Rethinking entrenched narratives about protected areas and human wellbeing in the Global South

Literature search

To search the literature on the social outcomes of PAs we combined two databases of evidence. First, we used a systematic map and database of 1043 studies published up to 2014 by McKinnon et al. [43] (available at https://natureandpeopleevidence.org), on the linkages between conservation interventions and human wellbeing in terrestrial and marine systems. We selected only peer-reviewed articles related to ‘area protection’ and/or ‘area management’ interventions in low- and lower middle-income countries only as designated by the World Bank (Supporting Information, World Bank Economies). We selected articles published after 2006 with a study date after 2003, to capture recent studies more reflective of people-centred approaches to PA conservation after the Durban Accord (2003) and the Millennium Ecosystem Assessment [44]. Our search resulted in a set of 285 relevant articles. These were screened on full text based on our exclusion criteria, reducing the set to 248 articles (Fig. 1; Supporting Information, Exclusion criteria).

Figure 1

Screening process and number of articles at each stage.

Second, we updated the database beyond 2014 with our own systematic literature search. We used the same search terms as McKinnon et al. [43], but limited the intervention search terms to those related to PAs and other area-based conservation measures, drawing upon terms used in a systematic review by Pullin et al. [20] of protected areas and supplementing these with our own. Using Web of Science, we limited the search to English language, peer-reviewed articles, published after 2014 (Supporting Information, Search terms). The search retrieved 7096 articles. These were imported into an EPPI-Reviewer 4 and screened based on our exclusion criteria, first on title and abstract, and second on full text, reducing the articles to 207. These were combined with the 248 articles identified from McKinnon et al. [43]. Duplicates were removed and 10 papers were excluded due to poor transparency of methods, resulting in a final set of 437 articles.

The 437 article abstracts (published between 2003 and 2017) were double screened for relevance to one or two narratives (with primary and secondary relevance agreed), resulting in 138 papers selected as relevant by two reviewers. Twenty papers were randomly selected from each set of papers per narrative for data extraction. Where fewer than 20 of the papers had primary relevance to one narrative, papers were selected from those that had secondary relevance. This was the case for N2 (one paper), N3 (one paper) and N5 (two papers). More papers were relevant to the Pro-poor (N1) and Participation (N4) narratives than to the remaining three narratives (Fig. 1). The location of PAs in the 100 papers were weighted towards Africa (63) and Asia (36), with only one paper from the Americas, and none from Oceania or Europe. This reflects both the disproportionate number of African and Asian countries categorised as low- and low middle-income (Supporting Information, World Bank economies), and publishing bias. Sixteen African countries and nine Asian countries are represented in the papers but with certain countries disproportionately represented: Tanzania (18); India (12); Nepal (10). Each paper was reviewed using a standard coding tool developed in Google Forms to extract and categorise the data relating to the study, PA, social outcomes and narratives (Supporting Information, Codebook).

Expert interviews

We carried out a total of 25 semi-structured interviews (either in person or on video call), including eight with academic researchers working on projects funded by the Ecosystem Services for Poverty Alleviation (ESPA) programme [45] and 17 with contacts of the authors working outside of academia. Interviewees were selected with the aim of achieving representation from different types of organisations across the globe, including international and in-country NGOs, state agencies and research organisations (Supporting Information, Non-academic interviewees) and for their experience in the governance of PAs and/or understanding their impacts. Interviewees were asked about their familiarity with each of the narratives and experience of their validity (Supporting Information, Interview questions). Interviews captured expert knowledge, long-term field experience and supported the identification and interpretation of key themes across the narratives.

Narrative synthesis

The publications that were randomly selected encompassed a range of designs, methods and data types (quantitative and qualitative), which was useful in exploring causal linkages, processes of change and contextual factors [46]. We assumed a level of quality through the peer-review processes of the journals, and used our expertise in the social sciences to assess the weight of evidence in support of the narratives in each paper which was categorised into strong (results fit the narrative with little deviation), partial (results are mixed or do not demonstrate the narrative in full) or none (results provide no support). Data from both the literature and interviews were combined in the analysis. A narrative synthesis aims to provide insight and deepen understanding rather than conventional systematic reviews which aim to answer specific questions [42]. We took a thematic synthesis approach [47], annotating and identifying themes within the extracted data and refining them in an iterative process. The findings are organised around these themes for each of the narratives in the text below and summarised in Table 1. The author carrying out the narrative review for each narrative reread the papers, extracted data, and interview transcripts, and the support categorisation and narrative text were discussed and agreed with the lead author.

