It’s Like Herding Monkeys into a Conservation Enclosure: The Formation and Establishment of the Jozani-Chwaka Bay National Park, Zanzibar, Tanzania

This manuscript examines a project that is representative of an emerging trend of new generation Integrated Conservation Development Projects in parts of Africa that combine socio-economic development with an emphasis on local institutional change. These ‘local’ projects are interlinked with global networks of conservation interests that provide technical expertise and resourcing. In the Jozani-Chawka Bay area, project planners brokered a community governance and bene ﬁ t sharing agreement that has been lauded as a watershed moment for conservation policy in Zanzibar. Key hurdles for establishing Zanzibar’s ﬁ rst national park, the Jozani-Chwaka Bay National Park, were limiting community access to customary forest resources, farmer-red colobus monkey con ﬂ ict, and setting up a supportive institutional arrangement. The con ﬂ ict resolution and institutional strategies adopted by the conservation planners with the aid of international funding provide insights that help explain why the project has been able to maintain a ‘fragile’ localised compliance with conservation goals at the Jozani-Pete village. This has been achieved despite lingering resentment over red colobus crop damage claims, and perceptions of insigni ﬁ cant conservation related bene ﬁ ts ﬂ owing to individuals and communities. This ﬁ nding raises broader concerns about whether containment strategies to ground fragile project arrangements are conducive to engendering the longer term support of local communities for new generation Integrated Conservation Development Projects.


INTRODUCTION
For almost 30 years there has been a concerted effort in conservation policy and practice to enrol local communities in large-scale protected area management. To enact these projects, international conservation advocates and states need workable solutions that satisfy the values and logic of international donors, while promising a brighter local future at the same time. The term, Integrated Conservation Development Project (ICDP) has been applied to describe these initiatives (Wells & McShane 2004;Adams et al. 2004). ICDPs in practice inevitably collide with an enormously divergent range of contingent, interrelated biophysical, moral, cultural, and practical issues that need to be handled if they are to prevail. Whereas in the past, the focus of these projects was on providing material trade-offs for those in and around buffer zones, in 'new' generation ICDPs, institutional dimensions are now seen to be equally important as development opportunities in forging mutually supportive links between delivering local benefi ts and protecting global values (Shackleton et al. 2010).
This article traces the path of a new generation ICDP through the experience of the Jozani-Chwaka Bay Conservation Area (JCBCA) project, which led to the establishment of the Jozani-Chwaka Bay National Park (JCBNP), the fi rst national park in Zanzibar, Tanzania. In new generation ICDPs community-based conservation techniques are being increasingly 'scaled up', promoted, and mainstreamed by global networks of non-government organisations, donors and international fi nancial institutions (Duffy 2009). Of particular interest to this study was to capture key events involved in the brokering of a benefi t sharing agreement, and strategies to implement community-based conservation governance. To do this, the differences between the rhetoric of institutional design and the implementation dynamics around participation, agenda setting, benefi ts, and ongoing management practices needed to be sorted out (Shackleton et al. 2010). Central to this purpose in examining the JCBCA project was to reveal how actors acquiesced to, resisted, manipulated, and transformed policy as it was negotiated into formal arrangements and then practice. Accordingly, the research presented here is directed towards the following concrete empirically focused questions: What is being negotiated and what strategies do actors use in pursuing their desired interests? How are agendas shaped, driven and infl uenced by the multiple actors involved in the project? How do 'squeezed' actors fi nd room for their agendas and goals, and how are these incorporated into the ongoing transformation of the project? Responding to these questions reveals the way power works to quell confl icts, align interests, and therefore enable projects to be realised.
The paper begins with a discussion which places the JCBCA project in the broader thinking and experience of contemporary ICDPs and community-based conservation projects. Then the methodological approach, inspired by actor-oriented theory, is elaborated and applied to examine actor strategies around key events in the formation of the JCBCA project. The focus here is on the way that opposing knowledge claims were used in settling the farmer-red colobus monkey (Procolobus kirkii) confl ict, which was at the centre of the local benefi t sharing arrangement. This dispute was over a claim by farmers in the area that they should receive compensation because of crop damage caused by the endangered red colobus monkey population. Negotiation of this agreement was crucial to the formation of the JCBCA in particular and is emblematic of global values versus local livelihoods struggles, which are commonly part of ICDP tensions. This is followed by a description of how the conservation project has infl uenced economic opportunities within the Jozani-Pete village, which is located in the buffer zone of the conservation area. Next there is a discussion, drawing on aspects of the ICDP and community conservation literature, that highlights the strategies used by conservation planners to settle key events at Jozani-Chawka. The paper concludes by suggesting that although local material benefi ts will always be important for the success of ICDPs, attainment of institutional legitimacy is a vexed issue involving compromise, but nevertheless vital for the formation of projects and their translation into practice.

