a large cumulative risk. Therefore, the long-term ecological benefits and risks of any ecological restoration project must both be considered. However, quantitative evaluation of the risk of afforestation in arid and semi-arid regions has been insufficiently studied. Here, we present a method for evaluating the risk associated with ecological restoration, using water shortages in artificial woodlands in China’s arid and semi-arid regions as an example of cumulative risk. We found an annual risk that amounted to 5174 RMB ha⁻¹ in 2014, which was 17% of the ecological service value of the forests. However, this cost depends on changes in the price, availability, and use of water in these regions. If ecological degradation occurs, it will trigger a series of serious consequences, and its cost may far exceed the expected benefits. Our inability to predict natural disasters such as drought and the problem of imperfect communication among stakeholders must be considered to achieve ecological restoration. The method described in this paper will provide theoretical support for future risk evaluations and guidance for the allocation of natural resources such as water, thereby increasing the likelihood of successful environmental management.     

Assessing costs and benefits of pest management on forested landbases in eastern and western Canada

To assist pest management planning, the Canadian Forest Service developed the Spruce Budworm Decision Support System (SBW DSS), which quantifies the timber supply impacts of protecting stands against spruce budworm (Choristoneura fumiferana Clem.) defoliation. We incorporated protection costs and timber product values in this system to evaluate economic aspects of spruce budworm control. The analysis allows pest managers to evaluate the degree to which the traditional volume protection priority objective corresponds to three economic criteria, namely the volume protected per dollar protection program cost, the benefit–cost ratio of the protection program, and the net present value of the protection program. Twelve alternative spruce budworm protection strategies were analyzed on Crown License 1 in New Brunswick and Prince Albert Forest Management Area (PAFMA) in Saskatchewan, based on a number of protection program extents and intensities. For both landbases under base-case market conditions, the largest, most intensive protection scenario provided the highest amount of volume saved and net present value (at 3.94 Mm³ and $39.98 M for PAFMA, and 4.04 Mm³ and $41.49 M for License 1, respectively) while smaller, less-intensive programs provided the highest benefit–cost ratios and volume protected per present value dollar of protection cost (at 8.22 and 0.52 m³/$ for PAFMA, and 10.26 and 0.65 m³/$ for License 1, respectively). Sensitivity analysis on product values and protection costs revealed that smaller, less-intensive programs could also produce the highest net present values when costs are higher and/or product values lower. These results highlight the conditions under which pest managers should consider deviating from their traditional strategy of maximizing volume saved to one that maximizes the economic returns of protection.     

An interval fuzzy chance-constrained programming model for sustainable urban land-use planning and land use policy analysis

In this study, an interval fuzzy chance-constrained land-use allocation (IFCC-LUA) model is developed for sustainable urban land-use planning management and land use policy analysis under uncertainty. This method is based on an integration of interval parameter programming (IPP), fuzzy flexible linear programming (FFLP) and chance-constrained programming (CCP) techniques. Complexities in land-use planning management system can be systematically reflected, thus applicability of the modeling process can be highly enhanced. The developed method is applied to planning land-use allocation practice in Nanjing city, China. The objective of the IFCC-LUA is maximizing net benefit from LUA system and the main constraints include investment constraints, land suitability constraints, water/power consumption constraints and wastewater/solid waste capacity constraints. Modeling results indicate that desired system benefit will be between [1.34, 1.74] × 10¹² yuan under the minimum violating probabilities; the optimized areas of commercial land, industrial land, agricultural land, transportation land, residential land, water land, green land, landfill land and unused land will be [290, 393] hm², [176, 238] hm², [3245, 4390] hm², [126, 170] hm², [49, 66] hm², [1241, 1679] hm², [102, 138] hm², [7, 10] hm² and [178, 241] hm². They can be used for generating decision alternatives and thus help decision makers identify desired land use policies under various system-reliability constraints of economic development requirement and environmental capacity of pollutant. Tradeoffs between system benefits and constraint violation risks can also be tackled.     

Evaluating the conformity of informal land subdivision with the planning law in Benin metropolis

Town planning law is a regulatory framework for guiding the spatial growth of cities. As a land subdivision regulation, it embraces all development parameters in urban areas. But locally adopted land subdivision standards appear incompatible with the existing Planning Law. Evaluating the conformity of informal land subdivision standards for residential and non-residential uses with those of the existing Planning Law, land subdivision attributes were collected from four neighbourhoods and compared with the minimum approved standards. Independent-sample t-test statistics for each of the selected neighbourhoods shows no significant difference in their Mean Scores. The significant values (two-tailed) in the t-test output tables of the selected communities were consistently above the required cut-off of α = 0.05 which suggest that informal land subdivision attributes statistically conform with those of the Planning Law. Key planning implications of unplanned land use patterns that have emerged and policy options to tackle them were suggested.     

A historical analysis of the drivers of loss and degradation of Indonesia’s mangroves

Mangroves have been systematically exploited in Indonesia since 1800, especially for the development of brackish water shrimp aquaculture (called ‘tambak’) and for timber harvesting. By the end of the 1960s, Indonesia is estimated to have lost more than 200,000 ha of its mangroves mostly in Java and Sumatra. The rate of mangrove loss started to dramatically increase in the 1970 when exploitation shifted to new areas outside Java, particularly in Kalimantan and Sulawesi, encouraged by government policies to boost timber production, followed by policies to expand tambak in 1980s and large scale tambak development triggered by increased shrimp price during Asian financial crisis in 1997. The result has been the loss of nearly 800,000 ha of mangroves in only 30 years, mostly now in the form of low productivity or abandoned tambaks. In recent years, timber harvesting activities in Indonesia's mangroves appear to have become more sustainable. Our analysis suggests that aquaculture will continue as the main driver of change in mangrove ecosystems in Indonesia followed by palm oil plantation. Failure to deal with the current low productivity of shrimp aquaculture in many parts of Indonesia will force shrimp producers to clear an estimated 600,000 ha more mangroves to make way for shrimp farms over the next two decades. However, with improvements in brackish water aquaculture productivity, halting palm oil concession to utilise mangroves, along with maintaining other mangrove use pressures at moderate levels, the net loss of mangroves in the next two decades could be reduced to around 23,000 ha.