Narrative 1: Conservation is pro-poor

This narrative asserts that as it is the poorest people who are most dependent on ecosystems for their livelihoods, biodiversity conservation through PAs can alleviate material poverty by securing provisioning ES such as food and fuel, and regulating services such as clean water [48,49]. This narrative would suggest that when there is loss of access to extractive uses economic benefits can come through tourism or payment mechanisms, for example, wildlife management areas (WMAs) are assumed to reduce poverty through increased income revenues from wildlife [50].

Of the 20 selected papers, three provided strong support for the narrative, with five showing no support or providing evidence against, and a further 12 showing some support but with mixed (positive and negative) or weak effects (Fig. 2). Interviewees were divided in their support (Fig. 3). The explanation for these divergent results rests on several factors. First, the extent that PAs are pro-poor centres on people’s access to ES and their benefits, in turn dependent on the management system which can range from strictly protected to community-managed areas. Although some services can benefit all across a landscape (e.g., flood protection), the negative impact of exclusion for other services was evident in our sampled papers in both terrestrial [51,52] and marine PAs [53]. The poor living in and around PAs are also more exposed to ecosystem ‘disservices’ from wildlife such as crop-raiding [52,54] which can have wide-ranging and hidden impacts such as on psychological health and education [55].

Figure 2

Level of support shown by articles for each narrative.

Figure 3

Interview responses on the validity of each narrative. n = 22 as three interviewees chose not to answer these closed ended questions in the interviews.

Nine of our interviewees questioned the logic of the narrative: the poor often do not benefit from ES from a PA, and in fact are more likely to lose out. The wealthy are better placed to benefit due to their higher capacity to capture resources and bypass access restrictions, especially if governance is weak. The papers that disaggregated data according to wealth supported this idea. For example, compared with poorer households, wealthy households participate more in Payment for Ecosystem Services (PES) schemes in PAs in Cambodia [56], benefit more in terms of food security from CBNRM in Tanzania [57] and access benefits from devolved forest management in Ethiopia [51]. Indigenous groups who are already socially marginalised are at particular risk of disproportionate harms if they are not given special protection, such as the Twa whose livelihoods and culture are intertwined with native forests in Rwanda [58]. The poorest and landless are more dependent on resources from PAs, and by necessity have to risk fines and imprisonment where there are legal restrictions [58,59]. Tourism benefits are also prone to elite capture without redistribution policies in place [55,56,60].

Where poor local residents are not excluded from the benefits of conservation, the papers showed limited evidence that PAs are a pathway out of poverty, a message reflected in the literature on linkages between ES and poverty alleviation [48,61]. PAs more readily act as a social safety net preventing further poverty. For example, those most reliant on income from the Chiradzulu Forest Reserve, Malawi, are among the poorest, who have little education, more dependents, fewer assets and are more likely to be women [62]. The provision of forest products to the poor from Kibale National Park, Uganda protects them against desperation sales of farmland and thus sinking deeper into poverty [63]. One paper in our sample showed neutral impacts on food security [64], and Canavire-Bacarreza and Hanauer [65] show an average reduction in poverty in municipalities in Bolivia that have at least 10% of their areas covered by PAs. These papers represent a growing body of robust quantitative research providing evidence that some PAs in the Global South can reduce poverty or at least do not necessarily increase it especially where there is tourism and or the PA is not strictly protected (e.g., [22,66,67]) but do not look beyond objectively measured average material poverty and health.

The papers in our sample that showed strong support for the narrative used variables and metrics centred on material wellbeing [62,63] with the exception of Canavire-Bacarreza and Hanauer [65], who measured average effects on a poverty index which incorporates education and health. Research that looked at changes in diverse aspects of wellbeing (e.g., non-use values, food security, empowerment) paints a more complex picture with gains in some variables and losses or no change in others [54,57,68]. Likewise, interviews suggested that the most important costs and benefits for wellbeing and local support for PAs may not be material, for example, cultural knowledge or a sense of autonomy. Gurney et al. [69] highlight this point: despite a positive impact on livelihood diversity and wealth from marine PAs in Indonesia, subjective wellbeing was negatively affected most likely due to increased conflict and unmet expectations.

To fully understand the impacts of PAs, consideration must be given to the wider spatial, temporal and socio-economic context. The effect of PAs may be relatively limited where there are strong drivers of poverty or development related to market access, land policy and population changes [52,56]. Dawson and Martin [58] highlight how positive outcomes for biodiversity and wellbeing are in part dependent on the governance of the wider landscape outside of PAs and therefore provision of alternative vital resources. Studies that investigate impacts at different scales show that the validity of the narrative can change through time and space with trade-offs involved. Those closest to PAs or in more accessible areas tend to access benefits derived from ES such as income [62] or tourism infrastructure [70], but are also exposed to the damage from wildlife [55]. Temporal dynamics affect how benefits are realised: for example, benefits may be felt most during implementation when funding is available [69], or conversely may take time to be realised [57]. Positive benefits from long-term sustainability involve time-lags and in the case of mangrove protection, counteracted immediate losses of resources but with uncertain trajectories [71].

