Institutional challenges to the conservation of Arabuko-Sokoke Coastal Forest in Kenya.
In the past decade, scholars and grassroots societies have increasingly attributed natural resource depletion to governance 'failures' not least by national governments (e.g. Ostrom 2008, Agrawal 2003, Ribot 2004, Lindsay 2004, Saunders 2014). As a response, interest and donor organisations have promoted different forms of community-based management strategies accompanied by changes in institutional setups and legal frameworks.
In the forest sector, Participatory Forest Management (PFM) is an umbrella term for some of the more inclusive natural resource management approaches. PFM is officially adopted in many African countries but with ambivalent results in terms of meeting the dual-objective of equitable poverty alleviation and forest conservation (see IIED Briefing 2016, Lund and Treue 2008, Pouliot and Treue 2013, Rasolofoson et al. 2015, Wardell and Lund 2006). Since the early 1990s, various projects have been piloted and promoted under the name of PFM in Arabuko-Sokoke forest in coastal Kenya (Mogaka 1991, Matiku et al. 2012a, Matiku et al. 2012b, Sande et al. 2009, Jackson and Naughton-Treves 2012). However, despite a long history of official and donor-supported decentralisation efforts in the area, PFM is still in a process of being defined rather than implemented.
Studies of forest decentralisation in developing countries report ambigouos effects on conservation and rural livelihoods. A vast amount of literature indicates a tendency for PFM projects to deliver different outputs than officially intended. Examples of indisputable success are few while evidences of failure and trade-offs between conservation and livelihood objectives are common (see Rasolofoson et al. 2015, Cox et al. 2010, Pagdee et al. 2006, Persha et al. 2011, Porter-Bolland et al. 2012, Pfliegner 2014). Among the key issues raised by scholars regarding decentralisation of forest management in Kenya are elite capture, lack of downward accountability and retention of meaningful powers by the state authority (see Thygesen et al. 2016, Chomba et al. 2015, Musyoki et al. 2016, Mutune and Lund 2016, Ongugo 2007).
Research into the effects of strategies promoted as PFM is accumulating but research into the reasons why officially intended and actual outputs often differ is wanting. Accordingly, this paper investigates how formal and informal institutional arrangements, as well as legal frameworks, influence PFM projects. This is approached by exploring two different but related PFM strategies in Arabuko-Sokoke; (i) decentralization, which we evaluate against the official dual-objective of forest conservation and poverty alleviation, and (ii) alternative income generating activities, in this case butterfly farming, which is intended to improve forest adjacent households' livelihoods and simultaneously provide an incentive for forest conservation.
STUDY SITE: THE ARABUKO-SOKOKE FOREST
Fieldwork for the study was carried out in Arabuko-Sokoke forest (Map 1), which covers an area of 41 600 ha and forms the biggest remaining protected area of the increasingly rare East African dry coastal forests. (Mbuvi et al. 2015, Myers et al. 2000). East African Coastal forests are included on the list of the ten most threatened biodiversity hotspots in the world, and they are habitats of an array of wildlife including rare, endemic, and globally threatened species (Habel et al. 2017, Glenday 2008, Muriithi and Kenyon 2002, Matiku et al. 2012). Likewise, much of the flora is either nationally or globally rare (ibid.). Solid scientific evidence on biodiversity and species developments in Arabuko-Sokoke forest is still wanting, but recent studies agree that the coastal dry forests of East-Africa are deteriorating and thus in critical need of conservation and restoration (see MacFarlane et al. 2015, Ndang'ang'a 2016).
Arabuko-Sokoke forest was originally declared a Crown Forest in 1932 and gazetted as a forest reserve in 1943. An additional 2 675 ha at Kararacha in the south-east was added in 1968. In 1977, about 4 300 ha within the forest, was declared a Strict Nature Reserve, and in 1979 this was extended by an additional 1 635 ha (SFMP 2002, UNESCO.n.d.). An electric fence was constructed around the forest in 2006 to reduce human-wildlife conflicts including crop raiding by elephants and wildlife attacks on livestock. The land around the forest is densely populated and cultivated by the Mijikenda community (1) distributed among 51 villages. In the 1990s, the population was estimated at 110,000 people with an annual growth rate of approximately 3.1% (Kilifi.go.ke, Gordon and Ayiemba 2003, Mbuvi et al. 2015, SFMP 2002). People who live around the forest are generally very poor and depend primarily on private plot subsistence farming.
METHODS AND DATA
The first author carried out three months of fieldwork in Arabuko-Sokoke during the period November 2016-January 2017. In collaboration with two local research assistants, a mix of qualitative and quantitative methods were applied to gather data on (i) the institutional setup and legal framework influencing forest management, (ii) the livelihood strategies and perceptions of forest adjacent communities, and (iii) the level of human activities inside the forest. To get insights on the institutional setup and associated legal framework, individual qualitative open-ended interviews were carried out with different actors involved with managing the forest, including employees of local NGOs, government officials, and local community members from villages surrounding the forest. Follow-up interviews were carried out with many of the informants to validate and crosscheck information (Table 1). Alongside the qualitative interviews, a total of 85 quantitative, as well as qualitative household-level interviews, were carried out. Of these, 53 respondents were current or previous butterfly farmers (Table 1). The inclusion of nonbutterfly farmers among interviewees allowed for qualitative insights into potential differences regarding forest use and perception among people who did and did not raise butterflies.
The qualitative data stem from semi-structured interviews, group discussions, field notes, and open-ended questions from the household survey. Except in 14 cases (2), all individual and household-level interviews were recorded and transcribed for subsequent analyses through coding. Drawing on 'grounded theory', coding is here understood as the process of "naming segments of data with a label that simultaneously categorises, summarises and accounts for each piece of data." (Charmaz 2014). Open coding was initiated during data collection to push for early identification of categories relevant to the research objective. Subsequently, more focused coding was practised. The first step was to read through all field notes, transcriptions from individual interviews, and results from the questionnaire survey to identify the most salient categories. Frequency tables were developed for many of the questions, (in the case of open questions, subject to coding), in an iterative process of 'data cleaning' (Charmaz 2014). To focus on the observations of interest to the research obective, less relevant categories were left aside (Angelsen et al. 2011:212). Finally, subcategories were developed, and relations between them highlighted. While the coding of the interviews and group discussions resembled pure qualitative analysis, the extraction of codes in the questionnaire survey also integrated quantitative analysis by a description of 'routine comments', whereas the qualitative analysis also, and especially, included a focus on rare individual statements. Routine comments were categorised in different themes such as 'forest activities of butterfly farmers' and 'participation in CFA committee', amongst others. The themes of highest relevance to this study were selected for further analysis. Apart from organising the data, this coding process shaped the analytical frame through which the analysis and theoretical discussion unfold.
Since the questionnaire survey also included quantitative elements, the explanatory data analysis was accompanied by a more descriptive approach. The descriptive part refers to the 'who', 'what', and 'where', while the explanatory analysis facilitated by the coding refers to the 'how', and the 'why,' i.e. the dynamics underlying the descriptive domain (Angelsen et al 2011:210). Shared quantities on the household economy were analysed in Excel with use of simple figures to visualise e.g. distribution as this data was insufficient for more in-depth economic analyses.
To get an impression of the level of illegal logging and hunting activities, which cannot be analysed through satellite imagery (see below), a forest disturbance survey was conducted by transect walks in three different parts of the forest. GPS coordinates were recorded for each cut stem or snare encountered, and details were entered in a standard Forest Survey Sheet provided by KWS, to allow for comparison with previous surveys. This sheet discerns between distinct types of traps and animals intended for these traps and between species of trees cut including the likely usage of the logs (carvings, poles, timber, etc.). Field notes were taken during all transects to provide details not captured in the survey sheet such as noise from logging and encountered charcoal production sites.
Research assistant I facilitated the survey as he had 15 years of experience with such work in the forest. We decided to apply the method he was familiar with, to be able to compare results across years. This method implied the identification of forest adjacent villages to keep track of covered areas. Then, village by village, identify and follow paths into the forest and record observations in the standard Forest Survey Sheet. The idea behind this method is to cover the entire web of human paths by rotating around the forest. Resource and time constraints, however, prevented a complete survey of the forest during the present study. Logging and hunting frequencies indicate the level of human activity in the forest, but they say little about its ecological condition. Given more time and plentiful financial resources, a biodiversity assessment should have complemented the forest disturbance survey.
