Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Human well-being impacts of terrestrial protected areas

Human well-being impacts of terrestrial protected areas Background: Establishing Protected Areas (PAs) is among the most common conservation interventions. Protecting areas from the threats posed by human activity will by definition inhibit some human actions. However, adverse impacts could be balanced by maintaining ecosystem services or introducing new livelihood options. Consequently there is an ongoing debate on whether the net impact of PAs on human well-being at local or regional scales is positive or negative. We report here on a systematic review of evidence for impacts on human well-being arising from the establishment and maintenance of terrestrial PAs. Methods: Following an a priori protocol, systematic searches were conducted for evidence of impacts of PAs post 1992. After article title screening, the review was divided into two separate processes; a qualitative synthesis of explanations and meaning of impact and a review of quantitative evidence of impact. Abstracts and full texts were assessed using inclusion criteria and conceptual models of potential impacts. Relevant studies were critically appraised and data extracted and sorted according to type of impact reported. No quantitative synthesis was possible with the evidence available. Two narrative syntheses were produced and their outputs compared in a metasynthesis. Results: The qualitative evidence review mapped 306 articles and synthesised 34 that were scored as high quality. The quantitative evidence review critically appraised 79 studies and included 14 of low/medium susceptibility to bias. The meta-synthesis reveals that a range of factors can lead to reports of positive and negative impacts of PA establishment, and therefore might enable hypothesis generation regarding cause and effect relationships, but resulting hypotheses cannot be tested with the current available evidence. Conclusions: The evidence base provides a range of possible pathways of impact, both positive and negative, of PAs on human well-being but provides very little support for decision making on how to maximise positive impacts. The nature of the research reported to date forms a diverse and fragmented body of evidence unsuitable for the purpose of informing policy formation on how to achieve win-win outcomes for biodiversity and human well-being. To better assess the impacts of PAs on human well-being we make recommendations for improving research study design and reporting. Keywords: National Park, Reserve, Community, Governance, Conservation, Poverty, Development, Biodiversity, Systematic review * Correspondence: a.s.pullin@bangor.ac.uk Centre for Evidence-Based Conservation, School of Environment, Natural Resources and Geography, Bangor University, LL57 2UW Bangor, Gwynedd, UK Full list of author information is available at the end of the article © 2013 Pullin et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Pullin et al. Environmental Evidence 2013, 2:19 Page 2 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Background available evidence on their human well-being impacts. The concept and practice of protecting areas for the That there have been and continue to be, some major purposes of conservation has been at the heart of con- negative impacts on local communities caused by the es- servation policy since its inception in the 19th Century. tablishment of some protected areas, is not in dispute. The idea that intervening to protect areas from human However, of particular interest to policy makers is the bal- activity is an effective way of conserving species and ance of positive and negative impacts on human livelihoods habitats and preventing habitat loss and species extinc- that arise from the PA establishment, the distribution of tion is arguably as pervasive today as it was when the these benefits and costs, and the factors that might cause first protected areas (PAs) were established (MEA 2005). this to vary. The central place of PAs in the conservation movement We report here on the conduct and outcome of a sys- has been reflected in the increase in both the number of tematic review of evidence for impacts on human well- PAs and the area of land and sea placed under protec- being arising from the establishment and maintenance tion. The proportion of total area of land under some of terrestrial PAs. Establishing the state of the evidence form of protection has now reached nearly 13% [1,2]. base through systematic review will inform decision mak- The process of protecting areas from the threats posed ing concerning future investment in PAs and future re- by human activity will by definition inhibit some human search needs. The review question was formulated by the actions and therefore has the potential to negatively impact Scientific and Technical Advisory Panel (STAP) of the human well-being. There are many historical records to Global Environment Facility (GEF). suggest that few PAs were uninhabited wildernesses before At the outset of the review the following broad cat- designation, and the early history of PAs, for example in egories were posed by STAP as a starting point for a the USA and East Africa, is of forced eviction and persecu- conceptual model concerning the potential impact of PA tion of local communities by colonial powers [3]. This sce- establishment (or change in PA status) on people and/or nario has continued in some countries with, in some cases, the communities of which they are part. These categor- the colonial powers being replaced by multinational corpo- ies and associated questions were used to guide develop- rations or international conservation Non-Governmental ment of specific evidence inclusion criteria (categories Organisations (NGOs) [4]. The problem of negative im- were subsequently modified based on an in-depth under- pacts of PAs on human well-being gained official recog- standing of the literature in order to code and present nition in the Convention on Biological Diversity (CBD) the available evidence, see methods). arising from the Rio Summit in 1992. The principle that PAs should do no harm to local people was only estab- Livelihood strategies lished at the World’s Park Congress in 2003 (at which, the Did the establishment or change in status of the PA or Durban Accord was proclaimed). Prior to 1992, the estab- management activities within the PA generate or decrease lishment plans of PAs did not normally have objectives specific production opportunities (e.g. more demand for concerning human well-being. labour, herding activities and associated products no longer However, negative impacts could be balanced by posi- viable, new demand for particular food, handicraft, services tive impacts as PAs may also improve human well-being or products etc.)? Did the PA influence (i.e. increase or de- and alleviate poverty [5]. By preventing conversion of crease) migration generally, and of particular social groups? natural habitats, PAs may improve the provision of some Has this differentially impacted (positively or negatively) valued ecosystem services to some users. For example, the most vulnerable groups in local communities (e.g. downstream farmers may benefit from conservation of women, children, poorest sectors of the community)? forested watersheds [6]. PAs may also directly introduce new livelihood options into a region through the expan- Social capital sion of tourism or research, or improvements to infra- Did the establishment and management of the PA affect structure may indirectly result in economic development. the development of social networks? Did it positively or Recently there has been considerable debate on whether, negatively impact education and capacity building, e.g. apart from their effects on global environmental benefits, by generating or decreasing opportunities for formal and/ the net impact of PAs on human well-being at local or re- or informal education? Has PA establishment differentially gional scales are positive or negative [7-10]. There is con- affected more vulnerable groups (e.g. women, children, cern that continuing with a policy of PA establishment poorest sectors within local communities) in a positive or could conflict with goals of poverty alleviation [11]. The negative way? CBD Aichi targets include a target of 17% of terrestrial and inland areas covered by well-managed PAs by 2020. Empowerment Future policy decisions on support for PA establishment Did the PA empower or disempower local communities and future management need to be informed by the best and any particular social groups? Were new organizations/ Pullin et al. Environmental Evidence 2013, 2:19 Page 3 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 institutional arrangements that represent the interests In this review we also recognise that the impacts of of communities and any particular social groups created PAs on human well-being are likely to be highly context or existing ones undermined? Have these organizations dependent and vary over the lifetime of the PA. For ex- developed activities aimed at improving their livelihoods ample, initial establishment may have significant and im- (e.g., legislation to support local livelihoods, land tenure, mediate negative impacts on the communities within the co-management of local resources, other social benefits) PA but in the longer term there may be positive impacts or have existing activities been negatively affected? on well-being. Similarly, spatial context of the PA and communities within and around it will influence impact. Human rights Some communities will be located in a PA, some in buf- Whilst recognising that the scope of human rights is very fer zones around the edge and others more distant. broad, in this review we focus on the following question; Other PAs may be close by and exert an influence. Pres- were the rights of any local stakeholders, affected either ence or absence of infrastructure may influence the positively or negatively by the PA (considering, e.g., rights spatial scale over which impacts occur. Impacts that are to education, adequate access to food, clothing, health, negative ‘locally’ may have to be balanced against im- choices)? pacts that are positive but experienced more widely or more remotely. Access to ecosystem goods and services and natural Given the diversity of PAs, contexts of establishment resources essential for well-being and of communities affected, this review did not aim to Did the PA have any positive or negative impact on ac- estimate a mean effect of PA establishment on local cess to ecosystem services and natural resources? For communities but to critically appraise the range of ef- example were there changes in the cost (in terms of fects that have been reported. The nature and scope of money, level of effort, or time) in obtaining firewood, the question posed in this review (including as it does is- clean water, and other resources/services? Was access sues of empowerment and social capital) suggested to to culturally significant places (e.g. sacred grounds) af- the Stakeholders and the Review Team that a combin- fected? Did self-sufficiency in food (by locally cultivating, ation of quantitative and qualitative evidence might pro- hunting, raising animals or gathering) or access to medi- vide the most reliable base with which to inform future cinal plants change? Has this been a consequence of the decision making. direct impact of the PA through legal prohibition of ac- cess or indirect as a consequence of changes in infra- Objective of the review structure and/or institutions? Have any of these positive The primary research question was: or negative impacts been disproportionably high or low What are the human well-being impacts of terrestrial on particular sectors of society? protected areas? Measuring impacts needs to take into account that ter- We aimed to synthesise the empirical evidence of posi- restrial PAs are varied in their status, management and tive, negative or neutral impacts of PAs on human well- objectives. For example, many protected areas will have being at the local to regional scales, with emphasis on been established with a primary aim of landscape or bio- local communities and contemplating as broad a defin- diversity conservation, possibly at a time when potential ition of well-being as possible (see categories to be con- impacts on local human well-being were not widely con- sidered below). sidered. Types of PAs have been categorised by the World We were also interested in two secondary questions; Commission on Protected Areas (WCPA) of the Inter- 1. How are costs and benefits distributed among and national Union for the Conservation of Nature (IUCN) within local communities living inside and in the buffer and are used to classify entries in the World Database of zones of PAs (by socio-economic status, gender, age etc.)? Protected Areas (http://www.protectedplanet.net/). The PA 2. How do costs and benefits vary with governance, re- categories reflect the range of management objectives of source tenure arrangements, and site characteristics? As different PAs, from strict protection of nature or wilder- the aim is to assess the human well-being impacts of PAs ness, conservation of ecosystems, protection of national in their current form and capture lessons for future inter- monuments or management of important habitat or spe- ventions, relevant evidence may be provided by studies cies (categories I-IV) through to objectives which place hu- that generate hypotheses about impact, as well as studies man use of the landscape much more centrally (categories that test hypotheses about impact. V-VI). Progressively, since 1992 the requirement of PAs to ‘Deliver benefits to resident and local communities consis- Methods tent with the other objectives of management’ has become Design of review a common objective [12] with the categorisation represent- An a priori protocol was established, peer reviewed and ing a gradation of increasing human intervention. posted on the CEE website by CEE [13]. This review was Pullin et al. Environmental Evidence 2013, 2:19 Page 4 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 divided into two separate processes; a qualitative synthe- Where studies are reported in other languages, rele- sis of people’s views, observations and related documen- vance was assessed initially from their titles and ab- tary evidence (led by the EPPI-Centre) and a synthesis of stracts (translated if necessary). Non-English language quantitative evidence of impact, including people’s atti- articles that could not be considered in this review were tudes and views (led by Bangor University). Henceforth recorded for future assessment (see Additional file 2). these will be referred to as ‘qualitative’ and ‘quantitative’ For the qualitative synthesis, non-English papers were as reflected in the protocol and the methods below. A sin- translated using Google Translate online translation tool. gle systematic search was conducted to identify potentially Google Translate was tested for accuracy using a sample relevant studies to answer the review questions. Initial report that included both and English and non-English screening of title was undertaken by Bangor University. abstracts. By translating the non-English abstract of the Further supplementary searching, screening, coding and paper and comparing with the English abstract given in synthesis were conducted independently by each of the re- the same paper we were able to see how well the tool view teams. Thus, the searching phase and initial screen- ‘understood’ and conveyed the same meanings in the text. ing by title were not separate, but thereafter the report Google Translate was found to be a fairly accurate tool, al- sections reflect the different approaches. though more accurate on some languages than others. Using Google Translate, non-English articles were trans- Searches lated then screened on full text and included or excluded. Search terms were selected to capture all information pertaining to PAs (the intervention) and relevant outcomes Estimating the comprehensiveness of the database search associated with human well-being. No search terms for the All search terms were included in a search string for- study population, in this case local communities, were matted according to requirements for searching in the used as these were not likely to be included in the title or Web of Knowledge database. A set of 18 references were abstract and therefore their incorporation in the search identified by the review team as being relevant to the might have risked excluding relevant studies. The use of review question and used as a ‘test library’ to check the asterisk denotes a wildcard character that prevents whether the search strings captured the expected stud- plurals or alternative word endings being excluded. The ies. The test library did not consist entirely of studies search terms are as follows: that would be included at every stage of the systematic review and data extraction. Instead, there were some Exposure: protected area*, nature reserve*, wilderness studies which were expected to be captured in the area*, national park*, natural monument*, natural search but did not include a suitable intervention, or feature*, management area*, world heritage site*, were not primary literature. Whilst it was not required biosphere reserve*, biodiversity conservation. for the database search to capture these articles, it was a Outcomes: poverty, human well*, socio-econom*, valuable test of how specific the search strings were to econom*, human health, livelihood*, social capital, the subject area whilst retaining the sensitivity required social welfare, empowerment, equity, ecosystem to capture all available literature. The balance between service*, perception*, attitude*. specificity and sensitivity is key to achieving a comprehen- sive search whilst avoiding capturing too many spurious Where the databases searched could accommodate all hits. The search string employed (see Additional file 1), search terms simultaneously, they were separated using consisting of the terms listed above, was the last of 16 iter- Boolean operators. The search terms are grouped by inter- ations and was arrived at using the test library to evaluate vention and outcome and we refer to these groups as ‘sets’. the search results returned by Web of Knowledge. The Boolean operator ‘OR’ was used to separate search terms within a set and the operator ‘AND’ used to separate Publication databases searched the two sets thus ensuring that the search returned refer- The search included the following computerized databases: ences including at least one term from each set. In many cases, the literature databases could not ac-  Web of Knowledge commodate all the search terms above and therefore  Scopus search strings had to be adapted as necessary. In some  Agricola cases, this meant using pairs of search terms, one from  CAB Abstracts each set, separated by ‘AND’. Occasionally, search terms  PubMed had to be entered individually and/or the database did not  Econlit accommodate Boolean operators, in these cases only the  Directory of Open Access Journals (DOAJ) search terms pertaining to the intervention were used. Full  LILACS (Latin American & Caribbean Health details of the search are recorded in Additional file 1. Sciences Literature – Spanish language) Pullin et al. Environmental Evidence 2013, 2:19 Page 5 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 A supplemental search was undertaken using an ex- level where not already included in the above database panded set of intervention and outcome terms to test and website searches. how many extra relevant articles they might provide. The following intervention and outcome terms were en- Search update tered into Web of Knowledge, and articles obtained were In May 2013 an update to the searches was undertaken assessed as above. to supplement the review with all relevant evidence pub- lished since the original search was carried out in October/ Exposure: ecotourism, eco-tourism, eco tourism, November 2011. wildlife tourism, trophy hunting, conservation The original search string (October 2011) was com- corridor*, community conservanc* bined with the supplemental search string (March 2012) Outcomes: natural resource, farm, enterprise*, human and entered into the database Web of Knowledge. Details AND migration, gender, discriminat* of this search string and the number of hits returned can be found in Additional file 1. Internet searches conducted Internet searches using online search engines were not Screening performed as part of this review. It was felt that the Inclusion criteria non-transparent and transient nature of search engine The following inclusion criteria were applied to those ar- functionality would result in an unacceptable lack of rep- ticles captured by the search. licability. We therefore focussed on an extensive search of specialist sources (see next section) to identify grey litera- Populations Human populations/communities currently ture and reduce possible publication bias. or previously living in or near terrestrial PAs. In order to avoid subjectivity in deciding what is ‘local’, studies were Specialist searches included if the subjects were any type of community A list of thirty-three relevant stakeholder organisations within the same country as the PA(s) that formed the was identified and their websites searched by members focus of the paper. of the qualitative and quantitative review teams and/or contacted by email with a request for relevant literature Exposure Establishment/implementation, presence or (Table 1). change in status of terrestrial PAs areas with IUCN clas- Websites were searched using a hierarchical approach, sifications I-VI as defined in the World Database of Pro- from the original search string down to individual words. tected Areas. Specifically, studies were included only Where this was not possible, the following individual where the collection of data on the impacts of PAs was terms were searched; national park*, protected area*, and undertaken during or after 1992. This included changes reserve*. Boolean operators and wildcards were used in status of PAs. The impacts of individual PAs were where possible. All returns were assessed except where treated independently where possible but in some cases the searches resulted in large numbers of results (i.e. > 100) studies looked at multiple PAs in such a way as to prevent when the first 50 returns were scanned for relevance. In the separation of impact. Projects that were established each case, results were assessed at title and then full text, within or surrounding PAs as a result of the existence of for relevance. Full texts were visually scanned for relevant that PA were also included. data, along with within-document searches for the follow- ing terms; park, protect, and reserve. Due to the disparate Types of study Studies were considered for inclusion in modes of operation of the website search engines, a wide this review if they satisfied one of the following criteria: range of approaches was necessary and is documented in Additional file 3. a) Evaluate the impact of PAs on human well-being (outcome evaluations which use the following study Bibliographic searches designs with appropriate comparators (i.e. a Where studies identified in the search used data re- comparable state with which the intervention or ported in earlier primary literature, the original reference exposure can be compared): Randomised was sought and included in the data extraction process. Controlled Trials (RCTs) / Controlled Trials (CTs), These references were catalogued in a separate Endnote control-intervention site comparisons, interrupted library. Five key reviews were identified from the above time series, Before-After/Control-Intervention searches. The references within the five relevant reviews (BACI) designs); were examined to identify pertinent articles for inclusion b) Reported economic valuation of welfare changes in in the review (for details, see Additional file 4). These monetary terms: based on stated or revealed references were assessed at title, abstract, and full text preferences or production function approaches, in Pullin et al. Environmental Evidence 2013, 2:19 Page 6 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Table 1 Relevant stakeholder organisations and their websites searched for relevant literature and contacted with a request for information Organisation Web Site Email Asociación Campesina e Indígena de Agroforestería http://www.acicafoc.org/ Yes Comunitaria Centroamericana Conservation International http://www.conservation.org Yes Cultural Survival www.culturalsurvival.org/ Yes Department for International Development http://www.dfid.gov.uk/ Yes Development Experience Clearinghouse (DEC) http://dec.usaid.gov/index.cfm Yes EcoTerra www.ecoterra.net/ Yes Eldis (Livelihoods Connect) http://www.livelihoods.org Yes Environment Knowledge Hub http://ekh.unep.org/ No European Tropical Forest Research Network http://www.etfrn.org/etfrn/index.html Yes First Peoples Worldwide www.firstpeoplesworldwide.org/ Yes Food and Agriculture Organisation http://fao.org Yes Forest Peoples Programme http://www.forestpeoples.org/ Yes GEF Evaluation Office http://www.thegef.org/gef/gef_Documents_Publications Yes GEF Small Grants Programme http://sgp.undp.org/ No Indigenous Knowledge http://indigenousknowledge.org/ Yes International Fund for Agricultural Development http://www.ifad.org/ Yes International Institute for Environmental Development http://www.iied.org Yes International Union for Conservation of Nature http://www.iucn.org Yes Nature Valuation and Financing Network http://www.naturevaluation.org Yes Overseas Development Institute http://www.odi.org.uk/ Yes Pacific Forestry Centre http://www.pfc.cfs.nrcan.gc.ca/ Yes Poverty and Conservation http://povertyandconservation.info Yes Poverty-Well-being Platform www.poverty-well-being.net Yes RECOFTC – The Center for People and Forests http://www.recoftc.org/site/ Yes Rights and Resources Initiative http://www.rightsandresources.org/ Yes Survival International www.survivalinternational.org/ Yes Tropenbos International http://www.tropenbos.org/ Yes United Nations Development Programme http://www.undp.org/ Yes United Nations Environment Programme http://www.unep.org/ Yes United States Agency International Development http://www.usaid.gov/ Yes Waldbau-Institut, University of Freiburg http://www.waldbau.uni-freiburg.de/ Yes World Bank http://web.worldbank.org Yes (http://documents.worldbank.org/curated/en/home) World Conservation Monitoring Centre http://www.unep-wcmc.org/ Yes which the comparator may be modelled and/or methods for which were adequately reported (such as implicit in survey responses (hypothetical); national census data). c) Sought to identify PA factors that influence human well-being or Additional inclusion criteria for quantitative evidence d) Sought evidence to support explanations or meaning Comparators Comparators were classified as temporal, of impact from people’s views about PAs and human spatial or modelled/hypothetical. Temporal comparators in- well-being, and observations or related documentary cluded time series, before and after, change over time (i.e. analysis single time point), and reported/perceived changes. Spatial comparators included dichotomous (i.e. inside/outside and We included studies that obtained data through direct near/far) and continuous (i.e. linear distance) variables. measurement, self-reported measures by respondents, BACI comparators included both spatial and temporal com- and other data sources that were not accessible, but the parisons. Studies may account for confounding variability Pullin et al. Environmental Evidence 2013, 2:19 Page 7 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 between comparator and exposure populations by using contained evidence of people’s views and/or quantitative matching techniques, testing for differences in confounding evidence of impact as described in the following sections. variables, or including these variables in statistical models. Whilst economic valuation studies may lack before- Screening articles for qualitative evidence after or matched site comparators, economic valuations Following preliminary screening by title (3.3.2) articles are inherently relative to a counter-factual (which may be were screened for evidence of people’sviews of PAsbyab- modelled or subjective). Not all of these inherent compar- stract and then full text. Any articles missing an abstract ators are suitable for this review, however. Many studies were accepted for full text screening. The inclusion and purport to value a PA. However, on closer inspection, they exclusion criteria were re-applied to the full texts (70.3% are often valuing the benefits of the ecosystem-protected were obtainable) and those that did not meet inclusion area complex as a single entity, and may present no infor- criteria were excluded. A coding tool was developed in mation with which to estimate the value of the ecosystem order to ascertain what relevant information was held in the absence of the PA as an appropriate comparator. within articles (see Additional file 5). ‘EPPI-Reviewer’ These studies were rejected on the grounds that they lack (Thomas et al. 2010) software was used for screening an appropriate comparator (the comparator is usually the using a single web location to house the documents and complete and sudden non-existence of the ecosystem, monitor progress of the review. Rejected articles and their which is deemed inappropriate for the purposes of this re- corresponding reasons are listed in Additional file 6. view). Our minimum inclusion criteria are that the study The systematic search and screen identified many studies makes some attempt to model what would happen to the relevant to the review question. The studies identified cov- ecosystem in the absence of the PA, for example, estimat- ered a broad range of geographical areas, types of PAs, and ing rates of degradation before the PA was established and sampled a diverse population group using a variety of extrapolating these into the future to calculate that portion qualitative research methods. The characterisation of stud- of the total value of the site which is attributable to the ies by methodological and contextual features provided the PA. Other economic valuation studies may value oppor- starting point for deciding which studies to include in the tunity costs of the PA, and in these instances the instant- qualitative synthesis (see section Studies included in the aneous loss of access to resources following establishment qualitative synthesis for further details). of the PA constitutes an appropriate comparator. Descriptive information about included studies was collected and presented as a ‘systematic map’ of research Outcomes Specific human-well-being indicators linked to into explanations and meaning of impact to people living those broad questions set out in the Background section in or near PAs. The ‘map’ provided a basis for informed (e.g. Livelihood strategies, social capital, empowerment, discussion and decision making between both review human rights, access to ecosystem services and natural re- teams about the focus of the qualitative synthesis which sources). Examples are measures of income, education, provides a detailed investigation of a more focused subset health and other mainstream socio-economic indicators; of this wider literature. As the synthesis was focused and World Bank and United Nations Development Programme narrowed down, a second set of inclusion criteria was de- (UNDP) human development index measures and mea- veloped and applied to the studies initially identified. sures that consider health, longevity, education, gender Understanding perceptions of the impact of protected equity, food security, livelihood diversity, subjective/re- areas requires appropriate research methods and full ported measures of well-being, resilience, measures of so- reporting of context. Further, understanding the mean- cial capital, and indicators of human rights. ing of protected areas for well-being requires qualitative data collection methods that allow people to express Additional exclusion criteria for qualitative evidence their views freely rather than merely responding to pre- Studies of people’s views were excluded if they: focused determined categories, and where the analysis provides solely on the development or validation of a measurement ‘thick’ or ‘rich’ findings. ‘Thick’ findings have been defined tool without also presenting views separately from the val- as a rich, detailed description of specifics (as opposed to idation of the tool, or reported trials or other outcome summary, standardization, generalization, of variables); “it evaluations, unless it was clear from the abstract that they captures the sense of what occurred and the drama of collected data about views as part of a process evaluation. events, thereby permitting multiple interpretations” [14]. We therefore selected for in-depth review those studies Preliminary screening process using qualitative research methods which elicited the Articles captured by the searches were stored in an End- views of people living in or near protected areas where: note library. Replicates were removed and titles examined for relevance to the inclusion criteria. Potentially relevant  The IUCN category, and the date this was assigned, titles were then separately screened as to whether they was known Pullin et al. Environmental Evidence 2013, 2:19 Page 8 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 The methods and date of data collection were near PAs. Framework synthesis does this because dimen- clearly reported sions of a framework for structuring the synthesis can be The analysis produced ‘thick’ or ‘rich’ findings identified by policy interests, research interests and con- Findings were linked to specific people (e.g. cepts emerging from the data [16]. Framework synthesis distinguishing ethnicities, employments or locations) also takes into account differences in context such as the different countries hosting PAs, the different cat- Screening articles for quantitative evidence egories of PA and the different times and stages in their Following preliminary screening by title (3.3.2) articles history that each area was studied. Based on framework were screened for quantitative evidence of impact by ab- analysis of primary research data [17], an initial concep- stract and then full text. Any articles missing an abstract tual framework, either built from assumptions held by were accepted for full text screening. Where information stakeholders or borrowed from related bodies of know- in titles or abstracts was insufficient, articles were ac- ledge, evolves during the synthesis as the reviewers be- cepted for screening at full text. come more familiar with the literature being reviewed A kappa test for consistency of decision regarding in- [16]. The five neat stages of framework analysis described clusion/exclusion at abstract level returned a moderate for primary research [18] are in practice a more iterative level of agreement (n = 100, kappa = 0.442, 95% C.I. 0.270 process for making sense of a pile of studies [19]. The to 0.614) between two independent reviewers (SED and process starts by delving into the abstracts then full texts ASP). Twenty-eight studies were treated differently by the to see the key issues and recurrent themes, some of which reviewers and these were re-read and the reasons for in- emerge from the data while others are purposely sought clusion or exclusion by each reviewer discussed. These in response to the review question or prior knowledge. discussions were used to come to consensus on the inter- Once most of the key themes are identified, each study is pretation of the titles and abstracts and informed the sub- coded with themes and the literature is sub-divided into sequent inclusion process. sections for in-depth analysis. The conceptual framework Full text screening resulted in the rejection of articles is refined in light of these themes, and coding continues it- that were not pertinent to the review in hand. Reasons for eratively but systematically as the framework becomes in- exclusion were recorded (Additional file 6). Relevant re- creasingly coherent and simultaneously accommodating views were identified for use in the bibliographic searching of the available data. Subsections of the framework shape phase described in Section 3.2. data tables under key themes in order to develop concise We attempted to obtain all articles deemed relevant summaries. Conclusions are drawn from the themes and during abstract-level screening, for assessment at full associations between them. text. Those articles that we failed to obtain within the limitations of time and resources of the review are listed Data extraction strategy in Additional file 2. Included studies were inspected in detail for the focus of their findings. The coding framework described above Qualitative synthesis of explanations and meaning (Additional file 5) was applied to each study. This frame- of impact work covered: health (mental health, physical health, safety There are three distinctive approaches to synthesising and wildlife conflict); socio-economic position (livelihood findings from research: aggregating the findings of very strategies, social capital, human rights, empowerment, en- similar studies where the key concepts are clearly defined vironmental capital, governance); inequalities; and context. in advance; configuring the findings of dissimilar studies Study text was extracted in the form of: participant data by investigating the implications of the differences in their (e.g. direct quotes); authors’ descriptions of findings; and methods, context and findings in order to define key con- authors’ conclusions, implementations and recommenda- cepts and develop theoretical understanding; and a com- tions. The coding framework was tested by two or more bination of the two which can be applied where studies reviewers independently coding successive studies, com- have important differences, but nevertheless share some paring and discussing how they applied the codes, and re- important key concepts of well-being [15]. fining the framework and definitions of codes until a We chose this third approach because some (but not shared understanding was achieved across the team. all) key concepts were decided and defined in advance: PAs with their clearly defined categories; populations Synthesis and outcomes of interest. Despite these concepts being Synthesis began by developing the initial key concepts identified by the funders, a synthesis method was re- into a coherent framework that could accommodate the quired that also allows new concepts to emerge from the main approaches to establishing and maintaining PAs. data in order for the impact on human well-being to be Figure 1 illustrates two contrasting approaches to terres- understood from the perspective of people living in or trial PAs. The first is governance models that are imposed Pullin et al. Environmental Evidence 2013, 2:19 Page 9 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Figure 1 Conceptual framework for structuring the qualitative synthesis. and enforced by external authorities and the second is of these three areas. Understanding the context of each participatory approaches such as community-based nat- study and the interplay between different concepts required ural resource management. The latter has been developed reviewers to return to the full report of each study, along- in response to tensions raised by restrictions and resettle- side the text extracted for each code, and distil the key ment having a deleterious impact on economic and envi- study characteristics and findings. The resulting coherent ronmental capital (agriculture, logging, tourism) and social text was placed appropriately within the emerging frame- capital (family and community relationships). In contrast, work (Figure 1). Where studies contributed substantial participatory approaches, within a regulatory framework, findings to different elements of the framework, these find- seek a vision of sustainability through building on social ings were distributed across the framework to allow each capital and good health. That vision of sustainability may element to draw on all relevant studies. As the framework be on a small scale, such as with IUCN category VI PAs, became more coherent with growing understanding of the where the aim is sustainability within the boundaries. Al- relationships between its different elements, the elements ternatively it may be on a wider scale, across IUCN cat- were reordered to present governance issues first, and then egories I or II and the wider area. These models are set well-being issues. The findings were summarised for each against a backdrop of an evolving consensus about human element of the framework, taking into account the govern- rights that began with protecting individual civil and polit- ance and time context of the studies. The contexts of the ical rights, followed by the support for economic, social studies contributing to the synthesis were recorded noting: and cultural rights and then setting these rights within a the location (name and country of the protected area); broader framework to harness the combined efforts of in- IUCN category (present category, years assigned and chan- dividuals, states and other bodies to build collective rights ged); governance (national government, local government, to self determination, heritage and equity [20]. This syn- private or community); and timing of study (year data was thesis aims to assess the extent to which these models and collected and the time lapse since the last assignment of their anticipated impacts are supported by perceptions of IUCN category). Finally, all studies were revisited for add- impact on human well-being held by people living in or itional evidence that appeared significant in light of emer- near protected areas and others working alongside them. ging findings. The framework was populated with research findings by dividing the studies into three overlapping subsets whose Synthesis of quantitative evidence findings clearly addressed: participation (human rights, em- Data extraction strategy powerment and governance), health (physical, mental and A preliminary data extraction phase was carried out dur- safety) and socio-economic position (wildlife conflict, liveli- ing full text screening, where each reported study was hood strategies, social capital and environmental capital). categorised in terms of which broad outcome types were Each set was taken by a different member of the team to represented in the quantitative data. These outcomes seed the synthesis with studies focussing on findings in one were categorised according to the coding framework Pullin et al. Environmental Evidence 2013, 2:19 Page 10 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 described in (and in detail in Additional file 5). Further- minimum possible value of 0 and a maximum of 8 more, details about the following were ascertained; the (Table 2). The four criteria were as follows: broad study methodology (i.e. self-reported data from close-ended questioning in interviews and questionnaires, i) Clarity of methods; the extent of detail given in the direct data collection, valuation or data taken from other study’s methodology regarding questionnaire design, sources such as national census documents and similar), survey implementation, replicate number and and whether there was a valid comparator. selection, and data analysis Comparator categories (i.e. no comparator, before-after, ii) Study design; the rigour of study design in terms of inside-outside, before-after-control-impacts, econometric) replicate number and sampling, location and choice were used to group the studies for a second stage of de- of replicates and controls, questionnaire design, and tailed data extraction. Only those studies that possessed survey implementation an appropriate comparator were taken on to this second iii)Appropriateness of analysis; the suitability of stage of data extraction, where details of the study meth- analytical techniques such as the implementation odology, potential effect modifiers or reasons for hetero- and choice of statistics, and comprehensiveness of geneity and relevant outcome data were obtained. For all analysis outcomes, averages (means/medians) and variability mea- iv) Implementation bias; the presence of one of the sures (standard deviation, standard error and confidence following biases in measurement and analysis; intervals) were obtained where possible, along with statis- Neyman bias (the intervention in question causes tical test results and sums of counts (i.e. percentages of re- systematic drop-out within the population, e.g. the spondents in agreement). most affected families move away), questioning bias (unbalanced/leading questions), potentially influential variables, recall bias (response affected by Study quality assessment memory). Details relating to study quality were extracted for each study to allow critical appraisal of relevance (external Susceptibility to bias scores were then combined with validity) with respect to the review question and reliabil- an objective weighting by basic study design in order to ity (internal validity). Critical appraisal was conducted in provide a categorical assessment (high, medium or low) two stages. Initially assessment was made as to whether of susceptibility to bias (Table 3). a study possessed one or more of the following: con- founding variables, within-study contradictory results, Potential effect modifiers and reasons for heterogeneity inappropriate implicit comparator, inappropriate popula- Data on potential effect modifiers that were extracted tion, inappropriate spatial comparator, inappropriate sta- from articles, included the following variables; protected tistics, inappropriate temporal comparator, inappropriate area studied, country, IUCN category, protected area time frame, insufficient detail in methods, data linked with size, date of establishment, and history of protection marine ecosystem protection, non-standardised data (in- prior to formal establishment, and residence history and comparable), protected area establishment too recent/not ethnicity of the study population. Where sufficient infor- yet established, data collection prior to 1992, extremely mation about IUCN category, protected area size and small sample size (e.g. one study, where wood extraction year of establishment was not presented in the article it- was quantified in only nine households with one interven- self, data were supplemented from the World Database tion and one control village), superseded by more recent on Protected Areas (WDPA) (http://www.protectedpla- study, unbalanced questioning/questioning bias, and un- net.net/). matched methods (see Additional file 6 for further detail). Studies with one or more of these characteristics were ex- Data synthesis and presentation cluded from the review. Extracted data were presented in narrative synthesis ta- Studies accepted following this first critical appraisal bles that summarise the studies’ aims, methodology, stage were then scored for reliability with respect to the study population, and results (Additional file 7). No fur- question using the concept of susceptibility to bias (the ther quantitative synthesis of data on outcomes was pos- extent to which a study is able to yield an unbiased esti- sible with the available data. mate of effect; [21]). Susceptibility to bias in study design and reporting can affect both internal validity (study qual- Results ity) and external validity (study generalisability). Review descriptive statistics Susceptibility to bias scores were assigned using a re- The main literature search was conducted between 11 peatable, but partially