Overall, our analysis suggests that it is possible for PAs to alleviate material poverty but the extent to which the PA will benefit the poor depends on a range of factors including restrictions to locally important ES (especially provisioning services), whether local people have the capability (related to wealth and status) to benefit from ES, and how the PA and wider landscape is governed.

Narrative 2: Poverty reduction benefits conservation

The idea that resource overexploitation is a response to poverty was first popularised amongst conservationists in the World Conservation Strategy of 1980 [72] and since then has formed the basis for an instrumental argument that poverty alleviation should be integral to conservation initiatives. This narrative, to varying extents, underpins ICDPs, alternative livelihoods and revenue sharing schemes from ecotourism. There are two principal rationales for such programmes: first, to provide economic substitutes that reduce reliance on natural resources and lessen environmentally damaging behaviours; and second, to increase local acceptance and support for conservation, creating positive change in attitudes and behaviours [73].

There was mixed support for this narrative in our sampled literature and our interviewees were divided on its validity. Several papers did show how schemes designed to improve people’s material wellbeing positively influenced attitudes towards conservation (e.g., [74,75]), but this did not extend to strong evidence of change in behaviour or biodiversity outcomes. Those papers which studied behaviour showed some effects on reported extractive activities which were small and inconsistent [76] or reflected potential confounding factors [75]. Ecological outcomes were not maintained in the longer term [77] or were not clearly linked to social improvements [78]. The relationship between conservation attitudes and behaviour is not straightforward, and the evidence highlighted the need to understand not only attitudes towards conservation but towards PA staff and conservation organisations which can be instrumental in creating support [74].

The experience of our interviewees suggests that the narrative is more valid when people perceive a direct link between the PA and benefits, they receive. This linkage can be achieved in two main ways; first where the livelihood intervention is materially dependent on effective conservation (ecotourism, agro-forestry and resource access), and second where there are economic incentives such as the conditionality of PES payments. In fact, the PES concept emerged as a counter-narrative to the assumption that support for local incomes automatically enhances conservation effectiveness, instead arguing that such support needs to be conditional on conservation performance [79]. Our sample included seven papers which looked at interventions in the former category, but positive effects were not more strongly evidenced than in other livelihood schemes. A case study provided by an interviewee documents one positive example: in the Amani Butterfly Project in northern Tanzania, successful butterfly farming relies on the existence of the PA natural forest and income from butterfly farming was positively associated with participation in forest conservation [80]. Farmers perceive a link between butterfly farming income and forest conservation, thus motivating behaviours such as tree planting and reporting of illegal activities. Although having a more logical basis, our interviewees suggested that in reality the socio-economic conditions conducive to such an arrangement are rare. The literature also suggests that these projects are no less susceptible to failures in implementation such as administrative delays, lack of technical support and unequal distribution of benefits which can all lead to erosion of trust and cooperation [81,82]. Our two sampled papers on PES, show that conditionality provides a better guarantee of positive environmental outcomes but impacts on poverty are dependent on the magnitude of payments which can often be small, and there is a tendency for benefits to be captured by elites [83,84].

Providing benefits is not a guarantee of attitude and behaviour change. In many cases, especially where time is not a limiting factor, these livelihoods will supplement rather than substitute resource extraction. Where there are big risks associated with conservation such as human–wildlife conflict, these may be a barrier to changing attitudes even where people are benefiting [85]. On the other hand, where there are large economic gains from alternatives, they may have the unintended consequence of exacerbating pressure on PAs by encouraging in-migration or reinvestment [86,87]. Livelihood decisions are driven by a range of factors beyond economic costs and benefits. Projects implemented with little regard to local community needs or cultural identities which may be closely tied to resource-dependent livelihoods such as fishing are more likely to fail [88]. In marine PAs in the Philippines, where economic expectations are not being met, this has led to negative attitudes towards conservation. Chaigneau and Brown [89] suggest in this case that it is more realistic and sustainable to emphasise non-material bequest and aesthetic values which also produce positive attitudes and action against illegal fishing.

Another key consideration is the differentiated nature of resource users. Although the poorest may be more dependent on natural resources, the wealthiest may be the heaviest extractors [90] and able to circumvent access restrictions [91]. As one interviewee pointed out, this creates a tension between strategies that will have the best outcomes for biodiversity and for poverty alleviation. Similarly, high natural resource dependency and lower social status for those in poverty restrict their ability to participate in poverty reduction programmes [81,92]. There are often larger forces at work in creating conservation problems at multiple organisational levels. Targeting only the livelihoods of local communities does not address wider drivers of unsustainable extraction such as fluctuating prices and political instability [90].