Research assistant 1 whom local and foreign NGOs, Kenya Forest Service (KFS), and Kenya Wildlife Service (KWS) had repeatedly hired to document timber and animal poaching, shared forest survey sheets with coordinates on cut stems and snares from the years 2008, 2009, 2014, and 2016. In addition to interviews and transect walks, changes in forest cover were investigated through Google Earth images and the forest-cover change facility of Global Forest Watch. However, such remote sensing analyses can only supplement the transect walks since forest cover change, per se, is not a sufficient indicator of environmental degradation. On satellite images, changes in species and size-classes through successive over-harvesting of large size commercially valuable trees can be 'disguised' by natural regeneration that maintains a canopy cover, albeit of increasingly smaller trees (Asner et al. 2005, Ndang'ang'a et al. 2016).
Forest conservation through decentralisation
In Kenya, the Forest Act of 2005 establishes the legal basis for Participatory Forest Management (PFM). In addition, it re-engineered the former Forest Department (FD) into the current Kenya Forest Service (KFS). The main difference between the two is that KFS has more on-site offices and administratively works as a parastatal with some degree of autonomy (Forest Act 2005).
KFS is supposed to facilitate community participation in forest management and collaborative management in general by including all relevant stakeholders in decision-making processes. Community members living within a 5 km range of the forest boundary are officially included as stakeholders (Forest Act 2005). Around Arabuko-Sokoke, the forest adjacent communities are administratively divided into three Community Forest Associations (CFAs); Jilore, Gede, and Sokoke, each of which is associated with a forest station. The official idea behind the establishment of CFAs is to use them as a way of including community members, living adjacent to the forest, in forest management decisions and thereby create awareness and incentives for forest protection. Subject to agreements with KFS, groups from forest communities who register with a CFA can engage in one or several of the conferred user rights listed in the 2005 Forest Act (box 1). Further, KFS holds the authority to operationalise user rights at local levels. In Arabuko-Sokoke, firewood collection, beekeeping, and butterfly farming, constitute the KFS-defined main use categories. Officially, pole wood collection is only allowed in cases of authorized house construction and reperation, under close supervision from KFS.
Neither the Forest Act (2005) nor the PFM guidelines (MENWR 2007) provide any details on modes of community participation, power distribution, and cost-benefit sharing mechanisms. The CFAs around Arabuko-Sokoke have no formal decision-making powers over the forest, and none of them have official forest revenue sharing agreements with the KFS. Accordingly, the CFAs are based on volunteer work, and they do not generate significant or stable incomes despite their legal responsibility to "protect, conserve, and manage" PFM forests and generally assist the KFS with monitoring and reporting (Forest Act 2005 Part IV, section 47).
The user rights and opportunity to participate in forest management through CFAs is formally enabled through the Forest Act (2005) but realised in practice through semi-legal frameworks at the local level: the CFA management plans, CFA management agreements, and CFA constitutions. The collaborative management approach is operationalised through the establishment of Arabuko-Sokoke Forest Management Team (ASFMT). ASFMT is intended to support collaboration between the four government institutions, Kenya Forest Service (KFS), Kenya Wildlife Service (KWS), Kenya Forestry Research Institute (KEFRI), and the National Museums of Kenya (NMK), together with relevant NGOs and the CFAs. However, by February 2017, management agreements, operational management plans, and constitutions were still in the process of being finalised. Moreover, the ASFMT agreement had not been reviewed since its establishment in 2002, i.e. before KFS and the three CFAs came into existence. According to many of the stakeholders and community members interviewed, the local KFS office had, in agreement with some of the traditional local elites (see below), attempted to postpone CFA elections, management agreements, constitutions etc. since the formation of CFAs and with some degree of success. During the fieldwork, two of the three CFAs, which were all formally established in 2007, had never conducted elections although this should happen every third year (Forest Act 2005).
The formal institutional setup is structured as shown in Figure 1. Nature Kenya was the only active NGO in Arabuko-Sokoke during the fieldwork. This setup implies a lot of uncertainties since the roles and responsibilities of the actors in the management team are not described anywhere in detail in any detail. In practice, the local KFS office determines the level of inclusion in decision-making, and it has unilateral control over official forest revenues. The power of KWS is limited to participation in monitoring and law enforcement while KEFRI and NMK primarily engage in monitoring and research activities. Most of the income-generating activities that are promoted as conservation strategies in the periphery of Arabuko-Sokoke forest are owned by one of these four government institutions. The butterfly farming initiative is organised under The Kipepeo Project managed and owned by a branch of NMK.
ASFADA is the Arabuko-Sokoke Forest Adjacent Dwellers' Association. Officially it represents the forest communities, but unlike the CFAs it is independent and not legally subordinated KFS.
Generally, the legal framework and institutional setup is very unclear. Most of the legal documents supposed to guide the collaborative management and community participation are either not finalised or under official 'review'. As a result, KFS is maintained as the main authority, unilaterally governing official forest use and revenue streams.
Mistrust and a lot of disappointment rotated within and between non-KFS actors. During interviews at village-level and with representatives from other institutions, it transpired that KFS is reluctant to share information on poaching reported by CFAs, which they are formally supposed to forward to the other PFM institutions. According to a key informant from ASFADA, this institutional environment constrained some and opened opportunities for others. He argued that the local 'royal family' from Sokoke CFA, which allegedly dominates the office positions, has an informal agreement with KFS on the flow of products and revenues from 'their' part of the forest, i.e. including organised illegal activities such as the carving industry, timber harvesting, and charcoal production. The reason, according to him, was that the community members piloting PFM had intensified their illegal use of the forest because their high expectations of benefiting from PFM had not been met; as he put it "they decided to benefit in another way". Following these statements, it seems that, in an attempt to regulate de facto harvesting levels (and unofficially tax this activity), KFS had struck a deal with the 'royal family'. Other informants shared the perspective of a link between disappointment over PFM and illegal activities. A key informant from KWS similarly pointed out that many of the people who are poaching today are "... community members who had high expectations to PFM and are now left with a feeling that they did not get what they were promised". The KFS forester from Gede likewise noted that "the communities feel like they have been cheated". The expectation to benefit from the forest also transpired during the household survey. As expressed by one of the community members recently engaged in a user group, "there is no benefit from the forest, it is protected. But we expect to get access to the resources; when we protect the forest, we expect to get permission to harvest materials - charcoal, firewood, poles, timber". This quote is very saying for the general confusion and disappointment among community members on the expected benefits of participating in management and conservation activities.
Staff from Nature Kenya and key informants from ASFADA believed that the absence of immediate benefits from participation had led to informal agreements between CFA executives and local KFS representatives. They emphasised that many previous and current CFA and ASFADA executives had informal, practical economic arrangements with KFS, including as the opportunity to get loans. According to the key informant from ASFADA such arrangements explained why many CFA and ASFADA executives were reluctant to reveal links between KFS and illegal activities in the forest; as he put it "even if you are a good leader, you cannot fight the hand, which is feeding you". Similar informal arrangements seem to surround community members employed as scouts assisting forest patrols. Members from the CFAs and Nature Kenya argued that because community scouts are not getting paid for their work, they show low compliance or make informal economic arrangements with KFS and/or accept bribes by poachers for not reporting illegal activities.
During data collection, people frequently referred to KFS as a corrupt and sometimes violent organisation i.e. in cases where a community member tried to go against or report on a KFS forest guard doing illegal activities. Informants from KFS expressed great frustration about what they argued were unsubstantiated rumours. Although the rumours are unlikely to reflect the reality in full, it is clear that KFS employees could get away with such practices under the, de facto, institutional arrangement. After completion of the fieldwork an expose, through ASFMT, of illegal forest activities including charcoal production resulted in the replacement of KFS staffs as well as death threats to people behind the expose. However, according to a well-informed local source, the fresh KFS staffs surprisingly quickly got recruited into the 'charcoal cartel'. This suggests that (i) KFS staffs do indeed benefit from illegal extraction of forest products; (ii) KFS is in cahoots with but hardly in control of well-established networks, which trade illegally sourced forest products; and (iii) such hidden institutional arrangements are, nevertheless, challenged through the complicated institutional setup, albeit thus with limited effect.
Overall, the unclear legal framework and institutional setup have resulted in unclear and contested distributions of powers and responsibilities vis a vis the forest resources including the extraction of forest products and the handling of forest revenues. According to the majority of community members and the local NGO, this is related to a poor representation of forest communities in the management team. With the 'royal family' as the most prominent example, the CFA members who are involved in management decisions tend to represent local elites. In this example elites who, de facto, albeit in collaboration with official forest authorities governed access to 'their' part of the forest before the introduction of PFM. The continuous postponement of CFA elections seems to have consolidated and with an air of legitimacy officialised these elite representatives' shared authority (together with KFS) over forest resources and revenues as it has allowed them to maintain their CFA office positions unchallenged (see also below).