subjective, set of four criteria, each October and 14 November 2011 with an update carried assessed on a scale of 0 to 2, and thus ranging from a out in May 2013. Pullin et al. Environmental Evidence 2013, 2:19 Page 11 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Table 2 Scoring criteria for subjective assessment of susceptibility to bias Assessment criteria 0 1 2 Clarity of methods All sample sizes provided, selection Some information regarding sample Sample size not given in full, sample selection method described, questionnaire size, selection method, or not stated, questionnaire design not discussed design fully disclosed questionnaire design provided Experimental design Good sample size, appropriate sampling Low effective sample size, poorly Very small sample size, pseudoreplication, regime, control and intervention well randomised design, control and non-random sample selection, control and matched, survey appropriately intervention choice not ideal intervention poorly matched/low implemented consistency in sampling over time Appropriateness of Confounding variables accounted for, Confounders only partly Significant confounding variables unaccounted analysis appropriate metrics reported accounted for/only some low for, inappropriate metrics reported, incorrect risk confounders ignored statistical analysis Implementation bias No identifiable bias reported/evident Low level bias present but ignored/ Strong bias present and unaccounted for strong bias accounted for The main search returned 13,248 articles (following unobtainable within the timeline and resource limita- removal of duplicate and marine-oriented articles) and tions of the review. These numbers are represented visu- title-level screening left a total of 1,710 articles. The ally in Figure 3. One hundred and fifty-seven of the search update returned 3,122 articles (following removal articles identified in the May 2013 search update were of duplicates) and title-level screening left a total of 204 obtainable, although five of these were in Chinese and additional articles. From this point forwards, screening could not be assessed (see Additional file 2). at abstract was independent for the two review sections. Following full text assessment 177 articles from the original search, 16 from the supplemental search and 17 Qualitative evidence from the update met the inclusion criteria and were sub- Abstract screening for the qualitative review resulted in sequently critically appraised. Following critical appraisal, the final inclusion of 30 studies as illustrated in Figure 2. articles/studies were rejected at full text for a wide range Four additional studies were included in an identical of reasons (see section Study quality assessment). A list of process following the search update in May 2013. these articles and associated reasons is provided in Add- itional file 6. A final set of 70 articles was included, report- Quantitative evidence ing on 79 studies and these are listed in Additional file 7. Abstract screening for the quantitative review resulted in the final inclusion of 79 studies as illustrated in Figure 3. Comparison of included studies from the two In addition to the above searches 2,345 articles were review processes identified by the supplemental search conducted on 23 The following statistics describe the 30 studies on quali- March 2012: 535 remained following title-level screen- tative evidence of people’s views and 79 on quantitative ing, and 171 following abstract-level screening. Twelve evidence of impact included in the synthesis. studies from this supplemental search were included fol- lowing critical appraisal. Study location Figure 4 displays the countries from Relevant organisation website searches conducted be- which data were collected for the included studies from tween 6 and 29 March 2012 yielded 94 relevant articles the two processes. Frequently studied countries include for the quantitative review. In addition to these searches, Cameroon, China, India, and Uganda for the qualitative 50 relevant articles were identified through bibliographic synthesis and India, Nepal, South Africa and Uganda for checks and secondary sources, yielding a total of 1,164 the ‘quantitative impacts’ review. potentially relevant articles. Of these articles, 76% were retrievable for full text assessment, whilst 275 were Category of PA studied The IUCN categories of the PAs (see below) examined in the included studies are Table 3 Categorization of susceptibility to bias shown in Figure 5. PAs predominantly fell under cat- Susceptibility to bias score egory II in both reviews, with 17 percent either unre- ported by WDPA or not present in the database for the Comparator 0-2 3-5 6-8 ‘quantitative impacts’ review (NB this group includes RCT* (randomised control trials) Low Low Medium only ‘not applicable’ for the qualitative synthesis, since BACI (before-after-control-intervention) Low Medium High studies were only included in this review if the protected Control-intervention/Before-after/ Medium High High area IUCN category was stated). A post-hoc decision to Econometric exclude category III PAs was made as they are small in *There were no RCT studies and this line is only provided for a theoretical perspective number and somewhat atypical of the set of categories Pullin et al. Environmental Evidence 2013, 2:19 Page 12 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Figure 2 Number of articles and studies at progressive stages of inclusion for qualitative synthesis. (specifically protecting natural monuments). Similar pat- Ib Wilderness Area -generally larger than Strict terns were observed in both review processes Figure 6 Nature Reserves, the main objectives of these areas displays PA establishment year, showing a peak in new is to provide an environment in which biodiversity PAs centered around the 1980s. and ecosystem processes (including evolution) are IUCN Protected Area Categories are described as follows: allowed to flourish or experience restoration if previously disturbed by human activity. Human use Ia Strict Nature Reserve - protected from all but light is limited, often allowing only those who are willing human use in order to preserve all geological and to travel of their own accord rather than via geomorphological features of the region and their established touristic activities. Wilderness areas can biodiversity, which is often dense and restricted to be classified as such only if they are devoid of exclusively to scientific monitoring, study or modern infrastructure, although they allow human education. Occasionally Strict Nature Reserves are activity to the level of sustaining indigenous groups of spiritual significance to surrounding communities living wilderness-based lifestyles. in which case the people are generally allowed to II National Park - provide protection for functioning continue the practice of their faith and may be ecosystems (similar to the objectives of Wilderness directly involved in the area's conservation and Areas), but tend to be more lenient with human management objectives, though perpetual human visitation and the supporting infrastructure. National intervention would more suitably be allocated to Parks are managed in a way that may contribute to categories IV or V. local economies through promoting educational and Pullin et al. Environmental Evidence 2013, 2:19 Page 13 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Figure 3 Number of articles and studies at progressive stages of inclusion for quantitative review. recreational tourism on a scale that will not reduce an identifiable species or habitat that requires the effectiveness of conservation efforts. The continuous protection. These protected areas will be surrounding areas of a National Park may be for sufficiently controlled to ensure the maintenance, consumptive or non-consumptive use, but should conservation and restoration of particular species nevertheless act as a barrier for the defence of the and habitats - possibly through traditional means - protected area's native species and communities to and public education of such areas is widely enable them to remain sustainable in the long term. encouraged as part of the management objectives. III Natural Monument or Feature - comparatively Habitat or Species Management Areas may exist as smaller areas, specifically allocated to protect a a fraction of a wider ecosystem or protected area natural monument and its surrounding habitats. and may require varying levels of active intervention Natural Monuments or Features can be natural in including - but not limited to - the prevention of the wholest sense, or include elements that have poaching, creation of artificial habitats, halting been influenced or introduced by humans. The latter natural succession and supplementary feeding should hold biodiversity associations or could practices. otherwise be classified as a historical or spiritual site, VI Protected Landscape/Seascape - cover entire though this distinction can be quite difficult to bodies of land or ocean which engage a range of for- ascertain. As such, the classification then falls into profit activities within the management plan. The two subcategories, those in which the biodiversity in main objective is to safeguard regions that have built uniquely related to the conditions of the natural up a 'distinct character' in regards to their feature, and those in which the current levels of ecological, biological, cultural or scenic value. biodiversity are dependent on the presence of the Protected Landscapes and Seascapes allow a higher sacred sites that have created an essentially modified level of sustainable interaction with -surrounding ecosystem. communities (such as traditional agricultural and IV Habitat/Species Management Area - focus on forestry systems) and should represent an integral more specific areas of conservation in correlation to balance between people and nature. Pullin et al. Environmental Evidence 2013, 2:19 Page 14 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Australia Peru Ecuador South America Quantitative Review Brazil Bolivia Qualitative Review United Stated of America Mexico North America Costa Rica Belize Sweden Slovakia Europe Norway Italy Greece Czech Republic Vietnam Thailand Pakistan Nepal Myanmar Asia Mongolia Indonesia India China Bhutan Bangladesh Zimbabwe Zambia Uganda Tanzania South Africa Nigeria Namibia Africa Mozambique Malawi Madagascar Kenya Ghana Gabon Ethiopia Cameroon Botswana 02468 10 Figure 4 Frequency of countries, arranged by continent, hosting the protected areas within included studies for the two reviews. VI Protected area with sustainable use of natural be developed to adapt to the diverse and possibly resources - based on a mutually beneficial growing range of interests that arise from the relationship between nature conservation and the production of sustainable natural resources. sustainable management of natural resources in (Text adapted from United Nations Environment correspondence the livelihoods of surrounding Programme (UNEP) World Conservation communities. A wide range of socio-economic Monitoring Centre (WCMC) website (http://www. factors are taken into consideration in creating local, unep-wcmc.org/iucn-protected-area-management- regional and national approaches to the use of categories_591.html; accessed 03/10/2012) natural resources. Though human involvement is a large factor in the management of these protected Study timescale Figure 7 displays the survey years for areas, developments are not intended to allow for the included studies for both reviews. A significant num- widescale industrial production. Governance has to ber of studies failed to report the survey year (applicable Pullin et al. Environmental Evidence 2013, 2:19 Page 15 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Quantitative Review Qualitative Review Ia Ib II III IV V VI Unknown/ not reported/ not applicable Figure 5 Frequency of all named PAs by IUCN category within included studies. only to the ‘quantitative impacts’ review, since this was Measured outcome Within the five key themes described an exclusion criterion for the qualitative synthesis). The in Section 1 (Livelihood Strategies, Human Rights, Social reduction in the number of surveys from 2005–2006 for Capital, Empowerment, and Access to Ecosystem Goods both reviews likely relates to a publishing delay. and Services), 24 individual outcomes were identified in an iterative process during screening. The reported outcomes were separated in order to facilitate the description and Further descriptive statistics for the review of quantitative analysis of similar, comparable data. The frequency of these evidence of impacts broad outcome types is shown in Figure 9. Attitude, in- Study comparator Figure 8 displays the range and fre- come, and wildlife conflict outcomes were most common. quency of comparators used in included studies. Spatial Exclusion of studies following critical appraisal resulted in comparators (site comparators and linear distance) were several outcomes identified during full title assessment be- common, along with reported change over time. ‘Before ing unrepresented by data namely, ‘empowerment’ and ‘reli- and after, control intervention’ (BACI) studies were very gion and spirituality’. Other outcomes are represented in rare, with only one instance of a full BACI study. Quantitative Review Qualitative Review Figure 6 Frequency of year of establishment of PAs within included studies. Pullin et al. Environmental Evidence 2013, 2:19 Page 16 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Quantitative Review Qualitative Review Figure 7 Frequency of year of PA survey for included studies. Surveys commence on stated year and are classified as ‘longitudinal study’ if survey spans more than two years. two studies or fewer; ‘gender equity’, ‘interactions with PA Qualitative synthesis of explanations and meaning of impact authorities’, ‘livelihood diversity’, ‘medicinal plants/animals’, Studies included in the qualitative synthesis ‘protect for future’,and ‘resettlement and displacement’. The criteria for selecting studies to include in the in-depth qualitative synthesis is provided in section Screening arti- cles for qualitative evidence. Although not statistically rep- Methodology of data collection Figure 10 displays the resentative of the 306 studies of people’s views identified frequency of different data sources in included studies, (characterised in the map see Additional file 8), studies showing that self reported measures formed the majority with ‘thick’ data were chosen for their ability to provide ex- of data reported (63 of the 79 included studies). planations and meaning of impact for people living in or Figure 8 Frequency of comparators used in included studies. Pullin et al. Environmental Evidence 2013, 2:19 Page 17 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Wildlife Conflict Attitude Income Econometric Food Environment quality Human development measure Employment Water resources Materials Health Education Resettlement and displacement Land access Infrastructure Gender equity Food/Materials 02468 10 12 14 16 Figure 9 Frequency of broad outcome types in included studies. near the PA. They also spanned the range of the broader The other stakeholders not well represented in the literature, in terms of countries frequently studied, as de- studies presenting ‘thick’ findings are visitors to pro- scribed in the map. However, poor reporting makes it diffi- tected areas. Studies seeking visitors’ views often did so cult to conduct further comparisons between the studies using highly structured questionnaires that provided lit- reviewed in-depth with the wider map in terms of IUCN tle understanding of what protected areas mean to them. categories and key dates. The one study of visitors included in the synthesis was The vast majority of included studies were by academic set in a category IV protected area, so was not typical of authors, with a small minority from NGOs, commercial most visitor surveys which, in the map, were most often organisations and protected area authorities. These differ- used in category II. ent perspectives are likely to shape the focus of studies This literature is thus a likely source for understanding and, consequently, their findings. people’s views, and setting those views in context, across Residents’ views were most often sought for IUCN cat- the IUCN categories, although less so for category IV. egory II, where there are strict restrictions on settlements, but rarely for IUCN category VI, where the aim is to bal- Summary findings of qualitative synthesis of explanations ance the aims of conservation and the needs of the local and meaning of impact population. Leaders’ views were most often sought about This section presents a summary of a full synthesis of IUCN category I protected areas, which does not allow findings from studies that provide explanations or ex- human settlements. There was also a noticeable absence plore the meaning of the impact of protected areas on of studies addressing the views of protected area author- people living in or near PAs. The full synthesis can be ities or employees in category IV (similar to the map). found in Additional file 9. It also draws on the perspec- tives of other stakeholders relevant to those experiences and relevant policy and historical documents. The syn- thesis is presented within a conceptual framework that was informed by conservation policies and refined by the emerging research literature (Figure 1). The synthesis process identified the implementation issues and characteristics of well-being that were associ- ated by research participants with either (a) tensions aris- ing from governance models imposed and enforced by external authorities or (b) a vision of sustainability sought through participatory management and empowerment 0 commonly known as community-based natural resource SRM ODS DDC DDC/ODS management. Figure 10 Frequency of data sources in included studies. DDC, The latter approach has been developed in response direct data collection by study authors; ODS, other data sources; to tensions raised by restrictions and resettlement having SRM, self-reported measure via survey responses. a deleterious impact on economic and environmental Pullin et al. Environmental Evidence 2013, 2:19 Page 18 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 (natural) capital and social capital. Participatory approaches, rules and local institutions and typically poor within a regulatory framework, seek a vision of sustainabil- communication ity through building on social capital and good health. That  Memories of forced or induced migrations vision of sustainability may be on a small scale, such as with negatively influence subsequent community IUCN category VI PAs, where the aim is sustainability responses to authorities within the boundaries. Alternatively it may be on a wider  Inadequate or non-existent compensation is a widely scale, across IUCN categories I or II and the wider area. held concern These models are set against a backdrop of an evolving  Externally imposed regulations were incompatible consensus about human rights that began with protecting with traditional regulations and did not take into individual civil and political rights, followed by the support account cultural and social diversity; respect for the for economic, social and cultural rights and then setting rules was greater where they were locally adapted these rights within a broader framework to harness the and allowed income-generating possibilities. combined efforts of individuals, states and other bodies to  A major challenge is distinguishing subsistence build collective rights to self determination, heritage and activities for a sustainable environment from larger equity [20]. This synthesis aims to assess the extent to scale industrial activities. which these models and their anticipated impacts are sup-  Success came from PAs staff having: prior ported by the perceptions of people living in or near pro- experience of working with locals; clear guidelines; tected areas and others working alongside them. extensive training in community development, The characteristics of the included studies are sum- gender issues and a variety of participatory marised in Table 4, and their findings are subsequently methodologies; meeting locals informally and synthesised in sections that match the conceptual frame- working with existing kinship networks. work (see Figure 11) to present a coherent narrative  All studies contributed some evidence related to which explores key themes within governance (source of governance. authority and nature of implementation) and then con- siders the impact on health and well-being in terms of Implementation problems The earliest source of ten- environmental (natural), economic and social capital. sion in the life course of protected areas can be contro- There is necessarily some overlap between the govern- versy about environmental risks posed by local inhabitants. ance issues and the impact on health and well-being. Studies in Australia [22], Nepal [23], Norway [24], Mexico The key messages for both are presented in boxes to [25], Indonesia [26] and USA [27] revealed local inhabi- help readers navigate the breadth and depth of the avail- tants appreciating areas for their aesthetic and spiritual able evidence. The contextual details of each study de- values as well as environmental products and economic scribed in terms of IUCN Category, the date the PA was and leisure opportunities. Mexican farmers in particular established, date of current PA status being assigned, value the land for its provision of food, water, wood and time between current status being assigned and data collec- other products, and they have developed farming styles tion are reported in Additional file 9. As the complexity of along a spectrum of reciprocal relationships between man the links between governance and well-being precludes and nature between wilderness and urbanisation [28]. Diaw simple distinctions between PA approaches and their im- [29] claims that a resettlement policy to establish a IUCN pacts, we focus first on governance and then on well-being, II category park in Cameroon in 1961 was driven by scien- acknowledging that there is necessarily some overlap. The tificmyths of apristineforestwhose protection wasin- key messages for both are presented in boxes to help compatible with indigenous residents despite historical readers navigate the breadth and depth of the available analysis showing that the current forest structure was the evidence. result of sustained use over centuries. Economic arguments favouring resettlement were flawed, with excessively strong Governance assumptions about tourism benefits, flood control, forest Matching study findings to the conceptual framework use, research discoveries, soil fertility and agricultural prod- facilitated their comparison and interpretation to reveal uctivity. Similarly, in Norway residents pointed to a lack of problems and potential solutions during implementation convincing scientific evidence supporting the need for cer- and instances of both benefit and harm. tain protective measures [24]. Residents of Utah claimed the land benefitted from how they had cared for it before it Portraying longstanding residents as an environmental was declared a protected area (Grand Staircase Escalante risk to pristine protected areas is controversial National Monument, IUCN II [27]). In Norway, local tour- PA regulations can flounder as a result of inaccurate ist firms were frustrated by the lack of opportunities to maps and poorly drafted legislation, and further playapart inthe protectedareamanagement(IUCN II). confusion arises from discrepancies between state They claimed that local expertise, based on generations of Pullin et al. Environmental Evidence 2013, 2:19 Page 19 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Table 4 Summary table of included studies in the qualitative synthesis Citation Protected area IUCN Date PA established Date of current Years since IUCN Year of data Governance model Country Category IUCN category category change/ collection establishment Allendorf et al. (2007) Royal Bardia NP II 1969 1976 14 1990 Government: federal/ Nepal national Almudi and Berkes (2010) Peixe Lagoon National Park V 1986 2001-2010 3 2005 & 2007 Cooperatively managed: Brazil collaboratively Bedunah and Schmidt Gobi Gurvansaikhan II 1993 1993 5-7 1998 – 2000 Information not found Mongolia (2004) Bizikova et al. (2012) Slovensky Raj Park II 1998 1998 8 2006 Government: federal/ Slovakia national Bolaane (2004) Moremi Game Reserve IV 1965 1965 32 1997 – 2001 Private Botwana Bruyere et al. (2009) Buffalo Springs Samburu II 1985 1985 20 2005 Government: local Kenya Castillo et al. (2005) El Vizcaíno IV 1988 1988 16 2004–2007 Government: federal/ Mexico national Sian Ka’an 2006 2008 3 Monarch Butterfly 2008 < 1 Davis (2011) Tarangire National Park II 1970 1970 35 2005 – 2007 Government: local Tanzania Diaw (2010) Korup National Park II 1961 1985 18 2003 Government: federal/ Cameroon national First Peoples (2006) Mgahinga National Park II 1930 1930 Not stated Government: federal/ Uganda national Bwindi National Park 1991 1991 14 2005 Government: federal/ Uganda national Awa Forest Reserve Zone VI 1988 1988 10-20 1998 – 2008 Community: indigenous Ecuador Gerritsen (2002) Sierra De Manantlan VI 1987 1987 6 – 11 1993 – 1998 Government: federal/ Mexico Biosphere Reserve national Hartter (2009) Kibale national park II 1932 1993 16 Uganda Haukeland (2011) Rondane National Park II 1962 1962 47 2009 Norway Jotunheimen National Park II 1980 1980 18 Government: national Herrold-Menzies (2011) Caohai Nature Reserve V Protected 1985 1985 13-16 1998-1999, 2000, Government China landscape/ 2000-2001 managed: local Seascape Hoole and Berkes (2010) Etosha National park II 1975 1975 31 2006 - 2007 Government Namibia Keskitalo and Lundmark Sarek National Park; Ib 1909 1909 95 years 2001 Government: federal/ Sweden (2010) national Stora Sjöfallet 1909 1909 95 years National Park; Abisko II 1920 1920 83 years National Park; Pieljekaise 1962 1982 22 years National Park; Vadvetjåkka Pullin et al. Environmental Evidence 2013, 2:19 Page 20 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Table 4 Summary table of included studies in the qualitative synthesis (Continued) National Park; Padjelanta National Park Haparanda-Sandskär Nature Reserve Lunstrum (2008) Limpopo National Park II 1979 1979 25 2004 – 2005 Cooperative: transboundary/ Mozabique collaborative Mbaiwa (2005) Moremi Game Reserve IV 1965 1965 38 1998, 2001, Private: for profit Botswana Mehring et al. (2011) Lore Lindu National Park II 1982 1982 24 2006 – 2008 Government: federal/ Indonesia national Milgroom and Limpopo National Park II 2001 2001 6 – 7 2007 – 2008 Cooperative: transboundary/ Mozabique Sperienburg (2008) collaborative Nguiffo (2001) Dja Wildlife Reserve IV 1950 1950 51 2001 Government: federal/ Cameroon national Ogra (2008) Rajaji National Park II 1983 1983 20 2003 – 2004 Information not found India Ormsby and Kaplan (2005) Masoala National park II 1997 1997 4 2001 Private/ non-profit Madagascar Petrzelka and Marquart- Grand Staircase V 1996 1996 0 1996 Government: federal/ USA Pyatt (2013) national Escalante National 10 2006 Monument Slater (2002) Qwaqwa National Park IV 1992 1992 6 – 7 1998 – 1999 Government: local South Africa Sletten et al. (2008) Mount Elgon NP II 1951 1951 47 2002 Government: federal/ Uganda national + collaborative Spenceley and Kruger National Park II 1926 1930 74 – 75 2000 - 2001 Information not found South Africa Goodwin (2005) Songorwa (1999) Selous Game Reserve IV 1905 1905 91 1996 Community: local Tanzania Stone and Wall (2004) Jianfengling National Forest V 1986 1986 16 years 2001 Government: local China Park and Diaoluoshan National Forest Park in Hainan Province, China Strickland-Munro and Purnululu National Park II 2003 2003 5 years 2008 Government: federal/ Australia Moore (2012) national + collaborative Stronza and Gordillo (2008) Kapawi Reserve; Madidi II Not stated Kapawi Reserve, 1991-2000 Not able to ascertain 2003 Cooperatively managed Bolivia National Park; Bahuaja Ecuador 1995: Madidi Ecuador 5 years:, Bolivia; Collaborative management Ecuador Peru Sonene National National Park 2000: Bahuaja 8 years:, Peru (various forms of Sonene National Park, Peru pluralist influence) Torri (2011) Sariska Tiger Reserve IV 1955 1955 46 2001 (pilot) Government: federal/ India 2007 national Yasuda (2011) Benoue National Park II 1968 1968 40 2004, 2009 Private: for profit Cameroon Pullin et al. Environmental Evidence 2013, 2:19 Page 21 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Figure 11 Summary findings from the qualitative synthesis. managing the area prior to its official status, should play an the definition, might outlaw common traditional practices important role in management [24]. such as the use of snares (metal wire), arrows (steel tipped) The second source of tension is the lack of clarity or rifles. The ban and uncertainty surrounding poorly de- in regulations and maps pertaining to protected areas. fined traditional hunting led to tension and mistrust be- Mehring et al. [26] investigated regulatory institutions in tween locals and conservation agents. In Uganda, the legal two villages in an Indonesian park established in 1982. agreement protecting the Mount Elgon National Park, In one, new regulations about forest land and products established in 1951, was flawed as it failed to refer accur- drawn up by the mayor and customary organisation were ately to maps or related by-laws, statutes or other docu- neither written down, nor completely implemented. There ments [31]. was support for state zoning of the Park to allow trad- Where regulations precluded living within an area, re- itional access to the forest for local people, but disagree- settlement could be forced or induced. In Cameroon, ment about the zone boundaries. Effective village sanctions forced migration and a violent confrontation prompted were considered important, but confusion about when to villagers to accept resettlement outside familiar territories, apply them appropriately arose from discrepancies between against the recommendations of earlier research [29]. state rules and local institutions. In Cameroon, ‘traditional Enacting laws to drive resettlement resulted in an inte- hunting’ was still allowed in 2001 the territories outside grated conservation and development plan that failed, protected areas (IUCN IV) so long as the products were leaving villagers bitter and sceptical. Expulsion to make for personal use, and not sold [30]. However, whether way for the privately managed Moremi Game Reserve in ‘traditional’ hunting referred to the people involved, the Botswana included huts being burnt down as residents weapons employed, or some other characteristic was not got loaded into trucks for relocation outside the reserve clear. Only allowing ‘traditional’ weapons, depending on [32]. Residents were forced to relocate (for a second time) Pullin et al. Environmental Evidence 2013, 2:19 Page 22 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 by the suspension of the provision of all social services but informed later (without being able to give a precise such as water supply, health facilities, shops, schools and date) of the existence of a conservation initiative by the communication services. authorities [30]. Although labelled as ‘voluntary,’ because the term ‘invol- Such problems are not restricted to developing coun- untary’ is politically problematic nationally and amongst tries. In Norway, several residents near a IUCN II category international donors, the resettlement from Limpopo Na- area found the process one-sided and undemocratic be- tional Park was widely recognised as ‘induced’ by planning cause national interests took precedence over local know- blight and economic decline [33]. Although consulted ledge [24]. Petrzelka and Marquart-Pyatt [27] describe the about resettlement, residents’ views were then disregarded. growing anger of residents and their diminishing trust in agencies to make good decisions about the management “Since the park was made we were supposed to leave. of the land after the Grand Staircase Escalante National Since they said that, people don’t construct houses, Monument was established in Utah, USA, in 1996 with no we don’t plant trees. This house was built in 2000 but prior consultation or publicity. Trust diminished further it was never really finished because the park came. over the next ten years later as residents saw roads closed There were trees but we stopped planting and the old and cattlemen’s leases rescinded, despite prior reassur- ones died. [papaya]. No one is investing, not to do ances to the contrary. Restrictions on visiting the park things for nothing. Even now that we have accepted to stoked anger amongst residents who felt ‘locked out of leave, the park does nothing” (p443). our backyard’,saying that ‘law enforcement is gun-toting like we’re a bunch of criminals’. Inadequate or non-existent compensation was a con- Even with management of Kenyan national reserves cern expressed in many studies; for loss of property or being delegated to a local level, and rangers and wardens land in 1960s [34], for access or use restrictions in India claiming to initiate and maintain dialogue, residents in 2007 [35], for environmental protection by owners of are disappointed with the processes of communication Swedish forests (IUCN Ib and II) in 2001 [36], for re- [40]. Most of the protected area staff considered their settlement in 2001 or for loss of crops or livestock in informal word of mouth network sufficient for com- 2007/8 Mozambique (IUCN II) [33], for personal injury municating with local communities about important or property damage from wildlife in South Africa (IUCN management decisions. II) in 2001 [37], and in Tanzania (IUCN IV) where there was no compensation policy at the time of the study in “Of course we cannot conserve this wildlife without 1996 [38]; or for loss of jobs or land in China (IUCN V the help of these communities. There must be that, a established 1986) [39]. For this last case, some compen- good relationship between the park and the sation had been made in the form of new homes, crop community. So we normally go to the [homes and seeds, lump sum payments, and subsidized education, villages], we have meetings with them, tell them that electricity, and water fees, but views differed on its na- these resources are also theirs, these are their ture and adequacy [39]. Residents of the Sariska Tiger resources.” (Park ranger, p55) Reserve (India, IUCN IV) were generally discouraged by staff from claiming compensation for restrictions to ac- In contrast, most community members thought that cess or use forest products [35]. Some villagers were communication between the reserves and communities never told they had a right to compensation. Another, was limited or non-existent, where decisions were made on asking for compensation, one villager was told by a without opportunities to provide input or ask questions. forest officer: In addition to the poor communication between au- thorities and residents, were the difficulties encountered “If you the villagers insist in living in the forest, then with communication between residents. The Qwaqwa be ready to accept as well all the consequences National Park (IUCN IV), South Africa had been estab- deriving from your choice. You could live elsewhere”. lished for the purpose of ecotourism; however, this did not suit the livelihoods of stockholding families and Opposition to the Selous Game Reserve in Tanzania others would have preferred the land to be subdivided grew amongst those losing crops and livestock to wildlife for agriculture. Amongst all this disagreement, some res- without compensation [38]. idents were more able than others to make their voices Poor communication between communities and au- heard, and family conflicts escalated as housing became thorities was typical. Residents of the Dja wildlife reserve more crowded because erecting new homes within the in Cameroon (IUCN IV, established 1951) reported be- park was forbidden [41]. In Lore Lindu National Park, ing “neither informed of nor invited to participate” Indonesia, the weak point for communication was be- (p.208) in their village becoming part of a protected area tween villagers and their leaders [26]. Although the Pullin et al. Environmental Evidence 2013, 2:19 Page 23 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 village leadership was active in the negotiations about activities for a sustainable environment from larger scale the park regulations many ordinary villagers had never industrial activities. As mentioned above, this challenge heard of the agreements. was seen in Cameroon where regulations failed to distin- In Australia, at the Purnululu National Park (IUCN II) guish clearly ‘traditional’ hunting methods for personal relationships between different indigenous groups were consumption from commercial hunting [30]. In Masoala so acrimonious that one group withdrew from the man- National park (IUCN II), Madagascar, residents acknowl- agement of the park [22]. In Slovakia, local authorities edged that some members of the community benefiting near Slovensky Raj Park (IUCN II) tried to make man- from illegal lemur hunting and timber harvest [43]: agement of the park a focus for building relationships and developing mutual trust between different groups ‘people [who] wanted easy money, especially the youth, [42]. Communication problems could be compounded so they went into the park to cut rosewood’ [p160]. by new regulatory arrangements being incompatible with traditional ways. For instance, very few Mexican farmers However, much greater damage was done by industrial applied for resource use permits because the formal bio- scale rosewood logging for international markets, and sphere (IUCN VI) rules competed with customary rules Park agents have limited legal powers over loggers. [28]. The formal rules were generic and did not take into account local variation in natural resource management. “people from all over come to this area to cut These mismatches created feelings of frustration: rosewood, there is no other way to get money than from valuable wood” (Park resident p.160). “The reserve is like a beautiful woman whom you cannot touch. It does not do you any good. The hills ‘[international] demand is driving the outside buyers are rich, but a poor man stays poor” [28] p205. of rosewood, and this is a much bigger issue than lemur hunting.’ (Park Manager, p.162). Contrary to tradition, only people living on the borders of Mount Elgon National Park (IUCN II) in Uganda were Similarly in Cameroon, the impact of residents hunt- given rights of access [31]. The new outsiders were required ing in Dja Wildlife Reserve (IUCN IV) to ensure a diet to pay the ‘insiders’ for access, even though half the insiders that includes animal protein is minor compared with in- thought outsiders should have equal rights of access. tensive industrial logging which opened up forest tracks Even where access was allowed, as in Permululu Na- and thereby provided access for well organised, commer- tional Park, Australia, costs of transportation across long cial poachers to use the tracks for transporting their distances over rough ground could be prohibitive [22]. game to city markets [30] p.208. In Indonesia Mehring et al. [26] attributed some of the Ironically, it was a combined forestry management and difficulties of incompatibility to the government’s indif- community development project in Ecuador that opened ference to cultural and social diversities when managing the eyes of indigenous people to the potential benefits of the Lore Lindu National Park (IUCN II). Indigenous logging; when profits were not what they had hoped for, people respected their own traditional informal rules they started making deals outside the community with that suited traditional use rights and sanctions at the village industrial loggers [44]. level. Elsewhere, more prosperous and ethnically diverse vil- Implementing regulations that have disadvantages for lagers, growing more cash crops, referred not to traditional local communities is challenging enough. The relation- institutions but to economic power structures, where there ship between residents and park officials in Masoala Na- was a widely spread laissez-faire attitude to resource use. tional park officials in Madagascar (IUCN II) was further With forest resources and agricultural land in short supply, damaged by absenteeism amongst staff who, unlike many locals, had the privilege of employment yet lacked villagers had no alternative to using the Park to extend their land. The State’s formal rules interacted with traditional in- training and clear job expectations, and had little inter- formal rules, leading to confusion and conflict. Migrants action with residents [43]. struggled to implement traditional informal rules, and indi- genous people failed to obey state-induced laws. Implementation solutions Blunt regulations imposed Traditional land ownership rights for indigenous popula- by external authorities have been widely disregarded so tions were also contested in Australia, both between local that protected areas have continued to be exploited on residents and protected area management, and amongst domestic and industrial scales. Studies have focused on local residents of different groups living near Purnululu efforts to improve communication, draw on indigenous National Park (IUCN II) [22]. knowledge and share decisions to combine community A major challenge to developing and implementing development with environmental conservation. They have regulations to protect areas is distinguishing subsistence had mixed success. Pullin et al. Environmental Evidence 2013, 2:19 Page 24 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 The Lore Lindu area in Indonesia was established as a wealth etc. Contrary to tradition, only people living on UNESCO Biosphere reserve in 1977 and a national park the park borders were given rights of access. The new (IUCN II) in 1993. Since then participatory approaches outsiders were required to pay the ‘insiders’ for access, have been advocated for managing Biospheres [45] and even though half the insiders thought outsiders should protected areas more widely [46]. Initial efforts to im- have equal rights of access. Conflicts arose from this pose external regulations failed and in the late 1990s, the situation and threatened the agreement’s endurance. In park authority, NGOs and village representatives began such sensitive situations, staff need the socio-cultural skills to negotiate Community Conservation Agreements [26]. to understand, interpret and interact with local people Within designated zones, village conservation councils about livelihoods, conflicts and challenges in appropriate were the bridge between the Park authority and the com- ways. Reports of misuse and corruption remained com- munity for planning, implementing, evaluating and report- mon. Nevertheless, collaborative arrangements improved ing the results of the Agreement. Despite the village relations and benefited biodiversity and livelihoods. leadership being active in the negotiations, communication Sletten’s findings in Mount Elgon Park (IUCN II) are between the Park authorities and the whole community supported by other studies. Elsewhere in Uganda sup- was poor, so many ordinary villagers had never heard of porting community’s transition from a hunter gatherer the agreements. The Agreements covered use of forest to a settled farming community in a culturally sensitive products and land and the village conservation councils way was more likely to result in community satisfaction were responsible for monitoring activities. The council and personal efficacy [44]. Training and capacity build- could employ punishments or sanctions, which were usu- ing by charities and NGOs led to an increase in skills ally based on village traditional rules. Insights into this and knowledge and new income generating activities. system came from NGO interviewees. A collaborative Two NGOs working with local people helped to organ- management approach aimed to minimise the gap between ise efforts around existing kinship networks and this the park management and the people, through participa- community reported the highest states of economic de- tion of local inhabitants and integration of local rules. Re- velopment compared to other communities. At the other spect for the rules was greater where they were ‘more end of the scale these communities were willing to sacri- practical’ having been locally adapted, and allowed income- fice their land claims to join relatives in other areas and generating possibilities. access charitable projects there leaving the settlements In Mount Elgon National Park, Uganda (IUCN II), as struggling to maintain a viable community. in other African countries, a similar ‘fortress management’ In Masoala National Park (IUCN II), Madagascar, resi- or ‘fence and fine policy’, based on systematic evictions, dents who were more familiar with Park staff viewed the exclusions and prohibition of using natural resources, met staff as well as the Park more favourably than residents increasing resistance [31]. Lack of success with ‘fence and who were unaware of staff or who had had negative inter- fine’ policies prompted approaches with greater partici- actions with Park agents [43]. Residents were confused by pation of local people in management and changes in reg- the different NGOs’ responsibilities and changing prior- ulations to legitimise sustainable use. Establishing the ities. A park manager and a local town official both consid- agreements was difficult even with guidelines and training ered community development as essential for maintaining for park staff; converting staff from law enforcers to com- a protected area. There was local support for protecting munity collaborative workers was difficult. Nevertheless, the park by providing community benefits through alter- meeting locals and getting to know them improved rela- native livelihoods. However, it is unclear from this whether tions. Some local people acknowledged that their initial the benefits essential for behaviour change were the intan- reluctance lessened as they met staff and learnt more gible empowerment benefits of community development, about the resource base; while a third of respondents did or the material benefits. not participate at all. In Selous Conservation Programme (IUCN category IV), Once established, the agreement provided greater clar- Tanzania, support from communities was greatest in areas ity over rights and duties, and opportunities for long where education and mobilisation campaigns had been term planning about livelihood strategies. However, as a conducted and benefits were beginning to be derived; find- legal document, the agreement was flawed as it failed to ings suggest that the majority of villagers supported the refer accurately to maps or related by-laws, statutes or project. The evidence showed that they were motivated to other documents [31]. Subsequently people were more join the conservation programme by promises of socioeco- positive towards the park, its resources and staff. How- nomic benefits [38]. ever, as the focus was on the park rather than the com- The arrival of western donors and NGOs in Caohai munity, people were sometimes organised according to Nature Reserve (IUCN V), China, in 1993 changed the what resources they collected, rather than by other so- focus from enforcement of resource regulations towards cially relevant criteria such as ethnicity, kinship, location, small-scale community development and outreach Pullin et al. Environmental Evidence 2013, 2:19 Page 25 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 programmes [47]. These included small grants and