Trading-off: A key ICDP strategy
Conservation planners have been criticised for using trade-off techniques as a means to sell or to gain acceptance of preconceived proposals (or even plans), rather than as a 'genuine' opportunity for a range of local actors or communities to choose development preferences more or less freely (Brown et al. 1998). Despite the multiple scales of interests that need to be aligned to support ICDPs, the targeted ecosystems and biodiversity values determine their bounded nature and hence the human populations that are then targeted by specifi c projects. Not surprisingly, 'trading-off' is invariably constrained by limiting economic activities in these areas to those that are deemed (by conservation planners 1 ) not to jeopardise conservation goals. Extractive practices, such as woodcutting and shifting agriculture are usually the targets of constraint or reform in this trade-off process, which is aimed at directing socio-economic change to non-extractive activities, such as those connected to tourism (Mvula 2001).
In East Africa and elsewhere, tourism has been the favoured alternative development strategy for ICDPs. Tourism has been promoted by conservation planners largely because it is seen to motivate people in and around protected areas to value biodiversity goals as the attraction that visitors come to experience. These projects often revolve around profi ling endangered or charismatic animals (such as the red colobus) to capture the global public's imagination (Garland 2008) with the intention of securing lucrative nature-based tourism markets and attracting international donor support (Kaimowitz & Sheil 2007). This approach creates tensions when species valued for their conservation status and their potential to draw tourism have long been considered pests by people living in and around project sites. This formation of tourism and biodiversity conservation indicates that tourism as a development strategy in ICDPs is not a grass roots movement motivated by local development concerns. Rather, it should be been seen as states acting strategically under the infl uence of structural adjustment 'liberalisations'. Under these political and economic conditions, emphasis has been placed on increasing foreign revenue earnings and delivering broader development opportunities within a competitive globalised market place. In bringing these projects to fruition, government and international NGO actors collaborate in planning and intermediary roles to pass enabling legislation, foster capacity building, facilitate market linkages, and attract seeding or start-up capital.
Past experiences tell us that there are substantial challenges to effectively developing locally benefi cial tourism in ICDPs, including constraints to fruitfully participate in tourism, personal, and infrastructure deficiencies (Britton 2004); structural exclusion by remote based tourism companies in contract with government leading to gross local leakage (Tosun 2006); and fl uctuating and fi ckle global tourism opportunities (Mowforth & Munt 2003). These challenges need to be overcome if tourism is to be a signifi cant, widespread and reliable source of income for people living in and around protected areas, and therefore part of a motivating strategy for positive involvement in ICDPs.
There is also empirical evidence showing that rural households will continue to practice production activities that are congruent with existing competencies and livelihoods and embedded in social relations rather than opt for simple economic substitutions (Brown 2003;Goodwin & Roe 2001). This suggests that trade-off processes alone do not adequately consider the local politics of projects, or the inertia of customary practice. They therefore often fail to adequately deal with the complexity or totality of institutional change required to effectively create conservation with development (Garnett et al. 2007). This is why new generation ICDPs are increasingly adopting strategies that set up local governance structures to give communities authority and responsibilities (Murphee 2009).

Building communities for new ICDPs
Until recently the emphasis in the ICDP literature has almost exclusively been on the trade-off process between conservation, resource use constraints and local benefi t sharing from tourism and related activities (Shackleton et al. 2010). However, in new generation ICDPs, community has been given more prominence. How community is construed and affected in its interaction with the state and international conservation interests has become an emerging topic of interest. Essential to progressing our understanding of this interaction and its effects are explanations of the micro-politics of attempts to build multi-scale institutional formations in support of ICDPs (Marfo 2007). As Walley (2004: 14) puts it when writing about a similar project on the nearby Mafi a Island, these "dynamics include the ways in which institutions are contested by target populations, how individuals and groups create alliances as well as suffer exclusion and the ways in which ideas of development, nature and participation are variously understood, appropriated, disputed and used".
Conservation planners and some parts of academia 2 have been charged with taking an overly optimistic view of the transformative potential of participation, especially when much of this effort is placed on building organisations and social capital in support of projects (Nelson & Wright 1995;Cleaver 1999Cleaver , 2000. These critics have focused on confl ict, unevenness of power, and how 'participation' and concepts of community have been used instrumentally to impose agendas of powerful international organisations on the poor (Cleaver 2002(Cleaver , 2007Kothari 2001;Blaikie 2006). Joining this critique, Ribot et al. (2006: 47) describe how grassroots groups and NGOs, despite having formalised community-based credentials may not be accountable to or representative of local people in a systematic manner. Perhaps this is not surprising given the way that trans-local actors commonly co-opt traditional institutional arrangements to forward conservation agendas. The participation of project target groups (communities in and around areas of conservation interest) is almost inevitably limited to the scope of an agenda framed elsewhere (such as international NGOs and government agencies setting the rules of the game). In concurring with this view, Cleaver (1999) argues that where incongruity arises between the project goals of intervention and participatory aspirations to empower a wide range of 'weak' contributors, it is overwhelmingly likely that intervening actors' goals will prevail. Greater institutional empowerment and responsibility granted to communities within new ICDPs is seen as a means to counter-balance extralocal control. However, without effective mechanisms to deal with the way that local politics engages with projects it is more than likely this attempt to reapportion infl uence will be in vain.
Community-based management in ICDPs commonly rely on administrative legal means to hold target communities in buffer zones accountable (Li 2007). Key to these formal institutional arrangements are the elaboration of rules, responsibilities, and rights that seek to regulate the way that people use local natural resources. The aim here is to 'normalize' the desired resource use behaviour through incentives, constraints, and deterrents (Ostrom 1990). Agrawal (2005) argues that the role of these self-regulatory institutions is to discipline and self-discipline participants in accord with conservation objectives. This view sees power not only in terms of hierarchical, top-down power of the state at a distance, but also through the routine of community conservation practices. Much of the focus in this literature is on the knowledge which supports the political rationale and means that structures the fi eld of action, in which the subject (those whose conduct is to be infl uenced) can act (Bulkeley 2005). Related to this, Blaikie (2006) argues that 'local knowledge', has not succeeded in negotiating with scientifi c knowledge used by experts, on an equal basis (e.g., conservation planners) to shape project dynamics. In conservation interventions, scientifi c knowledge has been accused of rendering highly political issues, such as claims about agricultural pests (e.g., red colobus-biology), resource use (e.g., forests-scientifi c forestry techniques), and control (e.g., governance arrangements-common pool resource theory), as technical management issues that can be 'made' devoid of politics through better scientifi c understanding (Li 2007). Forsyth (1998) asserts that the ambition is to cleanse ecological knowledge from its historical, social, and political context (in some ways the antithesis of local knowledge). It is only in this way that it can be re-embedded as 'neutral' scientifi c knowledge to order the politics of society in line with policy actors' preferences (Forsyth 1998). The essential point being that power defi nes what is (more) acceptable as knowledge (Fletcher 1992) thereby providing an unobtrusive means to shape and set agendas. This makes possible settling conflicts without becoming mired in endless and messy political negotiations. Key to understanding this empirically is to identify which voices are given authority to speak at crucial project junctures.
This discussion suggests that the complex interaction between actors, projects and practices, their intended and unintended outcomes, creates both the constraining and enabling framework for social action (Long 2001: 4). This has implications for revealing how conservation policy becomes transformed in practice as different actors manipulate and adapt intentions and goals into their own logic and interests.