Forest conservation through income-generating activities
In Arabuko-Sokoke, butterfly farming as a PFM strategy functions under the Kipepeo Project, which is administered by the National Museums of Kenya and was piloted in 1993 by Nature Kenya under a Global Environment Facility Small Grant of USD 50 000. The rationale of butterfly farming is to link participation, improved livelihoods, and forest conservation. By generating alternative but forest-dependent income, the local attitude towards conservation is expected to improve (Gordon and Aiemba 2003, kipepeo.org). In addition, the general assumption behind IGAs as a forest conservation strategy is that they will deflect people's assumed unsustainable forest use by occupying their time with profitable homestead activities. Most of the IGAs found in Arabuko-Sokoke are owned by government institutions. The National Museums of Kenya owns the butterfly farming, beekeeping, and some of the tree nurseries through the Kipepeo Project. Kenya Forest Service manages some IGAs dealing with charcoal production from forest plantations, though not many details were available, and Kenya Forestry Research Institute runs some tree nurseries. Community-owned IGAs seemed limited to minor eco-tourism activities.
The butterfly farming activity is considered sustainable because it depends on the forest without requiring much extraction of forest product; it is officially limited to butterfly larvae (3) and fodder plants. Once a forest user group is formed, it can register under the respective CFA and apply to KFS for individual member permits to collect butterflies in the forest. What this implies in terms of what they can collect (adult butterflies, pupae, larvae, fodder), how often, when during the year, and through which methods is unspecified. In theory, each user group has a representative, an assistant representative, a chairman, a vice chairman, a secretary, a vice secretary, and a treasurer. The standard breeding procedure is to catch females of the desired adult species in the forest with a net and put them in a breeding cage at home. In the breeding cages, the butterflies lay their eggs, which turn into larvae that feed on different plants until they reach the state of pupae. A representative regularly collects the pupae for export. Butterfly farmers who can afford large breeding cages and irrigate fodder plants do not frequently go to the forest because they can sustain a breeding population at their homestead. Poorer butterfly farmers tend to collect larvae instead of butterflies from the forest and feed them in plastic containers until they reach the state of pupae. The community members collect and breed butterflies individually, but they rely on the representative to deliver their pupae to the management office for weekly shipments to overseas markets, mainly the U.K. Today, 350 community members are listed in the butterfly farming project (Butterfly farmers list 2016). However, the number of user groups is decreasing, and many of the farmers do not have the necessary equipment and access to irrigation to maintain a year-round production. As a result, the Kipepeo project manager had recruited producers from outside Arabuko-Sokoke to sustain quality production while searching for additional export markets.
Responses in forest adjacent communities to decentralisation and income generating activities
PFM was piloted in Arabuko-Sokoke in three villages on the western side of the forest: Kahingoni, Dida, and Kafitsoni, which are all located in Sokoke CFA. Key informants from the community associations and Nature Kenya argued that, prior to the piloting of PFM, the same area had for generations been under informal leadership by a 'royal family', i.e. a local elite, that according to key informants through collusion with KFS representatives share the, de facto, control of parts of the western side of the forest. The informants also argued that this family made up the core of Dida Forest-Adjacent Area Forest Association (DIFAAFA), a community association representing the area where PFM was piloted. It was through DIFAAFA that community members were invited to develop Arabuko-Sokoke Forest Strategic Management Plan (SFMP 2002), guiding the further implementation of PFM in Arabuko-Sokoke. DIFAAFA is no longer vibrant, but according to the key informants, the 'royal family' still dominates the area including Sokoke CFA office bearer positions. Some of the households interviewed explicitly emphasised that the CFAs were associated with an elite, e.g. when asked if they participated in the CFA meetings and activities one of them answered "I am not invited, they only invite those who know each other, those from the royal family". The postponement of CFA elections, unfinished constitutions, and missing signatures in management agreements, has reproduced and consolidated local hierarchies and how (if) they represent the local communities remain highly unclear. Since their establishment, office positions in the CFAs are to a very large extent held by the same group of people, cf. above, which emerged as one of the main barriers to broad participation in the CFAs.
The absence of formal benefits from engaging in CFAs due to the lack of income sources and benefit sharing mechanisms with KFS constitutes another barrier. Interviews and investigations of the, de facto, institutional structures, revealed that the benefits obtainable from participating in the CFAs were either indirect through building social and political capital or more direct and immediate through informal personal arrangements with KFS. People could receive money for assisting KFS with unofficial assignments and be the first to know about upcoming forest use restrictions. For example, in January 2016 KFS unilaterally decided to put a temporal ban on firewood collection. The community executives were informed about the decision before its official enforcement allowing them to stock up firewood. 'Ordinary' community members argued they received no warnings. These and similar events were, according to key informants, part of the explanation why KFS and some of the CFA executives were reluctant to conduct the legally stipulated CFA elections as this might disturb these informal reciprocal relationships.
Participation in income-generating activities in the forest periphery is another way to be officially included in forest management since it implies the establishment of user groups that register under one of the CFAs. Unsurprisingly, however, the household survey among butterfly farmers and individual interviews with other stakeholders revealed that the potential benefits increase with production advancement which requires equipment and economic capital, all of which the poorer households are unlikely to have. Moreover, advancing from participation in butterfly farming to becoming active in the CFAs requires time and transport opportunities. Combined with the absence of formal benefits from CFA participation, this constitutes another barrier to the poorer households. As shown in the figures below, the income trend for butterfly farming is reverse L-shaped, i.e. a few earn a lot while the majority earn very little, and the household survey indicated that these few cases were often the representatives of the user groups or holding other office positions. A similar pattern characterised an income distribution investigation of butterfly farmers under the Kipepeo Project in 1999 (Gordon and Ayiemba 2003).
Effect on forest conservation
The patterns emerging from this fieldwork and other research on the forest condition and causes of forest degradation the Arabuko-Sokoke forest are somewhat bewildering. The biodiversity of Arabuko-Sokoke forest is high, but the forest has not been systematically monitored or documented which makes it difficult to establish a historical overview. Satellite images show that the forest cover has not changed significantly over the past 15 years (Globalforestwatch, n.d.) and an ocular interpretation of Google Earth images indicate an intact forest cover throughout the period 1975-2016 (Map 2.).
Available reports on the ecology of Arabuko-Sokoke provide evidence of degradation trends for flora as well as fauna, but they are indecisive on the magnitude, significance, and causes (Matiku 2002, Mbuvi 2007, Mbuvi 2015). Members of on-site institutions and community associations interviewed during fieldwork generally shared the view that forest resources in Arabuko-Sokoke were degrading. Likewise, 83% of the community members interviewed around the forest noted a decrease of big trees in the forest. A CFA chairman emphasised that "In ten or twenty years when we talk of Arabuko-Sokoke, we will be talking about bushes and not a forest". Similar statements came from people in KWS and KEFRI, who pointed out that hunting of animals and logging of big trees in combination with overharvesting of small size trees for firewood were the main threats to the forest ecology. The National synthesis from Nature Kenya and Birdlife International arrived at similar conclusions (Matiku et al. 2002, Birdlife International, n.d. b). Existing records of cut stems and snares in the forest (see below) might qualify or challenge the perceptions of forest and wildlife degradation, but these have never been officially processed or made publicly available.
Just a few meters from the fence as well as deeper inside the forest, logging sites, charcoal burning sites, and snares were easily and frequently identified during transect walks. The aggregate picture established from coordinates of recorded snares and cut stems for the years 2008, 2009, 2014, 2016, and 2017, suggests that these illegal activities were and still are common in the forest (Map 3.). Furthermore, the survey and GIS analyses indicate that no areas in the forest are left undisturbed, i.e. all zones appear affected by consumptive and formally illegal use (logging/cutting of big and small trees as well as snaring).
Google Earth analyses were used to supplement the spatial impression of Map 3 and produce Figures 4 and 5, which depict the number of incidents per 100 m and the average distance between logging and snaring incidents, respectively. Lines were drawn between coordinates to structure the transect paths and measure the distance between incidents. After measuring all paths for the years 2008, 2009, 2014, 2016, and 2017, the average distance between incidents was calculated for each path. Additionally, the average amount of incidents per 100m was calculated to support the maps and the analysis of average distance between incidents. These analyses should, however, be treated with caution because the methods of data collection behind the secondary data from previous survey years are not documented. Accordingly, we cannot know to which degree the results arise from differences in the intensity and mode of data collection over the years. This stated, the data suggest that the average distance between incidents varies across years, but not by any obvious trend or pattern. Hence, 2017 seems to represent 'business as usual' in comparison to previous years. However, it appears that when the number of cut stems is high for any given year, the number of snares is low and vice versa. As pointed out by KWS staff, a possible explanation is that areas with a lot of logging activity are not very attractive for animal poaching because of the noise and higher frequency of human activity. Figures 4 and 5 also show that cut stems occur in a more clustered pattern, which makes sense because this maximises the number of trees cut per hour and thus minimises the potential financial loss per cubic meter associated with getting caught red-handed. By contrast, snares are set further apart at lower frequencies, presumably because effective animal poaching requires a larger area than trees. Considering the forest is fenced and officially guarded against cutting of live trees as well as hunting, the occurrence of incidences presented in figure 4 is in any case rather high. The data on cut stems and snares are not separated into PFM and non-PFM zones. In sum, however, such activities are widespread and appear to have been so throughout the years, before as well as after the formal implementation of PFM.
