• Open Access

Acting Optimally for Biodiversity in a World Obsessed with REDD+


  • Editor Lauren Persha


REDD+ presents novel options for conservation in the tropics, yet it is unclear how biodiversity-focused organizations or actors should react to these carbon-focused opportunities. Here, we critically assess for the first time the expected outcomes of five contrasting scenarios of engagement between a biodiversity actor and REDD+. We discover that in the Berau regency, Indonesia, it is usually beneficial for a biodiversity actor to react in some way to REDD+, but the preferred reaction depends on whether a REDD+ project is already developing in the region, and the scale and type of conservation objectives. In general, from a strict biodiversity perspective, the most cost efficient reaction to the presence of REDD+ is to use biodiversity funds to protect areas neglected by REDD+. Our results demonstrate that if biodiversity actors fail to adapt the way they pursue conservation in the tropics, REDD+ opportunities could go largely untapped.


Tropical forests are among the planet's most important biological treasures (Wilson 1988; Myers et al. 2000), but rampant deforestation is eroding this biome (Pimm et al. 1995; Laurance 2007a). Biodiversity actors, including nongovernmental organizations, bilateral aid, and multilateral groups, are an integral part of the response to this tropical forest crisis (Armsworth et al. 2012), and together generate more than USD $6 billion annually to protect imperiled species and ecosystems (James et al. 1999).

Presently, a new mechanism and an additional set of actors are emerging to finance the conservation and management of tropical forests. These actors are supported by the carbon payment mechanism termed REDD+ (Reducing emissions from deforestation and forest degradation, plus the conservation, sustainable management and enhancement of forest carbon stocks; Venter & Koh 2012). Early commitments of “fast-start financing” for REDD+ currently total about US$4.5 billion, but future funds could eclipse current spending: The Green Climate Fund alone could raise up to $100 billion annually by 2020, a portion of which could be used for REDD+. The scale of these investments would fundamentally change conservation in the tropics (Venter et al. 2009; Busch et al. 2011).

In response to this opportunity, a suite of studies have suggested ways to secure and enhance positive outcomes for biodiversity from REDD+. Recommendations include monitoring biodiversity outcomes (Lindenmayer et al. 2012), drafting policy safeguards for biodiversity (Grainger et al. 2009; Putz & Redford 2009; Harvey et al. 2010), changing technical definitions to the benefit of biodiversity (Sasaki & Putz 2009), outlining biodiversity friendly management interventions (Stickler et al. 2009; Gardner et al. 2012), and ensuring the widest adoption of REDD+ measures (Busch et al. 2011; Strassburg et al. 2012). All of these suggestions are aimed at influencing the design and implementation of REDD+ policies. At its core, however, REDD+ must remain a carbon-focused mechanism, and its ability to adopt these policy suggestions may therefore remain limited. In stark contrast with the efforts to influence REDD+ policy, conservation science and theory has yet to explore how biodiversity actors could instead adapt their policy and behavior in response to the presence of REDD+ actors and opportunities.

Most conservation agencies undertake a systematic planning process to identify priority areas and actions that ensure their limited resources deliver the greatest benefits (Brooks et al. 2006; Moilanen et al. 2009). There are a number of conceivable ways that biodiversity actors could adapt their business-as-usual policy of prioritizing investments in light of emerging REDD+ schemes. For instance, it has been suggested that biodiversity actors could engage REDD+ actors by co-investing in forest protection (Venter et al. 2009). On the other hand, biodiversity actors could allow forests with high carbon content to be protected by REDD+, and then optimize their investments for biodiversity by focusing on high biodiversity but low-carbon areas (Miles & Kapos 2008), which resembles a free-riding strategy (Bode et al. 2011; Espinola-Arredondo & Munoz-Garcia 2011).

