4.1. Drivers of Change: REDD+
UN-REDD mainly focused on Free Prior and Informed Consent (FPIC) activities and community carbon monitoring, whereas FFI-REDD+ also tried to reshape the forest governance of the commune by introducing CBFM and restoring the villages’ customary forest boundaries [
48,
53].
Table 2 further elaborates on the conducted REDD+ activities in both research sites. The main difference between both projects is that UN-REDD employed a do-no-harm approach, meaning that it would not negatively affect local households, whereas FFI-REDD+ resembled an Integrated Conservation and Development Project (ICDP). Contrary to UN-REDD, FFI-REDD+ actively adopted a pro-poor approach, and included, besides carbon sequestration, other development goals in their planning.
Vietnam has been one of the original partner countries in the UN-REDD programme, since 2008. As part of Phase 1 of UN-REDD, two pilot districts in Lam Dong province were chosen to implement REDD+ activities. These activities had mixed outcomes (see: RECOFTC [
56] for a full evaluation). The FPIC activities carried out in Kala Tonggu (amongst other villages) consisted of four principles: all communities living in and/or around the forest need to be involved; FPIC activities need to be proactively undertaken; the heterogeneity of the communities needs to be acknowledged; and FPIC processes need to be driven by relevant beneficiaries [
53,
54]. During the FPIC process, several potential opportunities and risks were identified, and communicated to the local communities (see
Table 2). Representatives of the Food and Agriculture Organisation (FAO) and United Nations Development Programme (UNDP) proclaimed that, because REDD+ was not implemented in Lam Dong yet, the FPIC process has proven problematic. The people had little reason to reject REDD+, because the associated risks of REDD+ on local scale were unclear. Local people associated REDD+ mainly with receiving financial benefits for forest protection [
55].
As for FFI-REDD+ in Hieu, the project focused on four components (as mentioned in
Table 2). Stakeholder consultation involved capacity building of local authorities and agencies, as well as FPIC activities with local communities. FFI aimed to implement CBFM in 10 villages in Hieu, because Vi Chrinh already owned a Red Book for their community forest (808 ha) as result of a CBFM project by Japan International Cooperation Agency (JICA) in 2008. FFI-REDD+ wanted to restore around 18,000 ha of customary forestland in the commune. This resulted in a complicated negotiation process with provincial, district and commune government agencies and other stakeholders. This negotiation process is still on-going. FFI-REDD+ furthermore established carbon baselines and they conducted a carbon stock inventory in the commune. The communities in Hieu were not able to measure carbon by themselves yet, due to technical difficulties. However, they already established community forest management boards which patrolled the natural forest on a two-weekly or monthly basis, using GPS and skills acquired from the programme. Finally, environmental and social impact assessments were conducted. According to FFI, Hieu commune was estimated to lose 161.6 ha of forestland every year without the REDD+ project. According to this forecast, it would mean that in 30 years, the project site would lose 4688 ha of natural forestland in total [
48,
51].
4.2. Livelihood Strategies and Capabilities
While traditional livelihoods of the indigenous communities of Vietnam are often (wrongfully) characterised as fairly similar [
57], there are considerable differences between the K’ho of Kala Tonggu and the M’nam of Hieu. Many K’ho communities located near the Dong Nai River (such as Kala Tonggu) have practiced wet rice cultivation for more than a century now. These communities grew a variety of wet rice both regular and glutinous. The M’nam, on the other hand, have traditionally practiced swidden agriculture. The swidden fields were farmed for 2–3 consecutive years before it was left for fallow. The fallow period would take up 7 to 10 years. Besides rice, the M’nam also cultivated maize, yam, banana, sugar cane, pineapple, watermelon and squash [
58]. Another difference between the K’ho in Kala Tonggu and the M’nam in Hieu was that the former converted to Christianity during colonial times, whereas the latter kept their animist beliefs. Therefore, the K’ho were more or less well connected to “mainstream” society, whereas the M’nam remained relatively isolated.