Tracing conservation interventions
The purpose of the fi eldwork was to focus on key events that needed to be resolved to enable the formation of the JCBCA. Jozani-Pete village, located in the buffer zone surrounding the JCBNP, is used to examine the role of community-based natural resource management (CBNRM) in this project. Jozani-Pete was selected because its enrolment in the project has been the most contentious, and therefore points of confl icts and strategies adopted for resolution between different actors and their interests were most visible ( Figure 1). The problematisation forwarded by the conservation planners in support of the intervention was that biodiversity was being threatened by unregulated forest use (Abdalla & Kitwana 1997), and that a conservation project with a mixed bag of state and community-based resource use regulation softened by new tourism-related economic opportunities was required as the solution. To deliver the 'solution', an agreement needed to be brokered among the different interests to align and stabilise cooperative action towards forming the JCBCA.
I used an actor-oriented approach to focus attention on significant points of tension around the intervention. Examining confl ict in a fi eld of social action such as an ICDP, reveals different norms, aspirations, and capacity to infl uence practice (Long & van der Ploeg 1989: 1-2). Key to this was identifying the actors involved, their strategies and interests, and how successful they were in getting their agendas formalised and translated in practice (Mosse 2005). In the JBCBA, these included international donors, local organisations, local people, NGOs, and representatives from different government sectors and levels.

Methods
Information was gathered from contemporary and historical documentation as well as materials collected during eight weeks of fi eldwork in March-April 2009, which included observation, in-depth interviews and informal conversations with Jozani-Pete villagers and actors involved in the planning and management of the JCBCA project. The fi eldwork at Jozani-Pete was carried out with the assistance of a village fi eld guide and translator. Informants were asked about the key issues for them in the formation of the JCBCA. Included were questions around governance issues, points of confl ict as well as transformations related to personal livelihood initiatives and village life since the coming of the project. Individual indepth interviews were held with villagers (n=40). Sampling was also based on consideration of characteristics such as occupation, age, sex, and socio-economic position. Interviews were also held with members from key village organisations, including the Village Development Committee (n=3), Village Conservation Committee (n=4) and the Sheha 3 Advisory Group (n=3). During the initial part of the fi eldwork there was some effort placed on understanding what constituted power within the village to ensure the inclusion in the sample of differently stratifi ed groups within the village [a wealth ranking carried out at Kisakasaka, Zanzibar in a previous study by Saunders et al. (2010) gave additional insight here]. I conducted a further 15 interviews mostly in English with actors selected because of their involvement in JCBCA management, operation, or research, including Department of Commercial Crops, Fruits and Forestry (DCCFF) offi cers [JCBNP (n=6) and headquarters (n=3)], Cooperative for Assistance and Relief Everywhere (CARE) staff (n=2), other researchers (n=2), and Jozani Environmental Conservation Association (JECA) board members (n=2).

Zanzibar conservation in transition
The shifting position of Zanzibar in response to global economic pressures and opportunities directed through multinational agencies has given rise to the contemporary development direction focused on conservation, leisure, and nature-based tourism (Gössling 2003). However, it is only recently that Zanzibar's natural values have been recognised by international conservation interests 4 . The political and economic liberalisation reforms in the 1980s created the space for multilateral agencies, and international and domestic NGOs to intervene in Zanzibari affairs (Levine 2007) to direct the (re-)commodifi cation of nature and culture for conservation-managed tourism consumption. As much as these macroeconomic changes resulted in a more liberalised political and economic milieu, the nexus between people oriented conservation with tourism culminated with the passing of Zanzibar's Environmental Management for Sustainable Development Act and Forest Management and Conservation Act. Both these Acts enabled and supported the delegation of resource management authority to institutions or individuals not employed by government, including communities (Levine 2007). Nature-based tourism is dependent on the sustainable consumption of biodiversity values and is, therefore, often coupled with conservation projects. Aside from conservation itself, a key objective under social conservation programs was to reorient control and benefi t to those living in and around conservation sites. This approach was informed by the limits of strict administrative boundaries to protect mobile wildlife, and an acknowledgement that imbuing communities with conservation responsibility and incentives is more likely to result in more effi cient control. Flowing from the set up of these enabling conditions there have been a range of conservation projects in Zanzibar based on tourism, that have been sponsored by international development aid in conjunction with private investment capital (Levine 2007). In these arrangements, the Revolutionary Government of Zanzibar has performed a bridging role in mediating the investments by setting up new institutions with rules designed to foster tourism investment and clarify and secure property rights (Chachage 2000). In the terrestrial environment these strategies are aimed at shifting the socio-economic behaviour of rural Zanzibaris away from wood cutting (for fuelwood, charcoal, poles, and timber) and shifting agriculture to conservation related tourism. It was in this context that the red colobus was seen as the key to attracting both the international donor support and tourism market needed to underpin the set up and operation of the JCBCA project, and in doing so help produce conditions that would foster future development and prosperity that would benefi t the 'whole of Zanzibar'.

Design outline of the Jozani-Chwaka Bay Conservation Area
In the mid 1990s, conservation planners (CARE and DCCFF) proposed a strictly protected area (national park) surrounded by a buffer zone of eight village territories (subsequently enlarged to nine) where restricted and self-regulated community resource use was to be allowed. In the buffer zone, resource use was to be managed by Sheha nominated village conservation committees. To support and inform community regulation, the conservation planners oversaw the development of Resource User Management Agreements (RUMA) with each village conservation committee. Management was to be regulated in accord with these nine separately negotiated agreements that specifi ed resource zones, use rules and community roles, and responsibilities and rights for each village.
The Jozani Environmental Conservation Association was established by CARE 5 in the late 1990s as an umbrella, domestic NGO with representatives from the management committees of the nine villages in the buffer zone. The purpose of this multi-scaled institutional arrangement was to enable community needs and priorities to feed into the conservation area wide planning processes. The role of JECA was to provide a conduit between the JCBNP staff 6 and the villages and to oversee the work of the nine village management committees (and later to disperse revenue generated from the JCBNP according to a yet to be agreed upon distribution scheme).