a IV and VI; and both before and after the Durban micro-credit programme for farmers to start up micro- Accord enterprises in the hope that they would be less reliant  Communities expressed a tension between on the reserve’s natural resources, infrastructure devel- appreciating the environment and wanting to opment, environmental education, a community based protect it, and also needing to make immediate use natural resource management programme, and school of land or natural products fees for girls from poor families. This involved two em-  They could be encouraged to participate in further ployees who had extensive prior experience of working conservation measures where they could anticipate with farmers, and required extensive training in com- socio-economic benefits munity development, gender issues and a variety of par- Evidence synthesised from seven studies ticipatory methodologies. The result was many fewer [23,25,26,28,29,38,43] hostile confrontations between local people and nature reserve managers, the participation of local people in Even where residents have recognised that conserving conservation activities and farmers contrasting the a park (IUCN II) and its wildlife is valuable on an indi- nature reserve’s concern for local people with the indif- vidual, local, national and global scale for economic, ference of corruption of other government agencies. educational, recreational, aesthetic and environmental Farmers now work cooperatively with the reserve to reasons, and for future generations, they still lament the seek resolutions to their own problems, sometimes tak- economic limitations imposed by restrictions on access, ing the initiative to raise issues about road construction, extracting resources and grazing, and the dangers of wild sanitation improvements, and agro-forestry projects. animals [23]. Indeed, some communities refuted the The transformation from conflict to cooperation has need for resettlement, having managed the land (IUCN been dependent on funds from NGOs and donors, II) for centuries; this was confirmed by the authors’ his- which raises questions about the project’s sustainability. torical analysis and portrayal of a pristine forest whose Another successful example of cooperative manage- protection was incompatible with indigenous residents ment was on the margins of a category II park in as a scientific myth [29]. Mozambique where land values increased exponentially. A range of Mexican farming styles have developed along With the support of an NGO residents thrived, benefit- the spectrum of reciprocal relationships between man and ting materially from land titles, revenues and empow- nature (co-production) between wilderness and urbanisa- ered by the process of acquiring land titles and setting tion, with farmers valuing the land for its provision of boundaries [48]. food, water, wood and other products [28]. Despite appre- Participatory approaches to governance were not al- ciating protected areas for their aesthetic, environmental ways successful. Almudi and Berkes [49] investigated the products, economic opportunities and spiritual values relationship between a local fishing community and offi- [25,28] their positive attitudes were lost when conserva- cials responsible for the creation and maintenance of tion regulations (IUCN IV) competed with productive ac- Brazil’s Peixe Lagoon National Park. They took a par- tivities such as cattle ranching or growing imported ticular interest in the factors that could empower local varieties of fruit, or with personal safety [25]. fishers to ‘defend their rights to remain physically within Responses to plans for balancing conservation and the park and politically in the conservation policy process’ economic development appear to vary depending on (p.220). The authors also found that fisher communities where the benefits might be felt. For some, it was not an struggled to participate in discussions essential to securing interest in conserving wildlife that motivated participation their ‘long-term access to the resources for their livelihoods in a Community-based Wildlife Management Programme or to trigger the development of a PA co-management ar- (IUCN IV) but promises of socioeconomic benefits to rangement’ (p.225). The following quotes were provided as themselves [38]. Whereas other respondents were critical examples of the fishers experiences: of a programme for promoting economic activities along- The authors summarised two of the main barriers con- side maintaining ecosystems; here tourism was expected tributing to the fishers’ lack of empowerment as: weak to benefit entrepreneurs and rich family owners of coastal assistance for developing community organisational cap- lands with tourism potential [25]. acity and leadership; and lack of basic knowledge on laws and fisher rights. Access to land PAs are important to communities for grazing, Well-being agriculture, hunting, foraging and spiritual homes Environmental (natural) capital  Relocation and loss of control over land and Appreciation of protected areas other than for resources can result in resentment, poaching and economic benefit was found in IUCN categories II, antagonism Pullin et al. Environmental Evidence 2013, 2:19 Page 26 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Participating in the process of setting boundaries of a balanced diet without animal protein and protected and securing land rights can be empowering areas were still seen as the ideal place to carry out hunt- Evidence synthesised from nine studies ing throughout the year [30]. Where dependence on ac- [30,33,34,40,41,44,48,50,51] from IUCN categories II cess to products was high, collection continued despite a after the Durban Accord, and from categories V and ban as compensation was not always considered ad- IV before the Durban Accord equate [39]. Once introduced to community based management to Access restrictions to protected areas (IUCN II) had harvest, process and sell timber, Ecuadorian communi- implications for grazing cattle, hunting and collecting ties who were disappointed by high start up costs and natural products [33,40,51]. Local people would like graz- slow, small gains made deals with external commercial ing rights in the park (IUCN II), especially during drought, loggers to raise their profits [44]. In contrast, in Mexico, and the opportunity to visit traditional areas and burial where conservation is widely valued, natural products areas [51]. Where staff were mostly concerned about il- were used in a sustainable way [28]. legal grazing; they would commonly impound livestock, and fine owners or refer them to a local judicial ward; in Economic capital contrast most community members felt the harsh envi-  Before the Durban Accord (IUCN Ib and II), ronment justified grazing cattle in protected areas [40]. concerns focused on: reduced employment and tax Communities considered that threats to their environ- revenues; reliance on foreign aid without mental capital outweighed any potential economic bene- understanding its link with conservation; and fits of living near the national park [50]. Outside the Park unrealistic expectations of the economic benefits residents felt they would not have access to resources so of tourism. far essential to their livelihoods: access to agricultural  After the Durban Accord (IUCN II), concerns land, forest resources and grazing land [33]. Wealthy cat- focused on: the meager benefits of tourism; what tle owners were even less disposed to moving because cat- benefits there were not being shared equitably; with tle would need to compete for food and water with host indigenous groups or those less amenable to falling villages, and cattle theft was more common outside of the in line with new regulations missing out. park [33].  Concerns about lack of compensation were Access restrictions posed similar problems for commu- expressed before and after the Durban Accord about nities in or near PAs categorised as IUCN IV. Accessing IUCN categories II and IV, and about IUCN V the forest was important in order to obtain resources to before the Durban Accord. Conversely, in developed support local people’s livelihood and for feelings about the countries there was a growing welfare dependency. forest as ‘theirs’ [30]. Authors described relocation and Evidence of economic capital found in thirteen loss of control over land and resources resulting in resent- studies [23,29,34,36-41,53-56], ment, poaching and antagonism [34] and overcrowding resulted from restrictions on building new homes [41]. Before the Durban Accord, the impact of IUCN II parks More positive views were expressed where land values on the wealth of whole areas was seen in Scandinavia, had increased exponentially on a park’s margins (IUCN South America and Asia. Forest workers in Sweden asso- II). Here residents thrived, benefitting materially from land ciated environmental protection with lower levels of em- titles and revenues and felt empowered by the process of ployment and production in commercial forestry [36]. acquiring land titles and setting boundaries [48]. Politicians anticipated conservation policies leading to lower tax revenue and greater emigration [55]. In Nepal Resource use communities were developing a dependence on foreign Residents appreciated protected areas for their rich aid, with some people considering it as an expected source products of income and not necessarily associating its benefits ac- Resource use was common even where illegal crued with conservation efforts [23]. Evidence synthesised from seven studies Some people living in or near national parks (IUCN II) [28,30,35,39,43,44,52] mostly conducted before the were concerned about neighbours having unrealistic ex- Durban Accord and spanning IUCN categories II, V pectations of the economic benefits derived from tourism and VI. and eco-lodges [55]. Others were unaware that commu- nity development was one of a park’s primary objectives Residents viewed protected areas as rich sources of only initially [43]. food and other products [28,30,39,43,44,52]. They ac- Benefits of tourism to national parks (IUCN II) were knowledged that illegal resource use continued despite seen as meagre, and distribution of revenues from pro- bans [39,43,52]. Some park residents could not conceive tected areas was considered inequitable or of little benefit Pullin et al. Environmental Evidence 2013, 2:19 Page 27 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 to indigenous communities [40]. Park and eco-lodge staff work ethic and less frequent use of their native language tend to be wealthier than community residents [43,55]; [44,55]. Stronza and Gordillo [55] saw changes in social and local people felt overlooked for employment, with op- ties within communities when some began working in the portunities favouring neighbouring communities and those tourism. Locals who secured employment in eco-lodges deemed amenable to new regulations [29]. Similar con- (IUCN II) talked about their work limiting participation in cerns about few or unevenly shared benefits and oppor- gatherings traditionally employed to complete community tunities were expressed in other, IUCN IV, protected areas tasks, and how communities tended to charge eco-lodge [34,41]. employees more because they were richer – it was buying Communities across IUCN categories, before and after them out of their community responsibilities. Working in the Durban Accord, often considered as inadequate the eco-lodges opened villagers’ eyes to new opportunities and monetary or in kind compensation available for: forestry a wider social circle but this was at the cost of leaving their constraint; resettlement; loss of land, crops, livestock or family and community. More direct commitment to con- jobs; or personal injury or property damage [33-39]. servation in a national park (IUCN II) also caused family Environmental protection is associated with economic pressures where the Village Scout scheme took young men decline in high income countries. In Utah, USA, local away from their family and farming responsibilities [38]. residents saw more tourism but no economic growth as Elsewhere (IUCN V), investment in communities was a result of establishing a protected area [27]. In Slovakia, mentioned by park staff and by residents; this included the although residents living near Slovensky Raj National aim of establishing a special university training program to park (IUCN II) anticipated multifunctional forest man- prepare students to fill key park management positions agement as a source of employment and income gener- [39]. ation, in practice the socioeconomic situation worsened, Villagers emphasised the importance of social relations particularly for minority ethnic communities, with re- as part of their survival strategy and expressed appre- duced employment and changes in welfare support [42]. hension about the weakening and possible disappearance With their lack of experience and opportunities for in- of these linkages as a result of displacement. These rela- volvement in small businesses and local or regional plan- tions were particular important in times of struggle such ning, minority ethnic groups saw no viable economic as periods of drought [35 p.60]. Stronza and Gordillo options. [55] considered that communities rich in social capital The most positive findings about social and material may not only be better able to manage changes associ- benefits came from an NGO funded study with one aca- ated with ecotourism, but that such changes in social demic author and one author employed by an ecolodge, capital can collectively sustain local institutions, which although they too shared concerns expressed above [55]. may subsequently be critical of conservation efforts. In contrast, the community managed forests in Ecuador cre- Social capital ated new business relationships and improved the com- Changes in livelihood strategies have influenced the munity’s social standing with other indigenous groups in shape of households and the strength of social ties, the area [44]. and introduced new inequalities within communities. Although ethnic identities could strengthen social cap- PAs have been established in areas inhabited by various ital within groups, they more often created tensions be- ethnic groups. The pressures resulting from regulatory tween groups. Some ethnic groups were resentful as and economic changes have introduced tensions or they perceived others receiving community development exacerbated historical tensions between them. preferentially, even though the authors saw these com- Evidence about social capital in PAs was found in munity development efforts as culturally inappropriate eight studies [30,35,38,39,41,44,55] [44]. As noted above, migrants struggled to implement traditional informal rules, and indigenous people failed Slater [41] noted that households configured themselves to obey state-induced laws. Some of these difficulties in a way so as to maximize livelihood diversification; some- were attributed to the government’s indifference to cul- tural and social diversities when managing the Park [26]. times to the detriment of familial relationships. Households could be separated geographically by livelihoods, or over- At the level of implementing regulations, residents crowded because sharing dwellings allowed younger adults objected to the leniency of guards towards those who were wealthier or ethnically related [23]. to rely on the support of older adults claiming pensions. Residents saw a direct link between livelihood diversi- fication and changes in cultural traditions and traditional Health relationships amongst local people; both positive and nega-  New diseases associated with changes in lifestyles tive [44,55]. Making the transition to a settled agrarian life were attributed to forest evictions or changing from for some meant a decrease in the traditional communal a nomadic to sedentary existence. Pullin et al. Environmental Evidence 2013, 2:19 Page 28 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Accidents and injuries resulted from conflicts day labourers, tourism employees), tenure (landowner or between guards and residents; animal conflicts not), gender, education and wealth. resulted in increased workload and exhaustion as Differences in wealth accrued from ownership of land well as injuries or livestock have been influenced by wildlife conflict Sexual aggression was more common when women [38,54] and by PA regulations about land access or live- were less protected following necessary changes in stock numbers, including whether or not people com- working patterns or kinship ties. plied [41]. Smaller businesses and entrepreneurs less Evidence of health and protected areas was found in able to control shocks may bear the brunt of further re- five studies [35,38,41,44,54] strictions on the forest products [36]. Some of these individual differences have arisen at In one study, community members evicted from forest least in part from inequalities between ethnic groups or unanimously spoke of their exposure to new disease from discrimination between indigenous groups. For in- when integration with other groups began; authors con- stance, some indigenous people have discouraged immigra- firmed that the community, particularly children, were tion and excluded immigrants from community governance seriously affected by malaria which did not exist in the procedures [26]. Within and between ethnic groups, people forest, and that HIV/AIDS is also appearing [44]. Their have taken advantage of others who earn more from the forced transition to an agrarian society has cut them off presence of the PA [55] or from people struggling to make from their access to and knowledge of traditional medi- a living who sell land legitimately owned only to claim pro- cinal plants they previously used to stay healthy. Else- tected land illegally [26]. where, villagers reported the lack of access to basic Misunderstandings or prejudices about the histories or health services and Torri [35] confirmed that child mor- abilities of communities led authorities to pursue deci- tality was high in isolated forest villages, where common sions which those communities found damaging [32,33]. illnesses, easily treatable given basic medical facilities, Authorities also discriminated within communities by could lead to death. exercising policies that protected landowners but not Residents described how crop raiding by elephants lead other longstanding residents [41]; or by involving people to food shortages and greater workloads, especially amongst chosen for their age, wealth, education and position ra- women who suffered more from insect-borne diseases and ther than practical relevant knowledge [31]. Ethnic dis- heat exhaustion. When elephants had damaged water pipes, crimination has been introduced by PA legislation that women risked drowning when collecting water from unsafe forbad resource use by indigenous people but allowed sources [54]. First Peoples [44] reported women’s safety be- resource use by predominantly white landowners [51]. ing compromised as men worked further afield, and as Where ethnic discrimination predates establishment of women were drawn out of their homes for new roles and PAs, historic practices of favouritism have strengthened as that the erosion of kinship ties maybe reducing protection land has become more valuable and rare because much of from male sexual aggression [44]. it has been put aside for conservation [50]. Prejudice and nepotism have been the source of favouritism or corrup- Inequalities tion amongst PA employees [23]. PAs and residents’ responses to them have Some differences have arisen from commercial or con- exacerbated existing local ethnic tensions servation enterprises providing more earning power for Participation in PA governance has favoured people some, but not all, roles [56]. Some eco-tourism enter- already advantaged by their socio-economic position prises have a history of being poor employers of local PAs impact unequally on people depending their casual labour [32]. Financial inequalities were introduced socio-economic position, such as the size of their when spent funds resulted in financial support for some business, their legal land tenure or their gender but not others, and bank loans have been refused on the New tourism enterprises have tended to employ grounds of having an address in a protected area [44]. outsiders rather than locals Women were vulnerable to attack and injury from Evidence of inequalities arising from PAs was found men or wildlife [38,44,54]. Female heads of households in 15 studies [23,26,31-33,35,36,38,41,44,50,51,54-56] were vulnerable financially as they tried to balance paid labour with subsistence farming [41]. More may have People living in and near PAs differ in many ways. Some been learnt if some women had not been reluctant to ex- indigenous people living near each other come from dif- press their opinions to researchers [35]. ferent ethnic groups. Some PAs include indigenous com- munities, new immigrant communities and residents long Narrative synthesis of quantitative evidence established following historical migrations. Individuals dif- In this section we provide a narrative overview of all in- fer in terms of their occupations (e.g. subsistence farmers, cluded studies providing quantitative data on impacts. Pullin et al. Environmental Evidence 2013, 2:19 Page 29 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 The studies are divided into six subsections; produced it- the diversity of outcomes reported by all studies. Further- eratively based on the outcomes reported in the accepted more, we avoid vote-counting, where the sum of all nega- literature (see typology in Figure 12). We do not claim that tive, positive and neutral study results are calculated. these divisions are definitive or optimal but they do provide Vote-counting is unreliable because it assumes that a sig- a pragmatic breakdown of a complex body of evidence. A nificant finding is evidence that an effect is present and a summary of data presented in the included studies is pro- non-significant finding is evidence that an effect is absent. vided as Additional file 7. A more detailed set of data ex- This former statement is true, but the latter is not (see traction tables is available as supplementary material. Of Borenstein et al. [21] for further details of vote-counting). the 79 studies included in this synthesis 63 were catego- rized as having ‘high’,11 ‘medium’ and 3 ‘low’ susceptibility Ecosystem goods to bias. Additional file 10: Table S1 displays the 14 studies and 33 outcome measures that were categorised as having  17 studies identified low and medium susceptibility to bias. Additional file 11  Only two studies not highly susceptible to bias provides detail of critical appraisal and the basis for assign- (‘medium’ susceptibility to bias) ing susceptibility to bias for all 79 studies. In the following  Nyahongo et al. [57] found that meat and fish text we concentrate on results reported in these 14 studies. consumption increased with proximity to Serengeti Here wherewediscuss studies’‘susceptibility to bias’ we National Park refer to their categorisation resulting from our critical ap-  Sarker and Røskaft [58] found residents’ perception praisal. In some instances we also identify specific types of of timber and fuelwood benefits from four PAs in bias (defined in Table 5, below) and in others we refer to Bangladesh decreased with distance from the PA shortcomings of the study design and implementation that boundary implicitly increase the studies’ susceptibility to bias.  High susceptibility to bias in remaining studies It is important to note that since the majority of stud- related to low methodological detail, confounding ies were identified as being highly susceptible to bias, variables, and weak experimental design the results of these studies are not considered further, ei- ther individually or in general. These studies are unreli- Of the 17 studies reporting results on ecosystem able both alone and in concert. Because this group of goods, only two studies had designs that were not highly potentially biased studies is unreliable, we cannot sum- susceptible to bias (i.e. medium). Nyahongo et al. [57] marise their findings any more than we can discuss indi- reported the number of meat and fish meals consumed vidual results. However, for completeness we also consider by survey respondents to be significantly negatively Figure 12 Typology for structuring the quantitative synthesis of impacts of protected areas. Pullin et al. Environmental Evidence 2013, 2:19 Page 30 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Table 5 Definitions of bias mentioned in the narrative synthesis Recall bias Imperfect recollection of past events by respondents. Generally worsened by longer periods of recall Social desirability bias Tendency to respond to questioning in such as way as to be viewed favourably by others Questioning bias Questioner leading respondents to reply to questioning in a certain direction Neyman bias Arises from a time lag between exposure and sampling such that undetected drop-out of participants may occur before the study begins. Attrition bias A skew in results where participants are lost between measurements at two time points (potentially as a result of the exposure) during the study Optimism bias A belief by a respondent that they are less likely to experience a negative event relative to other respondents, or over-optimism on the part of analysts or interviewers, about the effects of a project. Hypothetical bias Failure of respondents to consider the true budget constraints in responding to financial questioning Strategic bias Tendency for respondents to alter their answers in an attempt to influence an event correlated with distance from Serengeti National Park whilst Bajracharya et al. [59], Mehta and Kellert [60], boundary. The authors’ statistics account for a range of Naughton-Treves et al. [61], Rinzin et al. [62], and Okello other potentially influential variables, contributing to the et al. [63] all measured accessibility and quality of water studies favourable susceptibility to bias rating (medium). resources, there was insufficient conformity to allow The study’s results indicate that meat consumption in- meaningful synthesis. creased with proximity to the park at a rate of 0.218 Other studies could not be used to infer reliably any (±0.052) meals per week per km, whilst fish consump- effect of a protected area for a range of reasons, includ- tion increased at a rate of 0.931 (±0.205) meals per week ing; (i) they recorded change over time with no spatial per km (assuming units in analyses are identical to units comparison and no adequate basis for determining whether described in the methodology). Sarker and Røskaft [58] the changes observed were attributable to the effect of the found that people inhabiting the areas surrounding four protected area; (ii) they made a spatial comparison, but lo- protected areas in Bangladesh identified more benefits cation relative to the protected area is clearly confounded from the protected area in the form of timber and fuel- with a range of other important site variables; (iii) the scale wood extraction closer to the park boundaries than fur- of the spatial comparison was too small to use it to infer ef- ther away, with an associated odds ratio of 1,000; i.e. “The fects of the protected area on the specific outcomes re- odds that respondents living closer to the protected area corded in the study; (iv) time elapsed between the creation boundary reported a higher level of benefit from timber of the protected area and the study was too short to infer and firewood because of the conservation programme any effect of the protected area. were 1,000 times greater than those for respondents living further away”. Livelihood strategies Thirteen studies used questionnaires and semi-structured interviews for data collection, resulting in a higher suscepti-  43 studies identified bility to bias since reporting by the respondents can be sub-  Eight studies not highly susceptible to bias (1 ‘low’ ject to recall or social desirability bias. Fourteen studies and 7 ‘medium’ susceptibility to bias) reported 39 different (but not all independent) outcomes  Four studies report poverty-related outcomes, finding related to food and materials, comprising a mixture of beneficial impacts of land protection in all cases foods obtained by hunting and gathering, agriculture and  Wildlife conflict is relatively well-studied (18 studies purchased, as well as other indicators such as land area and 47 outcomes; 5 studies with 7 outcomes under cultivation or grazing, availability of fodder, support ‘medium’ susceptibility to bias. The majority for agricultural development, dietary diversity, gathered reported significantly more problems with proximity fuelwood and other plant products including timber, and to protected areas than further afield also change in ownership of goods. Five studies reported  Lundgren [64] found no significant difference in seven different (but not all independent) outcomes related income growth or forestry/tourism sector to water resources: three studies related predominantly to employment as a result of protected areas in Sweden water quality (e.g. households relying on least safe water re-  Household income [65] and consumption [66] were sources) and two to water availability and supply. Only one found to increase with proximity to/within study reported one outcome related to medicinal plants/ protected areas animals. Studies reporting data on common themes differed sig- Studies in this set report on access to markets, employ- nificantly in the precise outcomes measured. For example, ment, income, livelihood diversity, human development Pullin et al. Environmental Evidence 2013, 2:19 Page 31 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 measures and wildlife conflict (Additional file 7). Wildlife  All studies are highly susceptible to bias due to lack conflict was the most frequently reported outcome meas- of methodological detail, non-random sample ure in this group, contributing 18 of the 43 studies and 47 selection, spillover, questioning bias and of the 101 outcomes. Only five of these 18 studies were uncontrolled confounding variables not highly susceptible to bias, all bar one [52] showing  The majority of outcomes relate to restrictions on significantly greater crop and livestock loss closer to pro- access or extraction tected areas. Two of these five studies employed question-  Two outcomes related to the perception of naires to elicit perceived disbenefits, whilst three studies relationships with park employees used observations of conflict. Lundgren [64] found no sig- nificant correlation between protected areas and income The six studies in this set report on 11 different (but growth or employment in either tourism or forestry sec- not all independent) outcomes related to land and re- tors in Sweden. Sims [66] reports higher consumption in source access, interactions with protected area author- regions with a high ‘share’ of PAs compared to a low ities, displacement and resettlement (Additional file 7). ‘share’ in Thailand. Richardson et al. [65] found house- All had a high risk of bias in their study design because holds within game management areas (GMAs) to have of a lack of detailed description of their methodologies. greater income across a number of Zambian PAs than In cases where the description was detailed, specific risks those outside GMAs. of bias were identified; replicates were non-randomly se- Four studies reported poverty-related measures (hu- lected [59,74] and spillover, questioning bias, and con- man development measure outcomes) as either poverty founding variables were not accounted for [59,75,76]. index or poverty headcount. All four studies found sig- nificant beneficial impacts of protected areas on poverty Health and safety alleviation in Costa Rica [67], Thailand [66,67] and Bolivia [68]. The study by Andam et al. [67] involved subsequent  Nine studies identified in-depth reanalysis in two later publications [8,69]. These  Only one study identified as not highly susceptible later analyses showed that along with protection alleviat- to bias (‘medium’ susceptibility to bias) ing poverty, poorer areas (measured at baseline) were  Korhonen et al. [77] found highly variable infant found to have greatest level of poverty reduction than mortality rates in and around Ramonafana National those that were less poor [8]; and that poverty alleviation Park in Madagascar, with slightly higher levels outside was also associated with characteristics that reduced the than inside the PA, although the pattern is not clear efficacy of deforestation prevention (i.e. where protection had been assigned to land that was unsuitable for agricul- Studies in this section look at health in the population ture, near major cities and infrastructure, and where agri- and access to health services (Additional file 7). Studies cultural employment is low) [69]. reported a limited range of outcomes relating to health The remaining studies were judged to have high sus- and safety, but the reliability of findings in general was ceptibility to bias and the following examples are only il- compromised due to various aspects of study design lustrative of the diversity. Foerster et al. [70] report that which make them highly susceptible to bias, such as a purchasing power is lower in villages closer to PAs in lack of comparator or non-random selection of study Gabon. Cardozo [71] conducted a questionnaire-based site sites or participants, or incomplete reporting, with no comparison of communities inside and outside Allpahuayo- details about selection of study populations, validity of Mishara National Reserve, Peru and reported changes in survey instruments or survey response rate, [73,78]. The income and livelihood diversity. Annual income from agri- only study in this group with ‘medium’ susceptibility to culture was lower inside the PA whereas income from do- bias, Korhonen et al. [77], report a case-study of repro- mestic animals and palm products was higher. Kayser et al. ductive health from a 10-year old Integrated Conserva- [72] report greater annual transfer of money to SMEs tion and Development Project (ICDP). The study was of through contracts financed by Addo Elephant National BACI design comparing purposively selected (for repre- Park, South Africa. Saayman & Saayman [73] provide data sentativeness) villages and municipalities within a 3 km on self-reported change in the business environment belt around Ranomafana National Park (the ‘peripheral around South African PAs. This provides weak evidence zone’) (7 villages, 6 municipalities) with those outside that opportunities for and turnover of business has in- this zone (6 villages, 4 municipalities). Data on modern creased as a result of the PA. contraceptives suggest an increase in use over time, a decrease with increasing distance from the park and Land access and restrictions variability in use within ‘park’ villages attributed by au- thors to varying access to ICDP activity, but also to edu- Six studies identified cational status of women, local culture and religious Pullin et al. Environmental Evidence 2013, 2:19 Page 32 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 beliefs. Fertility levels were estimates rather than direct people has increased’. Other studies report perceived im- measures and do not indicate any change over time or provements in infrastructure in and around PAs com- differences between park and ‘outside’ villages. Visits to pared to elsewhere or before establishment [59,78] but the health centre for pre-natal care increased over time this is not always the case [62]. in line with national trends and did not differ between park and ‘outside’ village residents. The data presented Attitudes towards PA and the benefits (or otherwise) for infant mortality were drawn from census data pre-1999 they provide and showed year-on-year variation and no clear differences between ‘park’ and ‘outside’ residents, although a slightly  24 studies identified higher mean over an 8 year period outside than inside the  Only one study identified as ‘medium’ susceptibility PA. Post-1999 data were health centre data and only re- to bias, all remaining studies are highly susceptible lated to health centre births and thus may be subject to to bias due to lack of methodological detail, bias if a higher percentage of more problematic births oc- confounding variables unaccounted for, and spillover curred at the health centre as opposed to in villages. from protected areas into controls Sarker & Røskaft [58] found attitude to PAs to be Society and development negatively associated with PA proximity Four studies failed to identify the location of ‘inside’ 13 studies identified populations, two studies failed to report the distance Two are not highly susceptible to bias (1 ‘low’ and 1 of controls from PAs, and in general controls were ‘medium’ susceptibility to bias) very close to PA boundaries (e.g. < 1 km and 2.5 km) Sheppard et al. [79] found a greater number of  Studies reported attitudes towards PAs, attitudes infrastructural developments inside Wechiau towards identified benefits/disbenefits from the PAs, Community Hippo Sanctuary in Ghana than outside and perception of environmental change Korhonen et al. [77] found no difference in the percentage of girls in primary schools either over Studies reporting attitudes were separated into two time or inside relative to outside Ramonafana distinct categories; attitudes towards the protected area National Park in Madagascar, but a possible slight or conservation, and attitudes concerning benefits or increase in the percentage of girls in the third grade disbenefits resulting from the protected area. All but one over time was found inside the PA of the 24 studies reporting attitudes fell into the category of ‘high susceptibility to bias’. The main issues identified Studies in this set include measures of education, em- in the critical appraisal were lack of detail in the powerment, infrastructure, recreation and social capital. methods, not accounting for important confounding vari- Thirteen studies reported data on 54 development-related ables, and, where spatial comparators were used, potential outcome measures. Two studies were classed as having low spill-over effects due to the intervention and ‘comparator’ (1) or medium (1) susceptibility to bias. Sheppard et al. [79] sites being close together. This latter issue was a particular found a greater number of developments in various aspects problem for studies looking at the effect of ‘distance from of infrastructure inside Wechiau Community Hippo Sanc- the PA boundary’ on attitudes. Thirteen studies examined tuary in Ghana than outside. Korhonen et al. [77] found no spatial differences in attitude with respect to distance from difference in the percentage of girls in primary schools over the protected area. However four studies failed to report time and no difference inside relative to outside Ramona- the location of the intervention population inside the fana National Park in Madagascar, but they did find a slight protected area. Two studies [58,80] regressed attitude increase in the percentage of girls in the third grade over scores against distance from the protected area, but the time inside the PA (but not outside). distances involved were not stated. Control populations All remaining studies were judged to be of ‘high’ sus- in ‘inside-outside’ studies were generally very close to the ceptibility to bias. There is weak evidence for improve- protected area boundary, for example 2.5 km (Sekhar ment in education provision following PA establishment 1998) and < 1 km [71]; although it was 50 km in the study in terms of increased number of schools and perceptions of Bonaiuto et al. [81]. of improvement [62,72,78]. A questionnaire based on re- ported change conducted by Saayman and Saayman [73] Attitudes towards the protected area Fourteen articles in multiple South African PAs produced variable re- reported a wide range of general attitudes towards the sponses (in most, but not all, the majority agreed) to PA. Thirteen of these gave respondents’ statements (usu- statements such as ‘participation in community activities ally a mix of positive and negative statements) and pre- has increased’, ‘the pride that the residents have in their sented data on the percentage agreeing or disagreeing town has improved’, ‘the opportunities to meet new with each statement. One study [81] presented composite Pullin et al. Environmental Evidence 2013, 2:19 Page 33 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 attitudinal scores made up of responses to a series of studies were of low quality predominantly due to a lack questions which were not presented in the article. of detailed methodology and shortcomings of the experi- In the only study that was judged not to be highly sus- mental design. For example, Cihar and Stankova [90] ceptible to bias, Sarker and Røskaft [58] found that re- lacked a true before-and-after comparison and generated spondents from around four parks in Bangladesh had ‘before data’ by asking informants to recall the past. This negative attitudes towards the protected areas, and that is clearly open to recall bias. negative attitudes decreased with distance from each protected area. The remaining studies were of ‘high’ sus- Economic valuation studies ceptibility to bias. Bonaiuto et al. [81] reported that re- gional identity and place attachment were higher inside  10 studies identified the Tuscan Archipelago National Park in Italy, but that  One study judged as not highly susceptible to bias specific and general attitude scores towards the pro- (‘medium’) tected area were lower relative to a control group of re-  Four groups of studies found: cost-benefit analyses; spondents 50 km away. Jim and Wu [82] noted that a stated preference studies; stated preference combined higher proportion of people living on the boundary of with a distance comparator; and reported direct Shimentai Nature Reserve in China ‘disliked’ the park financial losses from a PA (e.g. fines/foregone income) than those living 4 km from its boundary. Finally,  These studies do not have real comparators (with Shrestha and Alavalapati [80] observed a positive correl- the exception of the distance comparator): instead ation between positive attitude and distance from Koshi they are hypothetical, and as a result are highly Tappu Wildlife Reserve in Nepal. susceptible to bias (e.g. optimism bias) In contrast to the above studies which found a positive  Studies are too heterogeneous and open to bias to relationship between distance from the park and attitudes, permit meaningful quantitative synthesis of valuations Cardozo [71], Gubbi et al. [83], Infield and Namara [84] and Sekhar [85] reported higher positive and lower nega- Studies in this section reported welfare impacts in tive attitudes inside protected areas than outside. Other monetary terms. Economists usually hold that individual studies found no statistically significant or observable dif- well-being is not directly and cardinally measurable, nor ference in attitudes between inside and outside protected comparable between individuals or time periods e.g. area [86-89] or over time [90]. [94]. However, changes in an individual’s well-being as a result of a PA’s existence can be expressed in terms of Attitudes towards named benefits/disbenefits result- the amount of money needed to render that individual ing from the park Thirteen studies reported data con- indifferent to the existence of the PA (the aggregation of cerning respondents’ attitudes towards named benefits such monetary amounts across individuals is common in or disbenefits resulting from the protected area. Again, applied economics, but deeply problematic). all studies were classed as highly susceptible to bias. Ite Ten studies were included that estimated well-being [89] found fewer respondents close to Cross River impacts of protected areas in monetary terms. Nine of National Park in Nigeria to believe that they have benefited these were categorized as having high susceptibility to from the protected area than those 5 km away, although bias, whilst one was categorised as medium susceptibility a third group of respondents 7.5 km from the protected to bias. area showed an intermediate perception of benefits. Jim Shrestha et al. [95] used a contingent valuation survey and Wu [82] reported no significant difference in the with a stratified random sample of 160 households perception of benefits from Shimentai Nature Reserve, within c. 6 km of Koshi Tappu Wildlife Reserve, Nepal, China, between respondents inside and those 4 km from to estimate their willingness to accept the PA in terms of the protected area, whilst significantly more respondents foregone resources. They found substantial local one- inside than outside claimed to have felt losses as a result time costs of 11,776.70 Nepali Rupees per household of the reserve. (1994/1995). Other studies found no evidence that respondents felt A major reason for excluding economic studies (see that either negative or positive impacts resulted from the Additional file 6) was that they measured the well-being protected area [75,86,88,91,92]. impacts of ecosystems within PAs, but did not isolate A smaller category of studies reported respondents’ the impact of the PA itself. One study [96] carried out perceptions of environmental change as a result of the a contingent valuation survey of willingness to accept protected area; with respondents in one study predomin- compensation for costs of the protected area’s presence antly not perceiving a change [90] and two studies on traditional pasture land, with respondents within the reporting the majority to have perceived an increase protected area and further away. The remaining studies in environmental ‘appearance’ [73,93]. However, these included only hypothetical or ‘modelled’ comparators. In Pullin et al. Environmental Evidence 2013, 2:19 Page 34 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 all cases this is done more or less explicitly by the ana- disaggregating costs and benefits by local, national and lysts themselves, but many studies also required respon- international groups. Both the effects of the PA and the dents to mentally construct hypothetical comparators, in counter-factual (no PA) were modelled, though little detail order to answer stated preference surveys. Such con- is presented and the evidence upon which the modelling structed comparators can be useful and indeed essential is based is often rather weak. They estimated that local when “real” comparators (RCTs, BACI etc.) are unavail- populations would lose from the establishment of the able (they may also be used in conjunction with such re- park, but that this would become a net gain if develop- search designs). However, they are vulnerable to a number ment projects associated with the park succeeded in rais- of potential biases, such as optimism bias, strategic bias, ing local incomes. At the national level, there would be a and hypothetical bias (see Table 5 for definitions). net loss due to the protected area. Without real comparators, direct evidence is lacking Seven studies used stated preference techniques to on the effects of the PA on individuals. Instead these must elicit estimates of welfare gains or losses. Four studies be predicted, using whatever information and opinion is [7,98-100] used contingent valuation to estimate regional available to the analyst or respondent. Numerous assump- or national populations’ willingness to pay for existing tions must necessarily be made, and will not always be protected areas in India, Brazil, China and Greece re- explicitly stated. This leaves these studies open to well- spectively, all four studies indicating generally positive recognised biases. For example, cost-benefit analyses are welfare impacts