Livelihood-based interventions continue to attract significant donor funding [93]. While improving livelihoods is a good thing in its own right, and can foster improved relationships and trust between communities and conservationists [94], there is a lack of evidence that this will inevitably result in improved ecological outcomes. In designing these projects, there is a need to understand the drivers of unsustainable resource extraction, the livelihood profiles of communities and the priorities of resource users. In theory, projects that link livelihoods to biodiversity and local people and/or involve conditionality are more likely to succeed in terms of ecological outcomes, but this may involve trade-offs with poverty alleviation.

Narrative 3: Compensation neutralises costs of conservation

This narrative accepts that there are unavoidable local costs to conservation in the form of access restrictions and human–wildlife conflict and assumes that these can be effectively offset thus fulfilling the ‘do no harm’ principle [95]. Compensatory approaches such as payments for harm caused by wildlife, resettlement, revenue sharing and development schemes, are driven not only by social justice concerns but also by efforts to reduce conflict and create positive attitudes towards conservation [96,97]. Increasingly, conservation is funded by major international donors who have explicit commitments to safeguard against negative social impacts and compensate for economic losses [98].

None of the reviewed literature was strongly supportive of this narrative, with only five papers providing some evidence that compensation is supported by local communities and at least partially offsets costs. The reasons related to both the compensation itself and the way in which schemes are implemented. First, the assumption that material compensation is commensurate with losses incurred from PAs is problematic. Compensation is often considered insufficient and not reflective of market values. In our sampled literature this was the case for compensation provided for a range of impacts including livestock loss [99,100], constraints on forest activities [101] and crop-raiding [102]. Material compensation is incommensurate with cultural losses. For example, although Twa communities received material benefits from revenue sharing from the Bwindi National Park, they have lost social freedoms and cultural heritage associated with hunting [103]. In Madagascar, many older households would be unwilling to stop the practice of swidden agriculture (tavy) in exchange for compensation, due to its socio-cultural value [104].

Material and monetary compensation is often provided for restricted access to land and displacement by PAs but may not account for material and non-material wellbeing losses. For example, land in resettlement villages was not perceived to be of comparable quality or quantity to that lost due to displacement from Suklaphanta Wildlife Reserve in Nepal, causing increased workloads, limited social interactions and reduced subjective wellbeing [105]. Land has cultural meaning, and places are intertwined with a sense of security, belonging, spirituality and identity that cannot be substituted [106,107]. Nevertheless, if community needs and aspirations are met, it is possible that resettlement can be carried out in a way that does not undermine people’s rights and wellbeing. For example, due to declining pastoral productivity and conflict with tigers, resettlement was the preferred option for the Gujjars in Nepal if it was associated with enhanced benefits including larger resettled land sizes, strengthened property rights and improved housing [108].

Although there was a mixture of views among our interviewees on the validity of this narrative, those that agreed were cautious in their support due to the difficulties in quantifying the meaning that livelihood practices hold, the practical challenges in administering compensation and unfulfilled promises made by government agencies. But several respondents explained how compensation can play an important role and provide a level of legitimacy for PA interventions, where there are tangible losses such as to livestock and agriculture. In a human–wildlife conflict compensation scheme in India, despite numerous shortcomings, respondents still supported a reformed compensation approach where conflict cannot be avoided [100]. The prevalent view amongst our respondents was that although not sufficient as a standalone approach, appropriate and timely compensation can be an important element of conservation if reinforced with greater engagement and recognition of costs. This should involve commitment that goes beyond the provision of one-off payments to include, for example, preventative measures to reduce human–wildlife conflict. However, two respondents raised the point that the whole idea of compensation removes power and incentives away from communities to manage ecosystems sustainably.

Even if compensation can work in theory, in practice schemes are often poorly implemented and administered. The process of claiming compensation can be long and tedious involving elaborate paperwork [99] and high transaction costs [100]. Where development projects are implemented, there can be a temporal mismatch whereby costs from resource access restrictions are immediate, but the benefits take time to emerge [101]. Limitations on the wildlife species included in compensation schemes or inappropriate methods to estimate compensation result in insufficient compensation [109]. Governments may fail to honour their commitments where compensation is not enshrined in policy or is associated with problems of corruption [100]. Inadequate or delayed compensation can develop deeply held grievances resulting in retaliatory killing of wildlife [110].