In practice, the lines between legal and illegal forest use and user categories are blurry and complex. Legal product collection requires a permit from KFS. In Arabuko-Sokoke, the permits include; (i) firewood collection (most common), (ii) butterfly collection (less common), and (iii) Eco-tourism (least common). All three activities are only legal if they are carried out within the permitted volume or frequency, e.g. only one headload of firewood per person, two times a week. Unfortunately, comprehensive data on the characteristics of forest users, actual forest uses, and what happens to forest products once extracted do not exist. However, impressions from the fieldwork indicated that the poorest are rarely directly involved in extracting, trading, or consuming the most valuable forest products. The local poor might extract valuable products, which are then picked up by middlemen who organise transportation to urban centres like Mombasa, Malindi, or Kilifi. The demand for various timber products in these commercial centres is high, and, e.g. carvings can be sold to tourists or transported for further distribution abroad. Some of the community members argued that KFS was involved in the illicit timber business with outside dealers while at the same time having an independent business on the side with products confiscated from community members, e.g. by selling confiscated poles to schools in the area.
Fines, prison, and KWS rules came up as routine comments to questions about hunting and consumption of bushmeat. Few community members were willing to share information on bushmeat hunting, whether their activities or knowledge about others. The answers from people who did share information varied greatly. Some argued that hunting was difficult, getting lucky just once or twice per month, and sometimes impossible. Different arguments against hunting came up during interviews including a decreasing wildlife population, that KFS outposts were too close, and the Church's rule against the consumption of forest products for food. Other households were more optimistic and argued that it was possible to bag one--five small, antelopes and/or wild pigs per week, depending on luck. Some even argued that it was possible to get three antelopes per day.
Burning charcoal is an illegal but widespread activity in and around Arabuko-Sokoke forest. The many selling points along forest near roads document that charcoal is a commercial product. Hundreds of bags of charcoal were observed within minutes of driving along the forest road. The household survey and individual interviews suggested that locally, charcoal burning is widespread albeit not a big but rather a supplementary income-generating activity. Transport of charcoal to commercial centres, on the other hand, was as a more serious income earner, which supports the alleged existence of charcoal cartels that buy from many small-scale rural producers and control down-stream trade towards urban end-users (c.f. above).
Key informants confirmed that some of the hardwood timber species in the forest are used for carvings in the tourist industry. According to community members, the carving industry is a well-hidden activity, done deep inside the forest to avoid KFS guards. Timber for furniture is also extracted, in this case first to carpentries. Intermediaries facilitate the business with whom private people or enterprises can place orders. Several of the households interviewed (48%) expressed that the timber business is impossible or difficult to access because of barriers such as a dwindling number of suitable trees, protection by forest guards, and high costs of bribing if caught red-handed. Some pointed out that it was a matter of negotiation and logging of timber could only be done if rent was paid to KFS forest guards.
Extraction of building poles is common and less sensitive, although this activity is normally not legal. Officially, permits to harvest building materials are granted on a need basis, e.g. to construct a new or repair an existing house. KFS claims to strictly control that the materials are used in accordance with the extraction permits as the rules stipulate, but many households indicated that it was a matter of negotiation. As one respondent summed it up: "Those who have money can talk to the forest guards".
Key informants pointed out that legal forest activities were frequently used to disguise illegal activities. For example, women who collect firewood often team up with their husbands, brothers, or sons doing logging. That way, the women can add logging residues to their firewood collection while taking turns in keeping an eye out for forest guards. The same informants argued that having someone at the forest fence to keep an eye out is a general strategy in illegal forest activities. Firewood and butterfly collection are legal activities that can be accompanied by illegal forest use when larger amounts than the permitted volumes of fuelwood are extracted and potentially commercialised, or if the collection of butterflies involves scouting for, e.g. valuable timber trees. Only 17% of the butterfly farmers believed that their forest use had changed since they joined the Kipepeo Project. Half of these argued a decrease in their forest activities while the other half argued an increase because they did not use the forest before joining the Kipepeo Project.
Accordingly, local people's relations to the forest resources are a good deal more complex than what the Kipepeo butterfly farming and similar kinds of IGA projects can directly address. Butterfly farming has the potential to generate high cash income for a few households while adding a small supplementing cash income to the majority (cf. Figure 2). However, the household survey indicated that the forest adjacent villagers are dependent on firewood from the forest for energy consumption when their farmland could not provide, while other forest products supplemented other sources of household income. Even when butterfly farming gave a great return, farmers continued to collect forest products. In short, fieldwork evidence suggests that increased household incomes do not reduce people's forest use.
Key informants argued that butterfly farmers have the potential to rival even the worst, i.e. most skilful poachers since they obtain a very thorough knowledge of the forest from going deeper than most people, to find valuable butterfly species. One of the community members argued that the butterfly farmers were the only ones who could still access valuable products in the forest. Research Assistant no. I pointed out that a consequence of (poor) butterfly farmers failing to earn enough through that activity was likely to be that they pretend to continue while in fact becoming even more effective timber and wildlife poachers than those currently engaged in such activities. The informant from KWS indicated a similar concern but added that the same risk applied to the other CFA members as well. She argued that the user groups, whether butterfly farmers or firewood collectors are moving along the same paths were cut stems, and snares are found, without reporting the incidents. Research assistant no. II agreed that local people were either involved in or agreed to keep quiet about illegal forest uses. In short, the official promotion of butterfly farming as a community-based forest conservation activity and income-generating activity under the management of Kipepeo and a supplementary element of PFM is nested in a political-economic reality where people do what they can get away with to benefit as much as possible from the forest resource. Hence, the intended engagement of butterfly farmers in general forest conservation has not materialised.
DISCUSSION AND CONCLUSIONS
Two different but overlapping strategies to conserve Arabuko-Sokoke are promoted under the concept of PFM; (i) decentralised forest management through structured involvement of forest adjacent communities in forest governance decisions, activities, and sharing of potential revenues and (ii) community-based 'income-generating activities' (butterfly farming) intended to combine poverty alleviation with legal and minimum impact forest use in an effort to promote a positive attitude towards forest conservation among local people. It was considered a bonus if this could serve to deflect local people's unsustainable forest uses. The sections below discuss the practical manifestations and outcomes of these concepts vis a vis their official and theory-based intentions.
Conditions for collective choice
Conducive conditions for collective choice are a prerequisite for realising the objectives of conservation and improved livelihoods through participation in forest management (e.g. Oakerson 1992, Agrawal 2003, Ribot 2004, Hobley 1996, Bruce 1999, Lindsay 2004, Ostrom 1998). The operational rules of the legal framework for participatory forest management form an essential part of these conditions. In Arabuko-Sokoke, the operational rules for forest management are problematic because they are top-down defined and sometimes unilaterally modified by KFS. This excludes forest adjacent communities from legally accessing, e.g. poles and charcoal that are of obvious importance to their livelihoods, and it excludes them from participating in modifying operational rules, e.g. to balance conservation and livelihood objectives of forest governance. Furthermore, the combination of a broadly phrased legislation and the ability of the local KFS office to constantly postpone local-level elections and finalisation of benefit sharing arrangements allow it to hijack rule-making processes.
Following Oakerson (1992), conditions for collective choice on the governance of Arabuko-Sokoke forest and associated revenue streams are absent. In reality, KFS has, partly in collusion with local elites, who have vested forest interests dating back to pre-PFM times, managed to monopolise the decision-making authority. Accordingly, the local KFS office upholds a, de facto, centralised governance regime where it can control official rule-making processes and maintain access to official and unofficial revenue streams. At the organisational level, this happens through local elites who control access to markets for illegally sourced forest products. The aggregated picture indicates that in exchange for sharing revenue streams with KFS staffs, members of these elites obtain and maintain influential positions in the PFM institutional setup. At lower levels, forest guards benefit directly through accepting bribes when detecting activities that KFS has decided to criminalise such as pole cutting for local construction purposes.