Reconstructing monkeys
The Jozani-Pete villagers have long practiced a broad range of economic activities. These include in descending order of importance agriculture, fi rewood cutting, petty trading, and lime making. Other less common activities have included hunting, quarrying, seaweed farming, and seafood collection. Like elsewhere in rural Zanzibar, farming in Jozani-Pete has been dominated by rain-fed agriculture for direct use and for sale in local markets. Most farmers in Jozani-Pete have also practiced shifting cultivation and wood cutting; livelihood activities seen by the JCBNP staff as a signifi cant threat to the biodiversity values of the area and used in the argument to establish a conservation area. One of the key impediments to the formation of the conservation area was the long-standing confl ict between monkeys and farmers. This confl ict was particularly acute in the Jozani-Pete area with common and voracious claims by farmers that monkeys were causing signifi cant crop damage.
The JCBCA project, with its emphasis on conservation, placed value on the red colobus ( Figure 2) because it is an endemic and endangered species and because of its potential as a tourism drawcard to generate income. Local distribution of income from monkey tourism could then be used to offset the perceptions and effects of greater resource use restrictions and the impact of monkey crop-raiding, and thereby gain local support for the project. In wildlife promotion and tourism terms the red colobus has been labelled the fl agship conservation species for Zanzibar and the key attraction that draws visitors to the JCBNP. The red colobus is a folivore that has a broad diet feeding on unripe fruit and leaves of common crops such as bananas, guavas, and coconuts, which are grown as perennial crops in usually small kinship based land holdings, known locally as shambas. The monkeys (around 2,500 in eight separate groups) reside in signifi cantly higher densities in the agricultural land adjacent to the JCBNP rather than in the surrounding forested areas (Siex & Struhsaker 1999). This land was external to the existing Forest Reserve, part of Jozani-Pete village territory and largely consists of shamba land. The author observed, and was also told by farmers and Jozani-Chwaka Bay National Park staff that red colobus groups, moving through the area in packs as big as 30, cause mechanical damage to other crops and in response there had been numerous incidents in the past where farmers have taken matters into their own hands to protect crops. In 2000, well after the conservation intervention was underway at Jozani-Pete, it was reported that a sub-population of red colobus was reduced from 200 to 30 by farmers, who killed (poisoned or shot) or chased them away because of crop damage (Zanzibar Commission for Natural Resources & CARE Tanzania 2000). It is not clear where this event occurred, but it underlines recent tension between farmers and monkeys (Nowak et al. 2009). This was and still is (in 2009) a charged issue. The issue at stake for DCCFF and its partners at the time of the formation of the JCBCA (mid 1990s to early 2000s) was to resolve ongoing confl ict between wildlife conservation (and its agenda of monkey tourism) and farmers in the area. The conservation planners sought to positively reconstruct monkeys as a source of revenue generated through conservation related tourism, in contrast to farmers (which include nearly all Jozani-Pete villagers) who saw them as agricultural pests that needed to be controlled because of their deleterious effects on crops. Economic concerns were not the only barrier for the conservation planners to overcome. In Kiswahili, the name for the red colobus is kima punju, which translates to 'poison monkey' and it has been suggested that this has associations with a kind of poison used by evil-doers. Punju is a legendary poison said to be made from snakes and other reptiles. The local etymology indicates that the red colobus is named after this mythical poison (M. Walsh pers. comm., March 1, 2010), which may enrich and supplement the economically based rationalisations of farmers' views towards them. It has been reported that local people believe that when the monkeys have fed in an area, trees and crops die. Dogs are thought to lose their fur if they eat the red colobus (Zanzinet 2004); McIntyre and McIntyre 2009; and mentioned in informal discussions with Zanzibaris). Another view put by Kombo (2005) is that red colobus are called 'poison monkey' because they feed on juvenile and immature fruit and leaves, which are toxic to other animals. The two differing explanations have different foundations, but are not necessarily mutually exclusive. As Walsh & Goldman (2007) describe in their discussion of attitudes towards the now thought to be extinct Zanzibar leopard (Panthera pardus adersi), some people in Zanzibar still have a strong belief in witchcraft, which can provide a convenient lens to view events and to incite action.
Regardless of the cultural basis of beliefs towards the red colobus, the negotiations over establishing the JCBCA revolved around the revenue sharing agreement to support village infrastructure as well as direct compensation to individual farmers whose land was to be enclosed within the JCBNP boundaries-which was a highly contentious issue. As part of this arrangement, regular payments to villagers were to be funded from visitor entrance fee revenue. This strategy, while discussed since 1996 (Goldman 1996), and supported by conservation planners directly involved in the project, was given additional political impetus after a public protest by local farmers who formed a domestic NGO called Umoja wa Wenye Mashamba Jozani. This protest received national media coverage on radio and newspapers when the JCBNP was being mooted. The situation became even more heated when the proposed JCBNP boundaries included the enclosure of Jozani-Pete shamba land.
Under the revenue sharing arrangement, the proposed eligibility for direct compensation was to be limited only to those farmers who had shambas that were to be used for monkey viewing by tourists and enclosed in the JCBNP's boundaries 7 . It was at this stage that the larger issue of direct compensation for all farmers in and around the JCBNP boundaries due to loss of crop yields from red colobus damage was sidestepped. Research fi ndings by Siex & Struhsaker (1999) infl uenced the handling of this issue. These fi ndings not only refuted the link between red colobus monkey browsing and smaller yields from coconut trees, but found that yields were increased through browsing. This fi nding provided the Zanzibar Government with the opportunity to effectively neutralise a hot political issue 8 . The JCBNP staff, members of JECA and other DCCFF staff advised in discussion around these events that a one-off payment for compulsorily acquiring the villager shambas was disregarded at the time because of the political heat it would have created-perhaps threatening international funding and ultimately the entire project. A senior DCCFF offi cial also suggested that it was also a way to motivate villagers to support the project given that this arrangement would create an ongoing reliance on revenue generated through tourism in the JCBNP. The same DCCFF offi cial believed that the international profi le of the project and the rather impressive involvement of large donors motivated exaggerated stories of local hardship caused by pest monkeys and resource use restrictions [a view also shared by Siex & Struhsaker (1999)]. The outcomes of this is that now around the southern boundary of the JCBNP there are shambas situated along side each other, equally affected by monkey crop damage claims, where one farmer is eligible for compensation and the other is not. This is because rather than compensating all farmers for monkey crop damage, which was the scope of the ambit before the Siex & Struhsaker (1999) study, it was now possible with the backing of scientifi c knowledge to compensate only those farmers whose land was to be included in the JCBNP, albeit on different grounds. This compensation was to be granted on the basis of restricted agricultural use and because DCCFF were using the farmers' land to conduct monkey tourism, not because of monkey-caused crop damage. All of the ineligible farmers interviewed at Jozani-Pete not directly involved in conservation exhibited anger and frustration over the compensation scheme. It was seen as unjust, and they blamed 'conservation' and the way it has been implemented by DCCFF 9 . Nor do the compensated owners of shambas in the JCBNP feel like they have been adequately compensated. The three informants interviewed with shambas now in the JCBNP, believed that the compensation was not suffi cient to cover the loss of income caused (as they saw it) by increased monkey damage and restrictions on land use 10 . They estimated that annual revenue, even after allowing for compensation had been reduced between 30% and 50% because of the restrictions on land use. During interviews with farmers there were two themes of complaints about monkeys since the JCBCA was set up. One was a perception of an increased monkey population in the area and the other related to their inability to act to do anything about the monkeys causing damage because of the introduction of heavy penalties for purposely harassing, injuring, or killing monkeys 11 . Siex & Struhsaker's (1999) study fi ndings were important in underpinning DCCFF's position of decoupling red colobus monkeys from crop damage and therefore from the JCBCA project. Finding otherwise would have given more clout to landholders located outside the JCBNP boundaries demanding compensation for crop damage claimed to be caused by monkeys. Thus, as a prominent member of JECA and a Jozani-Pete resident told me, it would have created an unmanageable situation, 'where anyone, anywhere in Zanzibar could apply for crop damage compensation'. This was a reference to the 5,000 farmers in the buffer zone who claimed to be adversely affected by monkeys causing crop damage.