The differences, as described above, can still be traced back today. The on-farm livelihood strategies of the K’ho consisted of coffee growing (Coffea canephora) (98.0%), wet rice cultivation (98.1%), and livestock rearing (28.0%). Besides coffee, these villagers did not grow any other cash crops. While currently the M’nam in Hieu also engaged in wet rice cultivation (98.1%) and livestock rearing (51.9%), a majority of the households still practiced swidden agriculture as a subsistence activity (73.1%). Only a few households in Hieu grew coffee (11.5%) and cash crops (11.5%–mainly industrial cassava). The off-farm livelihood strategies of both research sites consisted of forest monitoring (96.0% and 86.5% for Kala Tonggu and Hieu respectively), collecting Non-Timber Forest Products (NTFPs) (50.0% and 57.7%), and being a seasonal land worker (26.0% and 9.6%). Additionally, 65.4% and 13.5% of the households in Hieu were involved in logging for housing and hunting respectively. Households in Kala Tonggu, on the other hand, claimed to completely abstain from participating in these activities. In terms of non-farm livelihood strategies, only a few households in both research sites were working in business and services, or for the local government.
The findings above indicate that the households of Kala Tonggu were notably more integrated in the market economy and also less dependent on the natural forests than their counterparts in Hieu. The average income of Kala Tonggu was therefore significantly higher than Hieu–2,500,000 VND ($109.97 USD) a month for the former and 807,692 VND ($35.53 USD) for the latter. The most common source of household income in Kala Tonggu came from coffee smallholding (averagely 20,479,592 VND ($900.90 USD) a year), whereas most earnings in Hieu were derived from wet rice cultivation (3,373,356 VND or $148.95 USD). The average contribution of forest monitoring to the total yearly household income in Kala Tonggu and Hieu has been minimal—935,111 VND ($41.14 USD) and 609,189 VND ($26.80 USD), respectively. These earnings came from forest contracting, PFES, and/or the REDD+ project funding. Even though the financial benefits of forest monitoring were small, most people still decided to participate in forest monitoring.
Table 3 explores the reasons why some households in both research sites did not conduct certain livelihood activities. Households in Kala Tonggu stated that they did not practice (potentially) environmentally degrading activities, such as logging for housing and selling, collecting NTFPs, and hunting and catching animals, because of government policy. The households of Kala Tonggu were less dependent on the natural forest, and thus more willing to adhere to government policy and regulations. This also explained the second most-mentioned reason why people did not engage in these activities: they considered it a personal choice. The same counted for intensive cultivation. It was clear that, for now, wet rice cultivation and planting coffee were sufficient for the villagers to maintain their livelihoods.
No land or Red Book was for many households in Hieu the most mentioned reason why they did not practice shifting and intensive cultivation, and coffee growing (only half of all households owned a Red Book as opposed to 84% in Kala Tonggu). No one mentioned that they could not conduct swidden agriculture because of government policy. It is condoned in Hieu, because of food insecurity and lack of livelihood alternatives. On the other hand, the capability to cut trees for housing and selling, collect NTFPs, and hunt and catch animals was hindered by government policy and regulations. Therefore, there was a growing awareness among households in Hieu that the natural forest was not free for everyone to exploit anymore.
Land tenure and government policy were the two most important factors influencing households’ livelihood capabilities and strategies. Even though both communities mentioned that their livelihoods improved during the past twenty years (most important livelihood changes included: higher income, more stability and more food security), it can be inferred that the livelihood strategies and capabilities in both research sites were significantly different from each other. The coffee smallholders of Kala Tonggu were relatively well integrated in the market economy and depended far less on the natural forest, than the swiddeners of Hieu, who relied on the natural forest for subsistence. Additionally, the households in Hieu continued to participate in environmentally degrading livelihood activities, which could have serious implications for the success of REDD+ and carbon sequestration. It remains unclear how FFI-REDD+ will solve these interconnected problems related to food insecurity, deforestation and forest degradation in an inclusive and sustainable manner.
Based on the findings above, we reason that the less a community depends on the surrounding forest, the more likely REDD+ will succeed. Currently, the coffee smallholders of Kala Tonggu had little reason to engage in environmentally degrading livelihood activities due to their dependence on coffee smallholding and on-farm livelihood alternatives. This also explains why many households in Kala Tonggu did not vote during the FPIC process as they had initially little vested interest in REDD+ (
Table 2). Furthermore, the direct financial benefits of the REDD+ activities have been minimal until now, even though most households participated in forest monitoring. It was expected in both research sites that REDD+ would eventually pay the households for their efforts. Therefore, both communities saw forest monitoring rather as a future investment than a livelihood strategy. This, however, could severely jeopardise the success of REDD+ if community expectations will not be met after carbon payments or in the near future.
