Deforestation trajectories and economic drivers
Cambodia has undergone significant forest loss in recent decades—with 2.6 million hectares of forest cover loss occurring since 2001, equating to 29.5% of forest cover7 and 1.45 billion tonnes of CO2 emissions8. The deforestation rates have increased by 76% in the last decade (2011–2021) compared to the previous (2001–2010; Fig. 1b)7. We find forest loss has occurred within three distinct Phases demonstrated by changepoint analysis: (1) Phase 1: steady rise from 2000 to 2009 (average = 0.82%/year), (2) Phase 2: peak years from 2010 to 2013 (average = 2.3%/year), (3) Phase 3: moderate phase from 2014 to 2021 (average = 1.6%/year). Whilst the annual rate of deforestation has declined since the Phase 2, Cambodia currently has the highest country-level annual rate of forest loss globally7, illustrating the relentless deforestation spreading across the landscape. Critically, much of this forest loss and degradation is occurring in mature primary forests (Fig. 1b), which hold significant carbon and are home to rich biodiversity and keystone species17,18,19.
This deforestation in Cambodia has been attributed to the widespread development of Economic Land Concessions (ELCs), the expansion of numerous agricultural frontiers and relentless illegal logging20,21,22. These drivers have been abetted by the establishment of an extensive national road network (Fig. 1a)20—developed to promote economic growth and urban–rural connectivity23. The majority (88.4%) of these roads have been funded by foreign governments (the People’s Republic of China: 38.5%, Japan: 37.9%, and Republic of Korea: 12.0%)18—all of whom have established land concessions within Cambodia’s borders24 through the allocation of state land into private land for long-term industrial plantations22,25. The expansion of ELCs across Cambodia (average addition of 105,000 ha/year of ELC land since 1998) has been directly attributed to up to 40% of the country’s deforestation21, with further indirect impacts due to encroachment into rural community lands (indigenous areas, community forests, subsistence agricultural fields). This results in landlessness, poverty, and land conflicts, forcing communities to migrate in search of arable land, further contributing to the growing degradation and destruction of forests22,26,27,28,29.
Strategic government intervention
Protected areas expanded across Cambodia in 1993 following a royal decree26; the legal details of which were delineated in the 2008 Protected Areas Law, introducing protected categories, wildlife corridors and strict laws prohibiting development9. While over 80 protected areas currently exist covering 35% of Cambodian land10, they are still under substantial threat30. In further efforts to curb deforestation, the Royal Government of Cambodia ordered the suspension of new ELCs and revocation of a subset of existing ELCs in 2012 (Order 01BB)31. This resulted in a reduction of ELCs from a peak of ~ 2.1 million ha in 2012 to ~ 1.6 million ha from 2014 onward (Fig. 1b), with a significant positive correlation between the quantity of land classified as ELCs and the country-level deforestation rate (R = 0.87, p < 0.0001), hinting at the positive impact of Order 01BB on the trajectory of Cambodia’s deforestation rate.
This order was followed by a strategic plan to establish and safeguard a network of protected areas in 201630 (Fig. 1b). However, forest concessions in Cambodia tend to be located adjacent to protected areas—the latter of which have limited physical boundaries (fences and signs), insufficient on-the-ground enforcement, as well as a significant lack of investment and active management—enabling land grabbing, laundering of timber through legitimate ELCs, as well as the illegal harvest and transport of wood by armed groups17,30,31. In all three protected area blocks, we find significant correlations between the deforestation in local ELCs and adjacent protected areas (R = 0.67–0.90, p < 0.0001)—providing evidence of illegal ELC-originating deforestation beyond concessionary boundaries, as reported by environmental groups11. ELCs bring roads and other infrastructure to new areas, making previously isolated forests more accessible for secondary deforestation. Whilst Graham et al.5 confirmed that protected areas are more effective at reducing forest loss than non-protected control areas, protected status alone does not solve the root of the problem: economic needs and aspirations.
Within the three protected area blocks and buffer zones, all have demonstrated recent significant increases in the rate of annual deforestation (Supplementary Table 2). In two of the three cases, the REDD + projects (KSRP and SCRP) were 158% more effective at protecting against forest loss compared to the adjacent protected areas. Prior to REDD + status, these project areas had a history of international NGO presence (albeit with limited financial resources) and relatively lower background deforestation rates compared to the country average (Supplementary Tables 1, 2). Keo Seima has been classified as protected since 2002 (Category V: Conservation Landscape) with subsequent conservation status upgrades until it was officially named a Wildlife Sanctuary (Category IV) in 2016. On the other hand, despite historic NGO presence, Southern Cardamom was only named a National Park (Category II) in 2016. While these projects did not eliminate forest loss, they decreased the average rate of deforestation by 50% from the project onset onwards.
Following REDD + initiation, KSRP and SCRP invested significant resources from carbon revenue into the development of extensive patrolling and security programs backed by law enforcement rangers— including the hiring of local community members who are often left at the frontier of land conflicts32,33 in addition to government-backed rangers. This investment has resulted in robust enforcement activity: for example, > 32,000 patrols have been conducted in SCRP since the REDD + project initiated in 2015, resulting in the confiscation of 6940 chainsaws, removal of 231,061 snares, seizure of 3382 logs and the rescue of nearly 2800 live animals (Supplementary Table 3). Whilst Wildlife Alliance was present in the area since 2002, the REDD + project provided substantial financial support and capacity, resulting in an average of a threefold increase in all enforcement outcomes (Supplementary Fig. 3). Furthermore, additional carbon revenue was invested in developing an ecotourism venture at SCRP, which has welcomed nearly 20,000 guests since 2017, generating $767,000 in revenue for the 5685 participating community members (Supplementary Table 4). While ecotourism was initiated prior to the REDD + project, the additional support of carbon revenues resulted in a 2.3-fold increase in the annual income generated despite the impact of the global pandemic.
