Trust Building to Coordinate Collective Action in Reforestation Program

The study explores the way trust among agencies is established to coordinate collective action in rehabilitating protected areas, which have been utilized, commodified, and settled. Using an ethnography approach, the fieldwork was conducted in the villages surrounding 2 protected areas of West Lampung and South Lampung Districts in Lampung Province of Indonesia. There are several factors which hinder trust building process i.e. past experiences in relation to eviction from protected areas, forest policies which are not consistent, forest status which is protected areas, and the attitude of forest officers which consider land users as has no responsibility for conservation. Among those factors, forest policies which discursively and materially incorporate trust-building are the main factors which may help forest land rehabilitation process. Trust building process through negotiation where prejudice is turned into understanding among agencies still offer the possibility for forest rehabilitation efforts in the context of commodified landscape, agrarian change, and migration. However, negotiation is established through 'give and take' mechanisms, trial and error, and a learning process. Landscape transformation where forest land rehabilitation occurs relies on the 'art' of 'negotiation' at a local level.


Introduction
Lampung Province has a complex migration history, which can not be separated from colonial and post-colonial modernization and development projects. Many people, particularly from Java Island, migrated to Lampung to work for plantation during the colonial time. Besides, a various period of migration of people within or from outside Sumatera Island also occurs with various motives and objectives: government-sponsored migration and spontaneous migration (Elmhirst 2001).
There is a link between modernisation, development and migration, and the implications for environmental changes (Hugo 2008). Burns (2016) mentioned that "modernisation referred to development or change towards modern economic, political, and social systems". The concern of modernisation is economic growth. For example, the establishment of a large-scale plantation to accelerate economic growth. Further impact of the development is the flow of migration from relatively undeveloped areas to more developed areas both through government programs and self-initiated migration. The establishment a large scale plantation in Sumatera Island during the colonial time needed many workers. Therefore, the Dutch Colonial move many people from Java Island to Sumatera to work in the plantations. After Indonesia independence, both government programs, and self-initiated migration occurs for development program in the areas. As the consequences, large forested areas were cleared to establish a plantation.
The impact of modernization, development, and migration associated with forest land management could be seen in Lampung Province. Coffee mosaics can be found in almost parts of Lampung Province. It is without the exception the protected areas which have been commodified with coffee plantation. The landscape appearance does not show the protected areas should be.
During the Reform era which started in 1998, land clearing has become worse. To deal with forest land clearing in designated forest areas, reforestation program became a priority program in state forests. It was without the exception in the protected areas in the study areas of West Lampung and South Lampung.
Reforestation in the protected areas has been conducted even during the New Order era under President Soeharto (1966−1997). During the New Order era in the 1990s, when reforestation took place, communities who managed protected areas both in West Lampung and South Lampung were evicted from protected areas. Elephants and army force were used to uproot and to destroy coffee plantation during the harvesting time. In West Lampung, hydropower was built funded by the World Bank. Many were given compensation to work to build the hydropower. In South Lampung, much out-migrate to look for other livelihood strategies. After the New Order regime, Reforestation was continued to be implemented.
Reforestation program under community forestry (hutan kemasyarakatan) scheme has been implemented in the study areas in both West and South Lampung starting from the end of the New Order regime (the late 1990s). Through reforestation program, communities were given access to degraded land for their livelihoods by forming farmer groups, while at the same time communities were required to re-grow tree on bare lands and grasslands.
The inclusion of the community in protected area management is termed a "rights-based approach" (RBA). RBA combines conservation objectives, human rights protection, and economic security (Tapscott 2012). Tapscott (2012) further argues the use rights together with management and utilization rights will be able to achieve economic growth as well as social, political, and economic objectives. RBA can help negotiation process. The rights include culture, development, and livelihoods among others (Campese 2009). While socio-cultural aspects of forest management through community involvement is given importance in the formulation of forest policy, there have been other emerging features in forest management. Migration and commodified landscape are increasingly becoming phenomena in association with protected area management without the exception in the study areas. Migration indicates that formal institution becomes a more loose formation because of in and out migration in the areas. Commodified landscape in the other hands indicates global market as the driver of the choice of commodity planted. With the fragile form of institution and commodified landscape in the study areas, this research aims to explore the possibility of rehabilitating state forest land.