4.3. Governance Systems and Actors
Twenty years ago, the village patriarch and elderly were the most powerful institutions in the research villages—being in charge of land and forest management, conflict resolution, spiritual and social affairs, and supra-village relations and politics. Even though they were still present in the research villages, these customary institutions mainly served a ceremonial function now. Nonetheless, most households stated that customary institutions continued to play an important role. Village affairs were at the time of research handled by the village headman and secretary—formal government representatives on local level. They were governed by the Commune People’s Committee (CPC), which was, on its turn, regulated by the District People’s Committee (DPC) and higher level government authorities. Both UN-REDD and FFI-REDD+ did not involve customary institutions in their respective programmes, and, subsequently, further consolidated the power of formal institutions in the research sites.
The majority of households in both research sites claimed to be involved in REDD+. Only 14.8% of the villagers did not participate in REDD+ because of health or personal reasons. However, all interviewed villagers had heard about REDD+ and most households asserted that the Community Forest Management Boards (CFMBs), village headmen, patriarchs, government agencies and outside organisations involved them in REDD+ activities.
Figure 3 illustrates the forest governance systems in both research sites after REDD+ implementation. In Kala Tonggu, the central government through its REDD+ institutions—Vietnam REDD+ Office (VRO), REDD+ Network and the Sub-Technical Working Groups (STWGs)—was responsible for policy making, technical support and overall management (
Figure 3a). Lam Dong province and its REDD+ taskforce were responsible for implementing REDD+ in Di Linh district. The District People’s Committee (DPC), together with its Forest Protection Department (FPD), and in cooperation with Bao Thuan State Forest Enterprise (SFE) and Hieu Commune People’s Committee (CPC), implemented REDD+ and CBFM in Kala Tonggu village. The CFMB of the village protected and monitored the community forest through its patrolling teams. This was done in cooperation with the commune forest rangers.
At the time of the research, Kon Tum province was not a UN-REDD province yet. Hence, the governance structure in Hieu was considerably different from Bao Thuan (
Figure 3b). The central government and Kon Tum Province People’s Committee (PPC) were responsible for general policy making and forest governance guidelines. Kon Plong DPC, Thach Nham Management Board (MB) and Mang La SFE owned most of the forestland in Hieu—3323.8 ha, 1814.3 ha and 12,516.9 ha, respectively. The DPC implemented its forest management policies and regulations through Hieu CPC, and was also responsible for forestland allocation. Thach Nham MB furthermore directly contracted households in the commune to monitor and patrol its natural forests for watershed protection. Each village in Hieu had a CFMB and corresponding patrolling teams. FFI’s role in Hieu’s forest governance was twofold. First, it paid and trained villagers to patrol and monitor the forests (unbroken double arrow). Second, it fulfilled a role as facilitator between the villages, Hieu CPC, Mang La SFE, Kon Plong DPC, Thach Nham MB, and Kon Tum PPC (broken arrows). The NGO aimed, through negations and capacity building, to persuade relevant actors to allocate forestland to local communities. FFI reported and shared its finding through the National REDD+ Network to improve REDD+ policies on the national level.
Based on
Figure 3 and the survey results, three important differences can be found between Kala Tonggu and Hieu. First, UN-REDD did not change the forest governance of Kala Tonggu and Bao Thuan commune on a local level. On the national level, various new institutions and actors related to REDD+ were introduced, but, on the local level, it was mainly implemented through Kala Tonggu’s existing forest governance infrastructure. This is in line with the findings of Trædal et al. [
59] and McElwee [
44], which conclude that market-based approaches, such as REDD+ and PFES, will most likely only reproduce existing institutional structures instead of restructuring them. On the contrary, FFI-REDD+ actively tried to reconstitute local forest governance structures by introducing CBFM. This resulted in the establishment of CFMBs and patrolling teams. The second difference is closely related to the first one: UN-REDD did not introduce new actors in Kala Tonggu village. Most REDD+ activities were carried out by either the DPC or CPC, and the REDD+ interlocutors only served the purpose of awareness raising and conducting FPIC activities. FFI, on the other hand, became an important actor in Hieu as it was not only financially and technically supporting local households to patrol and protect the community forests, but it was also directly influencing provincial, district and commune level agencies. Third, the role of power in Hieu was more pronounced than in Kala Tonggu. There were many land-related conflicts in Hieu between local households and Mang La SFE, between villages, and among households within a village. The SFE refused to give up its land for community forestry, whereas local households kept trying to enter and exploit it. Lack of land therefore caused a lot of tension in Hieu, and therefore, a vast majority of the households approved the efforts of FFI-REDD+ to reallocate natural forestland to the community (see also: To et al. [
43]). This strongly resembles the second layer of the livelihood trajectory approach—the role of power [
30]. Local households simply lacked the power and means to restore their, supposedly, customary boundaries. However, whether FFI is able to challenge and restructure existing power relations in Hieu commune remains open to question, as the negotiation process is still on-going at the time of writing this article.