Scant traces of 'alternative development'
This section describes the benefi t sharing agreement and presents evidence of the extent of the Jozani-Pete villagers' involvement in conservation related alternative income generating activities. The aim is to show how Jozani-Pete villagers have participated in the alternative income generating activities' 12 that have been explicitly initiated through or are part of the deliberate planning of the conservation project 13 . In the collection of data around this issue, informants were also asked to comment on how conservation had affected their livelihoods and what new initiatives they had or were participating in. The interviews were held in April 2009; no strict time limits were imposed, so informants referred generally to 'conservation' and related interventions around establishing the JCBCA, which commenced around 1995 and was formalised in 2004.
The revenue sharing benefi ts have changed slightly over the past fi ve years, but the proportional distribution has remained more or less the same. Under this retention, 18% of the visitor entrance fee revenue goes to a community development fund disbursed by the JECA among the nine villages in the buffer zone, 20% [called the (Pete) Mangroves Boardwalk Share] is shared between the Jozani-Pete farmers as compensation 14 and the Jozani-Pete community development fund, which is administered through the Village Development Committee, and the remainder goes to the JCBNP management and the Treasury of the Revolutionary Government of Zanzibar. The Jozani-Pete Mangroves Boardwalk is included as part of the admission fee to the JCBNP, even though it lies outside the JCBNP boundary; this is important because it entitles the Jozani-Pete village to a share of the entrance fee to the JCBNP 15 . Upon paying the entrance fee to the JCBNP, visitors are entitled to an interpretative walk through the mangroves accompanied by the JCBNP staff. This experience is usually bundled up with a monkey viewing and groundwater forest walk that is also facilitated by the JCBNP staff (usually, but not always, by volunteers from surrounding villages). The compensation was to be supplemented by alternative income generating activities aimed at reducing the Jozani-Pete village's reliance on extractive practices.
Many (n=30) of the informants when asked about the implications of the conservation area complained about the restrictions on access to forest products in the village commons (uwanda -village property) imposed by the conservation regulations, particularly related to fuel wood collection and charcoal making. A number of reasons were given by informants in support of this view. These included increased enforcement and monitoring in the former Jozani Forest Reserve areas (now JCBNP), increased restrictions placed on cutting wood in the dryland forest areas due to RUMA zoning, the establishment (and proximity) of the JCBNP and the conservation profi le that this brings, and the incorporation of former Jozani-Pete community forest land in the JCBNP. It seems from the consistent responses that nearly all fuel wood is sourced from the villagers' own shambas now, rather than common village areas or what was the former Jozani Forest Reserve.
Although 15 Jozani-Pete villagers were working at the JCBNP in 2009, only three ranger staff were paid employees; the remainder volunteers conducting visitor tours were dependent on tips from tourists for income. Volunteers, acting as visitor guides, wear a uniform similar to that of the professional ranger staff. Volunteers get some basic interpretative training and sign contracts of agreement with DCCFF stating that work performed does not entitle them to any monetary compensation. The income from tips varies widely and is affected by seasonal fl uctuations of visitor numbers to the JCBNP 16 . It was clear from observation and from talking to the visitors that some tourists were under the mistaken impression that volunteers were actually getting paid by DCCFF for conducting interpretation, probably because volunteers wear a uniform identical to that of the paid professional ranger staff, and also, in terms of interpretative walks, perform the same duties. Many of the volunteers were reluctant to discuss their working conditions with me 17 , but those that did (n=3) saw the volunteer role as a stepping stone to either professional work in conservation/tourism or were studying or engaged in other livelihood activities.
Another heralded income activity (CARE International in Tanzania 2008;Hartley & Rijal 2003) is making cultural products for the tourism and international market, however, only four women in Jozani-Pete were involved in this activity and they were not selling goods through the gift shop set up by JECA at the JCBNP Visitor Centre because they could get a 'better price' through other outlets such as MOTO 18 . The two women involved in this activity who were interviewed indicated that they used this income for funding special occasions, such as weddings, rather than as a reliable source of income.
Most Jozani-Pete residents interviewed were not directly involved in conservation related governance or associated income generating activities. They expressed a view of being 'resigned' to conservation and the JCBNP (n=25) and expressed resentment about the 'increased' restrictions that are placed on resource use believing that they are worse off now in terms of personal income opportunities. Representative of the response when asked about their view on the conservation initiatives in the area was, "conservation is good, but…"; even though the answer to this question varied, commonly there was a positive acknowledgement that the money from the income benefi t sharing scheme has provided support for the construction of village infrastructure, including health and school buildings 19 . This positive support was tempered by perceptions of loss of livelihoods for individuals and households because of increased conservation regulations and restrictions on resource (forest) use.
A savings and credit scheme was introduced by CARE to Jozani-Pete around the same time. This scheme was deemed by informants to be a success in its own right. That is, it facilitated a process that allowed many people (particularly women) to increase savings and have options to invest in business or personal expenditure. An NGO called Jozani Credit and Development Organisation was formed to help coordinate savings and credit activities in the JCBCA. I was advised that most villagers actively participated in this project. The project invested seed funds in terms of a matching loan to increase the group capital where every member of numerous small groups that had been set up could borrow the same amount of cash for individual investment in small business activities. The scheme was coupled with small business training. Apart from a few very poor villagers, all of the Jozani-Pete village informants interviewed were members of the scheme. Although none of the respondents linked this scheme to the JCBCA or to conservation more generally, they spoke positively about it and commonly viewed it as having helped them leverage their savings to enable investment in various items for household and business use. This suggests that linkages between savings and credit opportunities, and the conservation project were not readily apparent to villagers.
Although this section should by no means be regarded as a comprehensive survey of change in livelihoods since conservation in the Jozani-Pete village, the data indicates very little evidence of involvement in alternative development activities let alone any signifi cant benefi ts being provided by the conservation project at Jozani-Pete-clearly most villagers are still heavily involved in various forms of agriculture and animal husbandry, regardless of whether they are participating in the JCBNP related activities. This fi nding is consistent with a study undertaken by Salum (2009), which is the only academic publication to assess the benefi ts to villagers from JCBNP generated eco-tourism. A recent report (CARE International in Tanzania 2008: 5) also agrees saying, "... communities have engaged in fewer alternative income generating activities and increased quality and quantity to products that have reliable market in and outside their villages that fi t for forward contract system with positive impact on conservation measures".