Here, we critically assess the expected outcomes for biodiversity under five scenarios of a biodiversity actor adapting its prioritization policy in light of REDD+. We focus on a regional-scale case study in the Berau regency in Indonesian Borneo. We simulate the effects of different interactive behaviors on emissions reductions and biodiversity conservation using a prioritization tool developed for the purpose of conservation planning in Berau (Venter et al. 2012), and additional data on the distribution of biodiversity features in Berau (Paoli & Wells 2010).


This section introduces the Berau case study, and then details the biodiversity features considered and the prioritization approach we use to explore policy scenarios. For a more detailed account of Berau and the prioritization approach see (Venter et al. 2012).

Berau case study

A landscape-scale REDD+ program, dubbed The Berau Forest Carbon Program (BFCP), is developing in Berau. Also, The Nature Conservancy (TNC), one of the world's largest nongovernment conservation organizations, is active in Berau, both supporting the BFCP and investing independently in conservation outcomes.

The regency of Berau covers 2.2 million hectares and is divided into three land designation classes: (1) protected forest, (2) production forest, which includes natural forest timber harvesting and monoculture timber plantations, and (3) land designated for conversion to agriculture (Web Figure 1a). There are five landscape-scale management strategies being considered by the BFCP and TNC in Berau.

  1. In existing protected areas, Strategy (1) improve land management to reduce illegal logging, forest conversion and fire.
  2. In production forest, Strategy (2) implement reduced impact logging (RIL) techniques and Strategy (3) set aside some areas for carbon storage.
  3. In areas designated for agriculture, Strategy (4) retire oil palm permits in high carbon or biodiversity areas and Strategy (5) identify alternative sites for oil palm permits that are both agriculturally suitable, and low in carbon or biodiversity.

A spatially explicit approach was developed to help prioritize among these five strategies and across locations in Berau (Venter et al. 2012). The approach uses the decision support software Marxan with Zones (Watts et al. 2009) to identify a spatial network of conservation actions that minimizes the costs of meeting prespecified planning targets (Moilanen et al. 2009). The basic spatial units for the analysis are 28,059 equal area hexagonal planning units measuring 1 km across. Where planning units were bisected by a land management boundary, such as an oil palm permit, the planning unit was split in two along that boundary. Marxan with Zones identifies which of the five strategies should be applied in each planning unit, based on the quantity of each conservation feature present in each planning unit (WebTable 1, WebFigure 2c-g), constraints on the strategies that can be deployed in each planning unit (WebFigure 2a), and the costs of applying each strategy in each planning unit. We estimated the opportunity, start-up, and ongoing management costs of the five REDD+ strategies (Table 1, WebFigure 2b). When calculating the opportunity costs for lands leased for private use, we included the opportunity costs of lost employment, lost tax revenues, and lost profits to the permit holder. For lands without private use rights, we included lost employment and tax revenues that could be derived if a permit were to be granted. We did not account for the displacement or “leakage” of competing land uses into areas not covered by conservation strategies.

Table 1. Data inputs used to calculate the start-up, opportunity and ongoing management costs of REDD+ strategies being considered by the Berau Forest Carbon Partnership. All values given as a onetime upfront cost in 2009 US$. We calculated opportunity costs as the net present value from 30 years of lost profits using a discount rate of 10.5%, one typically used by Indonesian companies (SMART-Tbk 2010). Management costs were estimated as the cost of an endowment capable of financing management into perpetuity, calculated using Indonesia's inflation rate of 3.56% and bond rate of 7.75% (www.bi.go.id)
  1. a

    Annual tax opportunity cost per hectare = (a · b−1) · (c), where (a) is the $260 million average annual assessed forest taxes in Indonesia between 2003 and 2006 (HRW 2009), (b) is the 20 million m3 national average annual reported volume of wood consumed by Indonesia's forestry industry between 2003 and 2006 (HRW 2009), (c) is the 45.2 m3/ha harvested on average using conventional logging techniques in Berau (FAO 1995; Bertault & Sist 1997; Hinrichs et al. 2002). Tax revenues are calculated using a 30 year average logging cycle.