The most important characteristic that both research sites had in common was lack of genuine community ownership. The communities did have a community Red Book or were in the process of having one. Moreover, the CFMBs were allowed to enter and patrol the forest, deal with violators, participate in afforestation activities, and respective communities could legally collect NTFPs in their community forest. However, according to households’ statements related to governance systems and actors, the community was largely absent in the rule and decision making processes concerning forest management nor were they able to engage in commercial activities (selling their forestland or using it as collateral). Generally, the community had to adhere to the stipulated rules of the government in order not to have their Red Book revoked. Indeed, forest governance in Vietnam was far from being truly “decentralised” [
43,
44,
45].
4.4. Resource Systems and Units
The resource system in Kala Tonggu village and Bao Thuan mainly consisted of degraded evergreen broad-leaved forest; deciduous forest; mixed bamboo and rattan forest; and coniferous forest, consisting of
Pinus kesiya and
Dipterocarpus obtusifolius trees. The community forest of Kala Tonggu was divided in medium (330 ha, accounting for 66%) and poor evergreen broad-leaved forest (170 ha, 34%). The latter was targeted for reforestation and rehabilitation [
42]. Hieu’s resource system, on the other hand, had a much higher level of biodiversity. It mainly consisted of subtropical evergreen broad-leaved forests. Some forest patches on its mountain peaks were mixed evergreen broad-leaved and subtropical coniferous forests containing genera such as
Fokienia, natural Pinus, and
Dacrydium. Hieu accommodated around 35 rare fauna and 12 rare flora species [
48].
These findings reflect the types of exploited resource units, for both selling and subsistence, in both research sites. The households of Kala Tonggu primarily collected firewood, mushrooms, wild vegetables and bamboo shoots, and they caught or hunted for mice, birds and deer. Compared to Bao Thuan, the resource units in Hieu were much more varied. The collected NTFPs included: firewood, various types of mushrooms, bamboo shoots,
Centella asiatica, honey,
Anoectochilus setaceus and wild vegetables. Furthermore, local households caught mice, birds, ferrets, squirrels and frogs. People in Hieu, furthermore, revealed that they logged for a relatively high variety of tree species. Some of these species, such as
Aquilaria crassna,
Dalbergia tonkinensis Prain, and
Erythrophleum fordii, are rare and listed on the International Union for Conservation of Nature (IUCN) Red List of Threatened Species as critically endangered or vulnerable [
60].
The exploitation of the resource system was often influenced by formal and/or customary forest classifications. The former consisted of special-use (e.g., a national park), protection and production forests, whereas the latter included watershed protection, ghost, and shifting cultivation forests. Ghost forests are traditionally well-protected due to community spiritual beliefs and perceptions that they are guarded by powerful spirits and ghosts of deceased [
43]. Furthermore, the community forest was usually part of (degraded) protection forest. Because most households in Hieu were informed by FFI-REDD+ where their community forest would be located, they already started patrolling and monitoring it. The households in the respective research sites were asked whether they were aware of the location, boundaries and respective size of aforementioned forest classifications (
Table 4).