of the PAs on these broad populations. known to suffer from optimism bias, especially when con- A fifth study, Ascuito et al. [101], similarly estimated ducted by groups with an interest in the project: in the case local willingness to pay for a fire prevention programme of PAs this may be the government or conservationists in an existing protected area again finding positive wel- more generally. For example, the results of Kremen et al. fare impacts. Two studies [95,96] used contingent valu- [97] are heavily dependent on optimistic assumptions made ation to estimate local populations’ willingness to accept about the efficacy of development interventions planned to restrictions on livelihoods imposed by existing protected accompany the PA: no evidence is presented on whether areas in Ethiopia and Nepal respectively, indicating these interventions indeed had the effects assumed by the negative welfare impacts of the PA. In the case of Jemal authors, since the analysis was conducted ex ante. [96], surveys were carried out with respondents inside Stated preference studies are known to suffer from and further away from the PA, and found that people both hypothetical bias and strategic bias on the part of within the PA were less willing to accept compensation respondents. Hypothetical bias may lead respondents to than those further away, suggesting negative welfare im- overstate their willingness to pay (WTP) for goods or pacts increased with proximity to the PA. Finally, Abbot services provided by a PA, because they fail to consider and Mace [102] present data on fines levied on local their true budget constraint. Respondents may also behave people for illegally harvesting fuel-wood in Lake Malawi strategically: beneficiaries may overstate their willingness NP. These fines were levied by the PA but no informa- to pay for a PA, in order to increase the likelihood of its tion is provided on areas outside the PA. establishment if they suspect they will not be required to contribute to it, or understate their WTP if they suspect Inequalities that this will result in lower user fees. Those who expect Assessing the impact of PAs on health or social inequal- to lose from PA establishment may overstate the amount ities would require either individual sound studies with that they would require to receive in compensation for the justifiable subgroup analyses, or a set of comparable establishment of the PA (i.e. their willingness to accept studies which describe in detail the socioeconomic pos- (WTA) the PA) to reduce the likelihood of its establish- ition of the populations studied [103,104]. Neither was ment, or increase compensation payments. Alternatively available from the extant literature. surveys may under-estimate opportunity costs if the activ- ities concerned are considered sensitive or of dubious le- Meta-synthesis of qualitative and quantitative evidence gality, and are likely to be under-reported. The results of In attempting to bring together the findings of the quali- stated preference studies are also known to be sensitive to tative and quantitative reviews it is important to reflect the information provided by surveyors, and the precise on the differences in their philosophies. The qualitative formulation of the questions. This renders them vulner- synthesis is essentially formative and attempts to form a able to the same optimism bias noted above. picture of how PAs are perceived to impact on human Methodologically, the included studies fell into three well-being. As such it can form a template for empirical groups: cost-benefit analyses, stated preference studies, investigation and hypothesis testing. The synthesis of and reported direct financial losses from a PA. One study, quantitative evidence is more summative and attempts Kremen et al. [97] carried out an ex ante cost-benefit ana- to test hypotheses of impact. In consequence we should lysis of the establishment of Masoala NP, Madagascar, not expect the meta-synthesis to be a simple matching Pullin et al. Environmental Evidence 2013, 2:19 Page 35 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 of similar studies or outcomes. In this section we sum- of impact only two studies were not highly susceptible to marise the findings of the qualitative synthesis and ask bias, showing that meat and fish consumption was greater whether the quantitative evidence of impacts can inform in proximity to a PA and that timber and fuelwood bene- the questions raised by these findings or whether it sug- fits were more frequently appreciated nearer another PA. gests something different. Economic capital Views expressed on impacts of PAs Governance on economic capital are generally negative, with the ex- The qualitative synthesis reveals a number of factors that ception of some views on the benefits of ecotourism. In can lead to negative views and impacts of PA establish- contrast the quantitative evidence of impact from three ment: lack of clarity in regulations and boundaries; dis- studies on livelihood strategies was neutral to positive crepancies between state rules and local institutions; in terms of poverty reduction. In particular, there were forced migration, inadequate or non-existent compensa- concerns in Sweden amongst foresters about sustaining tion; and poor communication between communities employment and amongst politicians about sustaining and authorities and government indifference to cultural tax revenue in the presence of regulations. However, and social diversities. Negative views on impact of man- these concerns were not upheld by a quantitative assess- agement can arise from poor relationships between resi- ment of impact in the same country. All but one of the dents and park officials. Views on how to lessen negative economic valuation studies suffered from high suscepti- impacts or achieve positive impacts include: rules that bility to bias and therefore add limited reliable quantita- are locally adapted or based on traditional rules; greater tive evidence to this issue. clarity over rights and duties; planning focussed on com- munity livelihoods as well as the park; appropriate cap- Social capital The qualitative synthesis suggests devel- acity building; and empowerment through the process of opment associated with PAs can exacerbate ethnic ten- acquiring land titles and setting boundaries. The existence sions through perceived preferential treatment of some of these views enables hypotheses to be generated on how communities. There may be a relationship between exist- to achieve change in impact. The synthesis of quantitative ing social capital and ability to adapt to new circumstances. measures of impact shows that these hypotheses are yet to Quantitative evidence of impact on social capital is mixed. be tested. What is absent from the evidence base is a Thereissomeevidenceofpositiveimpact of land protec- quantitative comparison of costs and benefits to local tion on poverty alleviation and on housing and infrastruc- people of different forms of PA governance. ture but also of increasing incidence of wildlife conflict. Well-being Health Views expressed on health of local populations Environmental (natural) capital The qualitative syn- are predominantly negative, including exposure to dis- thesis presents a range of positive and negative attitudes ease, wildlife conflict and women’s safety. Quantitative among local populations towards PAs. Alongside an ap- studies of impact of PAs on health and safety are notable preciation and desire to protect the environment were by their absence. concerns about reliance on those same areas to maintain economic livelihoods. Although the qualitative literature Discussion provides evidence of difference views, quantitative evi- Historical accounts of establishment of PAs provide evi- dence to estimate the scale and reach of those views was dence that substantial negative impacts on local popula- not extractable since all but one study were highly sus- tions can occur and have occurred. Forced displacement ceptible to bias, of communities is a recurring theme in the narrative concerning negative impacts of PAs e.g. [4]. This review Access to land The qualitative synthesis revealed two does not seek to question this historical narrative. Commu- very different scenarios in terms of access to land. The nity development and infrastructural improvements in close first is resentment at loss of access and the second is proximity to PAs has also been documented and suggests benefit from acquisition and value of land on the PA that PA establishment can be positive (i.e. win-win solutions margin. All quantitative studies of impact of PAs on land for biodiversity and human well-being are possible). Estab- access and restrictions were highly susceptible to bias lishment of PAs will inevitably lead to impacts on local, and and so the current evidence does not allow the magni- possibly regional, populations but the challenge is to im- tude of these scenarios to be assessed. prove our capacity to predict which factors will influence the balance of positive and negative impacts. In this review Resource use A range of positive and negative views were we have attempted to provide an assessment and charac- found concerning PAs as a source of natural resources and terization of the range of positive and negative impacts in ecosystem goods. In the synthesis of quantitative evidence the period following the Rio Summit and establishment of Pullin et al. Environmental Evidence 2013, 2:19 Page 36 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 the CBD. It was not the objective of this review to revisit manner, the findings from international studies of peo- this history prior to 1992. The review also attempts to col- ple’s views about the impact of protected areas on their late evidence on the factors that modify impact, either in a lives. To reduce the likelihood of missed studies, sensi- positive or negative direction. tive searches of bibliographic databases were supple- mented by other methods to seek out less easily found Comparison of qualitative and quantitative evidence literature such as unpublished reports from topic rele- Although thequantitativeevidenceisinsufficientto draw vant websites. Studies providing thick descriptive data conclusions about the scale of either positive or negative spanning the different categories of IUCN protected impacts of protected areas on well-being, it was possible to areas and before and after the Durban Accord offered an synthesise understandings from qualitative studies about excellent source for synthesising understanding about how positive and negative changes in well-being can arise how protected areas impact on people’s lives. from establishing and implementing regulations to protect During the reading and re-reading of individual studies the natural environment, with or without simultaneous in- we found that data often encapsulated the complexity of vestment in community development. living in or near protected areas, touching on a multipli- The qualitative synthesis has identified a number of city of interrelated themes. Within the limitation of time themes in the ways governance of protected areas affect hu- and resources for this study we have only been able to man populations well-being and how PAs are viewed. Some present these themes fairly superficially without explor- of these themes, such as the impact of land protection on ing fully all their interconnections. forestry sector employment in Sweden, are reflected in the The disparate, fragmented literature limits our ability impacts assessed in studies considered in the quantitative to test the comprehensiveness of the search. In reviewing review. Other themes, however, have not been rigorously such a broad and interdisciplinary question it has been a assessed in the quantitative literature, for example, novel significant challenge to test all the possible sources of rele- diseases resulting from changes in lifestyle, increased work- vant material; nor is it simple to measure what proportion load and heat exhaustion due to crop raiding, safety risks of the relevant articles we have been able to access with felt by women as men worked further afield and as women the time and resources available. Limiting our search to themselves were drawn out of their homes for new roles. English-language articles may be significant. Diversity of the literature has limited any assessment Reasons for heterogeneity of extent of publication bias. Whilst we have attempted The identification of variables that influence whether posi- to minimise publication bias by employing a systematic tive or negative impacts will occur would be desirable for search strategy, we have no way of testing for publica- supporting decision making on the process of establish- tion bias in the literature we obtained. The selective na- ment and subsequent management of PAs. Unfortunately ture of many studies, in terms of the type of impacts the nature of the evidence provides little opportunity to investigated, is also a potential source of bias, since re- analyse differences in impact among different PAs (see lim- searchers may ‘cherry-pick’ (possibly inadvertently) those itations below). impacts most likely to show a particular effect. Mode of governance is commonly viewed as a key variable determining impacts of PAs and this is sup- Limitations of the evidence base ported by the qualitative synthesis in which many narra- Although the quality of the studies was sufficient to tives are available on different aspects of governance. draw out their findings to explain how different impacts However, rigorous tests of governance as an effect modi- may arise, many of the studies failed to report ad- fier are absent. A similar lack of quantitative evidence is equately their methods of data collection and analysis. apparent for the following questions: There is a potential in any studies looking at the quan- titative impact of protected areas that respondents will What practices repeatedly lead to negative impacts, and bias their responses in an effort to influence protected area which ones seem to be recurrently improving people’s governance. Some questionnaire-based studies attempted well-being? to minimise this strategic bias by clearly stating that inter- Are some of these practices becoming more/less view and questionnaire results would be used solely for re- common with time? search purposes, other studies did not acknowledge this Are any costs or benefits associated with particular potential bias or attempt to reduce it. Of the 305 outcome types of PAs (e.g. size or location)? measures extracted from 79 included studies, 92 outcome measures (30%) involved ‘reported changes’. For these data Strengths and limitations of the review the comparator is implicit in the respondent’sreply;they This is the first systematic review of which we are aware are reporting a change over time due to the protected area. that attempts to identify and synthesise, in a transparent Whilst these results are relevant (externally valid) to the Pullin et al. Environmental Evidence 2013, 2:19 Page 37 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 review in hand, they can be susceptible to significant recall There is a lack of primary studies estimating impact bias and questioning bias (elements of internal validity). of PAs on human well-being using direct measurement Several studies attempted to retrospectively elicit opinions techniques in a BACI format. In addition to a generally and attitudes towards protected area establishment many high susceptibility to bias, very few studies employed ro- years after the event, which is similarly open to substantial bust comparators over appropriate time frames in order sources of bias. to maximise evidence linking protected areas to ob- For these reasons, results in the form of reported served human well-being impacts. Only one included changes which involve significant recall should be viewed study used a full BACI design to account for spatial and with caution, and studies critiqued in depth for potential temporal confounding variables. Only three studies used sources of bias. We attempted to account for these sources direct data collection. of bias during critical appraisal using our ‘susceptibility to We found a surprisingly small number of studies on bias’ scoring system. health of populations. Only nine studies reported data Sixty-six of the 79 studies accepted following critical on human health impacts of protected areas. This is sur- appraisal collected data in the form of self-reported mea- prising since the majority of articles in this review pur- sures. Fifty-six of these articles failed to provide details port to measure human well-being. Difficulties in ethical of the questionnaires given to respondents, and only two approval for human study may account in part for the articles provided a copy of the survey instrument in full paucity of health studies. [76,105]. Without details of the questioning involved in these surveys it is difficult to assess questioning bias. Review conclusions Variation in all the question elements (PECO; popula- Implication for policy/management tion, exposure, comparator, outcome) and the high de- The evidence base provides a range of possibilities to in- gree of specificity in outcome measures identified in this form but little evidence to support decision making on review provides problems for synthesis. In particular, how to maximise positive impacts of PAs on human well- studies based on self-reported measures commonly asked being. The diversity of studies and of outcomes measured, very specific questions that could not then be synthesised together with the diversity (or lack of clear signal) in the along with other similar outcomes. Similarly, a high degree data suggests that impacts of PAs are highly context of variability in the choice and design of comparators pre- dependent. However, the evidence base is insufficient to vented synthesis. In some studies, the inside-outside com- provide any power with which to predict impacts on well- parison was open to many confounding factors that cloud being from a knowledge of their context. It logically fol- the link between protected area presence and impacts. lows that there is an insufficient evidence base to identify This highlights the difficulty of balancing minimisation of circumstances/variables/effect modifiers that might lead to spillover effects, whereby the comparator population is greater or lesser impact. At present, the available evidence close enough to feel the effects of the exposure, and con- base is failing to inform policy on the progress (or lack of trol of non-target variables. Whilst some studies accounted it) being made, since 1992, toward lessening negative and for this problem by including confounding variables in promoting positive impact of PAs on human well-being. statistical models, many others did not. Furthermore, very few studies examined differences in environmental condi- tions between the comparator and exposure populations. Implication for research High susceptibility to bias in most studies limits ability The nature of the research reported to date forms a di- to attribute outcomes/impacts to presence of PAs. Forty- verse and fragmented evidence base that is insufficiently five studies were excluded during critical appraisal due to developed to reliably inform future policy decisions flaws in experimental design and data analysis, or due to a (recognising that many included studies did not set out lack of methodological detail. However, many studies in- to address the review question). Many studies appear to cluded after the first stage of critical appraisal also failed to have been conducted opportunistically and lack baseline account for confounding variables, selected replicates in a measures. There is no evidence of a strategic approach non-random manner, and used opportunistic methodology. or strategic investment to this field of research beyond The most frequently occurring factor that affected the sus- individual research group initiatives. If a sufficient evidence ceptibility to bias score in included studies, however, was a base is to be formed then there is a need for concerted failure to appropriately report their methodology. Signifi- programme of research rather than an uncoordinated short cant details such as recall period, response rate, item pool term opportunistic approach. balance and order, sample selection process, sample size, The diversity of outcome measures and the consequent and sample location were not disclosed in a large number difficulty for synthesis suggests a need for use of standard of cases. Together, these factors limit the ability to attribute indicators of human well-being that allow comparison the reported impacts to protected areas. among studies and meaningful synthesis of evidence. Pullin et al. Environmental Evidence 2013, 2:19 Page 38 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Comparative research needs to progress from PA/no a payoff between maximising similarity and minimising PA to PA type A/PA type B comparisons. Comparisons spillover (the overflow of impacts from the intervention should be made between potential proximate causes of into the nearby comparator). Statistical tests can help to positive or negative impacts when the ultimate cause confirm similarity across intervention and comparator is PA establishment/management. This review suggests populations, and descriptive variables can be included in some of the candidates to be governance models, exist- models that test for the significance of the intervention in ing social capital, cultural diversity and poverty index. order to account for differences that might occur. It would be helpful to research efforts for funders to find consensus on minimum standards for methodolo- Replication gies, for both qualitative and quantitative evidence, that Care must be taken to ensure that there is an appropriate provide improved quality and thus reliability of data. The trade-off between a study’s accuracy and its precision. large proportion of included studies that suffered high When combining many studies in a synthesis, more accur- susceptibility to bias is an indicator of such a need and ate results are preferable to more precise ones. For ex- also an indicator that scarce research resources are not be- ample, a study that measures daily resource extraction ing used effectively. over a year in ten households from one intervention and one comparator village is less likely to reflect the true im- Recommended study design pact of the intervention than a study that measures daily In order to better assess the impacts of protected areas resource extraction over a month from 12 intervention on human well-being we make the following recommen- and 12 comparator villages. This spectrum is not clear-cut, dations for future research study design and reporting; however, and the allocation of resources to pseudorepli- cation (improving precision) and true replication (improv- Methodological detail; Studies must report sufficient ing accuracy) must be considered carefully. Indeed, the details regarding the location of sample sites (in scale at which conclusions will be drawn defines what is relation to the protected area boundaries in particular), pseudoreplication and what is true replication, and this the degree of replication, the data collection tool definition may be different for the author and the system- (e.g. quote questions posed to respondents in atic reviewer. questionnaires), the method of sample selection (e.g. random or purposeful), and the times and duration of Statistics sampling. This is not an exhaustive list, and sufficient Statistics, both in summarising results and analysing pat- detail must be provided to allow the sampling to be terns, must be used with great care. We recommend that repeated. Where information cannot fit within a statistician be consulted during experimental design in published articles these details should be provided in order to optimise design for analysis. The use of models supplementary material. that account for changes in non-target variables across Baseline assessment; Where changes following temporal and spatial scales are recommended, but tests establishment or change in protected area governance for differences in confounders between intervention and are being investigated, adequate baselines must be comparator populations are also appropriate. Where in- assessed. Although this is difficult and requires formation can be presented in summary statistics (e.g. planning prior to the intervention, full ‘before-after- mean/median and standard deviation/confidence inter- control-intervention’ (BACI) study design is vital to vals) this will aid future meta-analysis. account for confounding temporal and spatial confounding factors. By assessing baselines, any Additional files differences between intervention and comparator populations can be compared relative to the starting Additional file 1: Search Strategy. Details of the development of the finalised search string, including scoping and testing against list of conditions to strengthen the evidence towards causation. recommended key articles, and recording of search results from all databases, including duplicate removal and library file creation. Matched controls Additional file 2: Unobtainable and Un-translated. A list of articles ‘Control’ or ‘comparator’ populations are vital to enable that could not be obtained in full text, along with a list of foreign language articles that were not translated for consideration in the conclusions to be drawn about impacts in the absence of quantitative review. the intervention. A reliable comparison requires that as Additional file 3: Web site searches. A list of organisational web sites many other variables describing the environment are held searched for material, along with search terms, hits returned and actions constant or matched between comparator and interven- taken for the quantitative review. tion populations, allowing only the intervention to change Additional file 4: Bibliographic Searches. Details of five relevant review article bibliographies searched for supplemental material in the in an ideal situation. In practice this is very difficult (and quantitative review. why baseline assessment is important), and there is often Pullin et al. Environmental Evidence 2013, 2:19 Page 39 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 areas in the developing world: Economic valuation of Morro do Diabo Additional file 5: Coding Tool. The coding tool used to assess articles State Park, Atlantic Rainforest, São Paulo State (Brazil). Ecol Econ 2008, at full text for the qualitative review, including the coding of outcomes 66:359–370. used in both reviews. 8. Ferraro PJ, Hanauer MM, Sims KRE: Conditions associated with protected Additional file 6: Excluded. Lists of articles excluded at full text assessment area success in conservation and poverty reduction. Proc Natl Acad Sci from both the qualitative and quantitative reviews along with reasons. 2011, 108:13913–13918. Additional file 7: Narrative Synthesis. Full narrative synthesis tables for 9. West P, Igoe J, Brockington D: Parks and peoples: the social impact of 79 included studies in quantitative review. protected areas. Annual Reviews in Anthropology 2006, 35:251–277. 10. Sutherland WJ, Adams WM, Aronson RB, Aveling R, Blackburn TM, Broad S, Additional file 8: Supplementary Descriptive Statistics for Ceballos G, Cote IM, Cowling RM, Da Fonseca GAB, et al: One hundred ‘Qualitative Synthesis’. Descriptive statistics for the 305 articles included questions of importance to the conservation of global biological in the systematic map of the first stage of analysis of the qualitative review. diversity. Conserv Biol 2009, 23:557–567. Additional file 9: Qualitative Synthesis. The full synthesis of 30 studies 11. Adams WM, Aveling R, Brockington D, Dickson B, Elliott J, Hutton J, Roe D, considered in the qualitative review, from which the summary synthesis Vira B, Wolmer W: Biodiversity conservation and the eradication of was produced. poverty. Science 2004, 306:1146–1149. Additional file 10: Table S1. Summary table of included studies and 12. Dudley N: Guidelines for Applying Protected Area Managment Categories. their measured outcomes in the quantitative review that were scored as Gland Switzerland: World Conservation Union; 2009. having ‘low’ or ‘medium’ susceptibility to bias (Susc. to Bias). 13. Pullin AS, Bangpan M, Dalrymple S, Dickson K, Healey JR, Hockley N, Jones JPG, Knight TM, Oliver S: Human well-being impacts of terrestrial Additional file 11: Details of the information considered during protected areas? CEE protocol 11–009 Collaboration for Environmental critical appraisal from each of the 79 studies included in the Evidence: wwwenvironmentalevidenceorg/SR11009html; 2012. quantitative review. 14. Neuman L: Social Research Methods: Qualitative and Quantitative Approaches. New York: Allyn and Bacon; 1997. Competing interest 15. Gough D, Thomas J, Oliver S: Clarifying differences between review There are no potential conflicts of interest to report. designs and methods. Systematic Reviews 2012, 1:28. 16. Oliver SR, Rees RW, Clarke‐Jones L, Milne R, Oakley AR, Gabbay J, Stein K, Authors' contributions Buchanan P, Gyte G: A multidimensional conceptual framework for ASP managed and planned the conduct of the SR. SO managed the conduct analysing public involvement in health services research. Health Expect of the qualitative section. SD and NRH undertook searches. Quantitative 2008, 11:72–84. Review; SD and NRH screened for inclusion. NRH carried out data extraction 17. Ritchie J, Spencer L: Qualitative data analysis for applied policy research. and critical appraisal. JPGJ, JRH, NH and TK assisted in planning, In Analyzing Qualitative Data. Edited by Bryman A, Burgess RG. London: interpretation and synthesis. Qualitative Review; MB, KD, HH SO and CV Routledge; 1994:305–329. undertook screening, data extraction and synthesis. All authors contributed 18. Pope C, Ziebland S, Mays N: Qualitative research in health care: Analysing to writing and approved the final manuscript. qualitative data. BMJ: British Medical Journal 2000, 320:114. 19. Thomas J, Harden A, Newman M: Synthesis: combining results systematically and appropriately. In Introducing systematic reviews. Edited Acknowledgments by Gough D, Oliver S, J. T. London: Sage Publications; 2012. We thank Sandra Diaz and Tom Hammond for their help in administering the progress of this review. 20. Vasak K: Human Rights: A Thirty-Year Struggle: the Sustained Efforts to give Force of law to the Universal Declaration of Human Rights. UNESCO Courier 30:11. Paris: United Nations Educational, Scientific, and Cultural Organization; 1977. Sources of support 21. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR: Introduction to meta- This review was funded by the Scientific and Technical Advisory Panel of the analysis. Chichester, UK: Wiley; 2011. Global Environment Facility. 22. Strickland-Munro J, Moore S: Indigenous involvement and benefits from tourism in protected areas: a study of Purnululu National Park and Author details Warmun Community, Australia. J Sustain Tour 2013, 21:26–41. Centre for Evidence-Based Conservation, School of Environment, Natural 23. Allendorf TD, Smith JLD, Anderson DH: Residents’ perceptions of Royal Resources and Geography, Bangor University, LL57 2UW Bangor, Gwynedd, UK. Bardia National Park, Nepal. Landscape and Urban Planning 2007, 82:33–40. Evidence-informed Policy and Practice Information and Coordinating Centre (EPPI-Centre), Institute of Education, University of London, 20 Bedford Way, 24. Haukeland JV: Tourism stakeholders’ perceptions of national park WC1H 0NR London, UK. School of Environment, Natural Resources and management in Norway. J Sustain Tour 2011, 19:133–153. Geography, Bangor University, LL57 2UW Bangor, Gwynedd, UK. 25. Castillo A, Magana A, Pujadas A, Martínez L, Godínez C: Understanding the interaction of rural people with ecosystems: A case study in a tropical Received: 9 January 2013 Accepted: 17 July 2013 dry forest of Mexico. Ecosystems 2005, 8:630–643. Published: 28 October 2013 26. Mehring M, Seeberg-Elverfeldt C, Koch S, Barkmann J, Schwarze S, Stoll- Kleemann S: Local institutions: Regulation and valuation of forest use— Evidence from Central Sulawesi, Indonesia. Land Use Policy 2011, 28:736–747. References 27. Petrzelka P, Marquart-Pyatt S: “With the Stroke of a Pen”: Designation of 1. Jenkins CN, Joppa L: Expansion of the global terrestrial protected area the Grand Staircase Escalante National Monument and the Impact on system. Biol Conserv 2009, 142:2166–2174. Trust. Hum Ecol 2013, 41:285–297. 2. Bertzky B, Corrigan C, Kemsey J, Kenney S, Ravilious CCB, Burgess N: Protected 28. Gerritsen PRW: Diversity at stake: a farmers’ perspective on biodiversity and Planet Report 2012: Tracking progress towards global targets for protected areas. conservation in western Mexico. Mexico: Wageningen Universiteit Holanda; 2002. Cambridge, UK: IUCN, Gland, Switzerland and UNEP-WCMC; 2012. 29. Diaw MC, Tiani AM: Fences in our heads: A discourse analysis of the 3. Brockington D, Igoe J, SCHMIDT‐SOLTAU K: Conservation, human rights, Korup resettlement stalemate. J Sustain For 2010, 29:221–251. and poverty reduction. Conserv Biol 2006, 20:250–252. 30. Nguiffo S: One forest and two dreams: The constraints imposed on the 4. Dowie M: Conservation refugees: The hundred-year conflict between global Baka in Miatta by the Dja Wildlife Reserve. In Indigenous people and conservation and native peoples. Cambridge USA: MIT Press; 2009. protected areas in Africa. Edited by Nelson J, Hossack L, Moreton-in-Marsh. 5. Turner WR, Brandon K, Brooks TM, Gascon C, Gibbs HK, Lawrence KS, UK: Forest Peoples Programme; 2003:195–214. Mittermeier RA, Selig ER: Global biodiversity conservation and the 31. Sletten M, Vedeld P, Kaboggoza J: To co-operate or not to co-operate? a alleviation of poverty. BioScience 2012, 62:85–92. study of collaborative management planning in Mount Elgon National Park. 6. Kramer RA, Sharma NP, Munasinghe M: Valuing tropical forests: methodology Uganda: NORAGRIC working paper; 2008. and case study of Madagascar. Madagascar: World Bank Publications; 1995. 7. Adams C, Seroa da Motta R, Ortiz RA, Reid J, Ebersbach Aznar C, de Almeida 32. Bolaane M: The impact of game reserve policy on the River BaSarwa/ Sinisgalli PA: The use of contingent valuation for evaluating protected Bushmen of Botswana. Social Policy & Administration 2004, 38:399–417. Pullin et al. Environmental Evidence 2013, 2:19 Page 40 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 33. Milgroom J, Spierenburg M: Induced volition: Resettlement from the 59. Bajracharya SB, Furley PA, Newton AC: Impacts of community-based Limpopo National Park, Mozambique. J Contemp Afr Stud 2008, 26:435–448. conservation on local communities in the Annapurna Conservation Area, 34. Mbaiwa JE: Wildlife resource utilisation at Moremi Game Reserve and Nepal. Biodivers Conserv 2006, 15:2765–2786. Khwai community area in the Okavango Delta, Botswana. 60. Mehta JN, Kellert SR: Local attitudes toward community-based J Environ Manage 2005, 77:144–156. conservation policy and programmes in Nepal: a case study in the 35. Torri MC: Conservation, relocation and the social consequences of Makalu-Barun Conservation Area. Environ Conserv 1998, 25:320–333. conservation policies in protected areas: Case study of the Sariska Tiger 61. Naughton-Treves L, Alix-Garcia J, Chapman CA: Lessons about parks and Reserve. India. Conservation and Society 2011, 9:54. poverty from a decade of forest loss and economic growth around 36. Keskitalo ECH, Lundmark L: The controversy over protected areas and Kibale National Park, Uganda. Proc Natl Acad Sci 2011, 108:13919–13924. forest-sector employment in Norrbotten, Sweden: forest stakeholder 62. Rinzin C, Vermeulen WJV, Wassen MJ, Glasbergen P: Nature Conservation perceptions and statistics. Soc Nat Resour 2009, 23:146–164. and Human Well-Being in Bhutan An Assessment of Local Community Perceptions. J Environ Dev 2009, 18:177–202. 37. Spenceley A, Goodwin H: Nature-based tourism and poverty alleviation: impacts of private sector and parastatal enterprises in and around 63. Okello MM, Buthmann E, Mapinu B, Kahi HC: Community opinions on Kruger National Park, South Africa. Current Issues in Tourism 2007, wildlife, resource use and livelihood competition in Kimana Group Ranch 10:255–277. near Amboseli, Kenya. Open Conservation Biology Journal 2011, 5:1–12. 38. Songorwa AN: Community-based wildlife management (CWM) in 64. Lundgren T: Environmental protection and impact on adjacent Tanzania: Are the communities interested? World Dev 1999, 27:2061–2079. economies: evidence from the Swedish mountain region. 39. Stone M, Wall G: Ecotourism and community development: case studies Growth and Change 2009, 40:513–532. from Hainan, China. Environ Manage 2004, 33:12–24. 65. Richardson RB, Fernandez A, Tschirley D, Tembo G: Wildlife Conservation in 40. Bruyere BL, Beh AW, Lelengula G: Differences in perceptions of Zambia: impacts on rural household welfare. World Dev 2012, 40:1068–1081. communication, tourism benefits, and management issues in a 66. Sims KRE: Conservation and development: Evidence from Thai protected protected area of rural Kenya. Environ Manage 2009, 43:49–59. areas. J Environ Econ Manag 2010, 60:94–114. 41. Slater R: Between a rock and a hard place: contested livelihoods in 67. Andam KS, Ferraro PJ, Sims KR, Healy A, Holland MB: Protected areas Qwaqwa National Park, South Africa. Geogr J 2002, 168:116–129. reduced poverty in Costa Rica and Thailand. Proc Natl Acad Sci 2010, 42. Bizikova L, Nijnik M, Kluvanková-Oravská T: Sustaining multifunctional 107:9996–10001. forestry through the developing of social capital and promoting 68. Canavire-Bacarreza G, Hanauer MM: Estimating the impacts of bolivia’s participation: a case of multiethnic mountain communities. protected areas on poverty. World Dev 2012, 41:265–285. Small-Scale Forestry 2012, 11:301–319. 69. Ferraro PJ, Hanauer MM: Protecting Ecosystems and Alleviating Poverty 43. Ormsby A, Kaplin BA: A framework for understanding community with Parks and Reserves: ‘Win-Win’or Tradeoffs? Environ Resour Econ 2011, resident perceptions of Masoala National Park, Madagascar. Environ 48:269–286. Conserv 2005, 32:156–164. 70. Foerster S, Wilkie DS, Morelli GA, Demmer J, Starkey M, Telfer P, Steil M: 44. First Peoples Worldwide: Okiciyab: To help each other-Promoting best Human livelihoods and protected areas in Gabon: a cross-sectional practices in Indigenous community Development. 2006. Final report, June. comparison of welfare and consumption patterns. Oryx 2011, 45:347–356. 45. ESCO: Biosphere reserves: The Seville Strategy and the Statutory Framework of 71. Cardozo M: Economic displacement and local attitude towards protected the World Network. Paris: UNESCO; 1996. area establishment in the Peruvian Amazon. Geoforum 2011, 42:603–614. 46. World Parks Congress: The Durban Action Plan. Durban, South Africa: World 72. Kayser D, Sobrevila C, Ledec G: From Planning to the Implementation of a Parks Congress; 2003. Successful Conservation and Socio-Economic Model. South Africa: Report from 47. Herrold-Menzies M: From Adversary to Partner: The Evolving Role of The World Bank and Addo Elephant National Park; 2011. Caohai Nature Reserve in the Lives of Reserve Residents. Canadian 73. Saayman M, Saayman A: Regional development and national parks in Journal of Development Studies/Revue canadienne d’études du développement South Africa: Lessons learned. Tour Econ 2010, 16:1037–1064. 2006, 27:39–50. 74. Ninan KN: Non Timber Forest Products and Biodiversity Conservation: 48. Lunstrum E: Mozambique, neoliberal land reform, and the Limpopo a Study of Tribals in a Protected Area in India. In Conserving and Valuing National Park. Geogr Rev 2008, 98:339–355. Ecosystem Services and Biodiversity: Economic, Institutional and Social 49. Almudi T, Berkes F: Barriers to empowerment: Fighting eviction for Challenges. Edited by Ninan KN. London: Earthscan; 2012:99–112. conservation in a southern Brazilian protected area. Local Environment 75. ETFRN NEWS 39/40: Globalisation, localisation and tropical forest 2010, 15:217–232. management. http://www.etfrn.org/etfrn/newsletter/pdf/etfrnnews3940.pdf. 50. Davis A: ‘ Ha! What is the benefit of living next to the park?’ Factors 76. Phamtrong N, Swan S: An investigation into the drivers of forest limiting in-migration next to Tarangire National Park, Tanzania. encroachment in Bidoup Nui Ba National Park, Vietnam. Vietnam: Report to Conserv Soc 2011, 9:25–34. WWF Greater Mekong - Vietnam Country Programme; 2009. 51. Hoole A, Berkes F: Breaking down fences: Recoupling social–ecological 77. Korhonen K, Rahkonen O, Hemminki E: Implications of integrated nature systems for biodiversity conservation in Namibia. Geoforum 2010, 41:304–317. conservation for human reproductive health: a case study from 52. Hartter J: Attitudes of rural communities toward wetlands and forest Ranomafana National Park, Madagascar. Dev South Afr 2004, 21:603–621. fragments around Kibale National Park, Uganda. Hum Dimens Wildl 2009, 78. Mishra C: Socioeconomic transition and wildlife conservation in the Indian 14:433–447. Trans-Himalaya. Journal of the Bombay Natural History Society 2000, 97:25–32. 53. Bedunah DJ, Schmidt SM: Pastoralism and protected area management in 79. Sheppard DJ, Moehrenschlager A, McPherson JM, Mason JJ, Pattenden J, Mongolia’s Gobi Gurvansaikhan National Park. Dev Chang 2004, 35:167–191. Khandker S, Koolwal G, Biggs B, King L, Basu S: Ten years of adaptive 54. Ogra MV: Human–wildlife conflict and gender in protected area community-governed conservation: evaluating biodiversity protection borderlands: a case study of costs, perceptions, and vulnerabilities from and poverty alleviation in a West African hippopotamus reserve. Uttarakhand (Uttaranchal), India. Geoforum 2008, 39:1408–1422. Environ Conserv 2010, 37:270–282. 55. Stronza A, Gordillo J: Community views of ecotourism. Ann Tour Res 2008, 80. Shrestha RK, Alavalapati JRR: Linking conservation and development: An 35:448–468. analysis of local people’s attitude towards Koshi Tappu wildlife reserve, 56. Yasuda A: The Impacts of Sport Hunting on the Livelihoods of Local Nepal. Environ Dev Sustain 2006, 8:69–84. People: A Case Study of Bénoué National Park, Cameroon. Soc Nat Resour 81. Bonaiuto M, Carrus G, Martorella H, Bonnes M: Local identity processes and 2011, 24:860–869. environmental attitudes in land use changes: The case of natural 57. Nyahongo J, Holmern T, Kaltenborn BP, Roskaft E: Spatial and temporal protected areas. J Econ Psychol 2002, 23:631–653. variation in meat and fish consumption among people in the western 82. Jim C, Xu SSW: Stifled stakeholders and subdued participation: Serengeti, Tanzania: the importance of migratory herbivores. interpreting local responses toward Shimentai Nature Reserve in South Oryx 2009, 43:258. China. Environ Manage 2002, 30:327–341. 58. Sarker A, Røskaft E: Human attitudes towards the conservation of 83. Gubbi S, Linkie M, Leader-Williams N: Evaluating the legacy of an protected areas: a case study from four protected areas in Bangladesh. integrated conservation and development project around a tiger reserve Oryx 2011, 45:391–400. in India. Environ Conserv 2008, 35:331. Pullin et al. Environmental Evidence 2013, 2:19 Page 41 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 84. Infield M, Namara A: Community attitudes and behaviour towards conservation: an assessment of a community conservation programme around Lake Mburo National Park, Uganda. Oryx 2001, 35:48–60. 85. Sekhar NU: Crop and livestock depredation caused by wild animals in protected areas: the case of Sariska Tiger Reserve, Rajasthan, India. Environ Conserv 1998, 25:160–171. 86. Fiallo EA, Jacobson SK: Local Communities and Protected Areas: Attitudes of Rural Residents Towards Conservation and Machalilla National Park, Ecuador. Environ Conserv 1995, 22:241–249. 87. Harada K: Attitudes of local people towards conservation and Gunung Halimun National Park in West Java, Indonesia. J For Res 2003, 8:271–282. 88. Hartter J, Goldman AC: Life on the Edge: Balancing Biodiversity, Conservation, and Sustaining Rural Livelihoods around Kibale National Park, Uganda. Focus Geogr 2009, 52:11–17. 89. Ite DUE: Community perceptions of the Cross River national park, Nigeria. Environ Conserv 1996, 23:351–357. 90. Cihar M, Stankova J: Attitudes of stakeholders towards the Podyji/Thaya River Basin National Park in the Czech Republic. J Environ Manage 2006, 81:273–285. 91. Alexander SE: Resident attitudes towards conservation and black howler monkeys in Belize: the Community Baboon Sanctuary. Environ Conserv 2000, 27:341–350. 92. Rugendyke B, Son NT: Conservation costs: Nature-based tourism as development at Cue Phuong national park, Vietnam. Asia Pacific Viewpoint 2005, 46:185–200. 93. Saayman M, Saayman A, Ferreira M: The socio-economic impact of the Karoo National Park. Koedoe 2009, 51:1–10. 94. Just RE, Hueth DL, Schmitz A: The Welfare Economics of Public Policy: A Practical Approach to Project and Policy Evaluation. Cheltenham, UK: Edward Elgar; 2004. 95. Shrestha RK, Alavalapati JRR, Seidl AF, Weber KE, Suselo TB: Estimating the local cost of protecting Koshi Tappu Wildlife Reserve, Nepal: a contingent valuation approach. Environ Dev Sustain 2007, 9:413–426. 96. Jemal H, Bogale A, Hagedorn K: Welfare loss for pastoralists due to wildlife protection areas: the case of Awash National Park, Ethiopia. Quarterly Journal of International Agriculture 2008, 47:217–238. 97. Kremen C, Niles J-O, Dalton MG, Daily GC, Ehrlich PR, Fay JP, Grewal D, Guillery RP: Economic incentives for rainforest conservation across scales. Science 2000, 288:1828–1832. 98. Han F, Yang Z, Wang H, Xu X: Estimating willingness to pay for environment conservation: a contingent valuation study of Kanas Nature Reserve, Xinjiang, China. Environ Monit Assess 2011, 180:451–459. 99. Maharana I, Rai S, Sharma E: Environmental economics of the Khangchendzonga National Park in the Sikkim Himalaya, India. GeoJournal 2000, 50:329–337. 100. Jones N, Iosifides T, Evangelinos KI, Florokapi I, Dimitrakopoulos PG: Investigating knowledge and perceptions of citizens of the National Park of Eastern Macedonia and Thrace, Greece. International Journal of Sustainable Development & World Ecology 2012, 19:25–33. 101. Asciuto A, Schimmenti E: Comparison between Willingness-to-Pay expressed by a panel of forestry experts and by a sample of non-expert respondents in a pilot survey conducted for a contingent valuation study. Mediterranean Journal of Economics, Agriculture and Environment 2005, 4:42–51. 102. Abbot JIO, Mace R: Managing protected woodlands: fuelwood collection and law enforcement in Lake Malawi National Park. Conserv Biol 2001, 13:418–421. 103. Kavanagh J, Oliver S, Lorenc T, Caird J, Tucker H, Harden A, Greaves A, Thomas J, Oakley A: School-based cognitive-behavioural interventions: A Submit your next manuscript to BioMed Central systematic review of effects and inequalities. Heal Sociol Rev 2009, 18:61–78. and take full advantage of: 104. Petticrew M, Tugwell P, Kristjansson E, Oliver S, Ueffing E, Welch V: Damned if you do, damned if you don’t: subgroup analysis and equity. J Epidemiol Community Health 2012, 66:95–98. • Convenient online submission 105. Reid R, Stone M, Whitely T: Economic value of wilderness protection and • Thorough peer review recreation in British Columbia. (Victoria BC): FRDA Working Paper; 1995. • No space constraints or color figure charges doi:10.1186/2047-2382-2-19 • Immediate publication on acceptance Cite this article as: Pullin et al.: Human well-being impacts of terrestrial • Inclusion in PubMed, CAS, Scopus and Google Scholar protected areas. Environmental Evidence 2013 2:19. • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Environmental Evidence Springer Journals

Loading next page...