There is significant evidence of distributional inequity in compensation programmes. Development programmes may not reach those experiencing the greatest costs from PAs, but instead cluster around village and tourist centres, exacerbating economic inequalities [101,111]. There are often barriers to the most vulnerable groups accessing compensation. Households receiving compensation tend to be larger and wealthier [100], more food secure, better socially connected and live in more accessible areas [112]. Women and the poor face greater difficulty in accessing compensation as they lack official title to land, awareness of schemes, literacy, time and familiarity with bureaucratic procedures [100,105]. Even where monetary compensation reaches the poor, they may not have the capacity to reinvest in buying land and restoring livelihoods [113]. The result is that marginalised groups receive the least from compensation, if anything at all, even in cases where safeguarding procedures are in place to ensure the contrary [112].

In summary, the evidence rejects the idea that compensation as implemented is enough to substitute for experienced costs that often encompass non-material aspects of wellbeing and injustices. This does not mean that compensation is unnecessary, but it is rarely sufficient or commensurate. In addition, compensation mechanisms often do not work in practice, undermining social justice and support for conservation. Furthermore, our review suggests that there are situations in which compensation will never be commensurate with the loss incurred, thereby demanding greater openness to culturally appropriate alternatives.

Narrative 4: Participation is good for conservation

There are two inter-linked reasons why participation is assumed to be instrumental to effective conservation. Firstly, participation can empower local communities to govern resources sustainably, an argument that owes much to research into governing commons [114] and the value of local knowledge [115]. Secondly, participation may motivate local support and stewardship by providing economic and non-economic benefits [33]. In other words, this narrative holds that participation can provide both the opportunity and the motive for communities to support conservation.

Such a narrative has ensured that participation became a central tenet of mainstream PA governance policy [16]. In international law it has also been clarified that procedural rights (access to information, participation in decision-making and access to justice) need to be respected in the designation and management of PAs [15]. The participation narrative has not gone completely unchallenged: a counter-narrative emerged around the turn of the century, questioning the effectiveness of participatory and community-based conservation [25]. In development studies some proclaimed participation a ‘new tyranny’ that served to reinforce unequal power relations and state control [116].

In our sample of 20 articles, 18 were judged to support the narrative although only three showed a strong link between participation and ecological outcomes. This was reflected by the interviews, where all respondents agreed with the narrative, except two who remained neutral. This body of research largely confirms that participation contributes to both motivation and capacity to support conservation, but also qualifies this in terms of the range of benefits that can motivate local people and the quality of participation that is required to empower people. Motivations for participation appear to vary across cases, and across different social groups. Participation can be motivated by expected livelihood benefits [117–119], but there are also several cases in which participatory conservation fails to deliver livelihood benefits yet is still valued for other reasons such as improved social capital [120], and sense of control [121]. According to one study, material motives are more important to men, whilst social motives are more important to women [122]. Whilst women may value participation for non-economic reasons, they are often less able to participate, due to constraints on their time or social barriers to taking on public roles [117,119,123–125]. As confirmed by interviewees, participation can thus impose a social cost due to lost time or livelihoods that outweigh the benefits of participating, so transaction costs need to be minimised. On balance, the evidence confirms that the opportunity to participate in PA management is widely valued by local communities.

The studies reviewed show us that the linkage between participation and effective conservation is not contingent on delivering livelihood benefits but can arise from either satisfying other needs and interests and/or triggering community capacity to control resource use. For example, a forest co-management programme in Malawi was found to have no short- or medium-term effect on household incomes, but participating households still cleared less forest than non-participants [126]. Similarly, participatory forest management in Tanzania did not provide measurable gains in wellbeing but forest governance was improved by reviving the community’s capacity to exclude outsiders [121]. Whilst community rights may be sufficient to unlock local capacity to manage resources, a study of marine fisheries in Kenya found that community co-management rights only led to positive ecological outcomes in conjunction with the establishment of no-take marine reserves highlighting the need for conducive socio-economic conditions and institutional capacities of communities [127].

The reviewed studies show that local contexts lead to variation in what motivates participation and what communities can achieve with rights to participate. One finding that is consistent across all of the studies and confirmed in interviews, is that the quality of participation is crucial in determining both motive and capacity for conservation [128]. Participation is often tokenistic and superficial, and this is recognised by communities as constraining what they can achieve. Interviewees highlighted that meaningful participation means having the power to effect change regarding ecosystem governance. In a survey in the Taita Hills, Kenya, 33% of respondents identified the superficiality of participation as the greatest constraint on forest conservation [122]. In both of the negative cases in our sample, the quality of participation is a key factor in undermining benefits to communities although there is evidence that ecological outcomes are positive at least in the short-term due to access restrictions [129,130]. Four papers that were categorised as partially supportive showed that superficial participation had negative implications for sustainability. In the study of WMAs in Tanzania, participation was manipulative, disempowering and went hand in hand with demonstrable harm to local livelihoods [129]. However, as highlighted in the interviews, participation is an evolving process, and one that needs sufficient time and resources to allow people to build relationships and negotiation skills, a point evidenced in the broader literature [131,132].