In Kala Tonggu, households did not really recognise customary forest classifications anymore (
Table 4). They classified their coffee fields as shifting cultivation forest because during the late 1990s they used slash-and-burn techniques to cultivate these fields. Especially, during this period, deforestation and forest encroachment were persistent problems in Kala Tonggu due to lack of clarity of forest classifications and boundaries. Prospective coffee smallholders would easily label forestland as production forest because the government supported them, through subsidies and loans, to grow coffee. This led to a high deforestation rate in the village and commune. Nowadays, the formal forest boundaries have been well established. However, some discrepancies can be observed in the findings. Even though all, but one, households claimed to be involved in forest monitoring, only 64.0% actually knew where their community forest was located. Households also often confused the protection forest for their community forest, because the estimated size of the former was 500 ha, which actually is the size of the latter. When households were asked where they collected their NTFPs, 52.0% claimed to do this in the protection forest, being a formally restricted area. While this amount is still relatively high, twenty years ago local households admitted to log (60.0%) and collect NTFPs (88.0%) in the protection forest—signifying a positive change over time. Nowadays, households mainly collected NTFPs in the shifting cultivation forest (70.0%) and community forest (68.0%)—forest area where exploitation, to a certain degree, is legally allowed.
The situation in Hieu was different from Kala Tonggu (
Table 4). Most households knew where their ghost forest was located (88.2%), while only 51.0% knew the location of the protection forest. Those who claimed to know where these types of forest were located primarily exploited the protection forest for NTFPs (73.1%) and timber (46.2%), whereas the ghost forest has been left mostly untouched over the past 20 years. Only 11.1% of all households collected NTFPs in their ghost forest, whereas no one mentioned hunting or logging activities in this sacred forest. Protection forests and shifting cultivation forests were primarily perceived as forests for exploitation, whereas customary forests, such as ghost and watershed protection forests, were relatively well protected.
Both UN-REDD and FFI-REDD+ did not incorporate customary forest classifications in their respective programmes. For UN-REDD this has not been problematic as the households in Kala Tonggu did not recognise customary forest classifications anymore. To the contrary, in biodiversity rich Hieu customary forest classifications continued to play an important role. If REDD+ would only consider formal forest classifications, there is a high risk that it would fail in communes such as Hieu. These households were often uninformed about any formal arrangements. Therefore, in some cases, building upon formal arrangements alone would not be effective, and sometimes even counterproductive [
60].
How did the resource systems and units of both communities change over time?
Figure 4 illustrates changes in the perceived benefits from the natural forests over the past ten years in both research sites. The changes in Kala Tonggu were more or less clear: the villagers proclaimed that there were currently less derived benefits from the natural forest than ten years ago (
Figure 4a). The results in Hieu, on the other hand, were mixed (
Figure 4b). Households did not seem to agree whether the availability of NTFPs, fertile land and valuable wood species changed over time. A majority of the households in Hieu did see increased benefits derived from the natural forest in the form of protection against natural disasters and watershed protection. Concerning major drivers of change in derived benefits, both communities mentioned restrictive state policy, overexploitation of resource units, and decrease of forest area due to deforestation and illegal encroachment. These findings reveal that even biodiversity rich Hieu increasingly experienced problems of deforestation and forest degradation. It was for this reason that FFI-REDD+ selected Hieu as their project site [
51].
To what extent has REDD+ thus far contributed to change in the resource systems of both research sites? More than 90% of the households in Kala Tonggu claimed that they had less access to the natural forest after REDD+ implementation. While these households perceived their community forest as a new classification, they did not attribute that to REDD+ as they already participated in CBFM prior to this programme. Around half of the households in Hieu proclaimed to have less access to the natural forest because of REDD+ (46.2%), whereas others stated that it remained the same or even the opposite. Contrary to Kala Tonggu and with exception of Vi Chrinh, the establishment of community forests in Hieu can be directly attributed to FFI-REDD+. This raises an important issue: did CBFM, as introduced by FFI-REDD+, really restore the customary forest boundaries in Hieu, or was this another attempt by outside agents to reshape the M’nam’s local forest landscape?
4.5. Focal Action Situations and Outcomes
The focal action situations and socio-ecological outcomes of REDD+ implementation are hard to predict. In both research sites, carbon payments were yet to happen. This depends on the outcomes of the Conference of the Parties (COP) meetings (UN-REDD) and approval of CCB or VCS accreditation (FFI-REDD+). It is for this reason that households in both research sites were asked to comment on statements related to socio-ecological outcomes. This last part of the results section primarily concerns livelihood perceptions and consists of three interrelated dimensions: harvesting levels and information sharing after REDD+ (
Figure 5); households’ assessment of socio-ecological performance of REDD+; and the socio-ecological vulnerability context.