  2. b

    Annual employment opportunity cost per hectare = (d · e−1) · (c · f), where (d) is the 54,140 people employed in the legal forestry sector in 2005, the only year for which data are available (FRA 2010), (e) is the 29 million m3 reported volume of wood consumed by Indonesia's forestry industry in 2005 (HRW 2009), f is the $1443.27 per capita gross national income in Indonesia 2005 (World Bank 2010) and (c) is as above. Employment costs are calculated using a 30 year average logging cycle.

  3. c

    Because of the differences in logging volumes across locations in Southeast Asia and through time (Ruslandi et al. 2011), we limited our data collection to studies which were conducted in the last 10 years and based in Kalimantan. When converting logging volumes to profit per harvested hectare, we used the log prices and harvesting costs presented in Ruslandi et al. (2011).

  4. d

    In Berau, logging concessions operate on a 30 year logging cycle, which means that on average 1/30 of their total concession area is harvested each year. To convert logging profits to net present value we used,

    display math

    where Pyr is the expected harvest profits in year i.

  5. e

    Where new permits are granted in areas that store more carbon than will be sequestered through the growth of oil palms (40 MgC/ha) (Dewi et al. 2010), the new permit creates an opportunity cost of lost REDD+ revenues at a rate of $5 MgC−1 for each MgC that the area is expected to emit due to oil palm development.

All strategies that involve managing a forest strictly for carbon storageProgram start-up$51.69 per hectare(Wilson et al. 2009)
Protected area managementOngoing management$70.88 per hectare(McQuistan et al. 2006)
Managing forest for carbon in areas currently designated for logging and agricultureOngoing management$311.22 ha−1(McQuistan et al. 2006)
Managing production forest strictly for carbon storageOpportunity cost-taxa$177.21 per hectare(FAO 1995; Bertault & Sist 1997; Hinrichs et al. 2002; HRW 2009)
Managing production forest strictly for carbon storageOpportunity cost-employmentb$36.73 per hectare(Bertault and Sist 1997; FAO 1995; Hinrichs et al. 2002; HRW 2009; World Bank 2010)
Managing production forest strictly for carbon storage (selective logging, 30 year cycle)Opportunity cost- logging company profitsc$2125.20 per hectare; d$639.65 per hectare(Hinrichs et al. 2002; Ruslandi et al. 2011)
Managing production forest strictly for carbon storage (planned conversion to pulp wood plantation)Opportunity cost- logging company profits$1574.55 per hectare if forested $1109.32 per hectare if already deforested(IFCA 2008)
Reduced impact loggingProgram start-up$33.87 per hectare(IFCA 2008)
Managing forested agricultural land strictly for carbon storageOpportunity cost- palm oil company profits$6766.00 per hectare(Ruslandi et al. 2011)
Managing forested agricultural land strictly for carbon storageOpportunity cost- tax and employment$6772.70 per hectare((AAL-Tbk 2010; SMART-Tbk 2010; Ruslandi et al. 2011)
Establishing new oil palm permitOpportunity cost- REDD+ revenuese$5 MgC−1 

Biodiversity features

As surrogates for biodiversity values in Berau, we focused on the Bornean orang-utan (Pongo pygmaes morio) and areas identified as High Conservation Value (HCV) (Paoli & Wells 2010). Orang-utans are a flagship species and one of the primary targets of conservation efforts in Indonesia, particularly on the island of Borneo (Wich et al. 2008). Based on P. pygmaes morio's distribution data sourced from the World Atlas of Great Apes and their Conservation (Caldecott & Miles 2005), P. pygmaes morio is predicted to occur across 594,000 hectares of Berau (WebFigure 1d). For HCVs, we used spatial data from a landscape-scale mapping of HCVs across the East Kalimantan province (Paoli & Wells 2010), which includes the Berau regency. The mapping included HCVs 2.1 & 2.2, which focus on large landscapes with a capacity to maintain natural ecological processes, and HCV 3, which are rare or endangered ecosystems (WebTable 1 WebFigure 2e–g; Paoli & Wells 2010).