 
/lp/springer-journals/human-well-being-impacts-of-terrestrial-protected-areas-9lYozM2ayg

References (134)

Publisher
Springer Journals
Copyright
Copyright © 2013 by Pullin et al.; licensee BioMed Central Ltd.
Subject
Environment; Environmental Management; Ecosystems
eISSN
2047-2382
DOI
10.1186/2047-2382-2-19
Publisher site
See Article on Publisher Site

Abstract

Background: Establishing Protected Areas (PAs) is among the most common conservation interventions. Protecting areas from the threats posed by human activity will by definition inhibit some human actions. However, adverse impacts could be balanced by maintaining ecosystem services or introducing new livelihood options. Consequently there is an ongoing debate on whether the net impact of PAs on human well-being at local or regional scales is positive or negative. We report here on a systematic review of evidence for impacts on human well-being arising from the establishment and maintenance of terrestrial PAs. Methods: Following an a priori protocol, systematic searches were conducted for evidence of impacts of PAs post 1992. After article title screening, the review was divided into two separate processes; a qualitative synthesis of explanations and meaning of impact and a review of quantitative evidence of impact. Abstracts and full texts were assessed using inclusion criteria and conceptual models of potential impacts. Relevant studies were critically appraised and data extracted and sorted according to type of impact reported. No quantitative synthesis was possible with the evidence available. Two narrative syntheses were produced and their outputs compared in a metasynthesis. Results: The qualitative evidence review mapped 306 articles and synthesised 34 that were scored as high quality. The quantitative evidence review critically appraised 79 studies and included 14 of low/medium susceptibility to bias. The meta-synthesis reveals that a range of factors can lead to reports of positive and negative impacts of PA establishment, and therefore might enable hypothesis generation regarding cause and effect relationships, but resulting hypotheses cannot be tested with the current available evidence. Conclusions: The evidence base provides a range of possible pathways of impact, both positive and negative, of PAs on human well-being but provides very little support for decision making on how to maximise positive impacts. The nature of the research reported to date forms a diverse and fragmented body of evidence unsuitable for the purpose of informing policy formation on how to achieve win-win outcomes for biodiversity and human well-being. To better assess the impacts of PAs on human well-being we make recommendations for improving research study design and reporting. Keywords: National Park, Reserve, Community, Governance, Conservation, Poverty, Development, Biodiversity, Systematic review * Correspondence: a.s.pullin@bangor.ac.uk Centre for Evidence-Based Conservation, School of Environment, Natural Resources and Geography, Bangor University, LL57 2UW Bangor, Gwynedd, UK Full list of author information is available at the end of the article © 2013 Pullin et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Pullin et al. Environmental Evidence 2013, 2:19 Page 2 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Background available evidence on their human well-being impacts. The concept and practice of protecting areas for the That there have been and continue to be, some major purposes of conservation has been at the heart of con- negative impacts on local communities caused by the es- servation policy since its inception in the 19th Century. tablishment of some protected areas, is not in dispute. The idea that intervening to protect areas from human However, of particular interest to policy makers is the bal- activity is an effective way of conserving species and ance of positive and negative impacts on human livelihoods habitats and preventing habitat loss and species extinc- that arise from the PA establishment, the distribution of tion is arguably as pervasive today as it was when the these benefits and costs, and the factors that might cause first protected areas (PAs) were established (MEA 2005). this to vary. The central place of PAs in the conservation movement We report here on the conduct and outcome of a sys- has been reflected in the increase in both the number of tematic review of evidence for impacts on human well- PAs and the area of land and sea placed under protec- being arising from the establishment and maintenance tion. The proportion of total area of land under some of terrestrial PAs. Establishing the state of the evidence form of protection has now reached nearly 13% [1,2]. base through systematic review will inform decision mak- The process of protecting areas from the threats posed ing concerning future investment in PAs and future re- by human activity will by definition inhibit some human search needs. The review question was formulated by the actions and therefore has the potential to negatively impact Scientific and Technical Advisory Panel (STAP) of the human well-being. There are many historical records to Global Environment Facility (GEF). suggest that few PAs were uninhabited wildernesses before At the outset of the review the following broad cat- designation, and the early history of PAs, for example in egories were posed by STAP as a starting point for a the USA and East Africa, is of forced eviction and persecu- conceptual model concerning the potential impact of PA tion of local communities by colonial powers [3]. This sce- establishment (or change in PA status) on people and/or nario has continued in some countries with, in some cases, the communities of which they are part. These categor- the colonial powers being replaced by multinational corpo- ies and associated questions were used to guide develop- rations or international conservation Non-Governmental ment of specific evidence inclusion criteria (categories Organisations (NGOs) [4]. The problem of negative im- were subsequently modified based on an in-depth under- pacts of PAs on human well-being gained official recog- standing of the literature in order to code and present nition in the Convention on Biological Diversity (CBD) the available evidence, see methods). arising from the Rio Summit in 1992. The principle that PAs should do no harm to local people was only estab- Livelihood strategies lished at the World’s Park Congress in 2003 (at which, the Did the establishment or change in status of the PA or Durban Accord was proclaimed). Prior to 1992, the estab- management activities within the PA generate or decrease lishment plans of PAs did not normally have objectives specific production opportunities (e.g. more demand for concerning human well-being. labour, herding activities and associated products no longer However, negative impacts could be balanced by posi- viable, new demand for particular food, handicraft, services tive impacts as PAs may also improve human well-being or products etc.)? Did the PA influence (i.e. increase or de- and alleviate poverty [5]. By preventing conversion of crease) migration generally, and of particular social groups? natural habitats, PAs may improve the provision of some Has this differentially impacted (positively or negatively) valued ecosystem services to some users. For example, the most vulnerable groups in local communities (e.g. downstream farmers may benefit from conservation of women, children, poorest sectors of the community)? forested watersheds [6]. PAs may also directly introduce new livelihood options into a region through the expan- Social capital sion of tourism or research, or improvements to infra- Did the establishment and management of the PA affect structure may indirectly result in economic development. the development of social networks? Did it positively or Recently there has been considerable debate on whether, negatively impact education and capacity building, e.g. apart from their effects on global environmental benefits, by generating or decreasing opportunities for formal and/ the net impact of PAs on human well-being at local or re- or informal education? Has PA establishment differentially gional scales are positive or negative [7-10]. There is con- affected more vulnerable groups (e.g. women, children, cern that continuing with a policy of PA establishment poorest sectors within local communities) in a positive or could conflict with goals of poverty alleviation [11]. The negative way? CBD Aichi targets include a target of 17% of terrestrial and inland areas covered by well-managed PAs by 2020. Empowerment Future policy decisions on support for PA establishment Did the PA empower or disempower local communities and future management need to be informed by the best and any particular social groups? Were new organizations/ Pullin et al. Environmental Evidence 2013, 2:19 Page 3 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 institutional arrangements that represent the interests In this review we also recognise that the impacts of of communities and any particular social groups created PAs on human well-being are likely to be highly context or existing ones undermined? Have these organizations dependent and vary over the lifetime of the PA. For ex- developed activities aimed at improving their livelihoods ample, initial establishment may have significant and im- (e.g., legislation to support local livelihoods, land tenure, mediate negative impacts on the communities within the co-management of local resources, other social benefits) PA but in the longer term there may be positive impacts or have existing activities been negatively affected? on well-being. Similarly, spatial context of the PA and communities within and around it will influence impact. Human rights Some communities will be located in a PA, some in buf- Whilst recognising that the scope of human rights is very fer zones around the edge and others more distant. broad, in this review we focus on the following question; Other PAs may be close by and exert an influence. Pres- were the rights of any local stakeholders, affected either ence or absence of infrastructure may influence the positively or negatively by the PA (considering, e.g., rights spatial scale over which impacts occur. Impacts that are to education, adequate access to food, clothing, health, negative ‘locally’ may have to be balanced against im- choices)? pacts that are positive but experienced more widely or more remotely. Access to ecosystem goods and services and natural Given the diversity of PAs, contexts of establishment resources essential for well-being and of communities affected, this review did not aim to Did the PA have any positive or negative impact on ac- estimate a mean effect of PA establishment on local cess to ecosystem services and natural resources? For communities but to critically appraise the range of ef- example were there changes in the cost (in terms of fects that have been reported. The nature and scope of money, level of effort, or time) in obtaining firewood, the question posed in this review (including as it does is- clean water, and other resources/services? Was access sues of empowerment and social capital) suggested to to culturally significant places (e.g. sacred grounds) af- the Stakeholders and the Review Team that a combin- fected? Did self-sufficiency in food (by locally cultivating, ation of quantitative and qualitative evidence might pro- hunting, raising animals or gathering) or access to medi- vide the most reliable base with which to inform future cinal plants change? Has this been a consequence of the decision making. direct impact of the PA through legal prohibition of ac- cess or indirect as a consequence of changes in infra- Objective of the review structure and/or institutions? Have any of these positive The primary research question was: or negative impacts been disproportionably high or low What are the human well-being impacts of terrestrial on particular sectors of society? protected areas? Measuring impacts needs to take into account that ter- We aimed to synthesise the empirical evidence of posi- restrial PAs are varied in their status, management and tive, negative or neutral impacts of PAs on human well- objectives. For example, many protected areas will have being at the local to regional scales, with emphasis on been established with a primary aim of landscape or bio- local communities and contemplating as broad a defin- diversity conservation, possibly at a time when potential ition of well-being as possible (see categories to be con- impacts on local human well-being were not widely con- sidered below). sidered. Types of PAs have been categorised by the World We were also interested in two secondary questions; Commission on Protected Areas (WCPA) of the Inter- 1. How are costs and benefits distributed among and national Union for the Conservation of Nature (IUCN) within local communities living inside and in the buffer and are used to classify entries in the World Database of zones of PAs (by socio-economic status, gender, age etc.)? Protected Areas (http://www.protectedplanet.net/). The PA 2. How do costs and benefits vary with governance, re- categories reflect the range of management objectives of source tenure arrangements, and site characteristics? As different PAs, from strict protection of nature or wilder- the aim is to assess the human well-being impacts of PAs ness, conservation of ecosystems, protection of national in their current form and capture lessons for future inter- monuments or management of important habitat or spe- ventions, relevant evidence may be provided by studies cies (categories I-IV) through to objectives which place hu- that generate hypotheses about impact, as well as studies man use of the landscape much more centrally (categories that test hypotheses about impact. V-VI). Progressively, since 1992 the requirement of PAs to ‘Deliver benefits to resident and local communities consis- Methods tent with the other objectives of management’ has become Design of review a common objective [12] with the categorisation represent- An a priori protocol was established, peer reviewed and ing a gradation of increasing human intervention. posted on the CEE website by CEE [13]. This review was Pullin et al. Environmental Evidence 2013, 2:19 Page 4 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 divided into two separate processes; a qualitative synthe- Where studies are reported in other languages, rele- sis of people’s views, observations and related documen- vance was assessed initially from their titles and ab- tary evidence (led by the EPPI-Centre) and a synthesis of stracts (translated if necessary). Non-English language quantitative evidence of impact, including people’s atti- articles that could not be considered in this review were tudes and views (led by Bangor University). Henceforth recorded for future assessment (see Additional file 2). these will be referred to as ‘qualitative’ and ‘quantitative’ For the qualitative synthesis, non-English papers were as reflected in the protocol and the methods below. A sin- translated using Google Translate online translation tool. gle systematic search was conducted to identify potentially Google Translate was tested for accuracy using a sample relevant studies to answer the review questions. Initial report that included both and English and non-English screening of title was undertaken by Bangor University. abstracts. By translating the non-English abstract of the Further supplementary searching, screening, coding and paper and comparing with the English abstract given in synthesis were conducted independently by each of the re- the same paper we were able to see how well the tool view teams. Thus, the searching phase and initial screen- ‘understood’ and conveyed the same meanings in the text. ing by title were not separate, but thereafter the report Google Translate was found to be a fairly accurate tool, al- sections reflect the different approaches. though more accurate on some languages than others. Using Google Translate, non-English articles were trans- Searches lated then screened on full text and included or excluded. Search terms were selected to capture all information pertaining to PAs (the intervention) and relevant outcomes Estimating the comprehensiveness of the database search associated with human well-being. No search terms for the All search terms were included in a search string for- study population, in this case local communities, were matted according to requirements for searching in the used as these were not likely to be included in the title or Web of Knowledge database. A set of 18 references were abstract and therefore their incorporation in the search identified by the review team as being relevant to the might have risked excluding relevant studies. The use of review question and used as a ‘test library’ to check the asterisk denotes a wildcard character that prevents whether the search strings captured the expected stud- plurals or alternative word endings being excluded. The ies. The test library did not consist entirely of studies search terms are as follows: that would be included at every stage of the systematic review and data extraction. Instead, there were some Exposure: protected area*, nature reserve*, wilderness studies which were expected to be captured in the area*, national park*, natural monument*, natural search but did not include a suitable intervention, or feature*, management area*, world heritage site*, were not primary literature. Whilst it was not required biosphere reserve*, biodiversity conservation. for the database search to capture these articles, it was a Outcomes: poverty, human well*, socio-econom*, valuable test of how specific the search strings were to econom*, human health, livelihood*, social capital, the subject area whilst retaining the sensitivity required social welfare, empowerment, equity, ecosystem to capture all available literature. The balance between service*, perception*, attitude*. specificity and sensitivity is key to achieving a comprehen- sive search whilst avoiding capturing too many spurious Where the databases searched could accommodate all hits. The search string employed (see Additional file 1), search terms simultaneously, they were separated using consisting of the terms listed above, was the last of 16 iter- Boolean operators. The search terms are grouped by inter- ations and was arrived at using the test library to evaluate vention and outcome and we refer to these groups as ‘sets’. the search results returned by Web of Knowledge. The Boolean operator ‘OR’ was used to separate search terms within a set and the operator ‘AND’ used to separate Publication databases searched the two sets thus ensuring that the search returned refer- The search included the following computerized databases: ences including at least one term from each set. In many cases, the literature databases could not ac-  Web of Knowledge commodate all the search terms above and therefore  Scopus search strings had to be adapted as necessary. In some  Agricola cases, this meant using pairs of search terms, one from  CAB Abstracts each set, separated by ‘AND’. Occasionally, search terms  PubMed had to be entered individually and/or the database did not  Econlit accommodate Boolean operators, in these cases only the  Directory of Open Access Journals (DOAJ) search terms pertaining to the intervention were used. Full  LILACS (Latin American & Caribbean Health details of the search are recorded in Additional file 1. Sciences Literature – Spanish language) Pullin et al. Environmental Evidence 2013, 2:19 Page 5 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 A supplemental search was undertaken using an ex- level where not already included in the above database panded set of intervention and outcome terms to test and website searches. how many extra relevant articles they might provide. The following intervention and outcome terms were en- Search update tered into Web of Knowledge, and articles obtained were In May 2013 an update to the searches was undertaken assessed as above. to supplement the review with all relevant evidence pub- lished since the original search was carried out in October/ Exposure: ecotourism, eco-tourism, eco tourism, November 2011. wildlife tourism, trophy hunting, conservation The original search string (October 2011) was com- corridor*, community conservanc* bined with the supplemental search string (March 2012) Outcomes: natural resource, farm, enterprise*, human and entered into the database Web of Knowledge. Details AND migration, gender, discriminat* of this search string and the number of hits returned can be found in Additional file 1. Internet searches conducted Internet searches using online search engines were not Screening performed as part of this review. It was felt that the Inclusion criteria non-transparent and transient nature of search engine The following inclusion criteria were applied to those ar- functionality would result in an unacceptable lack of rep- ticles captured by the search. licability. We therefore focussed on an extensive search of specialist sources (see next section) to identify grey litera- Populations Human populations/communities currently ture and reduce possible publication bias. or previously living in or near terrestrial PAs. In order to avoid subjectivity in deciding what is ‘local’, studies were Specialist searches included if the subjects were any type of community A list of thirty-three relevant stakeholder organisations within the same country as the PA(s) that formed the was identified and their websites searched by members focus of the paper. of the qualitative and quantitative review teams and/or contacted by email with a request for relevant literature Exposure Establishment/implementation, presence or (Table 1). change in status of terrestrial PAs areas with IUCN clas- Websites were searched using a hierarchical approach, sifications I-VI as defined in the World Database of Pro- from the original search string down to individual words. tected Areas. Specifically, studies were included only Where this was not possible, the following individual where the collection of data on the impacts of PAs was terms were searched; national park*, protected area*, and undertaken during or after 1992. This included changes reserve*. Boolean operators and wildcards were used in status of PAs. The impacts of individual PAs were where possible. All returns were assessed except where treated independently where possible but in some cases the searches resulted in large numbers of results (i.e. > 100) studies looked at multiple PAs in such a way as to prevent when the first 50 returns were scanned for relevance. In the separation of impact. Projects that were established each case, results were assessed at title and then full text, within or surrounding PAs as a result of the existence of for relevance. Full texts were visually scanned for relevant that PA were also included. data, along with within-document searches for the follow- ing terms; park, protect, and reserve. Due to the disparate Types of study Studies were considered for inclusion in modes of operation of the website search engines, a wide this review if they satisfied one of the following criteria: range of approaches was necessary and is documented in Additional file 3. a) Evaluate the impact of PAs on human well-being (outcome evaluations which use the following study Bibliographic searches designs with appropriate comparators (i.e. a Where studies identified in the search used data re- comparable state with which the intervention or ported in earlier primary literature, the original reference exposure can be compared): Randomised was sought and included in the data extraction process. Controlled Trials (RCTs) / Controlled Trials (CTs), These references were catalogued in a separate Endnote control-intervention site comparisons, interrupted library. Five key reviews were identified from the above time series, Before-After/Control-Intervention searches. The references within the five relevant reviews (BACI) designs); were examined to identify pertinent articles for inclusion b) Reported economic valuation of welfare changes in in the review (for details, see Additional file 4). These monetary terms: based on stated or revealed references were assessed at title, abstract, and full text preferences or production function approaches, in Pullin et al. Environmental Evidence 2013, 2:19 Page 6 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Table 1 Relevant stakeholder organisations and their websites searched for relevant literature and contacted with a request for information Organisation Web Site Email Asociación Campesina e Indígena de Agroforestería http://www.acicafoc.org/ Yes Comunitaria Centroamericana Conservation International http://www.conservation.org Yes Cultural Survival www.culturalsurvival.org/ Yes Department for International Development http://www.dfid.gov.uk/ Yes Development Experience Clearinghouse (DEC) http://dec.usaid.gov/index.cfm Yes EcoTerra www.ecoterra.net/ Yes Eldis (Livelihoods Connect) http://www.livelihoods.org Yes Environment Knowledge Hub http://ekh.unep.org/ No European Tropical Forest Research Network http://www.etfrn.org/etfrn/index.html Yes First Peoples Worldwide www.firstpeoplesworldwide.org/ Yes Food and Agriculture Organisation http://fao.org Yes Forest Peoples Programme http://www.forestpeoples.org/ Yes GEF Evaluation Office http://www.thegef.org/gef/gef_Documents_Publications Yes GEF Small Grants Programme http://sgp.undp.org/ No Indigenous Knowledge http://indigenousknowledge.org/ Yes International Fund for Agricultural Development http://www.ifad.org/ Yes International Institute for Environmental Development http://www.iied.org Yes International Union for Conservation of Nature http://www.iucn.org Yes Nature Valuation and Financing Network http://www.naturevaluation.org Yes Overseas Development Institute http://www.odi.org.uk/ Yes Pacific Forestry Centre http://www.pfc.cfs.nrcan.gc.ca/ Yes Poverty and Conservation http://povertyandconservation.info Yes Poverty-Well-being Platform www.poverty-well-being.net Yes RECOFTC – The Center for People and Forests http://www.recoftc.org/site/ Yes Rights and Resources Initiative http://www.rightsandresources.org/ Yes Survival International www.survivalinternational.org/ Yes Tropenbos International http://www.tropenbos.org/ Yes United Nations Development Programme http://www.undp.org/ Yes United Nations Environment Programme http://www.unep.org/ Yes United States Agency International Development http://www.usaid.gov/ Yes Waldbau-Institut, University of Freiburg http://www.waldbau.uni-freiburg.de/ Yes World Bank http://web.worldbank.org Yes (http://documents.worldbank.org/curated/en/home) World Conservation Monitoring Centre http://www.unep-wcmc.org/ Yes which the comparator may be modelled and/or methods for which were adequately reported (such as implicit in survey responses (hypothetical); national census data). c) Sought to identify PA factors that influence human well-being or Additional inclusion criteria for quantitative evidence d) Sought evidence to support explanations or meaning Comparators Comparators were classified as temporal, of impact from people’s views about PAs and human spatial or modelled/hypothetical. Temporal comparators in- well-being, and observations or related documentary cluded time series, before and after, change over time (i.e. analysis single time point), and reported/perceived changes. Spatial comparators included dichotomous (i.e. inside/outside and We included studies that obtained data through direct near/far) and continuous (i.e. linear distance) variables. measurement, self-reported measures by respondents, BACI comparators included both spatial and temporal com- and other data sources that were not accessible, but the parisons. Studies may account for confounding variability Pullin et al. Environmental Evidence 2013, 2:19 Page 7 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 between comparator and exposure populations by using contained evidence of people’s views and/or quantitative matching techniques, testing for differences in confounding evidence of impact as described in the following sections. variables, or including these variables in statistical models. Whilst economic valuation studies may lack before- Screening articles for qualitative evidence after or matched site comparators, economic valuations Following preliminary screening by title (3.3.2) articles are inherently relative to a counter-factual (which may be were screened for evidence of people’sviews of PAsbyab- modelled or subjective). Not all of these inherent compar- stract and then full text. Any articles missing an abstract ators are suitable for this review, however. Many studies were accepted for full text screening. The inclusion and purport to value a PA. However, on closer inspection, they exclusion criteria were re-applied to the full texts (70.3% are often valuing the benefits of the ecosystem-protected were obtainable) and those that did not meet inclusion area complex as a single entity, and may present no infor- criteria were excluded. A coding tool was developed in mation with which to estimate the value of the ecosystem order to ascertain what relevant information was held in the absence of the PA as an appropriate comparator. within articles (see Additional file 5). ‘EPPI-Reviewer’ These studies were rejected on the grounds that they lack (Thomas et al. 2010) software was used for screening an appropriate comparator (the comparator is usually the using a single web location to house the documents and complete and sudden non-existence of the ecosystem, monitor progress of the review. Rejected articles and their which is deemed inappropriate for the purposes of this re- corresponding reasons are listed in Additional file 6. view). Our minimum inclusion criteria are that the study The systematic search and screen identified many studies makes some attempt to model what would happen to the relevant to the review question. The studies identified cov- ecosystem in the absence of the PA, for example, estimat- ered a broad range of geographical areas, types of PAs, and ing rates of degradation before the PA was established and sampled a diverse population group using a variety of extrapolating these into the future to calculate that portion qualitative research methods. The characterisation of stud- of the total value of the site which is attributable to the ies by methodological and contextual features provided the PA. Other economic valuation studies may value oppor- starting point for deciding which studies to include in the tunity costs of the PA, and in these instances the instant- qualitative synthesis (see section Studies included in the aneous loss of access to resources following establishment qualitative synthesis for further details). of the PA constitutes an appropriate comparator. Descriptive information about included studies was collected and presented as a ‘systematic map’ of research Outcomes Specific human-well-being indicators linked to into explanations and meaning of impact to people living those broad questions set out in the Background section in or near PAs. The ‘map’ provided a basis for informed (e.g. Livelihood strategies, social capital, empowerment, discussion and decision making between both review human rights, access to ecosystem services and natural re- teams about the focus of the qualitative synthesis which sources). Examples are measures of income, education, provides a detailed investigation of a more focused subset health and other mainstream socio-economic indicators; of this wider literature. As the synthesis was focused and World Bank and United Nations Development Programme narrowed down, a second set of inclusion criteria was de- (UNDP) human development index measures and mea- veloped and applied to the studies initially identified. sures that consider health, longevity, education, gender Understanding perceptions of the impact of protected equity, food security, livelihood diversity, subjective/re- areas requires appropriate research methods and full ported measures of well-being, resilience, measures of so- reporting of context. Further, understanding the mean- cial capital, and indicators of human rights. ing of protected areas for well-being requires qualitative data collection methods that allow people to express Additional exclusion criteria for qualitative evidence their views freely rather than merely responding to pre- Studies of people’s views were excluded if they: focused determined categories, and where the analysis provides solely on the development or validation of a measurement ‘thick’ or ‘rich’ findings. ‘Thick’ findings have been defined tool without also presenting views separately from the val- as a rich, detailed description of specifics (as opposed to idation of the tool, or reported trials or other outcome summary, standardization, generalization, of variables); “it evaluations, unless it was clear from the abstract that they captures the sense of what occurred and the drama of collected data about views as part of a process evaluation. events, thereby permitting multiple interpretations” [14]. We therefore selected for in-depth review those studies Preliminary screening process using qualitative research methods which elicited the Articles captured by the searches were stored in an End- views of people living in or near protected areas where: note library. Replicates were removed and titles examined for relevance to the inclusion criteria. Potentially relevant  The IUCN category, and the date this was assigned, titles were then separately screened as to whether they was known Pullin et al. Environmental Evidence 2013, 2:19 Page 8 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 The methods and date of data collection were near PAs. Framework synthesis does this because dimen- clearly reported sions of a framework for structuring the synthesis can be The analysis produced ‘thick’ or ‘rich’ findings identified by policy interests, research interests and con- Findings were linked to specific people (e.g. cepts emerging from the data [16]. Framework synthesis distinguishing ethnicities, employments or locations) also takes into account differences in context such as the different countries hosting PAs, the different cat- Screening articles for quantitative evidence egories of PA and the different times and stages in their Following preliminary screening by title (3.3.2) articles history that each area was studied. Based on framework were screened for quantitative evidence of impact by ab- analysis of primary research data [17], an initial concep- stract and then full text. Any articles missing an abstract tual framework, either built from assumptions held by were accepted for full text screening. Where information stakeholders or borrowed from related bodies of know- in titles or abstracts was insufficient, articles were ac- ledge, evolves during the synthesis as the reviewers be- cepted for screening at full text. come more familiar with the literature being reviewed A kappa test for consistency of decision regarding in- [16]. The five neat stages of framework analysis described clusion/exclusion at abstract level returned a moderate for primary research [18] are in practice a more iterative level of agreement (n = 100, kappa = 0.442, 95% C.I. 0.270 process for making sense of a pile of studies [19]. The to 0.614) between two independent reviewers (SED and process starts by delving into the abstracts then full texts ASP). Twenty-eight studies were treated differently by the to see the key issues and recurrent themes, some of which reviewers and these were re-read and the reasons for in- emerge from the data while others are purposely sought clusion or exclusion by each reviewer discussed. These in response to the review question or prior knowledge. discussions were used to come to consensus on the inter- Once most of the key themes are identified, each study is pretation of the titles and abstracts and informed the sub- coded with themes and the literature is sub-divided into sequent inclusion process. sections for in-depth analysis. The conceptual framework Full text screening resulted in the rejection of articles is refined in light of these themes, and coding continues it- that were not pertinent to the review in hand. Reasons for eratively but systematically as the framework becomes in- exclusion were recorded (Additional file 6). Relevant re- creasingly coherent and simultaneously accommodating views were identified for use in the bibliographic searching of the available data. Subsections of the framework shape phase described in Section 3.2. data tables under key themes in order to develop concise We attempted to obtain all articles deemed relevant summaries. Conclusions are drawn from the themes and during abstract-level screening, for assessment at full associations between them. text. Those articles that we failed to obtain within the limitations of time and resources of the review are listed Data extraction strategy in Additional file 2. Included studies were inspected in detail for the focus of their findings. The coding framework described above Qualitative synthesis of explanations and meaning (Additional file 5) was applied to each study. This frame- of impact work covered: health (mental health, physical health, safety There are three distinctive approaches to synthesising and wildlife conflict); socio-economic position (livelihood findings from research: aggregating the findings of very strategies, social capital, human rights, empowerment, en- similar studies where the key concepts are clearly defined vironmental capital, governance); inequalities; and context. in advance; configuring the findings of dissimilar studies Study text was extracted in the form of: participant data by investigating the implications of the differences in their (e.g. direct quotes); authors’ descriptions of findings; and methods, context and findings in order to define key con- authors’ conclusions, implementations and recommenda- cepts and develop theoretical understanding; and a com- tions. The coding framework was tested by two or more bination of the two which can be applied where studies reviewers independently coding successive studies, com- have important differences, but nevertheless share some paring and discussing how they applied the codes, and re- important key concepts of well-being [15]. fining the framework and definitions of codes until a We chose this third approach because some (but not shared understanding was achieved across the team. all) key concepts were decided and defined in advance: PAs with their clearly defined categories; populations Synthesis and outcomes of interest. Despite these concepts being Synthesis began by developing the initial key concepts identified by the funders, a synthesis method was re- into a coherent framework that could accommodate the quired that also allows new concepts to emerge from the main approaches to establishing and maintaining PAs. data in order for the impact on human well-being to be Figure 1 illustrates two contrasting approaches to terres- understood from the perspective of people living in or trial PAs. The first is governance models that are imposed Pullin et al. Environmental Evidence 2013, 2:19 Page 9 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Figure 1 Conceptual framework for structuring the qualitative synthesis. and enforced by external authorities and the second is of these three areas. Understanding the context of each participatory approaches such as community-based nat- study and the interplay between different concepts required ural resource management. The latter has been developed reviewers to return to the full report of each study, along- in response to tensions raised by restrictions and resettle- side the text extracted for each code, and distil the key ment having a deleterious impact on economic and envi- study characteristics and findings. The resulting coherent ronmental capital (agriculture, logging, tourism) and social text was placed appropriately within the emerging frame- capital (family and community relationships). In contrast, work (Figure 1). Where studies contributed substantial participatory approaches, within a regulatory framework, findings to different elements of the framework, these find- seek a vision of sustainability through building on social ings were distributed across the framework to allow each capital and good health. That vision of sustainability may element to draw on all relevant studies. As the framework be on a small scale, such as with IUCN category VI PAs, became more coherent with growing understanding of the where the aim is sustainability within the boundaries. Al- relationships between its different elements, the elements ternatively it may be on a wider scale, across IUCN cat- were reordered to present governance issues first, and then egories I or II and the wider area. These models are set well-being issues. The findings were summarised for each against a backdrop of an evolving consensus about human element of the framework, taking into account the govern- rights that began with protecting individual civil and polit- ance and time context of the studies. The contexts of the ical rights, followed by the support for economic, social studies contributing to the synthesis were recorded noting: and cultural rights and then setting these rights within a the location (name and country of the protected area); broader framework to harness the combined efforts of in- IUCN category (present category, years assigned and chan- dividuals, states and other bodies to build collective rights ged); governance (national government, local government, to self determination, heritage and equity [20]. This syn- private or community); and timing of study (year data was thesis aims to assess the extent to which these models and collected and the time lapse since the last assignment of their anticipated impacts are supported by perceptions of IUCN category). Finally, all studies were revisited for add- impact on human well-being held by people living in or itional evidence that appeared significant in light of emer- near protected areas and others working alongside them. ging findings. The framework was populated with research findings by dividing the studies into three overlapping subsets whose Synthesis of quantitative evidence findings clearly addressed: participation (human rights, em- Data extraction strategy powerment and governance), health (physical, mental and A preliminary data extraction phase was carried out dur- safety) and socio-economic position (wildlife conflict, liveli- ing full text screening, where each reported study was hood strategies, social capital and environmental capital). categorised in terms of which broad outcome types were Each set was taken by a different member of the team to represented in the quantitative data. These outcomes seed the synthesis with studies focussing on findings in one were categorised according to the coding framework Pullin et al. Environmental Evidence 2013, 2:19 Page 10 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 described in (and in detail in Additional file 5). Further- minimum possible value of 0 and a maximum of 8 more, details about the following were ascertained; the (Table 2). The four criteria were as follows: broad study methodology (i.e. self-reported data from close-ended questioning in interviews and questionnaires, i) Clarity of methods; the extent of detail given in the direct data collection, valuation or data taken from other study’s methodology regarding questionnaire design, sources such as national census documents and similar), survey implementation, replicate number and and whether there was a valid comparator. selection, and data analysis Comparator categories (i.e. no comparator, before-after, ii) Study design; the rigour of study design in terms of inside-outside, before-after-control-impacts, econometric) replicate number and sampling, location and choice were used to group the studies for a second stage of de- of replicates and controls, questionnaire design, and tailed data extraction. Only those studies that possessed survey implementation an appropriate comparator were taken on to this second iii)Appropriateness of analysis; the suitability of stage of data extraction, where details of the study meth- analytical techniques such as the implementation odology, potential effect modifiers or reasons for hetero- and choice of statistics, and comprehensiveness of geneity and relevant outcome data were obtained. For all analysis outcomes, averages (means/medians) and variability mea- iv) Implementation bias; the presence of one of the sures (standard deviation, standard error and confidence following biases in measurement and analysis; intervals) were obtained where possible, along with statis- Neyman bias (the intervention in question causes tical test results and sums of counts (i.e. percentages of re- systematic drop-out within the population, e.g. the spondents in agreement). most affected families move away), questioning bias (unbalanced/leading questions), potentially influential variables, recall bias (response affected by Study quality assessment memory). Details relating to study quality were extracted for each study to allow critical appraisal of relevance (external Susceptibility to bias scores were then combined with validity) with respect to the review question and reliabil- an objective weighting by basic study design in order to ity (internal validity). Critical appraisal was conducted in provide a categorical assessment (high, medium or low) two stages. Initially assessment was made as to whether of susceptibility to bias (Table 3). a study possessed one or more of the following: con- founding variables, within-study contradictory