Methods Location
The study area in West Lampung is located in Bukit Rigis of Sumberjaya Sub-district totaling an area of 8,295 ha (Levang et al. 2012). It was classified as Boschwezen (forest controlled by the Forest Department) by the Belsuit Resident (resident's regulations) during the Dutch Colonial administration in 1935. The area of protected forest in Sumberjaya provides conservation functions, watershed protection, and also water availability for the downstream hydropower facility (Way Besai Dam was constructed with World Bank funding and began operation in 1994).
Meanwhile, study area in South Lampung is the Grand Forest Park (Wan Abdul Rahman), which is located west of Bandar Lampung City. It covers an area of approximately 22,000 ha. Conservation areas fall into a higher category of protected area than that in West Lampung District. Within the Indonesian regulation, this higher category of protected area means limited activities. The area was designated as a conservation area for biodiversity also in` the Dutch colonial era by Lampong Resident Belsuit (Lampung Resident Regulation) Number 16 of 1925. The objectives of Wan Abdul Rahman Conservation Area are biodiversity conservation, ecosystem protection, watershed management, education, cultural support, recreation, and tourism. Figure 1 shows the location of study areas.
In the map, the research sites are one research site in the protected area of Bukit Rigis of Register 45B in Sumberjaya West Lampung and another area in the Great National Park of Wan Abdul Rahman of Register 19 in Gunung Betung of South Lampung.
The research was conducted from 2004−2010 and was exploratory. An exploratory approach is using unstructured questions to investigate research participants' experiences which are unclear and unknown. It does not start with particular phenomena to investigate (Schwab & Syed 2015).
An ethnographic approach is used since close contacts and interactions with research participants are needed to study a phenomenon (Watson 2011;Pfadenhauer & Grenz 2015). The ethnographic approach helps to explore insight of cultural processes, which will not be available by simply conducting interviews without the engagement with the communities in their everyday lives. The ethnographic approach also helps to uncover unspoken issues or silences, which are not always able to be expressed by research participants through formal interviews (Witasari 2010).
Ethnography is considered suitable for the research although only a part of the story can be revealed through the engagement of the researcher with the research participants (Watson 2011). In relation to the use of ethnography approach, Escobar (2001) argues that a more advanced theoretical framework is not necessarily required. Rather, more sensitivity is needed in capturing the inter-subjective process of shared experience of 'world-making', focusing on every day, immediate practical activity, and on the embodied and place-based practices, and social life. Ethnography approach has the ability to read historical condition, to produce situated and comparative situations, and to analyze across scale (Fortun 2012).
Research participants were chosen based on 'snowball sampling', where the next interviewee was chosen based on information from the previous one (Handcock & Gile 2011). As much as 67 land users' individual interviews, focus groups, and group discussions were used for data analysis. Other agencies interviewed included forest officers at various levels (11 people), local government (3 people), regional people's representative (1 person), academics (2 people), and NGO staff (2 people). Table 1 showed the list of the 86 research participants.
For the land users themselves, there were 6 selected areas that were: Trimurti, Sukarame, and Mekarjaya (West Lampung) and Talang Agung, Benjo, Semeru (South Lampung). The choice of research sites was based on the various use rights to protected areas, agencies who facilitate the forest management, and types of forest cover. The aim of the variation is to explore themes which are important to formulate policy in forest rehabilitation in various conditions.
In Benjo and Semeru of South Lampung, the legal framework for forest management is an annual contract. The extension of access is evaluated based on land user efforts in rehabilitating land annually. The annual contract would not be extended when unsound forest management is applied by land users e.g. using fire to clear land or unwilling to plant the areas with timber trees together with multi-purpose tree species (MPTS).
Trimurti and Talang Agung are the first farmer groups which are granted legal rights through community forestry decree in Lampung. In Talang Agung, farmer groups are facilitated by a local NGO, while in Benjo and Semeru network is established with forestry extension workers and other forest officers. In the study areas, reforestation which had been done during the New Order era (1980−1981) continued in 1997−1998 through MPTS. This was applied in the areas which had been converted into agricultural lands for seasonal crops i.e. vegetables and commodities.
The size of individual land granted varies. In West Lampung, one farmer may utilize only 0.25 ha, while others may utilize around 5 ha (Sukarame 2002). The size was not determined by the government but depended on the size of the state forest land previously cleared and claimed by land users. Some land users utilized 2 or 3 state forest land areas in different locations.