Bureaucratic resistance to forest decentralisation is not unique although it can take many forms. In Senegal, Ribot (2004) documented a similar case of how the Forest Department ignores new legislation that transfers powers over forest resources from the central to local governments (ibid). In Nepal, forest bureaucracies re-centralise powers over decentralised forest resources through legal-sounding rhetoric and complex technical rules that decentralised forest management plans must live up to (Basnyat and Treue et al. 2018, Nightingale 2005, Ojha 2006, Agrawal and Ribot 1999).
The official attempt to decentralise forest management, in Arabuko-Sokoke and Kenya in general, has deconcentrated power to a government parastatal, KFS, which has few if any incentives to share these powers with community-level institutions, the CFAs. Accordingly, the missing ingredient to make forest decentralisation work for CFAs around Arabuko-Sokoke is a mechanism that makes KFS downwardly accountable to local people as well as to the law. Only topdown political pressure on KFS at the national level can make this happen, e.g. through deadlines for holding CFA elections, signing of benefit sharing agreements, and credible legal prosecution including conviction and punishment of KFS staffs found guilty (c.f. Agrawal and Ribot 1999, Ribot 2004). Whether the latter is possible in practice is questionable, especially if official salary-levels and organisational norms of KFS (and possibly other branches of Kenya's bureaucracy) breeds a practice according to which the employees are 'expected' to generate unofficial incomes through their official positions.
No evidence suggests that community members, through CFAs, are included in important decisions. Important decision-making power understood here as the power to influence rules governing forest use (Oakerson 1992). To characterise the level of decentralisation Agrawal and Ribot (1999: 6) emphasise the importance of the power to "... make decisions about how a particular resource or opportunity is to be used". The allocation of responsibility without clearly defined rights and decision-making power also goes against Ostrom's (2008) third design principle for collective choice arrangements, that; "Most individuals affected by harvesting and protection rules are included in the group that can modify these rules." (ibid: 33). Considering other PFM cases from Kenya, the lacking participation and influence of local communities on forest governance is not uncommon, e.g. Mutune and Lund (2016: 48) found that the Forest Act enables CFAs to assist KFS, but that KFS sets the agenda. Using decentralisation to get free labour is a sign of poor local governance and at the same time a disputed, but not uncommon, incentive for decentralisation in Kenya, and other African countries (e.g. Pfliegner 2014, Appiah 2001, Brockington 2005, Ellis and Mdoe 2003, Topp-Jorgensen et al 2005, Poliout and Treue 2013, Carter and Gronow 2005). Reluctance to devolve control over valuable forest products such as timber has likewise been found to be a common feature (Thygesen et al 2016, Carter and Gronow 2005).
Firewood collection and alternative income-generating activities like butterfly farming are local communities' legal options to benefit from forest resources (MENR 2007, MENWR 2014). However, circumstances that justify KFS' withdrawal or modification of user rights are not specified. KFS has not interfered with butterfly farming as it has with firewood collection, but its unilateral decision-making authority appears to cause great insecurity for the local people around the Arabuko-Sokoke forest. Following Lindsay (2004) and considering the legal vacuum in which CFAs exist, this state of affairs poses another challenge to community participation. In the absence of legal security and exclusion from meaningful decision-making, community members might decide not to participate at all, or bypass official rules and seek security and access to benefits through informal arrangements much like what KFS seems to be doing at a higher level by colluding with local elites to maintain access to forest-based revenue flows (see, e.g. Ribot and Peluso 2003).
The mode and benefits of public participation also link to accountability relations. Downward accountability to local communities requires mechanisms for representation. An essential mechanism is democratic elections, which was surrounded by delays and conflicts in Arabuko-Sokoke. Without regular democratic elections, the CFAs cannot be expected to represent their constituencies or be downwardly accountable (c.f. Ribot 2004). Downward accountability is generally absent among the institutions officially implementing PFM in Arabuko-Sokoke forest. The NGOs are mostly accountable to their donor(s), the butterfly farming business depends on the market, and the KFS deliberately resists becoming downwardly accountable to CFAs.
Benefit-sharing and conflict resolution mechanisms are among the principles highlighted by scholars to be important for effective community associations (Ostrom 2008, Lindsay 2004, Ribot 2004). The CFAs in Arabuko-Sokoke are, however, dependent on government institutions and NGOs for financial and administrative resources. Accordingly, the CFAs' opportunity to work independently and enjoy the benefits of participation is greatly influenced by the fact that potential but unpredictable income is expected to cover actual and predictable expenses (Forest Act 2005, CFA PFMP, draft; CFA Constitution, draft). The community members who engaged in CFAs are still formally dependent on KFS for resources, equipment, and technical expertise to develop management plans.
People's level of participation depends on the perceived benefits of doing so as well as their ability to access certain activities and associated benefits. Lack of capital often defines barriers to access benefits. Anyone can do butterfly farming, but to make it a profitable undertaking requires capital and irrigation which to some extent is accessed through structural and relational mechanisms (c.f. Ribot and Peluso 2003). A study of ICDPs in Tanzania with butterfly farming as a case, also found that lack of access to capital, knowledge, and authorities reduced the potential to advance (Ndumbaro and Mutula 2017). In general, poor community members are at a disadvantage compared to the more wealthy who can use their capitals to negotiate more privileged access to key resources (c.f. Ribot and Peluso 2003:166). Statements from interviews with implementing institutions and community members strongly suggested that the successful butterfly farmers around Arabuko-Sokoke with access to VDFCC and CFA committees were people with resources to travel to and use time on meetings and other activities. As noted by Pfliegner (2014) most cases in Africa, India and Nepal show that "... transaction cost (i.e. attending meetings, forest monitoring, and patrolling), relative to benefits are higher for poorer households compared to medium-income and richer households". However, local people are generally denied access to the most valuable forest products like timber, and in the absence of legal benefits including rights to a share of revenue flows, illicit and informal negotiation-driven access mechanisms make a mockery of the officially intended incentive structure.
The slow pace of establishing conditions for collective choice and community participation may partly be attributed to the fact that these changes are implemented in a postcolonial centralised governance of natural resources. Thus, centralised organisations like the KFS are prone to pursue PFM implementation strategies that promote "dependency of communities on KFS rather than reciprocity between communities and KFS" which also nurtures a view of communities as "recipients of aid rather than partners in development." (Newmark and Hough 2000:589). Several scholars have argued that economic incentives are necessary to motivate and sustain community participation in forest management (Pfliegner 2014, Musyoki et al. 2016, Ravindranath et al. 2004, Topp-Jorgensen et al. 2005). However, observed participation without apparent material or cash reward is generally explained through gains in social capital (see, e.g. Gibson et al. 2003). In the case of Arabuko-Sokoke the two, i.e. economic and social capital, seem to reinforce each other.
Hobley's (1996) concept of 'hidden institutions' offers a convincing way to explain the dynamic influence of informal arrangements as a common but complex institutionalised element of decentralised forest governance. In Arabuko-Sokoke, local people have established hidden institutions either by bending user rights independently or by establishing informal agreements with KFS representatives. Rather than ensuring participation, the permit system seems intended to uphold appearances of conservation through regulated access while their real function is to criminalise certain (most) forest uses because this forces community members to invest in relations with KFS to gain, maintain and improve access (c.f. Ribot and Peluso 2003:154).
Lack of transparency and reluctance to share information added to the impression that implementing institutions want to hide certain things. Notably, the absence of peer-reviewed research on actual forest use patterns and thus drivers of degradation constitutes an essential knowledge gap. The little knowledge informants shared came through ambiguous statements by the government institutions, which sometimes argued that poor people illegally encroached the forest and sometimes denied that illegal activities in the forest were a problem at all. The absence of evidence on the forest's conservation status and ambiguous arguments on whether there are reasons for concern were also apparent in the KEFRI research reports. Hence, the extent, severity or even reality of forest degradation, underlying cause-effect relations, and thus what to do about it is kept open to endless speculation and contestation.
That KFS has gained control and authority through the temporary and uncoordinated state of PFM affairs probably explains its unwillingness to change these conditions, especially if KFS staffs also benefit financially from the unofficial harvest of forest products. The frequency of cut stems and snares in the forest indicates a weak and/or bribable rule enforcement system. By maintaining the current, de facto, centralised forest governance system, KFS upholds full control and thus maintains its ability hide its alleged involvement in illegal activities, i.e. the existence of hidden institutional arrangements (c.f. Wardell and Lund 2006: 1900). This stated, corruption is not necessarily equivalent to unsustainable management of forest resources, at least not in theory. Only reliable time series of biophysical data can generate evidence-based conclusions about the, de facto, governance regime's conservation outcomes. However, such data is (conveniently) absent.