DISCUSSION
Confl ict around people, parks and wildlife is not new in East Africa (Balmford et al. 2001;Lamarque et al. 2008). CARE's previous experience working in Uganda and elsewhere has helped forge the Jozani-Chawka Bay initiative into a formalised conservation area. The project design combined a benefi ts perspective (inspired from ICDP) with an institutional approach (inspired from CBNRM). This is a response to the main critique of the ICDP experience that cash incentives for those affected are not enough to ground projects (Garnett et al. 2007). Institutional building that recognises and legitimises affected actors' rights and responsibilities has become increasingly important in conservation projects. There was criticism that in early ICDPs people in local communities rarely had an opportunity to express their views or needs, and remained alienated from the management of protected resources and the alternative development initiatives (Larson et al. 1997). This experience suggested that it was not enough to only consider incentives provided by local benefi t sharing arrangements, but progressing ICDPs (as an effective conservation intervention strategy) would also require more explicit engagement with the dynamics of local interaction and concerns. This is a task that involves collaboration to build formal institutional structures to support conservation projects, such as community-based organisations and rules of resource use. Effective collaboration to build conservation institutions needs to acknowledge that 1) people pursue various interests in spaces opened up by project interventions, 2) some people are more constrained than others in pursuing these interests, and 3) project interventions will change the existing social and socio-environmental relations in some way (Long 1999).
A key event that limited the politics of the formation of the JCBCA was how the monkey-farmer confl ict was handled. Indeed by privileging scientifi c knowledge (Siex & Struhsaker 1999) over local knowledge (farmers' claims), the path was cleared for the policy justifi cation not to pay compensation to farmers for crop damage caused by red colobus monkeys 20 . This enabled a broader political controversy to be delimited from the agenda of the project. Interestingly, Siex & Struhsaker's (1999) study examined the implications of red colobus monkeys browsing on coconut trees and did not address possible impacts on other important agricultural tree species (e.g., bananas, mangos etc.) or the broader effects of mechanical damage from large groups of monkeys moving around. Although this fi nding of positive browsing effects has not altered Jozani-Pete residents' view of the red colobus as agricultural pests, it may have played a vital role in shoring up the participating international donor opinion at the time, concerning harmony in the project between conservation and local social justice. In practice, scientifi c knowledge was afforded greater value than local knowledge, despite the attention paid to local knowledge in the community conservation literature. In this case, even if scientifi c knowledge was justifi ably 'privileged knowledge', a mismatch still exists between the limitations of Siex & Struhsaker's (1999) fi ndings and how they were translated into a more generalised policy position. How are such disparate knowledge claims connected to actors' interests to be evaluated, if not through a political trade-off process? 21 Local knowledge claims were seen to be inseparable from strong vested interests and unfounded cultural beliefs, so these claims were easily devalued as parochial and irrational, but scientifi c knowledge was seen to be empirically informed, rational, and interest neutral, despite its obvious connection to policy actors' interests. The broader literature concerning the accuracy of the relationship between farmers' perceptions and 'actual' monkey damage to crops in different situations is unclear at best. This literature provides no direction here, except that some authors recommend that even perceptions of crop damage need to be dealt with in policy responses (Priston 2005). Siex & Struhsaker's (1999) research was politicised when it was transformed from limited and conditional fi ndings concerning the effects of red colobus monkey browsing on coconut tree yields to evidence for a policy position that denied a negative relationship amongst the red colobus, conservation, crops, and ultimately damage to local farmers' welfare. This exercise of power by policy actors, in delimiting the JCBCA project agenda has therefore played a key role in shaping how inhabitants around the JCBNP interact with, are affected by, and perceive, the project.
Mirroring ICDP fi ndings elsewhere (Goodwin & Roe 2001;Brown 2003), there was little evidence that economic substitutions have taken place because of the project-more that extraction based livelihoods have come under increasing strain. The lack of conservation related economic activities at Jozani-Pete is not surprising given the relative fertility of the soil in much of the area and its long history as an agricultural/ forest user village. It is also not surprising given the way tourism to the JCBNP is structured; visitation to the JCBNP is typically through short packaged tours (two hours maximum) operated by businesses either situated in Zanzibar City or in large foreign owned hotels in coastal resort areas. A visit to the JCBNP, like spice tours and other cultural attractions, acts as a supplement to the most common tourist pursuit-a beach holiday. At the JCBNP Visitor Centre there is no localised tourism agency or tourist accommodation or other infrastructure apart from a small gift shop (run by JECA) and a cafe. The consequence of this is that the only tourism revenue that is retained locally is that part of the visitor entrance fee that is administered through the benefi t sharing agreement.
According to Village Conservation Committee (VCC) members at Jozani-Pete, there is no formalised enforcement of resource rules. Even though the Jozani-Pete RUMA, established to guide resource management in the buffer zone, states legally recognised local resource use by-laws. When asked about this, the VCC members interviewed advised that enforcement was handled informally through face-to-face discussions and negotiations, without formally enforcing the sanctions specifi ed in the RUMA by-laws. This fi nding contrasts with stricter interpretations of common pool resource theory, which emphasises the primacy of enforcement (involving monitoring and sanctioning) for effi cient institutional functioning (Gibson et al. 2005). Reasons given for this revolved around a desire not to generate overt confl ict and divisions over conservation within the village, but also related to reciprocal respect and the importance given to maintaining kinship and other social relations and mutual obligations. The role of social institutions of kinship and networks in Swahili culture provide numerous functions beside friendship and recreation, including economic and political opportunities, land tenure and inheritance mechanisms, labour relations, food security, religious practices, and general support in times of ill-health (Walley 2004). Mosse (2003) and Cleaver (2000) have found similar motivations in different cultural settings in India and Zimbabwe respectively, and de la Torre-Castro (2006) has observed this in the context of community fi shing regulations in Zanzibar.
Even though positive economic benefi ts attributable to conservation were hard to fi nd, there were clear signs that people felt obligated or resigned to conservation's presence. Villagers were no longer actively undermining or resisting conservation in the area (e.g., through widespread illegal cutting, harassment or killing of the red colobus monkey population, or public protest-all forms of resistance that had occurred previously). The dynamics of interaction with conservation for villagers were far more complex and nuanced than overt acts of resistance. The importance of the institutional approach adopted by the conservation planners helps to explain this. The JCBCA planners grafted on to existing institutions by using by-laws and enrolling infl uential village actors, including the Jozani-Pete Sheha, in the conservation cause. Increased status, perhaps infl uence, and certainly opportunities to act and work as brokers and mediators were available for those already infl uential in the village who have embraced and participated in the conservation project. The relationship between the broader population of the Jozani-Pete village and bodies set up to represent their interests in, and regulation of resource use was unclear at best, thereby calling into question the democratic representativeness and accountability of these bodies, whose formation was largely sponsored by CARE with the membership overseen by the Village Sheha. The Sheha and the Sheha Advisory Group are not democratically elected by the members of the village, but appointed by regional government offi cials and considered by some to be political appointees (Tidemand 2003). This makes the linking of the community-based RUMA agreements rhetorically underpinned by participative democratic principles, to the locally legitimate, but hardly democratic, Shehia institutions, a noteworthy move by the conservation planners. This is a move where project effi ciency was traded-off against grassroots participation in governance. It is not the intention to forward an argument of exploitative elite capture (perhaps it is a type of benign elite capture), but it is merely an observation about the opposing tensions that have tangible implications, but are often glossed over in project evaluations.
Amongst the villagers there was widespread acknowledgement of the legitimacy of conservation goals, while challenging the limits that conservation had placed on their livelihood strategies. Although this section has emphasised the role of scientifi c knowledge in shaping trade-offs through the local benefi t sharing agreement, the way that the community-based institutions were grounded with local elites was also an important enabling factor for conservation at Jozani-Chawka Bay.