Our interviewees highlighted that the time, capacity and resources required for effective participatory processes often require the support of external agencies who can share the costs. For example, multi-community partnerships in marine PA sites in the Comoros, involving networks of communities, government and NGO actors, facilitated cooperation in fishery management ensuring all communities cooperated in fishery management on an equal footing [128]. Similarly, participation in marine PA sites in Indonesia was more extensive if management groups were supported by external institutions, such as through partnership with NGOs, academia and other community groups [133]. However, internal power structures will affect how participation takes place. Working through established customary governance arrangements is an effective route to establishing participatory conservation, but without mediation to steer negotiations towards inclusive governance, minority interests may get sidelined with repercussions for long-term sustainability [134].

In summary, this narrative is supported by our review, although there is limited evidence linking participation to ecological outcomes. Based on our findings we would qualify the narrative somewhat, such that participation supports PA effectiveness where it genuinely empowers communities and provides benefits that are locally valued and equitably distributed.

Narrative 5: Secure tenure rights for local communities support effective conservation

Secure tenure rights are increasingly considered an important foundation for attaining positive conservation outcomes as they may increase the local legitimacy of and participation in conservation governance, promote the sustainable use of resources and foster local environmental stewardship against internal and external pressures [34,135]. The scope of legitimate tenure rights is not limited to individual property rights, which are often afforded greater legal status. Prominent theories, frameworks and international policy guidance defining tenure specifically include multiple types of tenure, and pivotally for conservation practice this includes customary and communal regimes and institutions [136], which are often side-lined as they comprise ‘informal arrangements’ and ‘unwritten customs and practices’ [137]. Increasing attention to security of tenure rights in conservation policy has resulted in the enhanced inclusion of areas managed by local communities within the global PA network [35]. Indigenous peoples already manage more than a quarter of the world’s land area but may struggle to protect these areas due to weak rights [138]. Clear and secure tenure rights are also pivotal for policy instruments such as PES or Reducing Emissions from Deforestation and forest Degradation (REDD+) to determine who is eligible to receive benefits and who is responsible for meeting contractual obligations [139]. Although absent in the Millennium Development Goals, tenure rights appear in five of the Sustainable Development Goals [140].

Of the 20 sampled articles addressing this narrative, none provide opposing evidence while 11 provide strongly supportive evidence. Six of those show evidence of a positive association whereby recognition of tenure rights leads to enhanced social and ecological outcomes, whereas five exhibit a negative association through which violation of or insecurity caused to local communities’ tenure rights through externally driven conservation interventions produces negative social and ecological outcomes. A further eight studies provide partial support for the narrative but assume the positive or negative social impacts promote or harm conservation, respectively, without providing specific evidence. The one remaining study suggests that secure individual property rights enhance conservation, although without paying any attention to other forms of tenure or potential social impacts of favouring a formal, individual tenure system [141]. Interviewees were also largely supportive of this narrative.

This set of cases highlights the pivotal importance of both tenure security based on customary and communal systems and of the scope of local influence in governance processes. At the most basic level, negative associations tend to occur when conservation interventions negate user rights with no regard for local needs or customary and communal institutions. In this situation, when a hegemonic model of conservation overrides existing systems through which rights are allocated among local communities, imposed structures may have negligible legitimacy and be entirely disregarded so that conservation goals are not met [142–144]. For example, where conservation interventions recognise only legal or individual property rights as compatible with conservation rules and override customary and communal local institutions, this may favour more powerful local or non-local actors to the detriment of vulnerable groups including the poor, women and cultural minorities. Those requiring access to support livelihoods or engage in cultural practices may act in defence of their needs and rights by seeking to establish an alternative to exclusive conservation rules, often through negotiation with alternate authorities such as sympathetic local government officials, as described by Rahman et al. [145] in Bangladesh. Such a situation can open the door to elite collusion and capture because the conservation intervention triggers a renegotiation of tenure rights, threatening ecological integrity both outside of and within PAs [146,147]. Instances of negative social and ecological outcomes resulting from imposed tenure regimes were also evidenced through contemporary governance approaches such as REDD+ [148,149].