Hidden institutions blur the picture of, de jure, vs, de facto, governance and use patterns. The butterfly farming activity exemplifies the mechanisms through which hidden institutions can facilitate illegalities behind a veil of legality. Accordingly, the actual forest governance and resulting conservation, as well as some distributive economic, outcomes are mainly shaped by hidden institutions through which KFS representatives, quite contrary to the official objectives of PFM, upholds a high degree of control over who gains and maintains access to forest products and associated revenue streams.
These findings strongly indicate that decentralisation programmes need to explicitly focus on hidden institutions as they probably explain why so many forest decentralisation reforms generate questionable biophysical as well as social outcomes. Hidden institutions are not straightforward to map out, and fully replacing them with overt, transparent, and downwardly accountable institutions is hardly realistic. It was beyond the scope of this study to investigate the size of unofficial incomes that KFS employees appear to generate from Arabuko-Sokoke, how such revenues might be (upwardly) shared within KFS, and whether KFS employees' official salaries are reasonable and adequate in comparison to, e.g. the private sector. Internal investigations by KFS can, however, shed light over such questions. If KFS staffs are notoriously underpaid and a system according to which shares of unofficial forest revenues collected at forest-level trickle upwards in the hierarchy, then fundamental changes within KFS are needed to make any forest decentralisation programme deliver on its official objectives. If, on the other hand, KFS staffs do not desperately need to collect unofficial revenues and if the central state is willing and able to confront and control business networks that trade in illegally sourced forest products, then anti-corruption efforts might reduce the hidden institution to a manageable size while strengthening the CFAs through deliberate top-down interventions.
Potential conservation outcomes
In the case of Arabuko-Sokoke, the dual PFM objective of improved livelihoods and conservation through activities like butterfly farming does not (and arguably did not attempt to) address the fact that poor forest adjacent community members' extraction of forest products might not be the biggest threat to forest conservation. Barret and Arcese (1998) found that many integrated conservation and development projects fail to realise that those who participate (on paper) are not necessarily the people whose activities threaten conservation. Similar uncertainties surround Arabuko-Sokoke.
We find it is unlikely that PFM has reduced timber and animal poaching in the forest. It was not possible to conclude whether the forest condition had improved, worsened, or remained stable but even if conservation status data had been available, it would be difficult to separate the effects of PFM from other factors including the electrical fence or national-level political changes such as the Constitution of 2010.
From a rural participation and poverty alleviation perspective, a continuation of the current hidden institutional set-up is not ideal. Firstly, there is no evidence that the people who benefit from the human activities in the forest include the target group for PFM, i.e. the poor rural forest adjacent communities. Secondly, if the level of human encroachment and illegal extraction of forest products continues, then the local KFS authority can feel compelled to or strategically use such circumstances to withdraw all local user rights. From a conservation perspective, the status quo is not ideal either since a continued extraction of, e.g. big trees and bushmeat could deplete certain species. However, PFM has established a forum (ASFMT) where local communities and non-KFS employees have voiced their grievances and if not dismantled the hidden charcoal production institution, then managed to expose and shake it.
Butterfly farming can generate forest-based alternative income for local families, but no evidence suggests that it reduces logging, charcoaling, or animal poaching; it might as well have increased such activities. If non-destructive forest-based enterprises like butterfly farming are to be sustained, they need diversification strategies to be able to cope during periods when local producers fail to deliver or when the demand for certain species decreases. The project has tried to meet these needs by promoting other IGAs such as honey production, and by opening up for marketing of products from butterfly farmers outside of Arabuko-Sokoke. These findings resonate with Seppala (1998), who stated that economic diversification by forest product collection is done by the poor to survive, by the middle class to reduce risks, and by the wealthy households to gain profit from, e.g. niche markets.
In a world that "... insists that forest lands be managed for a multiplicity of benefits." (Hobely 1996: 13), trade-offs between objectives are inevitable. Studies of community-based and participatory forest management programmes and projects report both success and failure regarding their ability to reduce deforestation, support sustainable forest use, promote forest protection, and generally deliver positive conservation outcomes. Linking causes to effects in the study of PFM outcomes also remains a challenge (Rasolofoson et al. 2015, Cox et al. 2010, Pagdee et al. 2006, Porter-Bolland et al. 2012, Pfliegner 2014, Agrawal 2003, OECD.org). Nevertheless, promotion of conservation through PFM and other integrated approaches that 'promise' social and economic rewards have become a dominant rationale while recognising nature's intrinsic value is increasingly labelled as bio-centric and associated with unpopular top-down and state-centred approaches to nature conservation. However, a key argument for conserving nature for its own sake is that many conservation-cum-development projects base (growing) human needs on (not growing) natural resources (Barret and Arcese 1995:1077).
This argument holds in several ways for Arabuko-Sokoke forest. First, the actual forest governance that hidden institutions' turn PFM into does not deliver many official and tangible economic benefits to the majority of local people who cannot influence rules on harvesting and costs/benefit sharing arrangements. Second, the permit system for forest product collection ineffectively regulates harvesting levels but effectively establishes rent-seeking opportunities for KFS employees. Third, the forest area is not increasing, and its productive capacity is hardly increasing either.
Currently, narratives of harmony between people and nature appear successful ways of attracting donor support although they frequently fail on-the-ground, where political and institutional resistance, as well as economic realities, pose serious challenges. What is 'good' for nature conservation is not necessarily socio-economically beneficial for people who live close to the concerned natural resource, and what would be economically beneficial for local communities would often be a loss to forest bureaucrats and established networks with vested interests in trading illegally sourced forest products. Hence, national or international funding mechanisms are needed to establish positive financial incentives for local people as well as representatives of government agencies to conserve natural resources that cannot 'pay for themselves'. To work as intended, such payments for nature conservation require alterations of authority structures that fundamentally change the position of, e.g. KFS and networks in Kenya that trade illegal forest products. Such new structures must also offer forest bureaucrats reasonable and legitimate sources of income while efficiently and effectively punishing rent-seeking activities and trade in illegally sourced forest products. In most so-called developed countries effective forest and nature conservation work within such privately or publicly subsidised fiscal realities. However, this approach is normally considered out of the question in most developing countries where the doings of poor peasants, poorly paid public employees, and ingenious private enterprises are equally difficult to control. Nevertheless, after trying 'everything else', subsidised nature protection might well be the only imaginable way to conserve threatened biotopes in developing countries such as dry coastal forests in East Africa.
Meanwhile, forest conservation activities must operate within the zone of the 'adjacent possible'. In doing so, however, they should seek to change their political and economic context by deliberately promoting incentive structures that reduce or eliminate local people's as well as employees of official authorities involvement in unsustainable resource use without undermining their livelihoods. In the case of the Arabuko-Sokoke forest, the most cost-efficient approach might be to support and establish coalitions among stakeholders who have an interest in exposing hidden institutions and pushing for full implementation of PFM. Also, although more costly, systematic and credible monitoring of the forest condition, which is made publicly available, would incentivise KFS to keep, de facto, levels of product extraction within 'acceptable' levels. None of the above would improve the conditions of KFS employees unless they could financially gain or break even by substituting their current involvement in illegal harvesting activities with activities related to monitoring the forest condition. Given the availability of sufficient (donor) funding, it should be possible to design a technical project that includes this political side-objective.
We would like to thank all the on-site institutions who offered their time and patience to assist the first author during fieldwork: Nature Kenya, Kipepeo Project, Kenya Forestry Research Institute, Kenya Forest Service, Kenya Wildlife Service, ASFADA, and the CFAs. We would also like to share our deepest gratitude to the great number of community members and local experts in Arabuko-Sokoke who volunteered to participate in interviews and participatory exercises. Without their kindness and willingness to devote their time and share their thoughts, this study would not have been possible. With regards to data processing, we want to thank Professor Henrik Meilby who provided technical assistance and valuable feedback on the GPS data, and who also devoted his time to participate in the analysis and processing of this data in GIS. Furthermore, we would like to thank the three anonymous reviewers' valuable inputs and constructive critique to this paper. Lastly, thanks to the University of Copenhagen, which contributed financial resources to the conduction of fieldwork in Kenya.
AGRAWAL, A. 2003. Sustainable Governance of CommonPool Resources: Context, Methods, and Politics. Annual Review of Anthropology 32: 243-62.
AGRAWAL, A., RIBOT, J. 1999. Accountability in Decentralization: A Framework with South Asian and West African Cases. The Journal of Developing Areas 33(4), (Summer, 1999), pp. 473-502.
ANGELSEN, A., JAGGER, P., BABIGUMIRA, R., BELCHER, B., HOGARTH, N.J., BAUCH, S., BORNER, J., SMITH-HALL, C. and WUNDER, S. 2014. Environmental Income and Rural Livelihoods: A GlobalComparative Analysis. World Development 64(1): S12-S28.