CONCLUSION
Drawing on the Jozani-Chawka Bay experience to examine how new ICDP alliances are constructed in practice, this paper describes how some of the more subtle aspects of implementation practice can result in localised compliance with conservation goals, even if trade-off benefi ts are not signifi cant (at least in the short-term). The methodology used in this study traced both the key trade-off processes and institutional dynamics to explicate their relationship in the JCBCA project. This has proved useful in capturing aspects of tactical interactions that are not often viewed in an integrated way. The JCBCA project is presented here as a new generation ICDP, because it tackled both community management and material benefi ts within a project primarily driven by conservation goals. Despite these dual emphases, the project had a top-down character, where pay offs were used as a strategy to bind actors to conservation rather than to support a self-administered change in local livelihoods. This suggests that the reform ambitions within new ICDPs-to include community conservation traits of participation and deliberation-were used not so much as goals in themselves, but as a means to improve legitimacy and thereby strengthen long-term project viability. The JCBCA was established because it managed to align international conservation interests, Zanzibar Government aspirations, and local institutions' interests. A combination of generous international funding support and the willingness of government authorities to mediate the involvement of outside agencies to develop funding sources for institutional support have made the most of limited opportunities for local development from tourism related revenue. These efforts were enriched by strategic and persistent efforts to decentralise conservation management to communities in the JCBNP buffer zones. Whether this mix results in longer term institutional stability with increasing local reward, however, remains questionable-particularly, given the current dependence on an unstable and limited (locally benefi ting) international tourism market and uncertain long-term international donor funding to support the JCBNP management.
The privileged and exaggerated use of scientifi c knowledge as a basis to delimit the requirement of the JCBCA project to respond to demands by farmers for compensation from monkey damage to their crops refl ected an exercise of power to 'depoliticise' a deeply political issue. This proved critical to the formation of the JCBCA. This cleared the way for a brokering of a revenue sharing agreement, that was central to quelling stiff local resistance, and assuring project donors, and thereby supporting the appearance, at least, of the harmonisation of local development and conservation.
The co-option of locally powerful groups may well have been necessary from the conservation planner's perspective, but whether this serves the interests of the broader sections of communities around the JCBCA remains to be seen. While this approach is pragmatic, it confl icts with other project goals such as widespread participation and empowerment. Intimate and/ or reciprocal kinship and social relations also worked to bind reluctant actors to the new conservation oriented conditions, even when they perceived that they were individually or even collectively disadvantaged or felt marginalised by the new ICDP induced socio-economic circumstances and institutional arrangements.
Regardless of the tensions between institutional design and practice, and the limited success of new socio-economic opportunities, the synergy of these two approaches has somehow internalised a mindset of a conservation future. This has resulted in a containment type of situation where many of the Jozani-Pete villagers, are not yet of the view that conservation offers them a meaningful and positive future. Perhaps as Sayer & Campbell (2004) warn, it is prudent not to be too hasty in judging 'outcomes' in such complex transitional projects, where lag times for the realisation of promises are likely to be considerable. The aim of this study, however, was not to judge outcomes so much as to take on the important task of examining how different actors interact in new generation ICDPs to shape what outcomes are permissible and possible. In doing so, this paper has revealed that externally driven conservation projects-underpinned by the integrated discourses of biodiversity science, communal governance, and local resource fl ows-must adapt and compromise in dealing with the heterogeneous local political terrain and, even if this arduous passage is navigated, it is likely that only a tenuous foothold will be established. played important parts in producing this paper were Institute of Marine Sciences staff, Drs S.M. Mohammed and N. Jiddawi, JECA executive members and DCCFF staff, Mr Ali Mwinyi and Mr Ali Basha.