We distinguished between strategies in their effectiveness at conserving biodiversity features. In particular, strategies which entail full protection (strategies 1, 3, and 4) were considered to fully protect biodiversity features, and reduced impact logging (strategy 2) was assumed to be 85% effective, the lower estimate of species retention from a meta-analysis of the effects of selective logging (Putz et al. 2012). Oil palm plantations (strategy 5) were assumed to lack conservation value.

Policy simulations

We used the Marxan with Zones framework to simulate five plausible policy scenarios for how a biodiversity actor could alter how they prioritize conservation actions and locations in light of REDD+ (Figure 1). We assume that the REDD+ actor is entirely focused on emissions reductions and indifferent to biodiversity outcomes. In some cases, this is an oversimplification as REDD+ policy does include specific safeguards for biodiversity (Venter & Koh 2012). However, the purpose of these safeguards is to prevent biodiversity negative outcomes, such as the conversion of natural forests to plantations, which is not a proposed REDD+ strategy in Berau.

Figure 1.

Five policy scenarios for how a biodiversity actor could adapt to REDD+ opportunities when prioritizing conservation actions and locations.

The policy scenarios fall into two general cases (Figure 1). In the first case, a REDD+ actor is currently developing or has developed a REDD+ project in the region, and in the second case, no REDD+ project exists in the region (although Berau already has a developing REDD+ project, many other regions do not). In both cases, the biodiversity actor can ignore REDD+ developments and opportunities (policy scenarios 1 and 4). In the case where a REDD+ project already exists in the region, the biodiversity actor could alternatively react to it by pursuing active collaboration with the project (policy scenario 2), or instead they could choose to free-ride off the project (policy scenario 3). By securing habitats for native species, forest protection from REDD+ will inadvertently contribute to conservation targets. In the free-ride scenario, the biodiversity actor essentially identifies the areas that will be, or are likely to be, afforded protection by the REDD+ project, and then prioritizes their biodiversity investments into the remaining unprotected areas of the region, such that the combined protection from each actor ensures the biodiversity targets are met. In the case where no REDD+ project exists in the region, the biodiversity actor can invest their funds into creating a REDD+ project in the region that is then able to harness additional funds from REDD+ funding mechanisms (policy scenario 5).

For all policy scenarios, we assume that the biodiversity actor has fixed conservation targets, and they work within a given region, which in our case is the Berau regency. To determine if our results (which are the relative costs of meeting conservation targets under each policy scenario) are sensitive to target size, we tested both low and high targets for both the biodiversity and REDD+ actors, which were respectively set as securing 10% and 50% of each planning feature in the region. The REDD+ target is based on expected carbon emissions, which incorporates information on forest carbon stores and the vulnerability of areas to future land cover change (Dewi et al. 2010). Some biodiversity actors have very specific conservation objectives, such as securing populations of flagship species, while others have more general conservation objectives, such as protecting a portion of an eco-type or landscape. To determine if our results are sensitive to conservation objectives, we test both specific and general biodiversity objectives for each policy scenario. The specific objective is to protect 10% or 50% of the orangutan habitat in Berau, and the general target is to protected 10% or 50% of each HCV. Also, because biodiversity protection in scenario 5 occurs as a collateral outcome from the REDD+ project and therefore it could not be predicted, instead of using specific targets sizes we explored the full target space of this scenario by using a high powered computer to run Marxan with Zones 24,000 times for 1 billion iterations each time. Figure 1 and WebPanel 1 contain addition detail on how the policy scenarios were simulated within the Marxan with Zones framework. We do not present maps of our results because of the potential implications for the BFCP.