The evidence suggests that to establish appropriate tenure security and sufficient rights to foster effective local stewardship, locally supported institutions that may have formed over long timescales need to be embedded within conservation structures that give sufficient confidence they will endure. Clear positive examples among the sample studies of recognising rights by embedding local tenure institutions within conservation included the Kasigau Corridor REDD+ project in Kenya that recognised communal forest tenure regimes [150], and the engaged stewardship and mobilisation of resistance to unsustainable logging in Cambodia [151]. Where secure tenure supports local livelihoods and fosters effective local stewardship it can be particularly important to protect those governance structures. For example, weakening of rights for betel nut growers in Soppinabetta forests in the Western Ghats of India to control resource use on their land led to many selling it for extractive development [152].

It is also important to consider the extent of rights granted to different groups of people, looking beyond basic user rights to address rights of control and authority that determine who has power to make decisions about resource allocation and influence governance structures [153]. Provision of use rights alone may not be enough to prevent tenure insecurity arising, because if people fear those rights are likely to be removed or overruled and they lack any power to block those decisions, positive feedbacks for conservation may be foregone, as exemplified by Davis [154] for Maasai pastoralists impacted by a WMA in Tanzania. Indeed, three interviewees highlighted difficulties for pastoralist groups whose dynamic and seasonal customary systems of land and resource rights may be threatened through tenure formalisation processes. To nuance these findings further, several studies highlight the dangers of romanticising local institutions and their ability to govern natural resources adaptively and inclusively, particularly because management capacity and local cohesion may be lacking in the face of numerous drivers of social and institutional change at the local level [144,155].

Our review also revealed the need to explore not just tenure systems but perceptions about the security of tenure rights, which can be a key determinant of behavioural change, even when tenure arrangements appear stable. Local perceptions of tenure security can be highly influenced by past experiences of policies enacted by states or colonial powers, and conservation interventions can be perceived as extensions of them [156,157]. Even if conservation authorities are trusted, perceived tenure security may be weak where the central government has a record of overriding them, for example, through the proliferation of land concessions for commercial, infrastructure and extractive industries in Cambodia [151].

In summary, secure tenure rights can empower communities to sustainably manage resources and participate in effective ecosystem governance. However, respect for customary and communal access systems, and trust in the governance arrangements are critical for success.

Research ethics statement

Ethical approval was granted by the Departmental Research Ethics Committee of University College London’s Anthropology Department.

Consent for publication statement

The authors declare that research participants’ informed consent to publication of findings – including photos, videos and any personal or identifiable information – was secured prior to publication.

Conservation organisations

Evidence for narratives on conservation organisation websites (January 2018)

*Organisation operates internationally outside of Africa.

Interview validation

World Bank Economies

(March 2017)

Exclusion Criteria

Exclude on date: The study has a publication date before 2014.

Exclude on country: The study is outside of the list of included countries for the IMPACTS project, which focuses on low- and lower middle-income countries only (according to the World Bank).

Exclude on population: The subjects of the intervention are not discrete individuals, households, communities or national states.

Exclude on intervention: The study does not include the establishment or management of an area-based protected or conserved area intervention or associated policy or programme.

Exclude on outcome: The study does not observe, measure or describe human wellbeing indicators, outcomes, or impacts. The study only focuses on biophysical outcomes of conservation or solely examines how status or trends in human wellbeing affect conservation outcomes.

Exclude on study type: The study is a theoretical or conceptual article, modelling study, commentary, editorial or narrative review.

Search Terms

Protected area intervention terms

protected area* OR nature reserve* OR wilderness area* OR national park* OR natural monument* OR natural feature* OR management area* OR world heritage site* OR biosphere reserve* OR biodiversity conservation OR national reserve* OR conservanc* OR ecotourism OR corridor* OR community-based conservation OR payment for ecosystem services OR PES AND

Intervention adjacent terms

marine OR freshwater OR coastal OR forest* OR ecosystem* OR species OR habitat* OR biodiversity OR sustainab* OR ecolog* OR integrated OR landscape OR seascape OR coral reef* OR natural resource* AND

Outcome terms

wellbeing OR well-being OR well being OR ecosystem service* OR nutrition OR skill* OR empower* OR clean water OR livelihood* OR (food) security OR resilience* OR vulnerability OR (social) capital OR attitude* OR perception* OR (human) health* OR human capital OR (traditional knowledge or TEK) AND

Outcome adjacent terms

human* OR people OR person* OR community* OR household* OR fisher* OR collaborative

Codebook

Non-academic interviewees

Interview Questions

Please consider the following context and the widespread assumptions that follow. We would then like to ask about your experience, and your opinion on those assumptions, through a set of structured questions.

STUDY CONTEXT

• We are interested in understanding the processes involved in positive and negative social impacts of PAs on human wellbeing, how impacts may differ between groups, for example, according to gender and age, and how relationships between social and environmental outcomes may vary with context such as with governance type, economic and social factors.