The harvesting levels of the households in Kala Tonggu were lower after REDD+ implementation as the villagers stated that they and others made less use of the natural forests. Furthermore, information sharing regarding forest management and exploitation had improved since REDD+. (
Figure 5a) In Hieu, on the other hand, the villagers provided mixed statements regarding forest use of themselves and others (
Figure 5b). Contrary to Kala Tonggu, households shared less information with each other after REDD+. This could be explained by the fact that households in Hieu were being educated on REDD+ at the time of research. New information on carbon, climate change and so on could have given them the feeling that they knew less about their surrounding forests; hence, they shared less information with each other.
Generally, most households claimed to be happy and supportive of REDD+. As mentioned in
Section 4.2, they mainly perceived participating in REDD+ activities as an investment rather than livelihood. Households also claimed that they were consulted before REDD+ implementation and that local institutions, both formal and customary, were equally as supportive. Additionally, households confirmed that grievance mechanisms were in place. Other household statements regarding the social and ecological performance of REDD+, which were generally answered positively, included: fewer forestland-related conflicts; more derived benefits from the natural forest; REDD+ will improve the natural forest; and REDD+ is important in preventing natural disasters. However, both communities in the research sites asserted that REDD+ prevented them from making more use of the natural forest. As stated before, REDD+ restricted access to the natural forests of the local communities—this being more the case for Kala Tonggu than Hieu.
Most households in Kala Tonggu (64.0%) disagreed with the statement that REDD+ would improve their financial situation as opposed to a majority of the villagers of Hieu (86.5%) agreeing with the statement. Looking at the current marginal financial benefits of forest protection, this optimism could be a potential threat. Carbon payments have not been made yet; it might well be lower than expected. A study on REDD+ in Nepal has shown that carbon payments were barely enough to cover the transaction costs of the local communities [
61]. Moreover, 70% of the households in Kala Tonggu disagreed with the statement that REDD+ gave them more livelihood possibilities. This figure was for Hieu completely the opposite: 80.8% of the villagers saw more possibilities after REDD+. FFI-REDD+ focused more on co-benefits (improved forest governance, forestland allocation, etc.), whereas UN-REDD primarily dealt with future carbon payments. This allowed the households of Hieu to see the benefits of REDD+ from a broader angle, and thus they saw more possibilities.
Temporal dynamics of livelihoods are important predictors of socio-ecological outcomes. These dynamics are not only influenced by agency but also by the socio-ecological vulnerability context. While households in Kala Tonggu were relatively financially well-off, they were particularly vulnerable to fluctuating global coffee prices. The global Robusta coffee crisis, during late 1990s, severely affected coffee smallholders in Vietnam, including the households of Kala Tonggu. Coffee smallholding in Vietnam has only been profitable again since 2006 [
62]. Due to lack of social capital, many (prospective) indigenous coffee farmers also faced problems accessing loans or credit to establish or maintain their plantations. Besides facing problems with drought (96.0%) and alcohol addiction in the village (82.0%), this has been the second most-mentioned problem in Kala Tonggu village (82.0%). Therefore, while households in Kala Tonggu might not enter the natural forest to sustain their livelihoods now, they could be forced to do otherwise if coffee smallholding becomes unprofitable again. This could be detrimental to the success of REDD+ and carbon sequestration through sustainable management of forests.
In Hieu, most households stated to be vulnerable to: alcohol addiction (100%), flooding (98.1%), drought (78.8%), food insecurity (73.1%), and poverty (71.2%). The villagers, furthermore, faced problems with: lack of equipment for growing cash crops (88.5%); accessing credit or loans (82.7%); collecting NTFPs (69.2%); and hunting or catching animals (51.9%). The socio-ecological vulnerability context in Hieu is complex and multifaceted—influenced by both households’ dependence on natural resources and the commune’s growing integration in the market economy. Most likely, future carbon payments alone would not make these communities any less vulnerable, as the returns of REDD+ are predicted to be minimal. It is therefore important that the (future) co-benefits of REDD+ in Hieu also include: livelihood diversification, forestland allocation to households, and solutions to food insecurity in the commune (e.g., improved infrastructure for wet-rice cultivation or introducing sustainable forms of swidden agriculture).