The same situation occurred in South Lampung. However, areas managed by farmers were usually smaller than those in West Lampung. Some areas were overlappedmeaning that a land user may manage forest land in more than 1 location. For example land users who managed state forest land within the management of the farmer group in Benjo may cultivate other areas within Semeru's management. Land users in Semeru managed approximately up to 1 ha. Only a small number managed areas more than 1 ha. Around 65% of land users only managed state forest 171 Scientific Article ISSN: 2087-0469 Name of agencies Agents Note The forest department A high-level officer at the headquarters Previously working in community forestry area High-level officers at regional offices Middle-level officers at regional offices Low-level officers at regional offices  Table 1 Agencies interviewed JMHT Vol. 22, (3): 192-201, December 2016EISSN: 2089-2063 land.
The influential community leaders were usually chosen as the head of farmer groups. Meanwhile, the community members involved in illegal logging in the past were often appointed as the committee of farmer groups e.g. forest guards. The choice of farmer groups is determined by the information of other agencies basically in relation to the progress of activities. Membership is decided by land users whether they want to be a member or not. The characteristics of the farmer groups in the study areas are shown in Table 2.
Group meetings are arranged to encourage research participants to speak up. Through group meetings, dialogues among participants were facilitated. Dialogues focus on mutual understanding and relationship building. Through dialogues, various views can be explored. The emphasis is a difference, not a consensus (Pieczka & Wood 2013).
For this research, focus group discussions (FGDs) were arranged to identify the issues on the implementation of Reforestation Program. FGDs are group interviews to identify collective views on a specific topic, where participants interact with others as well as with the researcher (Ryan et al. 2014). In addition, FGDs are used to explore the construction and the negotiation of meaning (Moloney 2011;Stanley 2016). Representatives of each farmer groups were invited to attend the FGDs. During the discussions, however, there were notified that not all participants were empowered to speak out. The discussions were dominated by the farmer group leaders. There was a weakness of using focus group to identify the issues as what was found during the fieldwork. This means that FGDs do not always empower all participants to speak out as what mentioned by Moloney (2011) based on her experiences in her research.
Focus groups and interviews were analyzed as bricolage. Bricolage works to construct knowledge by using various methods. Each method supplement other methods. This approach goes beyond standardized methods. It is innovative and stimulates a unique knowledge development (McMillan 2015). Secondary data on farmer groups' documents, land use history, government policies documents, newspapers, research journals and other materials were also collected. The data was transcribed and analyzed.
Observations are used to explore transactional as well as material landscape as the manifestation of the dialectics between human-environment relationships and humanhuman relationships. The material landscape offers signs of cultural practices. Observation is the key method of qualitative inquiry. It integrates all senses: seeing, hearing, and feeling (Schwab & Syed 2015).
All data were triangulated. Triangulation means the use of various perspectives. It is used in both qualitative and quantitative research both to enhance a broader, deeper, and comprehensive understanding of the issues studied (Flick 2016;Kern 2016). However, triangulation is less used for the objectivity of interpretation (Flick 2016).

Results and Discussion Factors hindering trust building
The key themes identified from research to mobilize collective action in the rehabilitation efforts in protected areas are trust building processes and negotiating practices. The dominant interactions occur between land users and the Forest Authority because the Forest authority is perceived by land users to be the 'owner' of the protected areas.
Study on trust as a kind of social capital and a feature of social organization is popularized by Putnam. Putnam et al. (1993) mentioned that "social capital is a feature of social organization such as trust, norms, and networks that can improve the efficiency of society by facilitating coordinated   Putnam et al. (1993) furthermore argued that a trust is a form of social interaction that builds on positive interactions to lead towards collective action. Trust, therefore is important for the success of collaborative natural resource management (Stern & Coleman 2014). Trustworthiness 'lubricates' social life, making it work better.
In this research, trust as a kind of social capital as mentioned by Putnam above is crucial to collective action because individuals try to cooperate for the group interest instead of individual interest (Henry & Dietz 2011). "Trust is a central theoretical variable within common property regime (CPR) theory". The way trust established particularly between the government and communities motivated communities to coordinate collective action in rehabilitating state forest lands. Collective action is associated with institutional arrangement-as the core of CPR theory. The institutional arrangement is where individuals and or groups establish networks, social relations and social interactions in managing natural resources e.g. forests. It can be realized if trust can be established among agencies.
Trust is established through a process of relationship building and by determining the roles of each partner. Trust building processes are symbolic because they are reflective, part of a learning process encompassing give and take mechanisms as well as 'trial' and 'error' practices as identified in the research.