results, Potential effect modifiers and reasons for heterogeneity inappropriate implicit comparator, inappropriate popula- Data on potential effect modifiers that were extracted tion, inappropriate spatial comparator, inappropriate sta- from articles, included the following variables; protected tistics, inappropriate temporal comparator, inappropriate area studied, country, IUCN category, protected area time frame, insufficient detail in methods, data linked with size, date of establishment, and history of protection marine ecosystem protection, non-standardised data (in- prior to formal establishment, and residence history and comparable), protected area establishment too recent/not ethnicity of the study population. Where sufficient infor- yet established, data collection prior to 1992, extremely mation about IUCN category, protected area size and small sample size (e.g. one study, where wood extraction year of establishment was not presented in the article it- was quantified in only nine households with one interven- self, data were supplemented from the World Database tion and one control village), superseded by more recent on Protected Areas (WDPA) (http://www.protectedpla- study, unbalanced questioning/questioning bias, and un- net.net/). matched methods (see Additional file 6 for further detail). Studies with one or more of these characteristics were ex- Data synthesis and presentation cluded from the review. Extracted data were presented in narrative synthesis ta- Studies accepted following this first critical appraisal bles that summarise the studies’ aims, methodology, stage were then scored for reliability with respect to the study population, and results (Additional file 7). No fur- question using the concept of susceptibility to bias (the ther quantitative synthesis of data on outcomes was pos- extent to which a study is able to yield an unbiased esti- sible with the available data. mate of effect; [21]). Susceptibility to bias in study design and reporting can affect both internal validity (study qual- Results ity) and external validity (study generalisability). Review descriptive statistics Susceptibility to bias scores were assigned using a re- The main literature search was conducted between 11 peatable, but partially subjective, set of four criteria, each October and 14 November 2011 with an update carried assessed on a scale of 0 to 2, and thus ranging from a out in May 2013. Pullin et al. Environmental Evidence 2013, 2:19 Page 11 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Table 2 Scoring criteria for subjective assessment of susceptibility to bias Assessment criteria 0 1 2 Clarity of methods All sample sizes provided, selection Some information regarding sample Sample size not given in full, sample selection method described, questionnaire size, selection method, or not stated, questionnaire design not discussed design fully disclosed questionnaire design provided Experimental design Good sample size, appropriate sampling Low effective sample size, poorly Very small sample size, pseudoreplication, regime, control and intervention well randomised design, control and non-random sample selection, control and matched, survey appropriately intervention choice not ideal intervention poorly matched/low implemented consistency in sampling over time Appropriateness of Confounding variables accounted for, Confounders only partly Significant confounding variables unaccounted analysis appropriate metrics reported accounted for/only some low for, inappropriate metrics reported, incorrect risk confounders ignored statistical analysis Implementation bias No identifiable bias reported/evident Low level bias present but ignored/ Strong bias present and unaccounted for strong bias accounted for The main search returned 13,248 articles (following unobtainable within the timeline and resource limita- removal of duplicate and marine-oriented articles) and tions of the review. These numbers are represented visu- title-level screening left a total of 1,710 articles. The ally in Figure 3. One hundred and fifty-seven of the search update returned 3,122 articles (following removal articles identified in the May 2013 search update were of duplicates) and title-level screening left a total of 204 obtainable, although five of these were in Chinese and additional articles. From this point forwards, screening could not be assessed (see Additional file 2). at abstract was independent for the two review sections. Following full text assessment 177 articles from the original search, 16 from the supplemental search and 17 Qualitative evidence from the update met the inclusion criteria and were sub- Abstract screening for the qualitative review resulted in sequently critically appraised. Following critical appraisal, the final inclusion of 30 studies as illustrated in Figure 2. articles/studies were rejected at full text for a wide range Four additional studies were included in an identical of reasons (see section Study quality assessment). A list of process following the search update in May 2013. these articles and associated reasons is provided in Add- itional file 6. A final set of 70 articles was included, report- Quantitative evidence ing on 79 studies and these are listed in Additional file 7. Abstract screening for the quantitative review resulted in the final inclusion of 79 studies as illustrated in Figure 3. Comparison of included studies from the two In addition to the above searches 2,345 articles were review processes identified by the supplemental search conducted on 23 The following statistics describe the 30 studies on quali- March 2012: 535 remained following title-level screen- tative evidence of people’s views and 79 on quantitative ing, and 171 following abstract-level screening. Twelve evidence of impact included in the synthesis. studies from this supplemental search were included fol- lowing critical appraisal. Study location Figure 4 displays the countries from Relevant organisation website searches conducted be- which data were collected for the included studies from tween 6 and 29 March 2012 yielded 94 relevant articles the two processes. Frequently studied countries include for the quantitative review. In addition to these searches, Cameroon, China, India, and Uganda for the qualitative 50 relevant articles were identified through bibliographic synthesis and India, Nepal, South Africa and Uganda for checks and secondary sources, yielding a total of 1,164 the ‘quantitative impacts’ review. potentially relevant articles. Of these articles, 76% were retrievable for full text assessment, whilst 275 were Category of PA studied The IUCN categories of the PAs (see below) examined in the included studies are Table 3 Categorization of susceptibility to bias shown in Figure 5. PAs predominantly fell under cat- Susceptibility to bias score egory II in both reviews, with 17 percent either unre- ported by WDPA or not present in the database for the Comparator 0-2 3-5 6-8 ‘quantitative impacts’ review (NB this group includes RCT* (randomised control trials) Low Low Medium only ‘not applicable’ for the qualitative synthesis, since BACI (before-after-control-intervention) Low Medium High studies were only included in this review if the protected Control-intervention/Before-after/ Medium High High area IUCN category was stated). A post-hoc decision to Econometric exclude category III PAs was made as they are small in *There were no RCT studies and this line is only provided for a theoretical perspective number and somewhat atypical of the set of categories Pullin et al. Environmental Evidence 2013, 2:19 Page 12 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Figure 2 Number of articles and studies at progressive stages of inclusion for qualitative synthesis. (specifically protecting natural monuments). Similar pat- Ib Wilderness Area -generally larger than Strict terns were observed in both review processes Figure 6 Nature Reserves, the main objectives of these areas displays PA establishment year, showing a peak in new is to provide an environment in which biodiversity PAs centered around the 1980s. and ecosystem processes (including evolution) are IUCN Protected Area Categories are described as follows: allowed to flourish or experience restoration if previously disturbed by human activity. Human use Ia Strict Nature Reserve - protected from all but light is limited, often allowing only those who are willing human use in order to preserve all geological and to travel of their own accord rather than via geomorphological features of the region and their established touristic activities. Wilderness areas can biodiversity, which is often dense and restricted to be classified as such only if they are devoid of exclusively to scientific monitoring, study or modern infrastructure, although they allow human education. Occasionally Strict Nature Reserves are activity to the level of sustaining indigenous groups of spiritual significance to surrounding communities living wilderness-based lifestyles. in which case the people are generally allowed to II National Park - provide protection for functioning continue the practice of their faith and may be ecosystems (similar to the objectives of Wilderness directly involved in the area's conservation and Areas), but tend to be more lenient with human management objectives, though perpetual human visitation and the supporting infrastructure. National intervention would more suitably be allocated to Parks are managed in a way that may contribute to categories IV or V. local economies through promoting educational and Pullin et al. Environmental Evidence 2013, 2:19 Page 13 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Figure 3 Number of articles and studies at progressive stages of inclusion for quantitative review. recreational tourism on a scale that will not reduce an identifiable species or habitat that requires the effectiveness of conservation efforts. The continuous protection. These protected areas will be surrounding areas of a National Park may be for sufficiently controlled to ensure the maintenance, consumptive or non-consumptive use, but should conservation and restoration of particular species nevertheless act as a barrier for the defence of the and habitats - possibly through traditional means - protected area's native species and communities to and public education of such areas is widely enable them to remain sustainable in the long term. encouraged as part of the management objectives. III Natural Monument or Feature - comparatively Habitat or Species Management Areas may exist as smaller areas, specifically allocated to protect a a fraction of a wider ecosystem or protected area natural monument and its surrounding habitats. and may require varying levels of active intervention Natural Monuments or Features can be natural in including - but not limited to - the prevention of the wholest sense, or include elements that have poaching, creation of artificial habitats, halting been influenced or introduced by humans. The latter natural succession and supplementary feeding should hold biodiversity associations or could practices. otherwise be classified as a historical or spiritual site, VI Protected Landscape/Seascape - cover entire though this distinction can be quite difficult to bodies of land or ocean which engage a range of for- ascertain. As such, the classification then falls into profit activities within the management plan. The two subcategories, those in which the biodiversity in main objective is to safeguard regions that have built uniquely related to the conditions of the natural up a 'distinct character' in regards to their feature, and those in which the current levels of ecological, biological, cultural or scenic value. biodiversity are dependent on the presence of the Protected Landscapes and Seascapes allow a higher sacred sites that have created an essentially modified level of sustainable interaction with -surrounding ecosystem. communities (such as traditional agricultural and IV Habitat/Species Management Area - focus on forestry systems) and should represent an integral more specific areas of conservation in correlation to balance between people and nature. Pullin et al. Environmental Evidence 2013, 2:19 Page 14 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Australia Peru Ecuador South America Quantitative Review Brazil Bolivia Qualitative Review United Stated of America Mexico North America Costa Rica Belize Sweden Slovakia Europe Norway Italy Greece Czech Republic Vietnam Thailand Pakistan Nepal Myanmar Asia Mongolia Indonesia India China Bhutan Bangladesh Zimbabwe Zambia Uganda Tanzania South Africa Nigeria Namibia Africa Mozambique Malawi Madagascar Kenya Ghana Gabon Ethiopia Cameroon Botswana 02468 10 Figure 4 Frequency of countries, arranged by continent, hosting the protected areas within included studies for the two reviews. VI Protected area with sustainable use of natural be developed to adapt to the diverse and possibly resources - based on a mutually beneficial growing range of interests that arise from the relationship between nature conservation and the production of sustainable natural resources. sustainable management of natural resources in (Text adapted from United Nations Environment correspondence the livelihoods of surrounding Programme (UNEP) World Conservation communities. A wide range of socio-economic Monitoring Centre (WCMC) website (http://www. factors are taken into consideration in creating local, unep-wcmc.org/iucn-protected-area-management- regional and national approaches to the use of categories_591.html; accessed 03/10/2012) natural resources. Though human involvement is a large factor in the management of these protected Study timescale Figure 7 displays the survey years for areas, developments are not intended to allow for the included studies for both reviews. A significant num- widescale industrial production. Governance has to ber of studies failed to report the survey year (applicable Pullin et al. Environmental Evidence 2013, 2:19 Page 15 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Quantitative Review Qualitative Review Ia Ib II III IV V VI Unknown/ not reported/ not applicable Figure 5 Frequency of all named PAs by IUCN category within included studies. only to the ‘quantitative impacts’ review, since this was Measured outcome Within the five key themes described an exclusion criterion for the qualitative synthesis). The in Section 1 (Livelihood Strategies, Human Rights, Social reduction in the number of surveys from 2005–2006 for Capital, Empowerment, and Access to Ecosystem Goods both reviews likely relates to a publishing delay. and Services), 24 individual outcomes were identified in an iterative process during screening. The reported outcomes were separated in order to facilitate the description and Further descriptive statistics for the review of quantitative analysis of similar, comparable data. The frequency of these evidence of impacts broad outcome types is shown in Figure 9. Attitude, in- Study comparator Figure 8 displays the range and fre- come, and wildlife conflict outcomes were most common. quency of comparators used in included studies. Spatial Exclusion of studies following critical appraisal resulted in comparators (site comparators and linear distance) were several outcomes identified during full title assessment be- common, along with reported change over time. ‘Before ing unrepresented by data namely, ‘empowerment’ and ‘reli- and after, control intervention’ (BACI) studies were very gion and spirituality’. Other outcomes are represented in rare, with only one instance of a full BACI study. Quantitative Review Qualitative Review Figure 6 Frequency of year of establishment of PAs within included studies. Pullin et al. Environmental Evidence 2013, 2:19 Page 16 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Quantitative Review Qualitative Review Figure 7 Frequency of year of PA survey for included studies. Surveys commence on stated year and are classified as ‘longitudinal study’ if survey spans more than two years. two studies or fewer; ‘gender equity’, ‘interactions with PA Qualitative synthesis of explanations and meaning of impact authorities’, ‘livelihood diversity’, ‘medicinal plants/animals’, Studies included in the qualitative synthesis ‘protect for future’,and ‘resettlement and displacement’. The criteria for selecting studies to include in the in-depth qualitative synthesis is provided in section Screening arti- cles for qualitative evidence. Although not statistically rep- Methodology of data collection Figure 10 displays the resentative of the 306 studies of people’s views identified frequency of different data sources in included studies, (characterised in the map see Additional file 8), studies showing that self reported measures formed the majority with ‘thick’ data were chosen for their ability to provide ex- of data reported (63 of the 79 included studies). planations and meaning of impact for people living in or Figure 8 Frequency of comparators used in included studies. Pullin et al. Environmental Evidence 2013, 2:19 Page 17 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Wildlife Conflict Attitude Income Econometric Food Environment quality Human development measure Employment Water resources Materials Health Education Resettlement and displacement Land access Infrastructure Gender equity Food/Materials 02468 10 12 14 16 Figure 9 Frequency of broad outcome types in included studies. near the PA. They also spanned the range of the broader The other stakeholders not well represented in the literature, in terms of countries frequently studied, as de- studies presenting ‘thick’ findings are visitors to pro- scribed in the map. However, poor reporting makes it diffi- tected areas. Studies seeking visitors’ views often did so cult to conduct further comparisons between the studies using highly structured questionnaires that provided lit- reviewed in-depth with the wider map in terms of IUCN tle understanding of what protected areas mean to them. categories and key dates. The one study of visitors included in the synthesis was The vast majority of included studies were by academic set in a category IV protected area, so was not typical of authors, with a small minority from NGOs, commercial most visitor surveys which, in the map, were most often organisations and protected area authorities. These differ- used in category II. ent perspectives are likely to shape the focus of studies This literature is thus a likely source for understanding and, consequently, their findings. people’s views, and setting those views in context, across Residents’ views were most often sought for IUCN cat- the IUCN categories, although less so for category IV. egory II, where there are strict restrictions on settlements, but rarely for IUCN category VI, where the aim is to bal- Summary findings of qualitative synthesis of explanations ance the aims of conservation and the needs of the local and meaning of impact population. Leaders’ views were most often sought about This section presents a summary of a full synthesis of IUCN category I protected areas, which does not allow findings from studies that provide explanations or ex- human settlements. There was also a noticeable absence plore the meaning of the impact of protected areas on of studies addressing the views of protected area author- people living in or near PAs. The full synthesis can be ities or employees in category IV (similar to the map). found in Additional file 9. It also draws on the perspec- tives of other stakeholders relevant to those experiences and relevant policy and historical documents. The syn- thesis is presented within a conceptual framework that was informed by conservation policies and refined by the emerging research literature (Figure 1). The synthesis process identified the implementation issues and characteristics of well-being that were associ- ated by research participants with either (a) tensions aris- ing from governance models imposed and enforced by external authorities or (b) a vision of sustainability sought through participatory management and empowerment 0 commonly known as community-based natural resource SRM ODS DDC DDC/ODS management. Figure 10 Frequency of data sources in included studies. DDC, The latter approach has been developed in response direct data collection by study authors; ODS, other data sources; to tensions raised by restrictions and resettlement having SRM, self-reported measure via survey responses. a deleterious impact on economic and environmental Pullin et al. Environmental Evidence 2013, 2:19 Page 18 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 (natural) capital and social capital. Participatory approaches, rules and local institutions and typically poor within a regulatory framework, seek a vision of sustainabil- communication ity through building on social capital and good health. That  Memories of forced or induced migrations vision of sustainability may be on a small scale, such as with negatively influence subsequent community IUCN category VI PAs, where the aim is sustainability responses to authorities within the boundaries. Alternatively it may be on a wider  Inadequate or non-existent compensation is a widely scale, across IUCN categories I or II and the wider area. held concern These models are set against a backdrop of an evolving  Externally imposed regulations were incompatible consensus about human rights that began with protecting with traditional regulations and did not take into individual civil and political rights, followed by the support account cultural and social diversity; respect for the for economic, social and cultural rights and then setting rules was greater where they were locally adapted these rights within a broader framework to harness the and allowed income-generating possibilities. combined efforts of individuals, states and other bodies to  A major challenge is distinguishing subsistence build collective rights to self determination, heritage and activities for a sustainable environment from larger equity [20]. This synthesis aims to assess the extent to scale industrial activities. which these models and their anticipated impacts are sup-  Success came from PAs staff having: prior ported by the perceptions of people living in or near pro- experience of working with locals; clear guidelines; tected areas and others working alongside them. extensive training in community development, The characteristics of the included studies are sum- gender issues and a variety of participatory marised in Table 4, and their findings are subsequently methodologies; meeting locals informally and synthesised in sections that match the conceptual frame- working with existing kinship networks. work (see Figure 11) to present a coherent narrative  All studies contributed some evidence related to which explores key themes within governance (source of governance. authority and nature of implementation) and then con- siders the impact on health and well-being in terms of Implementation problems The earliest source of ten- environmental (natural), economic and social capital. sion in the life course of protected areas can be contro- There is necessarily some overlap between the govern- versy about environmental risks posed by local inhabitants. ance issues and the impact on health and well-being. Studies in Australia [22], Nepal [23], Norway [24], Mexico The key messages for both are presented in boxes to [25], Indonesia [26] and USA [27] revealed local inhabi- help readers navigate the breadth and depth of the avail- tants appreciating areas for their aesthetic and spiritual able evidence. The contextual details of each study de- values as well as environmental products and economic scribed in terms of IUCN Category, the date the PA was and leisure opportunities. Mexican farmers in particular established, date of current PA status being assigned, value the land for its provision of food, water, wood and time between current status being assigned and data collec- other products, and they have developed farming styles tion are reported in Additional file 9. As the complexity of along a spectrum of reciprocal relationships between man the links between governance and well-being precludes and nature between wilderness and urbanisation [28]. Diaw simple distinctions between PA approaches and their im- [29] claims that a resettlement policy to establish a IUCN pacts, we focus first on governance and then on well-being, II category park in Cameroon in 1961 was driven by scien- acknowledging that there is necessarily some overlap. The tificmyths of apristineforestwhose protection wasin- key messages for both are presented in boxes to help compatible with indigenous residents despite historical readers navigate the breadth and depth of the available analysis showing that the current forest structure was the evidence. result of sustained use over centuries. Economic arguments favouring resettlement were flawed, with excessively strong Governance assumptions about tourism benefits, flood control, forest Matching study findings to the conceptual framework use, research discoveries, soil fertility and agricultural prod- facilitated their comparison and interpretation to reveal uctivity. Similarly, in Norway residents pointed to a lack of problems and potential solutions during implementation convincing scientific evidence supporting the need for cer- and instances of both benefit and harm. tain protective measures [24]. Residents of Utah claimed the land benefitted from how they had cared for it before it Portraying longstanding residents as an environmental was declared a protected area (Grand Staircase Escalante risk to pristine protected areas is controversial National Monument, IUCN II [27]). In Norway, local tour- PA regulations can flounder as a result of inaccurate ist firms were frustrated by the lack of opportunities to maps and poorly drafted legislation, and further playapart inthe protectedareamanagement(IUCN II). confusion arises from discrepancies between state They claimed that local expertise, based on generations of Pullin et al. Environmental Evidence 2013, 2:19 Page 19 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Table 4 Summary table of included studies in the qualitative synthesis Citation Protected area IUCN Date PA established Date of current Years since IUCN Year of data Governance model Country Category IUCN category category change/ collection establishment Allendorf et al. (2007) Royal Bardia NP II 1969 1976 14 1990 Government: federal/ Nepal national Almudi and Berkes (2010) Peixe Lagoon National Park V 1986 2001-2010 3 2005 & 2007 Cooperatively managed: Brazil collaboratively Bedunah and Schmidt Gobi Gurvansaikhan II 1993 1993 5-7 1998 – 2000 Information not found Mongolia (2004) Bizikova et al. (2012) Slovensky Raj Park II 1998 1998 8 2006 Government: federal/ Slovakia national Bolaane (2004) Moremi Game Reserve IV 1965 1965 32 1997 – 2001 Private Botwana Bruyere et al. (2009) Buffalo Springs Samburu II 1985 1985 20 2005 Government: local Kenya Castillo et al. (2005) El Vizcaíno IV 1988 1988 16 2004–2007 Government: federal/ Mexico national Sian Ka’an 2006 2008 3 Monarch Butterfly 2008 < 1 Davis (2011) Tarangire National Park II 1970 1970 35 2005 – 2007 Government: local Tanzania Diaw (2010) Korup National Park II 1961 1985 18 2003 Government: federal/ Cameroon national First Peoples (2006) Mgahinga National Park II 1930 1930 Not stated Government: federal/ Uganda national Bwindi National Park 1991 1991 14 2005 Government: federal/ Uganda national Awa Forest Reserve Zone VI 1988 1988 10-20 1998 – 2008 Community: indigenous Ecuador Gerritsen (2002) Sierra De Manantlan VI 1987 1987 6 – 11 1993 – 1998 Government: federal/ Mexico Biosphere Reserve national Hartter (2009) Kibale national park II 1932 1993 16 Uganda Haukeland (2011) Rondane National Park II 1962 1962 47 2009 Norway Jotunheimen National Park II 1980 1980 18 Government: national Herrold-Menzies (2011) Caohai Nature Reserve V Protected 1985 1985 13-16 1998-1999, 2000, Government China landscape/ 2000-2001 managed: local Seascape Hoole and Berkes (2010) Etosha National park II 1975 1975 31 2006 - 2007 Government Namibia Keskitalo and Lundmark Sarek National Park; Ib 1909 1909 95 years 2001 Government: federal/ Sweden (2010) national Stora Sjöfallet 1909 1909 95 years National Park; Abisko II 1920 1920 83 years National Park; Pieljekaise 1962 1982 22 years National Park; Vadvetjåkka Pullin et al. Environmental Evidence 2013, 2:19 Page 20 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Table 4 Summary table of included studies in the qualitative synthesis (Continued) National Park; Padjelanta National Park Haparanda-Sandskär Nature Reserve Lunstrum (2008) Limpopo National Park II 1979 1979 25 2004 – 2005 Cooperative: transboundary/ Mozabique collaborative Mbaiwa (2005) Moremi Game Reserve IV 1965 1965 38 1998, 2001, Private: for profit Botswana Mehring et al. (2011) Lore Lindu National Park II 1982 1982 24 2006 – 2008 Government: federal/ Indonesia national Milgroom and Limpopo National Park II 2001 2001 6 – 7 2007 – 2008 Cooperative: transboundary/ Mozabique Sperienburg (2008) collaborative Nguiffo (2001) Dja Wildlife Reserve IV 1950 1950 51 2001 Government: federal/ Cameroon national Ogra (2008) Rajaji National Park II 1983 1983 20 2003 – 2004 Information not found India Ormsby and Kaplan (2005) Masoala National park II 1997 1997 4 2001 Private/ non-profit Madagascar Petrzelka and Marquart- Grand Staircase V 1996 1996 0 1996 Government: federal/ USA Pyatt (2013) national Escalante National 10 2006 Monument Slater (2002) Qwaqwa National Park IV 1992 1992 6 – 7 1998 – 1999 Government: local South Africa Sletten et al. (2008) Mount Elgon NP II 1951 1951 47 2002 Government: federal/ Uganda national + collaborative Spenceley and Kruger National Park II 1926 1930 74 – 75 2000 - 2001 Information not found South Africa Goodwin (2005) Songorwa (1999) Selous Game Reserve IV 1905 1905 91 1996 Community: local Tanzania Stone and Wall (2004) Jianfengling National Forest V 1986 1986 16 years 2001 Government: local China Park and Diaoluoshan National Forest Park in Hainan Province, China Strickland-Munro and Purnululu National Park II 2003 2003 5 years 2008 Government: federal/ Australia Moore (2012) national + collaborative Stronza and Gordillo (2008) Kapawi Reserve; Madidi II Not stated Kapawi Reserve, 1991-2000 Not able to ascertain 2003 Cooperatively managed Bolivia National Park; Bahuaja Ecuador 1995: Madidi Ecuador 5 years:, Bolivia; Collaborative management Ecuador Peru Sonene National National Park 2000: Bahuaja 8 years:, Peru (various forms of Sonene National Park, Peru pluralist influence) Torri (2011) Sariska Tiger Reserve IV 1955 1955 46 2001 (pilot) Government: federal/ India 2007 national Yasuda (2011) Benoue National Park II 1968 1968 40 2004, 2009 Private: for profit Cameroon Pullin et al. Environmental Evidence 2013, 2:19 Page 21 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Figure 11 Summary findings from the qualitative synthesis. managing the area prior to its official status, should play an the definition, might outlaw common traditional practices important role in management [24]. such as the use of snares (metal wire), arrows (steel tipped) The second source of tension is the lack of clarity or rifles. The ban and uncertainty surrounding poorly de- in regulations and maps pertaining to protected areas. fined traditional hunting led to tension and mistrust be- Mehring et al. [26] investigated regulatory institutions in tween locals and conservation agents. In Uganda, the legal two villages in an Indonesian park established in 1982. agreement protecting the Mount Elgon National Park, In one, new regulations about forest land and products established in 1951, was flawed as it failed to refer accur- drawn up by the mayor and customary organisation were ately to maps or related by-laws, statutes or other docu- neither written down, nor completely implemented. There ments [31]. was support for state zoning of the Park to allow trad- Where regulations precluded living within an area, re- itional access to the forest for local people, but disagree- settlement could be forced or induced. In Cameroon, ment about the zone boundaries. Effective village sanctions forced migration and a violent confrontation prompted were considered important, but confusion about when to villagers to accept resettlement outside familiar territories, apply them appropriately arose from discrepancies between against the recommendations of earlier research [29]. state rules and local institutions. In Cameroon, ‘traditional Enacting laws to drive resettlement resulted in an inte- hunting’ was still allowed in 2001 the territories outside grated conservation and development plan that failed, protected areas (IUCN IV) so long as the products were leaving villagers bitter and sceptical. Expulsion to make for personal use, and not sold [30]. However, whether way for the privately managed Moremi Game Reserve in ‘traditional’ hunting referred to the people involved, the Botswana included huts being burnt down as residents weapons employed, or some other characteristic was not got loaded into trucks for relocation outside the reserve clear. Only allowing ‘traditional’ weapons, depending on [32]. Residents were forced to relocate (for a second time) Pullin et al. Environmental Evidence 2013, 2:19 Page 22 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 by the suspension of the provision of all social services but informed later (without being able to give a precise such as water supply, health facilities, shops, schools and date) of the existence of a conservation initiative by the communication services. authorities [30]. Although labelled as ‘voluntary,’ because the term ‘invol- Such problems are not restricted to developing coun- untary’ is politically problematic nationally and amongst tries. In Norway, several residents near a IUCN II category international donors, the resettlement from Limpopo Na- area found the process one-sided and undemocratic be- tional Park was widely recognised as ‘induced’ by planning cause national interests took precedence over local know- blight and economic decline [33]. Although consulted ledge [24]. Petrzelka and Marquart-Pyatt [27] describe the about resettlement, residents’ views were then disregarded. growing anger of residents and their diminishing trust in agencies to make good decisions about the management “Since the park was made we were supposed to leave. of the land after the Grand Staircase Escalante National Since they said that, people don’t construct houses, Monument was established in Utah, USA, in 1996 with no we don’t plant trees. This house was built in 2000 but prior consultation or publicity. Trust diminished further it was never really finished because the park came. over the next ten years later as residents saw roads closed There were trees but we stopped planting and the old and cattlemen’s leases rescinded, despite prior reassur- ones died. [papaya]. No one is investing, not to do ances to the contrary. Restrictions on visiting the park things for nothing. Even now that we have accepted to stoked anger amongst residents who felt ‘locked out of leave, the park does nothing” (p443). our backyard’,saying that ‘law enforcement is gun-toting like we’re a bunch of criminals’. Inadequate or non-existent compensation was a con- Even with management of Kenyan national reserves cern expressed in many studies; for loss of property or being delegated to a local level, and rangers and wardens land in 1960s [34], for access or use restrictions in India claiming to initiate and maintain dialogue, residents in 2007 [35], for environmental protection by owners of are disappointed with the processes of communication Swedish forests (IUCN Ib and II) in 2001 [36], for re- [40]. Most of the protected area staff considered their settlement in 2001 or for loss of crops or livestock in informal word of mouth network sufficient for com- 2007/8 Mozambique (IUCN II) [33], for personal injury municating with local communities about important or property damage from wildlife in South Africa (IUCN management decisions. II) in 2001 [37], and in Tanzania (IUCN IV) where there was no compensation policy at the time of the study in “Of course we cannot conserve this wildlife without 1996 [38]; or for loss of jobs or land in China (IUCN V the help of these communities. There must be that, a established 1986) [39]. For this last case, some compen- good relationship between the park and the sation had been made in the form of new homes, crop community. So we normally go to the [homes and seeds, lump sum payments, and subsidized education, villages], we have meetings with them, tell them that electricity, and water fees, but views differed on its na- these resources are also theirs, these are their ture and adequacy [39]. Residents of the Sariska Tiger resources.” (Park ranger, p55) Reserve (India, IUCN IV) were generally discouraged by staff from claiming compensation for restrictions to ac- In contrast, most community members thought that cess or use forest products [35]. Some villagers were communication between the reserves and communities never told they had a right to compensation. Another, was limited or non-existent, where decisions were made on asking for compensation, one villager was told by a without opportunities to provide input or ask questions. forest officer: In addition to the poor communication between au- thorities and residents, were the difficulties encountered “If you the villagers insist in living in the forest, then with communication between residents. The Qwaqwa be ready to accept as well all the consequences National Park (IUCN IV), South Africa had been estab- deriving from your choice. You could live elsewhere”. lished for the purpose of ecotourism; however, this did not suit the livelihoods of stockholding families and Opposition to the Selous Game Reserve in Tanzania others would have preferred the land to be subdivided grew amongst those losing crops and livestock to wildlife for agriculture. Amongst all this disagreement, some res- without compensation [38]. idents were more able than others to make their voices Poor communication between communities and au- heard, and family conflicts escalated as housing became thorities was typical. Residents of the Dja wildlife reserve more crowded because erecting new homes within the in Cameroon (IUCN IV, established 1951) reported be- park was forbidden [41]. In Lore Lindu National Park, ing “neither informed of nor invited to participate” Indonesia, the weak point for communication was be- (p.208) in their village becoming part of a protected area tween villagers and their leaders [26]. Although the Pullin et al. Environmental Evidence 2013, 2:19 Page 23 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 village leadership was active in the negotiations about activities for a sustainable environment from larger scale the park regulations many ordinary villagers had never industrial activities. As mentioned above, this challenge heard of the agreements. was seen in Cameroon where regulations failed to distin- In Australia, at the Purnululu National Park (IUCN II) guish clearly ‘traditional’ hunting methods for personal relationships between different indigenous groups were consumption from commercial hunting [30]. In Masoala so acrimonious that one group withdrew from the man- National park (IUCN II), Madagascar, residents acknowl- agement of the park [22]. In Slovakia, local authorities edged that some members of the community benefiting near Slovensky Raj Park (IUCN II) tried to make man- from illegal lemur hunting and timber harvest [43]: agement of the park a focus for building relationships and developing mutual trust between different groups ‘people [who] wanted easy money, especially the youth, [42]. Communication problems could be compounded so they went into the park to cut rosewood’ [p160]. by new regulatory arrangements being incompatible with traditional ways. For instance, very few Mexican farmers However, much greater damage was done by industrial applied for resource use permits because the formal bio- scale rosewood logging for international markets, and sphere (IUCN VI) rules competed with customary rules Park agents have limited legal powers over loggers. [28]. The formal rules were generic and did not take into account local variation in natural resource management. “people from all over come to this area to cut These mismatches created feelings of frustration: rosewood, there is no other way to get money than from valuable wood” (Park resident p.160). “The reserve is like a beautiful woman whom you cannot touch. It does not do you any good. The hills ‘[international] demand is driving the outside buyers are rich, but a poor man stays poor” [28] p205. of rosewood, and this is a much bigger issue than lemur hunting.’ (Park Manager, p.162). Contrary to tradition, only people living on the borders of Mount Elgon National Park (IUCN II) in Uganda were Similarly in Cameroon, the impact of residents hunt- given rights of access [31]. The new outsiders were required ing in Dja Wildlife Reserve (IUCN IV) to ensure a diet to pay the ‘insiders’ for access, even though half the insiders that includes animal protein is minor compared with in- thought outsiders should have equal rights of access. tensive industrial logging which opened up forest tracks Even where access was allowed, as in Permululu Na- and thereby provided access for well organised, commer- tional Park, Australia, costs of transportation across long cial poachers to use the tracks for transporting their distances over rough ground could be prohibitive [22]. game to city markets [30] p.208. In Indonesia Mehring et al. [26] attributed some of the Ironically, it was a combined forestry management and difficulties of incompatibility to the government’s indif- community development project in Ecuador that opened ference to cultural and social diversities when managing the eyes of indigenous people to the potential benefits of the Lore Lindu National Park (IUCN II). Indigenous logging; when profits were not what they had hoped for, people respected their own traditional informal rules they started making deals outside the community with that suited traditional use rights and sanctions at the village industrial loggers [44]. level. Elsewhere, more prosperous and ethnically diverse vil- Implementing regulations that have disadvantages for lagers, growing more cash crops, referred not to traditional local communities is challenging enough. The relation- institutions but to economic power structures, where there ship between residents and park officials in Masoala Na- was a widely spread laissez-faire attitude to resource use. tional park officials in Madagascar (IUCN II) was further With forest resources and agricultural land in short supply, damaged by absenteeism amongst staff who, unlike many locals, had the privilege of employment yet lacked villagers had no alternative to using the Park to extend their land. The State’s formal rules interacted with traditional in- training and clear job expectations, and had little inter- formal rules, leading to confusion and conflict. Migrants action with residents [43]. struggled to implement traditional informal rules, and indi- genous people failed to obey state-induced laws. Implementation solutions Blunt regulations imposed Traditional land ownership rights for indigenous popula- by external authorities have been widely disregarded so tions were also contested in Australia, both between local that protected areas have continued to be exploited on residents and protected area management, and amongst domestic and industrial scales. Studies have focused on local residents of different groups living near Purnululu efforts to improve communication, draw on indigenous National Park (IUCN II) [22]. knowledge and share decisions to combine community A major challenge to developing and implementing development with environmental conservation. They have regulations to protect areas is distinguishing subsistence had mixed success. Pullin et al. Environmental Evidence 2013, 2:19 Page 24 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 The Lore Lindu area in Indonesia was established as a wealth etc. Contrary to tradition, only people living on UNESCO Biosphere reserve in 1977 and a national park the park borders were given rights of access. The new (IUCN II) in 1993. Since then participatory approaches outsiders were required to pay the ‘insiders’ for access, have been advocated for managing Biospheres [45] and even though half the insiders thought outsiders should protected areas more widely [46]. Initial efforts to im- have equal rights of access. Conflicts arose from this pose external regulations failed and in the late 1990s, the situation and threatened the agreement’s endurance. In park authority, NGOs and village representatives began such sensitive situations, staff need the socio-cultural skills to negotiate Community Conservation Agreements [26]. to understand, interpret and interact with local people Within designated zones, village conservation councils about livelihoods, conflicts and challenges in appropriate were the bridge between the Park authority and the com- ways. Reports of misuse and corruption remained com- munity for planning, implementing, evaluating and report- mon. Nevertheless, collaborative arrangements improved ing the results of the Agreement. Despite the village relations and benefited biodiversity and livelihoods. leadership being active in the negotiations, communication Sletten’s findings in Mount Elgon Park (IUCN II) are between the Park authorities and the whole community supported by other studies. Elsewhere in Uganda sup- was poor, so many ordinary villagers had never heard of porting community’s transition from a hunter gatherer the agreements. The Agreements covered use of forest to a settled farming community in a culturally sensitive products and land and the village conservation councils way was more likely to result in community satisfaction were responsible for monitoring activities. The council and personal efficacy [44]. Training and capacity build- could employ punishments or sanctions, which were usu- ing by charities and NGOs led to an increase in skills ally based on village traditional rules. Insights into this and knowledge and new income generating activities. system came from NGO interviewees. A collaborative Two NGOs working with local people helped to organ- management approach aimed to minimise the gap between ise efforts around existing kinship networks and this the park management and the people, through participa- community reported the highest states of economic de- tion of local inhabitants and integration of local rules. Re- velopment compared to other communities. At the other spect for the rules was greater where they were ‘more end of the scale these communities were willing to sacri- practical’ having been locally adapted, and allowed income- fice their land claims to join relatives in other areas and generating possibilities. access charitable projects there leaving the settlements In Mount Elgon National Park, Uganda (IUCN II), as struggling to maintain a viable community. in other African countries, a similar ‘fortress management’ In Masoala National Park (IUCN II), Madagascar, resi- or ‘fence and fine policy’, based on systematic evictions, dents who were more familiar with Park staff viewed the exclusions and prohibition of using natural resources, met staff as well as the Park more favourably than residents increasing resistance [31]. Lack of success with ‘fence and who were unaware of staff or who had had negative inter- fine’ policies prompted approaches with greater partici- actions with Park agents [43]. Residents were confused by pation of local people in management and changes in reg- the different NGOs’ responsibilities and changing prior- ulations to legitimise sustainable use. Establishing the ities. A park manager and a local town official both consid- agreements was difficult even with guidelines and training ered community development as essential for maintaining for park staff; converting staff from law enforcers to com- a protected area. There was local support for protecting munity collaborative workers was difficult. Nevertheless, the park by providing community benefits through alter- meeting locals and getting to know them improved rela- native livelihoods. However, it is unclear from this whether tions. Some local people acknowledged that their initial the benefits essential for behaviour change were the intan- reluctance lessened as they met staff and learnt more gible empowerment benefits of community development, about the resource base; while a third of respondents did or the material benefits. not participate at all. In Selous Conservation Programme (IUCN category IV), Once established, the agreement provided greater clar- Tanzania, support from communities was greatest in areas ity over rights and duties, and opportunities for long where education and mobilisation campaigns had been term planning about livelihood strategies. However, as a conducted and benefits were beginning to be derived; find- legal document, the agreement was flawed as it failed to ings suggest that the majority of villagers supported the refer accurately to maps or related by-laws, statutes or project. The evidence showed that they were motivated to other documents [31]. Subsequently people were more join the conservation programme by promises of socioeco- positive towards the park, its resources and staff. How- nomic benefits [38]. ever, as the focus was on the park rather than the com- The arrival of western donors and NGOs in Caohai munity, people were sometimes organised according to Nature Reserve (IUCN V), China, in 1993 changed the what resources they collected, rather than by other so- focus from enforcement of resource regulations towards cially relevant criteria such as ethnicity, kinship, location, small-scale community development and outreach Pullin et al. Environmental Evidence 2013, 2:19 Page 25 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 programmes [47]. These included small grants and a IV and VI; and both before and after the Durban micro-credit programme for farmers to start up micro- Accord enterprises in the hope that they would be less reliant  Communities expressed a tension between on the reserve’s natural resources, infrastructure devel- appreciating the environment and wanting to opment, environmental education, a community based protect it, and also needing to make immediate use natural resource management programme, and school of land or natural products fees for girls from poor families. This involved two em-  They could be encouraged to participate in further ployees who had extensive prior experience of working conservation measures where they could anticipate with farmers, and required extensive training in com- socio-economic benefits munity development, gender issues and a variety of par- Evidence synthesised from seven studies ticipatory methodologies. The result was many fewer [23,25,26,28,29,38,43] hostile confrontations between local people and nature reserve managers, the participation of local people in Even where residents have recognised that conserving conservation activities and farmers contrasting the a park (IUCN II) and its wildlife is valuable on an indi- nature reserve’s concern for local people with the indif- vidual, local, national and global scale for economic, ference of corruption of other government agencies. educational, recreational, aesthetic and environmental Farmers now work cooperatively with the reserve to reasons, and for future generations, they still lament the seek resolutions to their own problems, sometimes tak- economic limitations imposed by restrictions on access, ing the initiative to raise issues about road construction, extracting resources and grazing, and the dangers of wild sanitation improvements, and agro-forestry projects. animals [23]. Indeed, some communities refuted the The transformation from conflict to cooperation has need for resettlement, having managed the land (IUCN been dependent on funds from NGOs and donors, II) for centuries; this was confirmed by the authors’ his- which raises questions about the project’s sustainability. torical analysis and portrayal of a pristine forest whose Another successful example of cooperative manage- protection was incompatible with indigenous residents ment was on the margins of a category II park in as a scientific myth [29]. Mozambique where land values increased exponentially. A range of Mexican farming styles have developed along With the support of an NGO residents thrived, benefit- the spectrum of reciprocal relationships between man and ting materially from land titles, revenues and empow- nature (co-production) between wilderness and urbanisa- ered by the process of acquiring land titles and setting tion, with farmers valuing the land for its provision of boundaries [48]. food, water, wood and other products [28]. Despite appre- Participatory approaches to governance were not al- ciating protected areas for their aesthetic, environmental ways successful. Almudi and Berkes [49] investigated the products, economic opportunities and spiritual values relationship between a local fishing community and offi- [25,28] their positive attitudes were lost when conserva- cials responsible for the creation and maintenance of tion regulations (IUCN IV) competed with productive ac- Brazil’s Peixe Lagoon National Park. They took a par- tivities such as cattle ranching or growing imported ticular interest in the factors that could empower local varieties of fruit, or with personal safety [25]. fishers to ‘defend their rights to remain physically within Responses to plans for balancing conservation and the park and politically in the conservation policy process’ economic development appear to vary depending on (p.220). The authors also found that fisher communities where the benefits might be felt. For some, it was not an struggled to participate in discussions essential to securing interest in conserving wildlife that motivated participation their ‘long-term access to the resources for their livelihoods in a Community-based Wildlife Management Programme or to trigger the development of a PA co-management ar- (IUCN IV) but promises of socioeconomic benefits to rangement’ (p.225). The following quotes were provided as themselves [38]. Whereas other respondents were critical examples of the fishers experiences: of a programme for promoting economic activities along- The authors summarised two of the main barriers con- side maintaining ecosystems; here tourism was expected tributing to the fishers’ lack of empowerment as: weak to benefit entrepreneurs and rich family owners of coastal assistance for developing community organisational cap- lands with tourism potential [25]. acity and leadership; and lack of basic knowledge on laws and fisher rights. Access to land PAs are important to communities for grazing, Well-being agriculture, hunting, foraging and spiritual homes Environmental (natural) capital  Relocation and loss of control over land and Appreciation of protected areas other than for resources can result in resentment, poaching and economic benefit was found in IUCN categories II, antagonism Pullin et al. Environmental Evidence 2013, 2:19 Page 26 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Participating in the process of setting boundaries of a balanced diet without animal protein and protected and securing land rights can be empowering areas were still seen as the ideal place to carry out hunt- Evidence synthesised from nine studies ing throughout the year [30]. Where dependence on ac- [30,33,34,40,41,44,48,50,51] from IUCN categories II cess to products was high, collection continued despite a after the Durban Accord, and from categories V and ban as compensation was not always considered ad- IV before the Durban Accord equate [39]. Once introduced to community based management to Access restrictions to protected areas (IUCN II) had harvest, process and sell timber, Ecuadorian communi- implications for grazing cattle, hunting and collecting ties who were disappointed by high start up costs and natural products [33,40,51]. Local people would like graz- slow, small gains made deals with external commercial ing rights in the park (IUCN II), especially during drought, loggers to raise their profits [44]. In contrast, in Mexico, and the opportunity to visit traditional areas and burial where conservation is widely valued, natural products areas [51]. Where staff were mostly concerned about il- were used in a sustainable way [28]. legal grazing; they would commonly impound livestock, and fine owners or refer them to a local judicial ward; in Economic capital contrast most community members felt the harsh envi-  Before the Durban Accord (IUCN Ib and II), ronment justified grazing cattle in protected areas [40]. concerns focused on: reduced employment and tax Communities considered that threats to their environ- revenues; reliance on foreign aid without mental capital outweighed any potential economic bene- understanding its link with conservation; and fits of living near the national park [50]. Outside the Park unrealistic expectations of the economic benefits residents felt they would not have access to resources so of tourism. far essential to their livelihoods: access to agricultural  After the Durban Accord (IUCN II), concerns land, forest resources and grazing land [33]. Wealthy cat- focused on: the meager benefits of tourism; what tle owners were even less disposed to moving because cat- benefits there were not being shared equitably; with tle would need to compete for food and water with host indigenous groups or those less amenable to falling villages, and cattle theft was more common outside of the in line with new regulations missing out. park [33].  Concerns about lack of compensation were Access restrictions posed similar problems for commu- expressed before and after the Durban Accord about nities in or near PAs categorised as IUCN IV. Accessing IUCN categories II and IV, and about IUCN V the forest was important in order to obtain resources to before the Durban Accord. Conversely, in developed support local people’s livelihood and for feelings about the countries there was a growing welfare dependency. forest as ‘theirs’ [30]. Authors described relocation and Evidence of economic capital found in thirteen loss of control over land and resources resulting in resent- studies [23,29,34,36-41,53-56], ment, poaching and antagonism [34] and overcrowding resulted from restrictions on building new homes [41]. Before the Durban Accord, the impact of IUCN II parks More positive views were expressed where land values on the wealth of whole areas was seen in Scandinavia, had increased exponentially on a park’s margins (IUCN South America and Asia. Forest workers in Sweden asso- II). Here residents thrived, benefitting materially from land ciated environmental protection with lower levels of em- titles and revenues and felt empowered by the process of ployment and production in commercial forestry [36]. acquiring land titles and setting boundaries [48]. Politicians anticipated conservation policies leading to lower tax revenue and greater emigration [55]. In Nepal Resource use communities were developing a dependence on foreign Residents appreciated protected areas for their rich aid, with some people considering it as an expected source products of income and not necessarily associating its benefits ac- Resource use was common even where illegal crued with conservation efforts [23]. Evidence synthesised from seven studies Some people living in or near national parks (IUCN II) [28,30,35,39,43,44,52] mostly conducted before the were concerned about neighbours having unrealistic ex- Durban Accord and spanning IUCN categories II, V pectations of the economic benefits derived from tourism and VI. and eco-lodges [55]. Others were unaware that commu- nity development was one of a park’s primary objectives Residents viewed protected areas as rich sources of only initially [43]. food and other products [28,30,39,43,44,52]. They ac- Benefits of tourism to national parks (IUCN II) were knowledged that illegal resource use continued despite seen as meagre, and distribution of revenues from pro- bans [39,43,52]. Some park residents could not conceive tected areas was considered inequitable or of little benefit Pullin et al. Environmental Evidence 2013, 2:19 Page 27 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 to indigenous communities [40]. Park and eco-lodge staff work ethic and less frequent use of their native language tend to be wealthier than community residents [43,55]; [44,55]. Stronza and Gordillo [55] saw changes in social and local people felt overlooked for employment, with op- ties within communities when some began working in the portunities favouring neighbouring communities and those tourism. Locals who secured employment in eco-lodges deemed amenable to new regulations [29]. Similar con- (IUCN II) talked about their work limiting participation in cerns about few or unevenly shared benefits and oppor- gatherings traditionally employed to complete community tunities were expressed in other, IUCN IV, protected areas tasks, and how communities tended to charge eco-lodge [34,41]. employees more because they were richer – it was buying Communities across IUCN categories, before and after them out of their community responsibilities. Working in the Durban Accord, often considered as inadequate the eco-lodges opened villagers’ eyes to new opportunities and monetary or in kind compensation available for: forestry a wider social circle but this was at the cost of leaving their constraint; resettlement; loss of land, crops, livestock or family and community. More direct commitment to con- jobs; or personal injury or property damage [33-39]. servation in a national park (IUCN II) also caused family Environmental protection is associated with economic pressures where the Village Scout scheme took young men decline in high income countries. In Utah, USA, local away from their family and farming responsibilities [38]. residents saw more tourism but no economic growth as Elsewhere (IUCN V), investment in communities was a result of establishing a protected area [27]. In Slovakia, mentioned by park staff and by residents; this included the although residents living near Slovensky Raj National aim of establishing a special university training program to park (IUCN II) anticipated multifunctional forest man- prepare students to fill key park management positions agement as a source of employment and income gener- [39]. ation, in practice the socioeconomic situation worsened, Villagers emphasised the importance of social relations particularly for minority ethnic communities, with re- as part of their survival strategy and expressed appre- duced employment and changes in welfare support [42]. hension about the weakening and possible disappearance With their lack of experience and opportunities for in- of these linkages as a result of displacement. These rela- volvement in small businesses and local or regional plan- tions were particular important in times of struggle such ning, minority ethnic groups saw no viable economic as periods of drought [35 p.60]. Stronza and Gordillo options. [55] considered that communities rich in social capital The most positive findings about social and material may not only be better able to manage changes associ- benefits came from an NGO funded study with one aca- ated with ecotourism, but that such changes in social demic author and one author employed by an ecolodge, capital can collectively sustain local institutions, which although they too shared concerns expressed above [55]. may subsequently be critical of conservation efforts. In contrast, the community managed forests in Ecuador cre- Social capital ated new business relationships and improved the com- Changes in livelihood strategies have influenced the munity’s social standing with other indigenous groups in shape of households and the strength of social ties, the area [44]. and introduced new inequalities within communities. Although ethnic identities could strengthen social cap- PAs have been established in areas inhabited by various ital within groups, they more often created tensions be- ethnic groups. The pressures resulting from regulatory tween groups. Some ethnic groups were resentful as and economic changes have introduced tensions or they perceived others receiving community development exacerbated historical tensions between them. preferentially, even though the authors saw these com- Evidence about social capital in PAs was found in munity development efforts as culturally inappropriate eight studies [30,35,38,39,41,44,55] [44]. As noted above, migrants struggled to implement traditional informal rules, and indigenous people failed Slater [41] noted that households configured themselves to obey state-induced laws. Some of these difficulties in a way so as to maximize livelihood diversification; some- were attributed to the government’s indifference to cul- tural and social diversities when managing the Park [26]. times to the detriment of familial relationships. Households could be separated geographically by livelihoods, or over- At the level of implementing regulations, residents crowded because sharing dwellings allowed younger adults objected to the leniency of guards towards those who were wealthier or ethnically related [23]. to rely on the support of older adults claiming pensions. Residents saw a direct link between livelihood diversi- fication and changes in cultural traditions and traditional Health relationships amongst local people; both positive and nega-  New diseases associated with changes in lifestyles tive [44,55]. Making the transition to a settled agrarian life were attributed to forest evictions or changing from for some meant a decrease in the traditional communal a nomadic to sedentary existence. Pullin et al. Environmental Evidence 2013, 2:19 Page 28 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Accidents and injuries resulted from conflicts day labourers, tourism employees), tenure (landowner or between guards and residents; animal conflicts not), gender, education and wealth. resulted in increased workload and exhaustion as Differences in wealth accrued from ownership of land well as injuries or livestock have been influenced by wildlife conflict Sexual aggression was more common when women [38,54] and by PA regulations about land access or live- were less protected following necessary changes in stock numbers, including whether or not people com- working patterns or kinship ties. plied [41]. Smaller businesses and entrepreneurs less Evidence of health and protected areas was found in able to control shocks may bear the brunt of further re- five studies [35,38,41,44,54] strictions on the forest products [36]. Some of these individual differences have arisen at In one study, community members evicted from forest least in part from inequalities between ethnic groups or unanimously spoke of their exposure to new disease from discrimination between indigenous groups. For in- when integration with other groups began; authors con- stance, some indigenous people have discouraged immigra- firmed that the community, particularly children, were tion and excluded immigrants from community governance seriously affected by malaria which did not exist in the procedures [26]. Within and between ethnic groups, people forest, and that HIV/AIDS is also appearing [44]. Their have taken advantage of others who earn more from the forced transition to an agrarian society has cut them off presence of the PA [55] or from people struggling to make from their access to and knowledge of traditional medi- a living who sell land legitimately owned only to claim pro- cinal plants they previously used to stay healthy. Else- tected land illegally [26]. where, villagers reported the lack of access to basic Misunderstandings or prejudices about the histories or health services and Torri [35] confirmed that child mor- abilities of communities led authorities to pursue deci- tality was high in isolated forest villages, where common sions which those communities found damaging [32,33]. illnesses, easily treatable given basic medical facilities, Authorities also discriminated within communities by could lead to death. exercising policies that protected landowners but not Residents described how crop raiding by elephants lead other longstanding residents [41]; or by involving people to food shortages and greater workloads, especially amongst chosen for their age, wealth, education and position ra- women who suffered more from insect-borne diseases and ther than practical relevant knowledge [31]. Ethnic dis- heat exhaustion. When elephants had damaged water pipes, crimination has been introduced by PA legislation that women risked drowning when collecting water from unsafe forbad resource use by indigenous people but allowed sources [54]. First Peoples [44] reported women’s safety be- resource use by predominantly white landowners [51]. ing compromised as men worked further afield, and as Where ethnic discrimination predates establishment of women were drawn out of their homes for new roles and PAs, historic practices of favouritism have strengthened as that the erosion of kinship ties maybe reducing protection land has become more valuable and rare because much of from male sexual aggression [44]. it has been put aside for conservation [50]. Prejudice and nepotism have been the source of favouritism or corrup- Inequalities tion amongst PA employees [23]. PAs and residents’ responses to them have Some differences have arisen from commercial or con- exacerbated existing local ethnic tensions servation enterprises providing more earning power for Participation in PA governance has favoured people some, but not all, roles [56]. Some eco-tourism enter- already advantaged by their socio-economic position prises have a history of being poor employers of local PAs impact unequally on people depending their casual labour [32]. Financial inequalities were introduced socio-economic position, such as the size of their when spent funds resulted in financial support for some business, their legal land tenure or their gender but not others, and bank loans have been refused on the New tourism enterprises have tended to employ grounds of having an address in a protected area [44]. outsiders rather than locals Women were vulnerable to attack and injury from Evidence of inequalities arising from PAs was found men or wildlife [38,44,54]. Female heads of households in 15 studies [23,26,31-33,35,36,38,41,44,50,51,54-56] were vulnerable financially as they tried to balance paid labour with subsistence farming [41]. More may have People living in and near PAs differ in many ways. Some been learnt if some women had not been reluctant to ex- indigenous people living near each other come from dif- press their opinions to researchers [35]. ferent ethnic groups. Some PAs include indigenous com- munities, new immigrant communities and residents long Narrative synthesis of quantitative evidence established following historical migrations. Individuals dif- In this section we provide a narrative overview of all in- fer in terms of their occupations (e.g. subsistence farmers, cluded studies providing quantitative data on impacts. Pullin et al. Environmental Evidence 2013, 2:19 Page 29 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 The studies are divided into six subsections; produced it- the diversity of outcomes reported by all studies. Further- eratively based on the outcomes reported in the accepted more, we avoid vote-counting, where the sum of all nega- literature (see typology in Figure 12). We do not claim that tive, positive and neutral study results are calculated. these divisions are definitive or optimal but they do provide Vote-counting is unreliable because it assumes that a sig- a pragmatic breakdown of a complex body of evidence. A nificant finding is evidence that an effect is present and a summary of data presented in the included studies is pro- non-significant finding is evidence that an effect is absent. vided as Additional file 7. A more detailed set of data ex- This former statement is true, but the latter is not (see traction tables is available as supplementary material. Of Borenstein et al. [21] for further details of vote-counting). the 79 studies included in this synthesis 63 were catego- rized as having ‘high’,11 ‘medium’ and 3 ‘low’ susceptibility Ecosystem goods to bias. Additional file 10: Table S1 displays the 14 studies and 33 outcome measures that were categorised as having  17 studies identified low and medium susceptibility to bias. Additional file 11  Only two studies not highly susceptible to bias provides detail of critical appraisal and the basis for assign- (‘medium’ susceptibility to bias) ing susceptibility to bias for all 79 studies. In the following  Nyahongo et al. [57] found that meat and fish text we concentrate on results reported in these 14 studies. consumption increased with proximity to Serengeti Here wherewediscuss studies’‘susceptibility to bias’ we National Park refer to their categorisation resulting from our critical ap-  Sarker and Røskaft [58] found residents’ perception praisal. In some instances we also identify specific types of of timber and fuelwood benefits from four PAs in bias (defined in Table 5, below) and in others we refer to Bangladesh decreased with distance from the PA shortcomings of the study design and implementation that boundary implicitly increase the studies’ susceptibility to bias.  High susceptibility to bias in remaining studies It is important to note that since the majority of stud- related to low methodological detail, confounding ies were identified as being highly susceptible to bias, variables, and weak experimental design the results of these studies are not considered further, ei- ther individually or in general. These studies are unreli- Of the 17 studies reporting results on ecosystem able both alone and in concert. Because this group of goods, only two studies had designs that were not highly potentially biased studies is unreliable, we cannot sum- susceptible to bias (i.e. medium). Nyahongo et al. [57] marise their findings any more than we can discuss indi- reported the number of meat and fish meals consumed vidual results. However, for completeness we also consider by survey respondents to be significantly negatively Figure 12 Typology for structuring the quantitative synthesis of impacts of protected areas. Pullin et al. Environmental Evidence 2013, 2:19 Page 30 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Table 5 Definitions of bias mentioned in the narrative synthesis Recall bias Imperfect recollection of past events by respondents. Generally worsened by longer periods of recall Social desirability bias Tendency to respond to questioning in such as way as to be viewed favourably by others Questioning bias Questioner leading respondents to reply to questioning in a certain direction Neyman bias Arises from a time lag between exposure and sampling such that undetected drop-out of participants may occur before the study begins. Attrition bias A skew in results where participants are lost between measurements at two time points (potentially as a result of the exposure) during the study Optimism bias A belief by a respondent that they are less likely to experience a negative event relative to other respondents, or over-optimism on the part of analysts or interviewers, about the effects of a project. Hypothetical bias Failure of respondents to consider the true budget constraints in responding to financial questioning Strategic bias Tendency for respondents to alter their answers in an attempt to influence an event correlated with distance from Serengeti National Park whilst Bajracharya et al. [59], Mehta and Kellert [60], boundary. The authors’ statistics account for a range of Naughton-Treves et al. [61], Rinzin et al. [62], and Okello other potentially influential variables, contributing to the et al. [63] all measured accessibility and quality of water studies favourable susceptibility to bias rating (medium). resources, there was insufficient conformity to allow The study’s results indicate that meat consumption in- meaningful synthesis. creased with proximity to the park at a rate of 0.218 Other studies could not be used to infer reliably any (±0.052) meals per week per km, whilst fish consump- effect of a protected area for a range of reasons, includ- tion increased at a rate of 0.931 (±0.205) meals per week ing; (i) they recorded change over time with no spatial per km (assuming units in analyses are identical to units comparison and no adequate basis for determining whether described in the methodology). Sarker and Røskaft [58] the changes observed were attributable to the effect of the found that people inhabiting the areas surrounding four protected area; (ii) they made a spatial comparison, but lo- protected areas in Bangladesh identified more benefits cation relative to the protected area is clearly confounded from the protected area in the form of timber and fuel- with a range of other important site variables; (iii) the scale wood extraction closer to the park boundaries than fur- of the spatial comparison was too small to use it to infer ef- ther away, with an associated odds ratio of 1,000; i.e. “The fects of the protected area on the specific outcomes re- odds that respondents living closer to the protected area corded in the study; (iv) time elapsed between the creation boundary reported a higher level of benefit from timber of the protected area and the study was too short to infer and firewood because of the conservation programme any effect of the protected area. were 1,000 times greater than those for respondents living further away”. Livelihood strategies Thirteen studies used questionnaires and semi-structured interviews for data collection, resulting in a higher suscepti-  43 studies identified bility to bias since reporting by the respondents can be sub-  Eight studies not highly susceptible to bias (1 ‘low’ ject to recall or social desirability bias. Fourteen studies and 7 ‘medium’ susceptibility to bias) reported 39 different (but not all independent) outcomes  Four studies report poverty-related outcomes, finding related to food and materials, comprising a mixture of beneficial impacts of land protection in all cases foods obtained by hunting and gathering, agriculture and  Wildlife conflict is relatively well-studied (18 studies purchased, as well as other indicators such as land area and 47 outcomes; 5 studies with 7 outcomes under cultivation or grazing, availability of fodder, support ‘medium’ susceptibility to bias. The majority for agricultural development, dietary diversity, gathered reported significantly more problems with proximity fuelwood and other plant products including timber, and to protected areas than further afield also change in ownership of goods. Five studies reported  Lundgren [64] found no significant difference in seven different (but not all independent) outcomes related income growth or forestry/tourism sector to water resources: three studies related predominantly to employment as a result of protected areas in Sweden water quality (e.g. households relying on least safe water re-  Household income [65] and consumption [66] were sources) and two to water availability and supply. Only one found to increase with proximity to/within study reported one outcome related to medicinal plants/ protected areas animals. Studies reporting data on common themes differed sig- Studies in this set report on access to markets, employ- nificantly in the precise outcomes measured. For example, ment, income, livelihood diversity, human development Pullin et al. Environmental Evidence 2013, 2:19 Page 31 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 measures and wildlife conflict (Additional file 7). Wildlife  All studies are highly susceptible to bias due to lack conflict was the most frequently reported outcome meas- of methodological detail, non-random sample ure in this group, contributing 18 of the 43 studies and 47 selection, spillover, questioning bias and of the 101 outcomes. Only five of these 18 studies were uncontrolled confounding variables not highly susceptible to bias, all bar one [52] showing  The majority of outcomes relate to restrictions on significantly greater crop and livestock loss closer to pro- access or extraction tected areas. Two of these five studies employed question-  Two outcomes related to the perception of naires to elicit perceived disbenefits, whilst three studies relationships with park employees used observations of conflict. Lundgren [64] found no sig- nificant correlation between protected areas and income The six studies in this set report on 11 different (but growth or employment in either tourism or forestry sec- not all independent) outcomes related to land and re- tors in Sweden. Sims [66] reports higher consumption in source access, interactions with protected area author- regions with a high ‘share’ of PAs compared to a low ities, displacement and resettlement (Additional file 7). ‘share’ in Thailand. Richardson et al. [65] found house- All had a high risk of bias in their study design because holds within game management areas (GMAs) to have of a lack of detailed description of their methodologies. greater income across a number of Zambian PAs than In cases where the description was detailed, specific risks those outside GMAs. of bias were identified; replicates were non-randomly se- Four studies reported poverty-related measures (hu- lected [59,74] and spillover, questioning bias, and con- man development measure outcomes) as either poverty founding variables were not accounted for [59,75,76]. index or poverty headcount. All four studies found sig- nificant beneficial impacts of protected areas on poverty Health and safety alleviation in Costa Rica [67], Thailand [66,67] and Bolivia [68]. The study by Andam et al. [67] involved subsequent  Nine studies identified in-depth reanalysis in two later publications [8,69]. These  Only one study identified as not highly susceptible later analyses showed that along with protection alleviat- to bias (‘medium’ susceptibility to bias) ing poverty, poorer areas (measured at baseline) were  Korhonen et al. [77] found highly variable infant found to have greatest level of poverty reduction than mortality rates in and around Ramonafana National those that were less poor [8]; and that poverty alleviation Park in Madagascar, with slightly higher levels outside was also associated with characteristics that reduced the than inside the PA, although the pattern is not clear efficacy of deforestation prevention (i.e. where protection had been assigned to land that was unsuitable for agricul- Studies in this section look at health in the population ture, near major cities and infrastructure, and where agri- and access to health services (Additional file 7). Studies cultural employment is low) [69]. reported a limited range of outcomes relating to health The remaining studies were judged to have high sus- and safety, but the reliability of findings in general was ceptibility to bias and the following examples are only il- compromised due to various aspects of study design lustrative of the diversity. Foerster et al. [70] report that which make them highly susceptible to bias, such as a purchasing power is lower in villages closer to PAs in lack of comparator or non-random selection of study Gabon. Cardozo [71] conducted a questionnaire-based site sites or participants, or incomplete reporting, with no comparison of communities inside and outside Allpahuayo- details about selection of study populations, validity of Mishara National Reserve, Peru and reported changes in survey instruments or survey response rate, [73,78]. The income and livelihood diversity. Annual income from agri- only study in this group with ‘medium’ susceptibility to culture was lower inside the PA whereas income from do- bias, Korhonen et al. [77], report a case-study of repro- mestic animals and palm products was higher. Kayser et al. ductive health from a 10-year old Integrated Conserva- [72] report greater annual transfer of money to SMEs tion and Development Project (ICDP). The study was of through contracts financed by Addo Elephant National BACI design comparing purposively selected (for repre- Park, South Africa. Saayman & Saayman [73] provide data sentativeness) villages and municipalities within a 3 km on self-reported change in the business environment belt around Ranomafana National Park (the ‘peripheral around South African PAs. This provides weak evidence zone’) (7 villages, 6 municipalities) with those outside that opportunities for and turnover of business has in- this zone (6 villages, 4 municipalities). Data on modern creased as a result of the PA. contraceptives suggest an increase in use over time, a decrease with increasing distance from the park and Land access and restrictions variability in use within ‘park’ villages attributed by au- thors to varying access to ICDP activity, but also to edu- Six studies identified cational status of women, local culture and religious Pullin et al. Environmental Evidence 2013, 2:19 Page 32 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 beliefs. Fertility levels were estimates rather than direct people has increased’. Other studies report perceived im- measures and do not indicate any change over time or provements in infrastructure in and around PAs com- differences between park and ‘outside’ villages. Visits to pared to elsewhere or before establishment [59,78] but the health centre for pre-natal care increased over time this is not always the case [62]. in line with national trends and did not differ between park and ‘outside’ village residents. The data presented Attitudes towards PA and the benefits (or otherwise) for infant mortality were drawn from census data pre-1999 they provide and showed year-on-year variation and no clear differences between ‘park’ and ‘outside’ residents, although a slightly  24 studies identified higher mean over an 8 year period outside than inside the  Only one study identified as ‘medium’ susceptibility PA. Post-1999 data were health centre data and only re- to bias, all remaining studies are highly susceptible lated to health centre births and thus may be subject to to bias due to lack of methodological detail, bias if a higher percentage of more problematic births oc- confounding variables unaccounted for, and spillover curred at the health centre as opposed to in villages. from protected areas into controls Sarker & Røskaft [58] found attitude to PAs to be Society and development negatively associated with PA proximity Four studies failed to identify the location of ‘inside’ 13 studies identified populations, two studies failed to report the distance Two are not highly susceptible to bias (1 ‘low’ and 1 of controls from PAs, and in general controls were ‘medium’ susceptibility to bias) very close to PA boundaries (e.g. < 1 km and 2.5 km) Sheppard et al. [79] found a greater number of  Studies reported attitudes towards PAs, attitudes infrastructural developments inside Wechiau towards identified benefits/disbenefits from the PAs, Community Hippo Sanctuary in Ghana than outside and perception of environmental change Korhonen et al. [77] found no difference in the percentage of girls in primary schools either over Studies reporting attitudes were separated into two time or inside relative to outside Ramonafana distinct categories; attitudes towards the protected area National Park in Madagascar, but a possible slight or conservation, and attitudes concerning benefits or increase in the percentage of girls in the third grade disbenefits resulting from the protected area. All but one over time was found inside the PA of the 24 studies reporting attitudes fell into the category of ‘high susceptibility to bias’. The main issues identified Studies in this set include measures of education, em- in the critical appraisal were lack of detail in the powerment, infrastructure, recreation and social capital. methods, not accounting for important confounding vari- Thirteen studies reported data on 54 development-related ables, and, where spatial comparators were used, potential outcome measures. Two studies were classed as having low spill-over effects due to the intervention and ‘comparator’ (1) or medium (1) susceptibility to bias. Sheppard et al. [79] sites being close together. This latter issue was a particular found a greater number of developments in various aspects problem for studies looking at the effect of ‘distance from of infrastructure inside Wechiau Community Hippo Sanc- the PA boundary’ on attitudes. Thirteen studies examined tuary in Ghana than outside. Korhonen et al. [77] found no spatial differences in attitude with respect to distance from difference in the percentage of girls in primary schools over the protected area. However four studies failed to report time and no difference inside relative to outside Ramona- the location of the intervention population inside the fana National Park in Madagascar, but they did find a slight protected area. Two studies [58,80] regressed attitude increase in the percentage of girls in the third grade over scores against distance from the protected area, but the time inside the PA (but not outside). distances involved were not stated. Control populations All remaining studies were judged to be of ‘high’ sus- in ‘inside-outside’ studies were generally very close to the ceptibility to bias. There is weak evidence for improve- protected area boundary, for example 2.5 km (Sekhar ment in education provision following PA establishment 1998) and < 1 km [71]; although it was 50 km in the study in terms of increased number of schools and perceptions of Bonaiuto et al. [81]. of improvement [62,72,78]. A questionnaire based on re- ported change conducted by Saayman and Saayman [73] Attitudes towards the protected area Fourteen articles in multiple South African PAs produced variable re- reported a wide range of general attitudes towards the sponses (in most, but not all, the majority agreed) to PA. Thirteen of these gave respondents’ statements (usu- statements such as ‘participation in community activities ally a mix of positive and negative statements) and pre- has increased’, ‘the pride that the residents have in their sented data on the percentage agreeing or disagreeing town has improved’, ‘the opportunities to meet new with each statement. One study [81] presented composite Pullin et al. Environmental Evidence 2013, 2:19 Page 33 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 attitudinal scores made up of responses to a series of studies were of low quality predominantly due to a lack questions which were not presented in the article. of detailed methodology and shortcomings of the experi- In the only study that was judged not to be highly sus- mental design. For example, Cihar and Stankova [90] ceptible to bias, Sarker and Røskaft [58] found that re- lacked a true before-and-after comparison and generated spondents from around four parks in Bangladesh had ‘before data’ by asking informants to recall the past. This negative attitudes towards the protected areas, and that is clearly open to recall bias. negative attitudes decreased with distance from each protected area. The remaining studies were of ‘high’ sus- Economic valuation studies ceptibility to bias. Bonaiuto et al. [81] reported that re- gional identity and place attachment were higher inside  10 studies identified the Tuscan