• We are interested in impacts on multi-dimensional human wellbeing and equity. Therefore, please consider the following aspects:

  • Material wellbeing including assets, income, food, livelihoods and subsistence activities

  • Health

  • Security – confidence in the future

  • Social relations, for example, within the community and between communities and PA authorities

  • Agency, empowerment and participation

  • Recognition of and impacts on cultural values and practices

  • Subjective wellbeing – how people feel about change

• By ‘protected areas’, we mean all kinds of protected and conserved areas in marine and terrestrial habitats. This includes the six IUCN Protected Area management categories, plus other effective area-based conservation measures, such as other types of indigenous, community and privately conserved areas; and areas that provide connectivity between designated protected areas. We are interested in the establishment, management processes and activities associated with these areas.

• Our focus is on PAs and communities in low- and lower middle-income countries (according to the World Bank), the countries that are the target of ESPA work.

We are structuring the analysis by investigating five key narratives in the data, which we have identified as underlying PA establishment and management actions. We are interested in whether these assumptions hold in real-life situations.

NARRATIVES

1. Ecosystem services narrative

Because poor people are disproportionately dependent on ecosystem services, protected areas that target those services will be pro-poor

This narrative sees PAs as a tool to reduce poverty. It is often the poorest people who are most dependent on ecosystems for their livelihoods and food security. By protecting these ES, PAs are thus important for delivering poverty reduction objectives by supporting a range of economic activities such as forestry, fisheries, agriculture and tourism as well as providing access to clean water and energy.

2. Poverty and conservation narrative

Because poor people are disproportionately dependent on ecosystem services, improving their material wellbeing will reduce pressure on protected areas

This narrative sees poverty reduction as a means to achieve PA conservation. This can occur through strategies such as alternative livelihoods, revenue sharing, investment in infrastructure and tourism, which are implemented to reduce reliance on natural resources and behaviours that might be environmentally damaging. The economic benefits generated by these strategies also increase local support for conservation. Incentive schemes such as PES that provide benefits to poor people are directly linked to PA conservation objectives, thus enhancing conservation success.

3. Compensation narrative

Unavoidable social costs of protected areas for poor people can be neutralised by providing appropriate compensation

Because poor people endure the costs of PAs, such as due to human–wildlife conflict (crop-raiding, livestock predation) or reduced access to PA resources for food, fuel or livelihoods, compensation schemes offset these costs and create more positive attitudes towards PAs.

4. Participation narrative

Participation in protected area governance is seen by communities as a positive social benefit and it is a route to effective conservation

This narrative sees participation of local people in PA decision-making and recognition of their values and interests as important factors contributing to long-term conservation success. Participation is valued by communities and increases the legitimacy of the PA and its policies. Participation increases communities’ sense of stewardship over PAs, builds capacity and creates greater incentives to harvest resources sustainably to ensure future access to benefits.

5. Secure tenure narrative

Secure land and resource tenure underpins improved conservation outcomes (social and ecological) in and around protected areas

This assumes that tenure insecurity is the cause of resource degradation. Secure land rights enable poor people to invest in resources and strategies that promote resilient livelihoods. It is also the foundation for the sustainable use of resources that supports long-term conservation outcomes.

We would like to ask you the following questions, and for you to share your thoughts based on your own experiences through your work on ESPA projects or other similar projects or research.

Overall:

1. From your ESPA project/other professional experience, can you give examples of a) positive and b) negative social impacts of PAs on the wellbeing of local people? Can you explain why these might have occurred? Where they have been negative? What has been done to reduce them?

2. From your own experience, what are the synergies you’ve found in PA conservation (i.e., different outcomes interact to improve the overall outcome) a) between social and ecological outcomes, and b) between different social outcomes? Can you give examples of each from your work?

3. What do you see as some of the critical trade-offs in protected area conservation, a) between social and ecological outcomes, and b) between different social outcomes? Can you give examples of each from your work? Are there difficult decisions that you’ve had to make, or you have seen others make, in regard to these trade-offs, and how they have been resolved?

Protected area narratives

For each of the narratives above can you/answer the following:

1. Is this assumption familiar to you? (circle as appropriate)

   

2. Where have you come across the idea?

3. Do you think the assumption has changed through time? How?

4. From your experience do you agree that the assumption is valid? (circle as appropriate)

   

5. If at all valid, can you provide examples? Where? In what kinds of PA? For whom?

6. If you do not think it is completely valid in what ways or in which contexts do you think it is not valid? Where and for whom? Please refer to specific examples such as case studies or papers.

7. Do you have an additional narrative or idea that you think guides PA management?