In the case where a REDD+ project exists in the region, the total cost of meeting conservation targets in Berau is always lower if the actors collaborate (policy scenario 2) instead of ignoring one another when prioritizing investments (Table 2 and WebTable 2). This is true both for the costs to the biodiversity actor, and the total cost to both actors: for the high conservation targets, collaboration (policy scenario 2) is $8 million cheaper for the biodiversity actor than the ignore scenario, and the total cost to both actors is $15 million lower (Table 2). These results are not sensitive to target size, nor are they sensitive to whether the biodiversity targets are general or specific (WebTable 2).

Table 2. The cost to the biodiversity actor for meeting their high targets (50%), and the total cost of meeting planning targets in Berau in the case where a REDD+ project is already developing with the target of reducing expected emissions by 50%. Total cost is the combined cost to the biodiversity and REDD+ actors of meeting planning targets. Numbers in bold are the policy scenarios for each set of target levels that result in the lowest total cost and the lowest cost to the biodiversity actor. The orangutan target is to protect 50% of the distribution of orangutans, and the HCV target is to protect 50% of the extent of each of 13 High Conservation Values
Biodiversity costIgnore (scenario 1)$16,203,890$162,550,093
 Collaborate (scenario 2)$8,613,309$119,217,504
 Freeride (scenario 3)$1,809,433$92,256,177
Total costIgnore (scenario 1)$439,114,159$585,460,362
 Collaborate (scenario 2)$423,932,998$498,795,184
 Freeride (scenario 3)$424,719,702$515,166,446

While it is better to collaborate than to ignore REDD+ in Berau, our results indicate that if a REDD+ project exists in the region, free-riding (policy scenario 3) allows the biodiversity actor to meet the same biodiversity targets with even further reduced spending outlays (Table 2 and WebTable 2). For low conservation targets, the cost reduction is drastic (89% lower than option 1, and 79% lower than option 2). However, it is important to note that the total cost of the conservation network is always more when the biodiversity actor free-rides (Table 2 and WebTable 2).

In the case where no REDD+ project exists in the region, the benefits of investing into creating a carbon-focused REDD+ project (policy scenario 5) depend both on the type of biodiversity objectives that are sought and their scale (Figure 2). The relative costs of creating a REDD+ project increases with increasing biodiversity targets. If biodiversity targets are low, creating a REDD+ project could be a realistic scenario for the biodiversity actor. But if biodiversity targets are greater than around 20%, the relative costs increase rapidly, especially for the more specific biodiversity objective of protecting orangutans. For very high target levels the relative costs start to fall again, which is an artifact of the conservation approaches (policy scenarios 4 and 5) converging as the system approaches near total protection.

Figure 2.

The relative cost of creating a REDD+ project in Berau (policy scenario 5). These are the costs of the REDD+ project relative to the costs of a traditional biodiversity project, where both achieve the same conservation targets. These costs can be viewed as the amount of surplus revenue that a REDD+ project would need to generate for each dollar invested from the biodiversity actor. Target levels are proportions of the total extent of each biodiversity feature in Berau. Each data point is the result of the best solution for the given target level from running Marxan with Zones 20 times for 1 billion iterations each time.


REDD+ presents novel opportunities for protecting biodiversity in developing countries (Laurance 2007b; Venter & Koh 2012), but these opportunities may not fully develop on their own. From a regional-scale case study in Indonesia, we discover that when in the presence of REDD+ projects, biodiversity actors must fundamentally reshape the way they prioritize their conservation investments to truly capitalize on REDD+.

From a strictly biodiversity perspective, our results indicate that the best way to adapt to the presence of a REDD+ project is to free-ride. That is, biodiversity actors should allow the REDD+ project to protect forests that are REDD+ priorities, and then use biodiversity funds to protect the remaining, usually lower carbon, areas. This is because actions and locations prioritized by the REDD+ actor provide considerable collateral benefit to the targets of the biodiversity actor, allowing the biodiversity actor to focus only on biodiversity features not sufficiently protected by the REDD+ actor.