1.
I am referring here to governments, NGOs, and multilateral agencies, who are the main initiators and mediators of ICDPs.
2. This refers to mainstream common pool resource theory theorists (Johnson 2004). The main purpose of this research is to provide a theoretical basis to inform CBNRM research (analysis) and project planning (policy).

3.
A Sheha is a village leader appointed by the government, and is thus regarded as a government offi cial rather than a democratically elected local government representative.

4.
Recognised as an important part of the Eastern Arc and Coastal Forest System (and Centre of Endemism in East Africa), and so recognised among the World's 25 Global Hot Spots (WWF-US 2003;Nahonyo et al. 2002).

5.
As an exit strategy by CARE; when its role in the setting up fi nished, a local NGO would be carrying on this role. JECA started out as an advisory group and become a formalised NGO in 1999.

6.
These are DCCFF staff employed in roles at the JCBNP.

7.
As a post script, in 2011 I was advised by DCCFF offi cials that the local share of revenue to communities distributed through the local benefi t sharing agreement has increased to 50%.

8.
A 1996 DCCFF report titled, Proposed Revenue Sharing Mechanism to Mitigate Farmer-Colobus Confl ict in the Jozani-Pete Area fl agged the importance of this study in providing the basis for parameters of the revenue sharing agreement (Zanzibar Commission for Natural Resources 1996).

9.
It was reported by the GEF in 2004 that 64% of farmers around the JCBCA said that their crops had been damaged by red colobus monkeys.
10. The factors cited for this included loss of charcoal making options, restrictions on planting perennial crops, greater restrictions on fuel wood cutting/collection, and increased crop damage by monkeys.
11. The penalties are imposed under Wildlife Laws.
12. These are deemed to be activities that are aimed at reducing livlihood dependance on local natural resources, such as park employment and tourism related income opportunities. I also included some information about the village and saving scheme here because it was specifi cally deployed as part of the ICDP.
13. Although the information I present here is directly linked with the JCBCA project, I did collect data on non-project related activities such as butterfl y farming, which was initiated by private interests in the area, but has received some fi nancing from the Critical Ecosystems Partnership Fund. The JCBNP staff were quite clear that the butterfl y farming project was not part of the DCCFF/CARE ICDP, which is the focus of this paper, although it could be argued that the butterfl y farming initiative only came about because of the existence of the JCBCA project. Having said that, about 26 Jozani-Pete residents were involved in butterfl y farming in 2009. The income opportunities from this activity varied widely depending on seasonality, time, quality of equipment, knowledge, and market. Indications from those involved were that for three to four months of the year working four hours a day one could earn between TZS 60,000 to 80,000 per month. Although the seasonal and other limits constrain the potential of this activity, it was seen in a positively light by the four Jozani-Pete villager butterfl y farmers interviewed.
14. This is estimated at USD 369 for each landholder per annum based on 20,000 international visitors per annum.
15. I was advised by local DCCFF staff that the average visitor numbers over the past fi ve years (2004)(2005)(2006)(2007)(2008)(2009) were 20,000 per annum. Based on the local benefi t sharing arrangement, these visitor numbers would have generated an estimated USD 12,800 per annum to Jozani-Pete Community Development Fund. A report commissioned by WWF/ DCCFF quotes visitor numbers at 16,484 per annum between the years, 2006-2008, which reduces the income to USD 10,550 per annum (Said 2008).
16. The main wet season is the low season and runs from March to the end of May.
17. Even though these people were volunteers they made sure that 'formal' permission was granted by the JCBCA Chief Park Ranger before they would agree to be interviewed.
18. MOTO translates as 'heat' or 'fi re' and is an inter-cultural initiative in Zanzibar to establish a network for local crafts cooperatives. It is set up as a competitive small industry for the sustainable development of the local economy, aiming to improve the livelihoods of, and empower, women in the rural communities.
19. I was told by several informants, including both villagers and policy actors, that Jozani-Pete's electricity supply improved as a by-product of the construction of the JCBCA Visitor Centre rather than as a specifi c goal of the project.
20. Despite this position being taken by DCCFF during the formation of the JCBCA and onwards, cracks still appear. Kombo (2005), a senior DCCFF offi cial acknowledged crop damage is caused by red colobus monkeys to farmers in the area and links compensation to the Mangrove Boardwalk revenue, which is revenue allocated to the Jozani-Pete village as community development funding, rather than to individual actors.
21. Agrawal (1995) questions the dichotomy generally assumed between indigenous knowledge and 'western' (or scientifi c) knowledge, and argues that these categories are (re)produced through networks of power. While the validity of scientifi c knowledge is subject to ongoing contest (via 'the post modern challenge'), how to weigh and value the differing logics and epistemologies of local knowledges in this sort of conservation intervention is the subject of even more disparate opinion.