The research results showed that there is a relationship between trust and other aspects of protected area management as shown in Figure 2. Figure 2 shows that, legal rights do not always result in the certainty. As mentioned by land users, although they were given legal rights, land users did not feel certain. There was an important factor which interacted in 2 ways with certainty to manage protected areas. The issue was trust. A feeling to be trusted by other agencies, in particular by forest department made land users feel certain to manage protected areas. On the other hands, land users felt certain in managing protected areas, if they trusted forest department through the policies which are pro-land users.
According to land users, trust building process is established through communication/interactions among agencies. It is particularly between land users and forest department since forest department is considered by land users to have authority in protected area management. According to land users, communication and interactions are important to avoid prejudice which often occurs between land users and forest department.
Communication/interactions facilitated negotiation to take place. For example: through meetings or visiting land users in the fields. Although negotiation process did not always end up in a compromise, land users thought it important to trust building process. Trust building process itself interplayed with their performance in terms of their commitment to manage protected areas sustainably. Figure 2 also showed several barriers to trust building as a process and their impact on forest management and community engagement as explored from the perspectives of land users: 1 Past experiences Personal histories determine whether trust can be established or not (Stern & Coleman 2014). Both land users in West and South Lampung had experienced traumatic experiences in the past, i.e. eviction from protected areas. The implementation of the reforestation program in the study areas during the Post-New Order era (1998 and beyond) have continued to weaken trust between government and community because of centralistic approach: "one fits all".
These past experiences are associated with securing people's livelihoods. The evictions took place in the 1990s. Traumatic experiences were more felt by land users in West Lampung compared to those in South Lampung. It is because many land users have no private land, meanwhile many land users still rely on the land for Scientific Article ISSN: 2087-0469 173 Figure 2 The relationship between trust and other aspects of protected area management.  JMHT Vol. 22, (3): 192-201, December 2016EISSN: 2089-2063 their livelihoods. In West Lampung, communities whose livelihoods relied on coffee were offered income generation through resettlement programs and jobs with the hydropower authority. The evictions took place during the peak harvest-occurring only once in the 7years productive life cycle of coffee. The coffee plantation was operated. Land users lost the last 7-years of investment in their main livelihood asset-coffee plantation. They had to start again and searched for activities to re-establish individual and collective identities established among them and with the environment.
In comparison, land users in the study areas in South Lampung had alternative livelihoods from market access after the eviction from the protected areas since the locations are closer to the provincial capital and market. Some land users also owned private land. Therefore, they could cultivate their private land and did need to leave their villages for their income.
Traumatic experiences may result in resistance to new programs offered by the government, even though there had been a shift from the 'heavy-handed' approach by the New Order regime to "down-to-earth" communication by regional forestry officers. The impact is in the process of establishing trust.
Land users also showed resistance to participate in the management of protected areas offered by the NGOs and other agencies. NGOs and other agencies are perceived by land users as being instruments of the outsiders. As Pretty (2003Pretty ( : 1913 notes: "…trust takes the time to build and is easily broken. When a society is pervaded by distrust or conflict, cooperative arrangements are unlikely to emerge". Trust building process started initially from asymmetrical power relationship between the state and society into "give and take" mechanism where land users gradually earned the trust of the government and gained more control of and authority to access protected areas at the time. Land users perceived that they would continue to be granted access to the protected areas if they conserved the areas. Some land users, however, perceived trust building as only a tool of the Forest Department to achieve its own goals. 2 Forest policies Despite the reforestation program has incorporated communities, reforestation is still often seen only as planting trees without understanding planting as a product of the socio-cultural process. History and culture influence and interplay with land use and resource management. However, history and culture are often neglected in resource management (Hibbard et al. 2008;Agatha 2016).
Furthermore, reforestation program in the study areas was implemented within the national policy. The policy was centralized. It was implemented nationally, without the exception in Lampung Province. This universal and standardized reforestation project hindered the regional authority from negotiating with land users. The regional authority was only the executing agency instead of the decision maker. Meanwhile, trust building is about negotiating rights and responsibilities which require the authority of regional forest officials to take a decision. For land users, the result was clear. There was difficult for negotiating rights and responsibilities because of the difficulties to establish trust between regional forest officials and land users. The regional forest officials, however, kept implementing the program as a form of loyalty to the central government.