However, from a broader conservation perspective that includes both biodiversity and REDD+ objectives, we draw very different conclusions. From this perspective, the best outcomes can be achieved when REDD+ and biodiversity actors actively collaborate. These findings are not sensitive to the level of protection or the type of biodiversity objectives considered. Our findings concord with those of Bode et al. (2011), who investigated the value of collaboration among biodiversity actors. In situations where it may not arise on its own, collaboration could be actively promoted by funding bodies which value both climate mitigation and biodiversity protection, or through direct negotiations between the actors.

If a region does not contain a REDD+ project, we find that using biodiversity funds to create a carbon-focused REDD+ project could deliver mixed results. If the targets are to protect up to 20% of the extent of biodiversity features, as is often the case in conservation, the REDD+ project would need to leverage between $1 and $6 for each dollar invested by the biodiversity actor, which in our opinion seems feasible. But if biodiversity targets are higher or very specific, such as protecting a population of an endangered species, the project would need to leverage up to $50 per dollar invested, in which case traditional conservation approaches may be more effective.

Because of Berau's diverse and often charismatic species, substantial forest cover and high rates of deforestation, the region is an obvious target for both biodiversity and REDD+ actors, as would be similar regions across the tropics. We therefore feel that our results may be generally applicable to other locations where REDD+ projects employing site-based interventions overlap with the actions of a biodiversity actor. Moreover, we believe that the Berau case study contains information relevant to REDD+ at a much larger scale.

At the global scale, tropical forests consistently emerge as a global priority for biodiversity conservation (Myers et al. 2000; Brooks et al. 2006). In light of REDD+, however, it has been suggested that biodiversity actors should divert their funding away from carbon-dense tropical forests, toward lower carbon biomes, such as savannahs and grasslands (Miles & Kapos 2008). This resembles our free-riding policy scenario, which we have illustrated can indeed lead to substantial increases in the efficiency of biodiversity conservation at the regional scale. On the other hand, we also found that collaboration can substantially reduce the cost of conserving the tropical forest biome, which could make tropical forests an even more attractive investment at the global scale relative to other conservation options. There would be value in future research to clarify how these divergent effects balance out at the global scale. In addition, exploring the relative benefits of free-riding versus collaboration for cases where the conservation target is a nonforest dependent species would be a valuable addition to our work.

We found that for biodiversity actors, free-riding is cheaper than collaborating with REDD+ actors. But we must note that our modeling assumes that areas and actions prioritized for REDD+ will actually translate into on the ground outputs. The implementation gap means that this may not be the case (Knight et al. 2008), and if REDD+ fails to deliver at scale, conservation targets could be seriously compromised by free riding behavior. Incorporating this uncertainty into studies of interactor engagement could conceivably reveal that the benefits of collaborating (which can reduce the probability of unexpected outcomes by improving information exchange), increase relative to the benefits of free-riding. Moreover, we also did not incorporate the benefits of agencies sharing their on-the-ground experience, contacts and infrastructure, all of which would increase the benefits of collaborative planning and implementation by decreasing overhead and transaction costs.

The actors and the funding mechanisms that drive tropical forest protection are changing. The imminent move toward a more carbon-focused conservation undoubtedly holds considerable potential to conserve biodiversity in one of the planet's most imperiled biomes. Yet our research clearly demonstrates that those concerned with the fate of biodiversity must not simply stand by as spectators to the expansion of REDD+ initiatives, but either engage the mechanism through novel partnerships, or instead shift their investments to areas that will remain unprotected by REDD+. Continuing with business-as-usual conservation in the tropics could mean that REDD+ benefits go largely untapped.


We thank Gary Paoli, Phillip Wells, Sonya Dewi, and Bronson Griscom for their input, and we thank The Nature Conservancy, The Australian Endeavour Program and The Canadian National Science and Engineering Research Council for funding.