APPIAH, M. 2001. Co-Partnership in Forest Management: The Gwira-Banso Joint Forest Management Project in Ghana. Environment, Development and Sustainability 3: 343-360.
ASNER, G.P., KNAPP, D.E., BROADBENT, E.N., OLIVEIRA, P.J.C., KELLER, M., SILVA, J.N. 2005. Selective logging in the Brazilian Amazon. Science 310: 480-482.
BARRET, C.B., ARCESE, P. 1995. Are Integrated Development and Conservation Projects (ICDPs) Sustainable?, World Development 23(7): 1073-84.
BASNYAT, B., TREUE, T., POKHAREL, R.K., LAMSAL, L.N., RAYAMAJHI, S. 2018. Legal-sounding bureaucratic re-centralisation of community forestry in Nepal, Forest Policy and Economics, Volume 91 Pages 5-18, ISSN 1389-9341
BROCKINGTON, D. 2005. The costs of conservation: monitoring economic change as a consequence of conservation policy at Mkomazi Game Reserve, Tanzania. In: K. Homewood (ed) Rural resources and local livelihoods in Africa. James Currey, Oxford.
BRUCE, J. 1999. Legal bases for the management of forest resources as common property, Community Forestry Note No. 14, FAO, Rome, 134 pp.
CARTER, J., GRONOW, J. 2005. Recent experience in collaborative forest management. A review paper, CIFOR Occasional Paper No. 43. Bogor, Indonesia, Center for International Forestry Research (CIFOR).
CHARMAZ, K. 2014. Constructing Grounded Theory. Second edition. Sonoma State University, Rohnert Park, USA, Chapter 5.
CHOMBA, S.W., NATHAN, I., MINANG, P.A., SINCLAIR, F. 2015. Illusions of empowerment? Questioning policy and practice of community forestry in Kenya. Ecology and Society 20(3): 2.
COX, M., ARNOLD, G., TOMAS, S.V. 2010. A review of design principles for community-based natural resource management. Ecology and Society 15(4): 38.
ELLIS, F., MDOE, N. 2003. Livelihoods and Rural Poverty Reduction in Tanzania, World Development 31(8): 1367-1384.
FISCHER, M., SHIVERLY, G. 2005. Can income Programs reduce tropical forest pressure? Income shocks and forest use in Malawi. World Development (33): 1115-1128.
GIBSON, C.C., MCKEAN, M.A., Ostrom, E. (Eds.) 2000. People and forests: Communities, institutions and governance: Cambridge, MA: The MIT Press.
GLENDAY, J. 2008. Carbon storage and emissions ofset potential in an African dry forest, the Arabuko-Sokoke Forest, Kenya. Environmental Monitoring and Assessment 142: 85-95.
GORDON, I., AYIEMBA, W. 2003. Harnessing Butterfly Biodiversity for Improving Livelihoods and Forest Conservation: The Kipepeo Project, The Journal of Environment Development 12: 82-98.
HABEL, J.C., CASANOVA, I.C.C., ZAMORA, C., TEUCHER, M., HORNETZ, B., SHAURI, H., MULWA, R.K., LENS, L. 2017. East African coastal forest under pressure. Biodiversity and Conservation 26: 2751-2758.
HOBLEY, M. 1996: Participatory Forestry: The Process of Change in India and Nepal, Rural Development Forestry Network, Overseas Development Institute, London, England, pp. 1-25
JACKSON, M.M., NAUGHTON-TREVES, L. 2012. Ecobursaries as incentives for conservation around Arabuko-Sokoke Forest, Kenya. Environmental Conservation 39: 347-356.
LINDSAY, J. 2004. Legal frameworks and access to common pool resources. Forests, Trees and Livelihoods 14: 263-279.
LUND, J.F., TREUE, T. 2008. 'Are we getting there? Evidence of decentralized forest management from the Tanzanian Miombo woodlands'. World Development 36(12): 2780-2800.
MACFARLANE, D.W., KINZER, A.T. BANKS, J.E. 2015. Coupled human-natural regeneration of indigenous coastal dry forest in Kenya. Forest Ecology and Management, 06.026.
MATIKU, P., MIRERI, C., OGOL, C. 2012a. Is participatory forest management (PFM) an asset or liability to local community households adjacent to Arabuko Sokoke Forest, Kenya? Journal of African Studies and Development 4: 96-104. Ostrich 81: 150-153.
MATIKU, P., MIRERI, C., OGOL, C. 2012b. Does participatory forest management change household attitudes towards forest conservation and management? African Journal of Environmental Science and Technology 6: 237-346.
MBUVI, M.T.E, WAMBUA, S., AYIEMBA, A.O. 2007. Assessing the social and economic costs and benefits of protected areas: The case of Arabuko-Sokoke forest reserve, Kenya. A report prepared for the Poverty and Environment Network, CARE International, Nairobi, September 2007.
MBUVI, M.T.E., WEKESA, C., LELEY, N., OTUOMA, J., MWADALU, R., NDALILO, L., OGWOKA, B., NJERI, L. ANDANJE, S., NYAMACHE, J., MALONZA, P.K., MULWA, R.K., MUTUNGA J., AMUR, A. 2015. Biodiversity status of Arabuko Sokoke Forest, Kenya, Unpublished.
MESHACK, C.K., AHDIKARI, B., DOGGART, N., LOVETT, J.C. 2006. Transaction costs of community-based forest management: empirical evidence from Tanzania. African Journal of Ecology 44(4): 468-477.
MOGAKA, H. 1991. Local utilisation of the Arabuko-Sokoke Forest Reserve. Unpublished Report, KIFCON, Nairobi.
MURIITHI, S., KENYON, W. 2002. Conservation of biodiversity in the Arabuko Sokoke Forest, Kenya. Biodiversity and Conservation 11: 1437-1450.
MUSYOKI, J.K, MUGWE, J., MUTUNDU, K., MUCHIRI, M. 2016. Factors influencing level of participation of community forest associations in management forests in Kenya. Journal of Sustainable Forestry 35(3): 205-216. MUTUNE, J.M., LUND, J.F. 2016. Unpacking the impacts of 'participatory' forestry policies: evidence from Kenya. Forest Policy and Economics 69: 45-52.
NEWMARK, W.D., HOUGH, J.L. 2000. Conserving Wildlife in Africa: Integrated Conservation and Development Projects and Beyond: Because multiple factors hinder integrated conservation and development projects in Africa from achieving their objectives, alternative and complementary approaches for promoting wildlife conservation must be actively explored. BioScience 50(7), 1 July 2000, Pages 585-592.
NDANG'ANG'A, P.K., BARASA, F.M., KARIUKI, M.N., MUORIA, P. 2016. Trends in forest condition, threats and conservation action as derived from participatory monitoring in coastal Kenya. John Wiley & Sons Ltd. African Journal of Ecology 54: 76-86.
NDUMBARO, F., MUTULA, S.M. 2017. Collaborative information behavior of butterfly farmers in Eastern Usambara Mountains, Tanzania, Malasyan Journal of Library & Information Science, [S.I.], v. 22, n. 2, p. 15-29, ISSN 1394-6234.
NIGHTINGALE, A.J. 2005. "The Experts Taught Us All We Know": Professionalisation and Knowledge in Nepalese Community Forestry. Antipode 37: 581-604.
OAKERSON, R.J. 1992. Analyzing the Commons: A Framework. In: Daniel W. Bromley et al. (eds) Making the Commons Work: Theory, Practice and Policy, pp. 41-59. San Francisco, CA: Institute for Contemporary Studies Press.
OJHA, H.R. 2006. Techno-bureaucratic Doxa and Challenges for Deliberative Governance: The Case of Community Forestry Policy and Practice in Nepal. Policy and Society 25(2): 131-175, ISSN 1449-4035.
ONGUGO, P.O. 2007. Participatory Forest Management in Kenya: Is there anything for the poor? Proceedings: International Conference on Poverty Reduction and Forests. September, 2007. Bangkok, Thailand.
OSTROM, E. 1998. A Behavioral Approach to the Rational Choice Theory of Collective Action: Presidential Address, American Political Science Association, 1997. The American Political Science Review 92(1): 1-22.
OSTROM, E. 2008. Design Principles of Robust PropertyRights Institutions: What Have We Learned? (2008). In: INGRAM, K.G., HONG, Y.-H. (eds.) Property rights and land policies. Cambridge, MA: Lincoln Institute of Land Policy, 2009.
PAGDEE, A., KIM, Y., DAUGHERTY, P.J. 2006: What makes community forest management successful: a meta-study from community forests throughout the world. Society and Natural Resources 19: 33-52.