Secondly, the evolution of forest policies at the national level also influenced trust building process at the local level. When the first time reforestation was launched, the percentage of timber required to be planted by land users was lower than the percentage of other trees to be planted. In South Lampung previously even timber which was planted by land users could be harvested. As the proportion of timber required to be planted was lower than other trees, land users were willing to participate in a reforestation program. However, along the time the percentage of timber became higher than that of other trees and no more timber could not be harvested. Since more proportion of timber was required to be planted, land users were uncertain about the consistency of forest policies -whether the policy would change in the future.
Policy changes and inconsistencies of the national reforestation project resulted in a significant distrust as well as skepticism about the government's commitment to supporting land users' livelihoods through access to protected areas. The evolution of reforestation policy where more timber required to be planted the reconstruction of the old conservation paradigm. The image of more timber to be planted means 'forest' to land users. This imagined 'forest' was considered by land users as a way to re-evict them from protected areas. Land users felt skeptical of participating in reforestation programthinking no benefit for their livelihoods as well as for the possibility of re-eviction from protected areas. According to land users, taller trees means no space for annual crops or in other words no livelihoods for them. The nature of conservation areas which results in physical displacement or eviction and economic displacement (restrictions which make the community not able to pursue their livelihoods) will threat people livelihoods (Brockington & Wilkie 2015). More conservationoriented reforestation program was also perceived by land users as a lack of trust by the government on existing land users' conservation knowledge and practices.
At a community level, there was also the impact caused by the evolution of reforestation policy. The farmer group leader was not willing to inform his members of the possibilities of benefit from timber as he was informed by the regional forestry officials. The reason was that he did not want to give 'uncertain hope'.
Thirdly, there was an impact on social cohesion within communities. Referring to Putnam et al. (1993), the implementation of forest policy has the potential to destroy the existing social cohesion by hindering collective action and changing the horizontal relationships within communities caused by reforestation funding.
Reforestation funding resulted in a divide between the Scientific Article ISSN: 2087-0469 communities because funding management was in the hands of 'elite' farmers or farmer group leaders. This divide increased distrust and undermined the existing social cohesion due to the lack of funding management transparency and accountability of leaders. 3 Forest status According to land users, protected area status means that no timber can be harvested as well as a limited autonomous decision making to utilize state-owned forests. Forest status in relation to reforestation policy which required a higher percentage of timber to be planted made land users unwilling to participate in a reforestation program. In other words, it was not only about the forest status, but more importantly changing policy which was more timber-oriented. This was perceived by land users as no support from the government in regards to their livelihoods.
Although protected areas were under the government's authority, not all land users wanted to claim protected areas as their own. Many acknowledged protected areas as a state forest. What they expected was long-term benefits from state forest management. This was particularly expected by land users who did not have private lands. 4 The attitudes of forest officers Forest guards' attitudes on the way the communication was established were considered by land users hinder trust-building between the forest department as an institution and land users. Instructive approaches by the forest guards, instead of dialogue was perceived by land users as no trust to land users to undertake responsible land management practices. Frequent monitoring by forest guards further increased the tension.
Extension workers also facilitated forest management in the field beside forest guards. Good communication by the extension workers to establish built trust was not always successful. This was caused by an imbalance between rights and responsibilities as the impact of the reforestation policy. Therefore, negotiation was absent. Land users had uncertainty since they could not give any input for decision making. They questioned the continuity of managing state forest land. Uncertainty led many land users to search for alternative livelihoods out of the villages. The impact was further deterioration of social relationship within the community as well as the deterioration of forest condition.
Overall, from various factors which contributed to trust building process the most significant factor was forest policy and this could also create other problems. For example, as discussed above, forest status which is protected areas can make communities participate in reforestation program as long as the policy accommodates livelihoods for land users.

Trust building processes involving other agencies
Trust building with other agencies (forest officers, the academics, the NGO staff, local government staff) was also analyzed to understand the way the trust was established and its contribution to the willingness of land users to participate in a reforestation program. This research showed that other agencies were skeptic towards land users' responsibilities. In other words, they did not trust land users. Through interviews with agencies including local government staff, NGO staff, national forest officers, and academics, many of them were pessimistic on sound forest management by land users.
In general, other agencies thought that land users only took benefit from forest land for economic objectives and did not care about conservation. Land users in particular migrants are often blamed for environmental problems, for example to legitimate government restrictions to land ownership claims. Meanwhile, the possible positive impacts of migration such as new agricultural techniques, socioeconomic benefits, and skilled labors are often undermined (Zommer & macDonald 2012).