In the final case, the TRP was 12% more effective than 50% of the surrounding protected areas and 79% less effective than the remaining protected areas. No relationship between the protected category level and effectiveness was found in this region. TRP and the surrounding PAs (Category IV: Prey Lang Wildlife Sanctuary and Beng Per Wildlife Sanctuary; Category VI: Northwest Biodiversity Corridor) are located within a deforestation hotspot with a 54% higher deforestation rate (2011–2021) than the Cambodia average, resulting in a total of 37.9% forest cover loss in the block since 2001 (Cambodia average = 29.5%). TRP has been classified at a low level of protection (26% of the area is Category VI: Community Forest) with no historic NGO presence and very limited carbon revenue (REDD + credits issued: 2018, first credits sold: 2021). Given the lack of both financial resources and foundational protection activity in TRP—combined with the extreme deforestation risk—it is unsurprising that this REDD + project has been less successful than the other REDD + initiatives (SCRP, KSRP) in Cambodia thus far. Despite the limited funding, TRP has initiated community-based patrols; however, backing by law enforcement limited and patroling remains restricted in scale at present34. Even if TRP carbon credits are sold at a stable rate going forward, the project has far less credits per unit area compared to the other REDD + project in Cambodia (TRP: 2.4 tCO2e/year/ha, SCRP: 6.2 tCO2e/year/ha, KSRP: 11.9 tCO2e/year/ha; Supplementary Table 5) due to the lower carbon density of TRP forests. Furthermore, there is often a delay between project initiation and investment of carbon finance due to administrative bottlenecks—thus, it is too soon to predict how successful TRP will be in the long-term. However, TRP is lacking certain characteristics that are associated with SCRP and KSRP, including: historic NGO presence, stable funding from carbon revenues, high-level protection status and effective patrolling to secure the area.
REDD + projects have faced criticism in the past—including citations of limited protection success35, weak community involvement (through Free, Prior and Informed Consent)36, insufficient sharing of carbon revenues with the communities37, risks of community displacement38 and a lack of community-run governance structure39—and not all have been successful35. However, those REDD + projects currently in operation in Cambodia hold the CCB (Climate, Community & Biodiversity) standard, demonstrating a commitment to achieving, and documenting through independent verification audits, net positive impacts to the local community and wildlife. Despite the TRP commitment to strong community involvement34, based on the lack of historic presence in the area and limited carbon revenues available for community sharing, the relationship and trust building with the community may take more time than KSRP and SCRP. Nevertheless, given the high deforestation risk in the area and presence of ELCs, it appears unlikely that the level of community involvement is a major factor leading to a lower success rate thus far at TRP.
The results presented highlight the significant deforestation encroaching upon PAs and REDD + project boundaries in Cambodia, with ELCs associated with PA deforestation. The Royal Government of Cambodia has taken important steps in mitigating the deforestation crisis in Cambodia through the halting of new ELCs26—resulting in a decrease in annual deforestation from 2014 to 2021 compared to the 2010–2013 peak.
The government also initiated a National REDD + Strategy in 201740, converting underfunded PAs into REDD + projects with the goal of developing a nested national REDD + system. In theory, carbon revenues would provide the adequate resources to enact rigorous boots-on-the-ground enforcement8 unavailable in national protected areas, and to provide material incentives to local government and the agents of deforestation through which the value of the standing forest could be comparable to the opportunity cost of forest conversion. If successful at scale, this strategy could work to reduce threats of non-permanence and leakage by transforming economic relationships between the people, the government and the forest at a national level, through a durable market-oriented policy and regulatory framework.
Thus far, the Southern Cardamom and Keo Seima REDD + projects demonstrated significantly higher effectiveness in reducing deforestation compared to adjacent PAs, while the Tumring REDD + project was less effective than 50% of the surrounding PAs. We hypothesize that a series of key enabling conditions are required for REDD + success in Cambodia, including: a history of on-site NGO presence and/or long-term conservation activities, rigorous boots-on-the-ground protection backed by law enforcement, stable funding from carbon revenues and a high-level of national protection status. Two new REDD + projects are currently being initiated in Cambodia in collaboration with the government: Northern Plains Landscape (NPL, initially comprising three PAs: Supplementary Table 6) and Phnom Samkos Wildlife Sanctuary (PSWS). Both of these project locations hold the above-mentioned enabling characteristics (Supplementary Table 7) and therefore appear to have favorable conditions in place for successful implementation.
Based on our findings, when implemented in association with favorable enabling conditions, REDD + initiatives have strong potential to ensure more effective long-term protection than national protected areas alone within high deforestation landscapes, jurisdictions, and countries.
Visit our sponsors
Wise (formerly TransferWise) is the cheaper, easier way to send money abroad. It helps people move money quickly and easily between bank accounts in different countries. Convert 60+ currencies with ridiculously low fees - on average 7x cheaper than a bank. No hidden fees, no markup on the exchange rate, ever.