PERSHA, L., AGRAWAL, A., CHHARTRE, A. 2011: Social and Ecological Synergy: Local rulemaking, Forest Livelihoods, and Biodiversity Conservation. Science (331): 1606-1608.
PFLIEGNER, K. 2014. Participatory Forest Management--Panacea or Pretence? The impacts of Joint Forest Management on forest quality, livelihoods and governance in the Uluguru mountains Tanzania, Scholar's Press 2014-02-10.
PORTER-BOLLAND, L., ELLIS, E.A., GUARIGUATA, M.R., RUIZ-MALLEN, I., NEGRETE-YANKELEVICH, S., Reyes-Garcia, V. 2012. Community managed forests and forest protected areas: an assessment of their conservation effectiveness across the tropics. Forest Ecology and Management 268: 6-17.
POULIOT, M., TREUE, T. 2013. Rural people's reliance on forests and the non-forest environment in West Africa: evidence from Ghana and Burkina Faso'. World Development 43: 180-193.
RASOLOFOSON, R.A., FERRARO, P.J., JENKINS, C.N., JONES, J.P.G. 2015. Effectiveness of community forest management at reducing deforestation in Madagascar. Biological Conservation 184: 271-277. DOI: 10.1016/j.biocon.2015.01.027
RAVINDRANATH, N.H., SUDHA, P., MALTHORA, K.C., PALIT, S. 2004. Sustaining Joint Forest Management in India. In: Joint Forest Management in India, Universities Press, Hydarabad.
RIBOT, J.C. 2004. Waiting for democracy: the politics of choice in natural resource decentralization. World Resources Institute, Washington, DC. 140p.
RIBOT, J.C., PELUSO, N.L. 2003. A theory of access, Rural Sociology 68(1): 153-181.
SANDE, S., Crewe, R.M., RAINA, S.K., NICOLSON, S.W., GORDON, I.J. 2009. Proximity to a forest leads to higher honey yield: another reason to conserve. Biological Conservation 142: 2703-2709.
SAUNDERS, F.P. 2014. The promise of common pool resource theory and the reality of commons projects. International Journal of the Commons 8(2): 636-656.
SEPPALA, P. 1998. Tanzanian Local Administration, A vehicle for democratic development? FAD Working Paper 3/98, Institute of Development Studies, University of Helsinki.
THYGESEN, S.H., LOBER, T., SKENSVED, E.M., HANSEN, C.P. 2016. Implementation of Participatory Forest Management in Kenya: A Case Study of Karima Forest, Department of Food and Resource Economics, Faculty of Science, University of Copenhagen, International Forestry Review 18(3): 357-368. 2016
TOPP-JORGENSEN, E., POULSEN, M.K., LUND, J.F., MASSAO, J.F. 2005. Community based monitoring of natural resource use and forest quality in montane forests and miombo woodlands in Iringa District, Tanzania. Biodiversity and Conservation 14(11): 2653-2677.
WARDELL, A.D., LUND, C. 2006. Governing access to forests in Northern Ghana: Micro-politics and the rent of non-enforcement. World Development 34(11): 1887-1906.
Butterfly farmers list 2016. Kipepeo Butterfly Project, National Museums of Kenya, Unpublished
CFA Constitution, draft: Standard Constitutional drafts Jilore; Sokoke; Gede Communtity Forest Associations, Arabuko-Sokoke forest, Kenya, Unpublished
Forest Act (2005) no. 7 available at http://www.kenyaforest-service.org/images/MMMB/forests%20act%20no.7%20of%202005.pdf
MATIKU 2002. The coastal forests of Kenya: Forests data, threats, socio-economic issues, values, stakeholders, challenges, strategies, investment and enabling environment, A national synthesis report for the development of the WWF-EARPO Eastern Africa Coastal Forests Ecoregion Programme, Nairobi.
MENR (2007): Participatory forest management Guidelines. Ministry of Environment and Natural Resources, Nairobi.
MENWR (2014): Forest Policy, 2014. Ministry of Environment, Water and Natural Resources, Nairobi.
PFMP, draft: Standard Draft Participatory Forest Management Agreement to be specified, entered and signed between Kenya Forest Service (KFS) and Jilore; Sokoke; and Gede Community Forest Association (CFA), Arabuko-Sokoke Forest, Kenya, Unpublished
SFMP (2002): Arabuko--Sokoke Strategic Forest Management Plan 2002-2007. Produced by ASFMT with financial support from EU, DFIDand BirdLife International.
Birdlife International: http://datazone.birdlife.org/site/factsheet/arabuko-sokoke-forest-iba-kenya/text Accessed 9.9.17 Accessed 03.05.2017
Birdlife International,b: http://datazone.birdlife.org/site/factsheet/arabuko-sokoke-forest-iba-kenya/details Accessed 9.9.17 Accessed 04.05.2017
Conservation International: https://www.conservation.org/NewsRoom/pressreleases/Pages/The-Worlds-10-Most-Threatened-Forest-Hotspots.aspx Accessed 08.05.2018
Globalforestwatch: http://www.globalforestwatch.org/map/11/-3.30/39.94/ALL/terrain/loss,forestgain?tab=analysis-tab&geostore=f4eae47c04df62fbcf55f761775a64bf&begin=2001-01-01&end=2016-01-01&threshold=30&dont_analyze=true Accessed 11.09.2017
IIED Briefing: http://pubs.iied.org/17368IIED Accessed 03.01.2017
Kilifi.go.ke: http://www.kilifi.go.ke/content.php?com=2& com2=4&com3=43#.Wb-IhIy0M2w Accessed 14.08.2017
Kipepeo.org: http://kipepeo.org/ Accessed 23.01.17
Sustainability writer: https://sustainabilitywriter.wordpress. com/2012/04/18/10-of-the-worlds-most-threatened-biodiversity-hotspots/ Accessed 04.05.2018
UNESCO: http://whc.unesco.org/en/tentativelists/5507/Accessed 29.08.2017
L.M. BUSCK-LUMHOLT (a) and T. TREUE (b)
(a) Autonomous University of Barcelona, ICTA-ICP, Carrer de les Columnes Campus UAB, Bellaterra, 08193, Spain
(b) University of Copenhagen, Section for Global Development, Rolighedsvej 25, 1958 Frb. C, Denmark
Email: firstname.lastname@example.org, email@example.com
(1) This is an umbrella term covering nine different tribes (also referred to as Giriama).
(2) These 14 interviews were documented through note-taking during the sessions.
(3) The Kipepeo Project has added beekeeping and honey to their activity portfolio.
Forest user rights in Forest Act, 2005
(a) collection of medicinal herbs;
(b) harvesting of honey;
(c) harvesting of timber or fuel wood;
(d) grass harvesting and grazing;
(e) collection of forest produce for community-based industries;
(f) ecotourism and recreational activities;
(g) scientific and education activities;
(h) plantation establishment through non-resident cultivation;
(i) contracts to assist in carrying out specified silvicultural operations;
(j) development of community wood and non-wood forest based industries; and (k) other benefits which may from time to time be agreed upon between an association and the Service
Source: Forest Act (2005)
TABLE 1 Summary of interview-based data collection Method Target Group(s) Group discussion Butterfly producers Semi-structured interviews Government Institutions On-site NGOs Community Associations Structured interviews Local community members Method Notes Group discussion * Mida C butterfly user group: 7 participants * Mida B butterfly user group: 5 participants * Maghangani butterfly user group: 5 participants Semi-structured interviews * Director, KEFRI (on-site) x 2 * Kipepeo Project Manager, NMK (on-site) x 3 * Forester Gede, KFS (on-site) x 2 * Forester Sokoke, KFS (Regional office, Kilifi) x 1 * Ecosystem Conservator, KFS (Regional office, Kilifi) x 1 * Researcher, KWS x 1 * Field coordinator (F.C.), Nature Kenya x 2 * Extension officer (E.O.), Nature Kenya x 1 * Chairman (C), ASFADA x 2 * Previous chairman (P.C.), ASFADA x 2 * Member (M), ASFADA x 2 * Chairman, Gede CFA x 1 * Chairman, Jilore CFA x 1 Structured interviews * Butterfly farmers (n=53) * Community members who do not produce butterflies (n=32)
|Printer friendly Cite/link Email Feedback|
|Author:||Busck-Lumholt, L.M.; Treue, T.|
|Publication:||International Forestry Review|
|Date:||Dec 1, 2018|
|Previous Article:||Getting forest science to policy discourse: a theory-based outcome assessment of a global research programme.|
|Next Article:||Forest ecosystem services in Nepal: a retrospective synthesis, research gaps and implications in the context of climate change.|