There was also a perception in relation to a stereotype on cultural practice: the dichotomy between native people and migrants. Land users are from various ethnicity and motivations. The majority of land users in the study areas are javanese, categorized as migrants in Lampung Province within the state's definition. Meanwhile, lampongers are perceived as native people in Lampung Province.
Relating to land use, migration is often associated with land use change (Carr 2009;Ouedraogo 2009). The loss of forest results from the establishment of new farms by migrants (Carr 2009). Zommer and macDonald (2012) mentioned that the impact of migration on the environment is determined by the market, institutions, and culture.
The consequence of migration in the study areas is that formal institution only cannot define who should benefit from protected areas. In and out migration from and to the areas could not be controlled. Therefore, both informal institutions through networks and formal institutions are important in defining who should benefit from protected areas. A common rule is that the benefits from forest areas should go to land users living in or close to forest areas and who are dependent on these lands for their livelihoods (Massiri et al. 2015). Having only one livelihood is rarely the 'norm'. Land users are not necessarily living close to the protected areas. They have multiple livelihood strategies which include migration.
Interviews with land users informed that the migrants and their ancestors came to Lampung at various times. Their ancestors came during the colonial time through transmigration programs and the establishment of plantations in Lampung. Many came as voluntary migrants. Later migration to the areas was also attracted by coffee plantation as land users said. The relatives of land users came from Java to cultivate coffee plantation-the promising global commodity. Capitalism together with globalization is implemented through cultural practices (Massey 2004). Migrants embodied capitalism characteristics and are blamed for land degradation.
There are struggles over cultural and symbolic meanings associated with migrant social relations. These struggles show the connection between land users to trust-building processes. The problematic colonial definition between migrants and indigenous people is continuously operating in this way within the study context. Discourses on identity are continuously used in Lampung Province even after the Post- JMHT Vol. 22, (3): 192-201, December 2016EISSN: 2089-2063 New Order era.
The land practice of the lampongers-the local people were perceived by other agencies more sustainable compared to that of migrants. Land practice associated with upland culture or shifting cultivation still mainly practiced by indigenous people is considered more sustainable compared to modernized cultivation practiced by migrants, which can bring the impact of forest clearing (Gibson-Graham 2016).
The stereotype on cultural practice was embedded in the communities' mind. For example, a migrant from Java said that if people from Java had come earlier, the landscape might have been different. He did not want to be blamed for forest clearing. A migrant descendent of a government-sponsored migration scheme called himself a 'local', not a migrant. Claiming to be local is a way to escape from 'migrant' identity which is associated with agents for forest clearing. According to him, self-initiated migrants, other 'migrants' and 'newcomers' who were perceived by him as 'migrants' were the real agents of deforestation. In fact, the study areas are comprised of complex communities with various motivation in managing protected areas. It could not be only binary between indigenous people versus migrants. It is also various economic level among land users.
What should be understood by forest officers and other agencies is that forest rehabilitation to transform the bare landscape into the forested landscape is not simply to plant trees, but it is the articulation of struggle of transforming overall livelihoods which include not only economic aspect but also socio-cultural aspects as well as the identity of caring nature. They should understand how this transformation process evolves and how local environmental knowledge emerge. Therefore, forest officers and other agencies should trust land users for their capability to manage protected areas. Setting the boundary between local people and migrants only hinders the outcomes of the reforestation program. Trust is a modality which influences the way forest policy is formulated.

Trust building process through communication and negotiation
Face-to-face interactions would help trust building process and negotiation, although it is not always successful. In addition, the role of other agencies is important to facilitate dialogue; negotiations on tree density, the numbers of trees to be planted, and the percentage of fruit trees and timber.
Land users asked for facilitation from NGO since they were not confident to negotiate with government directly. Some respondents said that they were unable to give voice during the New Order era. Therefore, it has been hard to initiate the dialogue without facilitation from NGO.
The impact of changing reforestation policies was the difficulties for field staff to nurture social relationships and trust with land users since the field staff could not ensure that the policies were consistent. In addition, centralized reforestation policy which was formulated by the Ministry of Forestry at the headquarter makes the regional forestry officers were difficult to negotiate with land users. This was because the regional forestry offices did not dare to change the policies. However, the willingness of the regional forestry officers to visit and to arrange the meetings with land users helped trust-building process despite a feeling of uncertainty in managing protected areas. The head of district forestry office often invited land users for dialogues, informing land users about successful farmers in planting trees by showing land uses the photographs and asked them to visit the areas of successful farmers. The response from land users was positive in terms of willingness to plant trees.
Strategy offered by forest officers and NGO with already commodified land was agroforestry where timber is planted together with other trees and crops. The approach is common to accommodate both livelihoods and conservation objectives. Various tree species which are planted together are expected to decrease market shock caused by the global market failure. More importantly, agroforestry approach also aims to reduce the side-effects of clearing the remaining forest areas (old growth). Changing land practice from monoculture to polyculture is a process, however since many of land users had no experiences before with polyculture. This change was continuously negotiated between land users and other agencies. One land user interviewed mentioned that he was surprised when he visited a location and found coffee planted together with other tree species. He was also surprised that coffee could grow well as he believed that coffee could not grow well together with other tree species.
Trust-building process for creating change through access and autonomy Trust is required for the success in rehabilitating protected areas with people's participation approach. As described above, the most important thing is that trust building process relies on the sensitivity of forest policies besides other factors i.e. historical experience, forest status, and the attitudes of forest guards. The sensitivity of forest policy is reflected through access and autonomy to land users.
In a classical property sense, 'access' is interpreted in terms of capital or an economic aspect, although there are social and cultural dimensions. Control over resources and the capacity to make the decision are important for selfdetermination (Moore 2012).
Access in terms of a bundle of rights to protected areas is still important to land users since most of the respondents still depend mainly on land and natural resources for their livelihoods. Access to protected areas helps to decrease outmigration to other areas as mentioned by land users. It is particularly for land users who have no private lands-making them more vulnerable to secure their livelihoods. Therefore, the policy to give access to protected areas is important to provide 'in-situ' livelihoods associated with social and cultural practices.
Access to protected areas results in increased responsibilities as perceived by land users. From the research, granting access decreased grassland clearing for coffee plantation. In addition, access to protected areas also helped to enhance local capacity to deal with internal conflicts in relation to access.
In relation to access, land users expected to be granted a longer period of access to manage protected areas. A longer period of access would provide them a change to balance rights and responsibilities which are livelihoods and conservation. Land users said that fruit trees and annual crops cannot be harvested in a relatively short time, let alone timber which is required to be maintained by the forestry JMHT Vol. 22, (3): 192-201, December 2016EISSN: 2089-2063 officers. A longer period of time would be sufficient for the transformation process of land use practice associated with livelihoods, social and cultural processes. It includes collective processes e.g. the establishment of community organizations, collective ideas of sound land practices.
Above all, land users emphasized the importance of social relationships, established from time to time with forestry personnel, as providing a more dependable outcome for them besides autonomy and a longer time for access. Social relationship is the foundation. Access through written legal rights perceived by land users is still not enough to guarantee effective management of the state forests. As the Figure 2 showed that legal rights do not always make land users feel certain in managing protected areas. The social relation could help trust building process in the absence of security and certainty in managing protected areas because of the memory of the local people in relation to negative past experiences, the evolution of forest policy, and forest status.

Conclusion
The trust-building process is the key theme to make land users willing to participate in reforestation program collectively. Being trusted by other agencies that land users concern about conservation also motivate land users to participate in reforestation. In another side, considering that other agencies, in particular the government, can be trusted is also important to make farmers responsible for collective benefits instead of pursuing only individual benefit. Based on the research finding, there are several reasons which hinder trust building process i.e. past experiences in relation to eviction from protected areas, forest policies which are not consistent, forest status which is protected areas, and the attitude of forest officers which consider land users as has no responsibility for conservation. Trust building process requires 'give and take', trial and error, and is expected to create an understanding among agencies. Established trust may offer the possibility for forest land rehabilitation in the context of commodified landscape, agrarian change, and migration. If reforestation program is understood not only simple as planting trees but also as the articulation of economic and socio-cultural identities, reforestation program may be successful in terms of people's livelihoods and forest land rehabilitation. The transformation of land practice embedded in socio-cultural identities needs to be accommodated through a longer period of access as well as authority. In addition, communication is very important in the negotiation process. Though there was a barrier on centralized policy where regional forestry officers had no authority in decision making, approach by regional forestry officers which was 'down-to-earth' assisted trust building process.