Archipelago National Park in Italy, but that  One study judged as not highly susceptible to bias specific and general attitude scores towards the pro- (‘medium’) tected area were lower relative to a control group of re-  Four groups of studies found: cost-benefit analyses; spondents 50 km away. Jim and Wu [82] noted that a stated preference studies; stated preference combined higher proportion of people living on the boundary of with a distance comparator; and reported direct Shimentai Nature Reserve in China ‘disliked’ the park financial losses from a PA (e.g. fines/foregone income) than those living 4 km from its boundary. Finally,  These studies do not have real comparators (with Shrestha and Alavalapati [80] observed a positive correl- the exception of the distance comparator): instead ation between positive attitude and distance from Koshi they are hypothetical, and as a result are highly Tappu Wildlife Reserve in Nepal. susceptible to bias (e.g. optimism bias) In contrast to the above studies which found a positive  Studies are too heterogeneous and open to bias to relationship between distance from the park and attitudes, permit meaningful quantitative synthesis of valuations Cardozo [71], Gubbi et al. [83], Infield and Namara [84] and Sekhar [85] reported higher positive and lower nega- Studies in this section reported welfare impacts in tive attitudes inside protected areas than outside. Other monetary terms. Economists usually hold that individual studies found no statistically significant or observable dif- well-being is not directly and cardinally measurable, nor ference in attitudes between inside and outside protected comparable between individuals or time periods e.g. area [86-89] or over time [90]. [94]. However, changes in an individual’s well-being as a result of a PA’s existence can be expressed in terms of Attitudes towards named benefits/disbenefits result- the amount of money needed to render that individual ing from the park Thirteen studies reported data con- indifferent to the existence of the PA (the aggregation of cerning respondents’ attitudes towards named benefits such monetary amounts across individuals is common in or disbenefits resulting from the protected area. Again, applied economics, but deeply problematic). all studies were classed as highly susceptible to bias. Ite Ten studies were included that estimated well-being [89] found fewer respondents close to Cross River impacts of protected areas in monetary terms. Nine of National Park in Nigeria to believe that they have benefited these were categorized as having high susceptibility to from the protected area than those 5 km away, although bias, whilst one was categorised as medium susceptibility a third group of respondents 7.5 km from the protected to bias. area showed an intermediate perception of benefits. Jim Shrestha et al. [95] used a contingent valuation survey and Wu [82] reported no significant difference in the with a stratified random sample of 160 households perception of benefits from Shimentai Nature Reserve, within c. 6 km of Koshi Tappu Wildlife Reserve, Nepal, China, between respondents inside and those 4 km from to estimate their willingness to accept the PA in terms of the protected area, whilst significantly more respondents foregone resources. They found substantial local one- inside than outside claimed to have felt losses as a result time costs of 11,776.70 Nepali Rupees per household of the reserve. (1994/1995). Other studies found no evidence that respondents felt A major reason for excluding economic studies (see that either negative or positive impacts resulted from the Additional file 6) was that they measured the well-being protected area [75,86,88,91,92]. impacts of ecosystems within PAs, but did not isolate A smaller category of studies reported respondents’ the impact of the PA itself. One study [96] carried out perceptions of environmental change as a result of the a contingent valuation survey of willingness to accept protected area; with respondents in one study predomin- compensation for costs of the protected area’s presence antly not perceiving a change [90] and two studies on traditional pasture land, with respondents within the reporting the majority to have perceived an increase protected area and further away. The remaining studies in environmental ‘appearance’ [73,93]. However, these included only hypothetical or ‘modelled’ comparators. In Pullin et al. Environmental Evidence 2013, 2:19 Page 34 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 all cases this is done more or less explicitly by the ana- disaggregating costs and benefits by local, national and lysts themselves, but many studies also required respon- international groups. Both the effects of the PA and the dents to mentally construct hypothetical comparators, in counter-factual (no PA) were modelled, though little detail order to answer stated preference surveys. Such con- is presented and the evidence upon which the modelling structed comparators can be useful and indeed essential is based is often rather weak. They estimated that local when “real” comparators (RCTs, BACI etc.) are unavail- populations would lose from the establishment of the able (they may also be used in conjunction with such re- park, but that this would become a net gain if develop- search designs). However, they are vulnerable to a number ment projects associated with the park succeeded in rais- of potential biases, such as optimism bias, strategic bias, ing local incomes. At the national level, there would be a and hypothetical bias (see Table 5 for definitions). net loss due to the protected area. Without real comparators, direct evidence is lacking Seven studies used stated preference techniques to on the effects of the PA on individuals. Instead these must elicit estimates of welfare gains or losses. Four studies be predicted, using whatever information and opinion is [7,98-100] used contingent valuation to estimate regional available to the analyst or respondent. Numerous assump- or national populations’ willingness to pay for existing tions must necessarily be made, and will not always be protected areas in India, Brazil, China and Greece re- explicitly stated. This leaves these studies open to well- spectively, all four studies indicating generally positive recognised biases. For example, cost-benefit analyses are welfare impacts of the PAs on these broad populations. known to suffer from optimism bias, especially when con- A fifth study, Ascuito et al. [101], similarly estimated ducted by groups with an interest in the project: in the case local willingness to pay for a fire prevention programme of PAs this may be the government or conservationists in an existing protected area again finding positive wel- more generally. For example, the results of Kremen et al. fare impacts. Two studies [95,96] used contingent valu- [97] are heavily dependent on optimistic assumptions made ation to estimate local populations’ willingness to accept about the efficacy of development interventions planned to restrictions on livelihoods imposed by existing protected accompany the PA: no evidence is presented on whether areas in Ethiopia and Nepal respectively, indicating these interventions indeed had the effects assumed by the negative welfare impacts of the PA. In the case of Jemal authors, since the analysis was conducted ex ante. [96], surveys were carried out with respondents inside Stated preference studies are known to suffer from and further away from the PA, and found that people both hypothetical bias and strategic bias on the part of within the PA were less willing to accept compensation respondents. Hypothetical bias may lead respondents to than those further away, suggesting negative welfare im- overstate their willingness to pay (WTP) for goods or pacts increased with proximity to the PA. Finally, Abbot services provided by a PA, because they fail to consider and Mace [102] present data on fines levied on local their true budget constraint. Respondents may also behave people for illegally harvesting fuel-wood in Lake Malawi strategically: beneficiaries may overstate their willingness NP. These fines were levied by the PA but no informa- to pay for a PA, in order to increase the likelihood of its tion is provided on areas outside the PA. establishment if they suspect they will not be required to contribute to it, or understate their WTP if they suspect Inequalities that this will result in lower user fees. Those who expect Assessing the impact of PAs on health or social inequal- to lose from PA establishment may overstate the amount ities would require either individual sound studies with that they would require to receive in compensation for the justifiable subgroup analyses, or a set of comparable establishment of the PA (i.e. their willingness to accept studies which describe in detail the socioeconomic pos- (WTA) the PA) to reduce the likelihood of its establish- ition of the populations studied [103,104]. Neither was ment, or increase compensation payments. Alternatively available from the extant literature. surveys may under-estimate opportunity costs if the activ- ities concerned are considered sensitive or of dubious le- Meta-synthesis of qualitative and quantitative evidence gality, and are likely to be under-reported. The results of In attempting to bring together the findings of the quali- stated preference studies are also known to be sensitive to tative and quantitative reviews it is important to reflect the information provided by surveyors, and the precise on the differences in their philosophies. The qualitative formulation of the questions. This renders them vulner- synthesis is essentially formative and attempts to form a able to the same optimism bias noted above. picture of how PAs are perceived to impact on human Methodologically, the included studies fell into three well-being. As such it can form a template for empirical groups: cost-benefit analyses, stated preference studies, investigation and hypothesis testing. The synthesis of and reported direct financial losses from a PA. One study, quantitative evidence is more summative and attempts Kremen et al. [97] carried out an ex ante cost-benefit ana- to test hypotheses of impact. In consequence we should lysis of the establishment of Masoala NP, Madagascar, not expect the meta-synthesis to be a simple matching Pullin et al. Environmental Evidence 2013, 2:19 Page 35 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 of similar studies or outcomes. In this section we sum- of impact only two studies were not highly susceptible to marise the findings of the qualitative synthesis and ask bias, showing that meat and fish consumption was greater whether the quantitative evidence of impacts can inform in proximity to a PA and that timber and fuelwood bene- the questions raised by these findings or whether it sug- fits were more frequently appreciated nearer another PA. gests something different. Economic capital Views expressed on impacts of PAs Governance on economic capital are generally negative, with the ex- The qualitative synthesis reveals a number of factors that ception of some views on the benefits of ecotourism. In can lead to negative views and impacts of PA establish- contrast the quantitative evidence of impact from three ment: lack of clarity in regulations and boundaries; dis- studies on livelihood strategies was neutral to positive crepancies between state rules and local institutions; in terms of poverty reduction. In particular, there were forced migration, inadequate or non-existent compensa- concerns in Sweden amongst foresters about sustaining tion; and poor communication between communities employment and amongst politicians about sustaining and authorities and government indifference to cultural tax revenue in the presence of regulations. However, and social diversities. Negative views on impact of man- these concerns were not upheld by a quantitative assess- agement can arise from poor relationships between resi- ment of impact in the same country. All but one of the dents and park officials. Views on how to lessen negative economic valuation studies suffered from high suscepti- impacts or achieve positive impacts include: rules that bility to bias and therefore add limited reliable quantita- are locally adapted or based on traditional rules; greater tive evidence to this issue. clarity over rights and duties; planning focussed on com- munity livelihoods as well as the park; appropriate cap- Social capital The qualitative synthesis suggests devel- acity building; and empowerment through the process of opment associated with PAs can exacerbate ethnic ten- acquiring land titles and setting boundaries. The existence sions through perceived preferential treatment of some of these views enables hypotheses to be generated on how communities. There may be a relationship between exist- to achieve change in impact. The synthesis of quantitative ing social capital and ability to adapt to new circumstances. measures of impact shows that these hypotheses are yet to Quantitative evidence of impact on social capital is mixed. be tested. What is absent from the evidence base is a Thereissomeevidenceofpositiveimpact of land protec- quantitative comparison of costs and benefits to local tion on poverty alleviation and on housing and infrastruc- people of different forms of PA governance. ture but also of increasing incidence of wildlife conflict. Well-being Health Views expressed on health of local populations Environmental (natural) capital The qualitative syn- are predominantly negative, including exposure to dis- thesis presents a range of positive and negative attitudes ease, wildlife conflict and women’s safety. Quantitative among local populations towards PAs. Alongside an ap- studies of impact of PAs on health and safety are notable preciation and desire to protect the environment were by their absence. concerns about reliance on those same areas to maintain economic livelihoods. Although the qualitative literature Discussion provides evidence of difference views, quantitative evi- Historical accounts of establishment of PAs provide evi- dence to estimate the scale and reach of those views was dence that substantial negative impacts on local popula- not extractable since all but one study were highly sus- tions can occur and have occurred. Forced displacement ceptible to bias, of communities is a recurring theme in the narrative concerning negative impacts of PAs e.g. [4]. This review Access to land The qualitative synthesis revealed two does not seek to question this historical narrative. Commu- very different scenarios in terms of access to land. The nity development and infrastructural improvements in close first is resentment at loss of access and the second is proximity to PAs has also been documented and suggests benefit from acquisition and value of land on the PA that PA establishment can be positive (i.e. win-win solutions margin. All quantitative studies of impact of PAs on land for biodiversity and human well-being are possible). Estab- access and restrictions were highly susceptible to bias lishment of PAs will inevitably lead to impacts on local, and and so the current evidence does not allow the magni- possibly regional, populations but the challenge is to im- tude of these scenarios to be assessed. prove our capacity to predict which factors will influence the balance of positive and negative impacts. In this review Resource use A range of positive and negative views were we have attempted to provide an assessment and charac- found concerning PAs as a source of natural resources and terization of the range of positive and negative impacts in ecosystem goods. In the synthesis of quantitative evidence the period following the Rio Summit and establishment of Pullin et al. Environmental Evidence 2013, 2:19 Page 36 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 the CBD. It was not the objective of this review to revisit manner, the findings from international studies of peo- this history prior to 1992. The review also attempts to col- ple’s views about the impact of protected areas on their late evidence on the factors that modify impact, either in a lives. To reduce the likelihood of missed studies, sensi- positive or negative direction. tive searches of bibliographic databases were supple- mented by other methods to seek out less easily found Comparison of qualitative and quantitative evidence literature such as unpublished reports from topic rele- Although thequantitativeevidenceisinsufficientto draw vant websites. Studies providing thick descriptive data conclusions about the scale of either positive or negative spanning the different categories of IUCN protected impacts of protected areas on well-being, it was possible to areas and before and after the Durban Accord offered an synthesise understandings from qualitative studies about excellent source for synthesising understanding about how positive and negative changes in well-being can arise how protected areas impact on people’s lives. from establishing and implementing regulations to protect During the reading and re-reading of individual studies the natural environment, with or without simultaneous in- we found that data often encapsulated the complexity of vestment in community development. living in or near protected areas, touching on a multipli- The qualitative synthesis has identified a number of city of interrelated themes. Within the limitation of time themes in the ways governance of protected areas affect hu- and resources for this study we have only been able to man populations well-being and how PAs are viewed. Some present these themes fairly superficially without explor- of these themes, such as the impact of land protection on ing fully all their interconnections. forestry sector employment in Sweden, are reflected in the The disparate, fragmented literature limits our ability impacts assessed in studies considered in the quantitative to test the comprehensiveness of the search. In reviewing review. Other themes, however, have not been rigorously such a broad and interdisciplinary question it has been a assessed in the quantitative literature, for example, novel significant challenge to test all the possible sources of rele- diseases resulting from changes in lifestyle, increased work- vant material; nor is it simple to measure what proportion load and heat exhaustion due to crop raiding, safety risks of the relevant articles we have been able to access with felt by women as men worked further afield and as women the time and resources available. Limiting our search to themselves were drawn out of their homes for new roles. English-language articles may be significant. Diversity of the literature has limited any assessment Reasons for heterogeneity of extent of publication bias. Whilst we have attempted The identification of variables that influence whether posi- to minimise publication bias by employing a systematic tive or negative impacts will occur would be desirable for search strategy, we have no way of testing for publica- supporting decision making on the process of establish- tion bias in the literature we obtained. The selective na- ment and subsequent management of PAs. Unfortunately ture of many studies, in terms of the type of impacts the nature of the evidence provides little opportunity to investigated, is also a potential source of bias, since re- analyse differences in impact among different PAs (see lim- searchers may ‘cherry-pick’ (possibly inadvertently) those itations below). impacts most likely to show a particular effect. Mode of governance is commonly viewed as a key variable determining impacts of PAs and this is sup- Limitations of the evidence base ported by the qualitative synthesis in which many narra- Although the quality of the studies was sufficient to tives are available on different aspects of governance. draw out their findings to explain how different impacts However, rigorous tests of governance as an effect modi- may arise, many of the studies failed to report ad- fier are absent. A similar lack of quantitative evidence is equately their methods of data collection and analysis. apparent for the following questions: There is a potential in any studies looking at the quan- titative impact of protected areas that respondents will What practices repeatedly lead to negative impacts, and bias their responses in an effort to influence protected area which ones seem to be recurrently improving people’s governance. Some questionnaire-based studies attempted well-being? to minimise this strategic bias by clearly stating that inter- Are some of these practices becoming more/less view and questionnaire results would be used solely for re- common with time? search purposes, other studies did not acknowledge this Are any costs or benefits associated with particular potential bias or attempt to reduce it. Of the 305 outcome types of PAs (e.g. size or location)? measures extracted from 79 included studies, 92 outcome measures (30%) involved ‘reported changes’. For these data Strengths and limitations of the review the comparator is implicit in the respondent’sreply;they This is the first systematic review of which we are aware are reporting a change over time due to the protected area. that attempts to identify and synthesise, in a transparent Whilst these results are relevant (externally valid) to the Pullin et al. Environmental Evidence 2013, 2:19 Page 37 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 review in hand, they can be susceptible to significant recall There is a lack of primary studies estimating impact bias and questioning bias (elements of internal validity). of PAs on human well-being using direct measurement Several studies attempted to retrospectively elicit opinions techniques in a BACI format. In addition to a generally and attitudes towards protected area establishment many high susceptibility to bias, very few studies employed ro- years after the event, which is similarly open to substantial bust comparators over appropriate time frames in order sources of bias. to maximise evidence linking protected areas to ob- For these reasons, results in the form of reported served human well-being impacts. Only one included changes which involve significant recall should be viewed study used a full BACI design to account for spatial and with caution, and studies critiqued in depth for potential temporal confounding variables. Only three studies used sources of bias. We attempted to account for these sources direct data collection. of bias during critical appraisal using our ‘susceptibility to We found a surprisingly small number of studies on bias’ scoring system. health of populations. Only nine studies reported data Sixty-six of the 79 studies accepted following critical on human health impacts of protected areas. This is sur- appraisal collected data in the form of self-reported mea- prising since the majority of articles in this review pur- sures. Fifty-six of these articles failed to provide details port to measure human well-being. Difficulties in ethical of the questionnaires given to respondents, and only two approval for human study may account in part for the articles provided a copy of the survey instrument in full paucity of health studies. [76,105]. Without details of the questioning involved in these surveys it is difficult to assess questioning bias. Review conclusions Variation in all the question elements (PECO; popula- Implication for policy/management tion, exposure, comparator, outcome) and the high de- The evidence base provides a range of possibilities to in- gree of specificity in outcome measures identified in this form but little evidence to support decision making on review provides problems for synthesis. In particular, how to maximise positive impacts of PAs on human well- studies based on self-reported measures commonly asked being. The diversity of studies and of outcomes measured, very specific questions that could not then be synthesised together with the diversity (or lack of clear signal) in the along with other similar outcomes. Similarly, a high degree data suggests that impacts of PAs are highly context of variability in the choice and design of comparators pre- dependent. However, the evidence base is insufficient to vented synthesis. In some studies, the inside-outside com- provide any power with which to predict impacts on well- parison was open to many confounding factors that cloud being from a knowledge of their context. It logically fol- the link between protected area presence and impacts. lows that there is an insufficient evidence base to identify This highlights the difficulty of balancing minimisation of circumstances/variables/effect modifiers that might lead to spillover effects, whereby the comparator population is greater or lesser impact. At present, the available evidence close enough to feel the effects of the exposure, and con- base is failing to inform policy on the progress (or lack of trol of non-target variables. Whilst some studies accounted it) being made, since 1992, toward lessening negative and for this problem by including confounding variables in promoting positive impact of PAs on human well-being. statistical models, many others did not. Furthermore, very few studies examined differences in environmental condi- tions between the comparator and exposure populations. Implication for research High susceptibility to bias in most studies limits ability The nature of the research reported to date forms a di- to attribute outcomes/impacts to presence of PAs. Forty- verse and fragmented evidence base that is insufficiently five studies were excluded during critical appraisal due to developed to reliably inform future policy decisions flaws in experimental design and data analysis, or due to a (recognising that many included studies did not set out lack of methodological detail. However, many studies in- to address the review question). Many studies appear to cluded after the first stage of critical appraisal also failed to have been conducted opportunistically and lack baseline account for confounding variables, selected replicates in a measures. There is no evidence of a strategic approach non-random manner, and used opportunistic methodology. or strategic investment to this field of research beyond The most frequently occurring factor that affected the sus- individual research group initiatives. If a sufficient evidence ceptibility to bias score in included studies, however, was a base is to be formed then there is a need for concerted failure to appropriately report their methodology. Signifi- programme of research rather than an uncoordinated short cant details such as recall period, response rate, item pool term opportunistic approach. balance and order, sample selection process, sample size, The diversity of outcome measures and the consequent and sample location were not disclosed in a large number difficulty for synthesis suggests a need for use of standard of cases. Together, these factors limit the ability to attribute indicators of human well-being that allow comparison the reported impacts to protected areas. among studies and meaningful synthesis of evidence. Pullin et al. Environmental Evidence 2013, 2:19 Page 38 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 Comparative research needs to progress from PA/no a payoff between maximising similarity and minimising PA to PA type A/PA type B comparisons. Comparisons spillover (the overflow of impacts from the intervention should be made between potential proximate causes of into the nearby comparator). Statistical tests can help to positive or negative impacts when the ultimate cause confirm similarity across intervention and comparator is PA establishment/management. This review suggests populations, and descriptive variables can be included in some of the candidates to be governance models, exist- models that test for the significance of the intervention in ing social capital, cultural diversity and poverty index. order to account for differences that might occur. It would be helpful to research efforts for funders to find consensus on minimum standards for methodolo- Replication gies, for both qualitative and quantitative evidence, that Care must be taken to ensure that there is an appropriate provide improved quality and thus reliability of data. The trade-off between a study’s accuracy and its precision. large proportion of included studies that suffered high When combining many studies in a synthesis, more accur- susceptibility to bias is an indicator of such a need and ate results are preferable to more precise ones. For ex- also an indicator that scarce research resources are not be- ample, a study that measures daily resource extraction ing used effectively. over a year in ten households from one intervention and one comparator village is less likely to reflect the true im- Recommended study design pact of the intervention than a study that measures daily In order to better assess the impacts of protected areas resource extraction over a month from 12 intervention on human well-being we make the following recommen- and 12 comparator villages. This spectrum is not clear-cut, dations for future research study design and reporting; however, and the allocation of resources to pseudorepli- cation (improving precision) and true replication (improv- Methodological detail; Studies must report sufficient ing accuracy) must be considered carefully. Indeed, the details regarding the location of sample sites (in scale at which conclusions will be drawn defines what is relation to the protected area boundaries in particular), pseudoreplication and what is true replication, and this the degree of replication, the data collection tool definition may be different for the author and the system- (e.g. quote questions posed to respondents in atic reviewer. questionnaires), the method of sample selection (e.g. random or purposeful), and the times and duration of Statistics sampling. This is not an exhaustive list, and sufficient Statistics, both in summarising results and analysing pat- detail must be provided to allow the sampling to be terns, must be used with great care. We recommend that repeated. Where information cannot fit within a statistician be consulted during experimental design in published articles these details should be provided in order to optimise design for analysis. The use of models supplementary material. that account for changes in non-target variables across Baseline assessment; Where changes following temporal and spatial scales are recommended, but tests establishment or change in protected area governance for differences in confounders between intervention and are being investigated, adequate baselines must be comparator populations are also appropriate. Where in- assessed. Although this is difficult and requires formation can be presented in summary statistics (e.g. planning prior to the intervention, full ‘before-after- mean/median and standard deviation/confidence inter- control-intervention’ (BACI) study design is vital to vals) this will aid future meta-analysis. account for confounding temporal and spatial confounding factors. By assessing baselines, any Additional files differences between intervention and comparator populations can be compared relative to the starting Additional file 1: Search Strategy. Details of the development of the finalised search string, including scoping and testing against list of conditions to strengthen the evidence towards causation. recommended key articles, and recording of search results from all databases, including duplicate removal and library file creation. Matched controls Additional file 2: Unobtainable and Un-translated. A list of articles ‘Control’ or ‘comparator’ populations are vital to enable that could not be obtained in full text, along with a list of foreign language articles that were not translated for consideration in the conclusions to be drawn about impacts in the absence of quantitative review. the intervention. A reliable comparison requires that as Additional file 3: Web site searches. A list of organisational web sites many other variables describing the environment are held searched for material, along with search terms, hits returned and actions constant or matched between comparator and interven- taken for the quantitative review. tion populations, allowing only the intervention to change Additional file 4: Bibliographic Searches. Details of five relevant review article bibliographies searched for supplemental material in the in an ideal situation. In practice this is very difficult (and quantitative review. why baseline assessment is important), and there is often Pullin et al. Environmental Evidence 2013, 2:19 Page 39 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 areas in the developing world: Economic valuation of Morro do Diabo Additional file 5: Coding Tool. The coding tool used to assess articles State Park, Atlantic Rainforest, São Paulo State (Brazil). Ecol Econ 2008, at full text for the qualitative review, including the coding of outcomes 66:359–370. used in both reviews. 8. Ferraro PJ, Hanauer MM, Sims KRE: Conditions associated with protected Additional file 6: Excluded. Lists of articles excluded at full text assessment area success in conservation and poverty reduction. Proc Natl Acad Sci from both the qualitative and quantitative reviews along with reasons. 2011, 108:13913–13918. Additional file 7: Narrative Synthesis. Full narrative synthesis tables for 9. West P, Igoe J, Brockington D: Parks and peoples: the social impact of 79 included studies in quantitative review. protected areas. Annual Reviews in Anthropology 2006, 35:251–277. 10. Sutherland WJ, Adams WM, Aronson RB, Aveling R, Blackburn TM, Broad S, Additional file 8: Supplementary Descriptive Statistics for Ceballos G, Cote IM, Cowling RM, Da Fonseca GAB, et al: One hundred ‘Qualitative Synthesis’. Descriptive statistics for the 305 articles included questions of importance to the conservation of global biological in the systematic map of the first stage of analysis of the qualitative review. diversity. Conserv Biol 2009, 23:557–567. Additional file 9: Qualitative Synthesis. The full synthesis of 30 studies 11. Adams WM, Aveling R, Brockington D, Dickson B, Elliott J, Hutton J, Roe D, considered in the qualitative review, from which the summary synthesis Vira B, Wolmer W: Biodiversity conservation and the eradication of was produced. poverty. Science 2004, 306:1146–1149. Additional file 10: Table S1. Summary table of included studies and 12. Dudley N: Guidelines for Applying Protected Area Managment Categories. their measured outcomes in the quantitative review that were scored as Gland Switzerland: World Conservation Union; 2009. having ‘low’ or ‘medium’ susceptibility to bias (Susc. to Bias). 13. Pullin AS, Bangpan M, Dalrymple S, Dickson K, Healey JR, Hockley N, Jones JPG, Knight TM, Oliver S: Human well-being impacts of terrestrial Additional file 11: Details of the information considered during protected areas? CEE protocol 11–009 Collaboration for Environmental critical appraisal from each of the 79 studies included in the Evidence: wwwenvironmentalevidenceorg/SR11009html; 2012. quantitative review. 14. Neuman L: Social Research Methods: Qualitative and Quantitative Approaches. New York: Allyn and Bacon; 1997. Competing interest 15. Gough D, Thomas J, Oliver S: Clarifying differences between review There are no potential conflicts of interest to report. designs and methods. Systematic Reviews 2012, 1:28. 16. Oliver SR, Rees RW, Clarke‐Jones L, Milne R, Oakley AR, Gabbay J, Stein K, Authors' contributions Buchanan P, Gyte G: A multidimensional conceptual framework for ASP managed and planned the conduct of the SR. SO managed the conduct analysing public involvement in health services research. Health Expect of the qualitative section. SD and NRH undertook searches. Quantitative 2008, 11:72–84. Review; SD and NRH screened for inclusion. NRH carried out data extraction 17. Ritchie J, Spencer L: Qualitative data analysis for applied policy research. and critical appraisal. JPGJ, JRH, NH and TK assisted in planning, In Analyzing Qualitative Data. Edited by Bryman A, Burgess RG. London: interpretation and synthesis. Qualitative Review; MB, KD, HH SO and CV Routledge; 1994:305–329. undertook screening, data extraction and synthesis. All authors contributed 18. Pope C, Ziebland S, Mays N: Qualitative research in health care: Analysing to writing and approved the final manuscript. qualitative data. BMJ: British Medical Journal 2000, 320:114. 19. Thomas J, Harden A, Newman M: Synthesis: combining results systematically and appropriately. In Introducing systematic reviews. Edited Acknowledgments by Gough D, Oliver S, J. T. London: Sage Publications; 2012. We thank Sandra Diaz and Tom Hammond for their help in administering the progress of this review. 20. Vasak K: Human Rights: A Thirty-Year Struggle: the Sustained Efforts to give Force of law to the Universal Declaration of Human Rights. UNESCO Courier 30:11. Paris: United Nations Educational, Scientific, and Cultural Organization; 1977. Sources of support 21. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR: Introduction to meta- This review was funded by the Scientific and Technical Advisory Panel of the analysis. Chichester, UK: Wiley; 2011. Global Environment Facility. 22. Strickland-Munro J, Moore S: Indigenous involvement and benefits from tourism in protected areas: a study of Purnululu National Park and Author details Warmun Community, Australia. J Sustain Tour 2013, 21:26–41. Centre for Evidence-Based Conservation, School of Environment, Natural 23. Allendorf TD, Smith JLD, Anderson DH: Residents’ perceptions of Royal Resources and Geography, Bangor University, LL57 2UW Bangor, Gwynedd, UK. Bardia National Park, Nepal. Landscape and Urban Planning 2007, 82:33–40. Evidence-informed Policy and Practice Information and Coordinating Centre (EPPI-Centre), Institute of Education, University of London, 20 Bedford Way, 24. Haukeland JV: Tourism stakeholders’ perceptions of national park WC1H 0NR London, UK. School of Environment, Natural Resources and management in Norway. J Sustain Tour 2011, 19:133–153. Geography, Bangor University, LL57 2UW Bangor, Gwynedd, UK. 25. Castillo A, Magana A, Pujadas A, Martínez L, Godínez C: Understanding the interaction of rural people with ecosystems: A case study in a tropical Received: 9 January 2013 Accepted: 17 July 2013 dry forest of Mexico. Ecosystems 2005, 8:630–643. Published: 28 October 2013 26. Mehring M, Seeberg-Elverfeldt C, Koch S, Barkmann J, Schwarze S, Stoll- Kleemann S: Local institutions: Regulation and valuation of forest use— Evidence from Central Sulawesi, Indonesia. Land Use Policy 2011, 28:736–747. References 27. Petrzelka P, Marquart-Pyatt S: “With the Stroke of a Pen”: Designation of 1. Jenkins CN, Joppa L: Expansion of the global terrestrial protected area the Grand Staircase Escalante National Monument and the Impact on system. Biol Conserv 2009, 142:2166–2174. Trust. Hum Ecol 2013, 41:285–297. 2. Bertzky B, Corrigan C, Kemsey J, Kenney S, Ravilious CCB, Burgess N: Protected 28. Gerritsen PRW: Diversity at stake: a farmers’ perspective on biodiversity and Planet Report 2012: Tracking progress towards global targets for protected areas. conservation in western Mexico. Mexico: Wageningen Universiteit Holanda; 2002. Cambridge, UK: IUCN, Gland, Switzerland and UNEP-WCMC; 2012. 29. Diaw MC, Tiani AM: Fences in our heads: A discourse analysis of the 3. Brockington D, Igoe J, SCHMIDT‐SOLTAU K: Conservation, human rights, Korup resettlement stalemate. J Sustain For 2010, 29:221–251. and poverty reduction. Conserv Biol 2006, 20:250–252. 30. Nguiffo S: One forest and two dreams: The constraints imposed on the 4. Dowie M: Conservation refugees: The hundred-year conflict between global Baka in Miatta by the Dja Wildlife Reserve. In Indigenous people and conservation and native peoples. Cambridge USA: MIT Press; 2009. protected areas in Africa. Edited by Nelson J, Hossack L, Moreton-in-Marsh. 5. Turner WR, Brandon K, Brooks TM, Gascon C, Gibbs HK, Lawrence KS, UK: Forest Peoples Programme; 2003:195–214. Mittermeier RA, Selig ER: Global biodiversity conservation and the 31. Sletten M, Vedeld P, Kaboggoza J: To co-operate or not to co-operate? a alleviation of poverty. BioScience 2012, 62:85–92. study of collaborative management planning in Mount Elgon National Park. 6. Kramer RA, Sharma NP, Munasinghe M: Valuing tropical forests: methodology Uganda: NORAGRIC working paper; 2008. and case study of Madagascar. Madagascar: World Bank Publications; 1995. 7. Adams C, Seroa da Motta R, Ortiz RA, Reid J, Ebersbach Aznar C, de Almeida 32. Bolaane M: The impact of game reserve policy on the River BaSarwa/ Sinisgalli PA: The use of contingent valuation for evaluating protected Bushmen of Botswana. Social Policy & Administration 2004, 38:399–417. Pullin et al. Environmental Evidence 2013, 2:19 Page 40 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 33. Milgroom J, Spierenburg M: Induced volition: Resettlement from the 59. Bajracharya SB, Furley PA, Newton AC: Impacts of community-based Limpopo National Park, Mozambique. J Contemp Afr Stud 2008, 26:435–448. conservation on local communities in the Annapurna Conservation Area, 34. Mbaiwa JE: Wildlife resource utilisation at Moremi Game Reserve and Nepal. Biodivers Conserv 2006, 15:2765–2786. Khwai community area in the Okavango Delta, Botswana. 60. Mehta JN, Kellert SR: Local attitudes toward community-based J Environ Manage 2005, 77:144–156. conservation policy and programmes in Nepal: a case study in the 35. Torri MC: Conservation, relocation and the social consequences of Makalu-Barun Conservation Area. Environ Conserv 1998, 25:320–333. conservation policies in protected areas: Case study of the Sariska Tiger 61. Naughton-Treves L, Alix-Garcia J, Chapman CA: Lessons about parks and Reserve. India. Conservation and Society 2011, 9:54. poverty from a decade of forest loss and economic growth around 36. Keskitalo ECH, Lundmark L: The controversy over protected areas and Kibale National Park, Uganda. Proc Natl Acad Sci 2011, 108:13919–13924. forest-sector employment in Norrbotten, Sweden: forest stakeholder 62. Rinzin C, Vermeulen WJV, Wassen MJ, Glasbergen P: Nature Conservation perceptions and statistics. Soc Nat Resour 2009, 23:146–164. and Human Well-Being in Bhutan An Assessment of Local Community Perceptions. J Environ Dev 2009, 18:177–202. 37. Spenceley A, Goodwin H: Nature-based tourism and poverty alleviation: impacts of private sector and parastatal enterprises in and around 63. Okello MM, Buthmann E, Mapinu B, Kahi HC: Community opinions on Kruger National Park, South Africa. Current Issues in Tourism 2007, wildlife, resource use and livelihood competition in Kimana Group Ranch 10:255–277. near Amboseli, Kenya. Open Conservation Biology Journal 2011, 5:1–12. 38. Songorwa AN: Community-based wildlife management (CWM) in 64. Lundgren T: Environmental protection and impact on adjacent Tanzania: Are the communities interested? World Dev 1999, 27:2061–2079. economies: evidence from the Swedish mountain region. 39. Stone M, Wall G: Ecotourism and community development: case studies Growth and Change 2009, 40:513–532. from Hainan, China. Environ Manage 2004, 33:12–24. 65. Richardson RB, Fernandez A, Tschirley D, Tembo G: Wildlife Conservation in 40. Bruyere BL, Beh AW, Lelengula G: Differences in perceptions of Zambia: impacts on rural household welfare. World Dev 2012, 40:1068–1081. communication, tourism benefits, and management issues in a 66. Sims KRE: Conservation and development: Evidence from Thai protected protected area of rural Kenya. Environ Manage 2009, 43:49–59. areas. J Environ Econ Manag 2010, 60:94–114. 41. Slater R: Between a rock and a hard place: contested livelihoods in 67. Andam KS, Ferraro PJ, Sims KR, Healy A, Holland MB: Protected areas Qwaqwa National Park, South Africa. Geogr J 2002, 168:116–129. reduced poverty in Costa Rica and Thailand. Proc Natl Acad Sci 2010, 42. Bizikova L, Nijnik M, Kluvanková-Oravská T: Sustaining multifunctional 107:9996–10001. forestry through the developing of social capital and promoting 68. Canavire-Bacarreza G, Hanauer MM: Estimating the impacts of bolivia’s participation: a case of multiethnic mountain communities. protected areas on poverty. World Dev 2012, 41:265–285. Small-Scale Forestry 2012, 11:301–319. 69. Ferraro PJ, Hanauer MM: Protecting Ecosystems and Alleviating Poverty 43. Ormsby A, Kaplin BA: A framework for understanding community with Parks and Reserves: ‘Win-Win’or Tradeoffs? Environ Resour Econ 2011, resident perceptions of Masoala National Park, Madagascar. Environ 48:269–286. Conserv 2005, 32:156–164. 70. Foerster S, Wilkie DS, Morelli GA, Demmer J, Starkey M, Telfer P, Steil M: 44. First Peoples Worldwide: Okiciyab: To help each other-Promoting best Human livelihoods and protected areas in Gabon: a cross-sectional practices in Indigenous community Development. 2006. Final report, June. comparison of welfare and consumption patterns. Oryx 2011, 45:347–356. 45. ESCO: Biosphere reserves: The Seville Strategy and the Statutory Framework of 71. Cardozo M: Economic displacement and local attitude towards protected the World Network. Paris: UNESCO; 1996. area establishment in the Peruvian Amazon. Geoforum 2011, 42:603–614. 46. World Parks Congress: The Durban Action Plan. Durban, South Africa: World 72. Kayser D, Sobrevila C, Ledec G: From Planning to the Implementation of a Parks Congress; 2003. Successful Conservation and Socio-Economic Model. South Africa: Report from 47. Herrold-Menzies M: From Adversary to Partner: The Evolving Role of The World Bank and Addo Elephant National Park; 2011. Caohai Nature Reserve in the Lives of Reserve Residents. Canadian 73. Saayman M, Saayman A: Regional development and national parks in Journal of Development Studies/Revue canadienne d’études du développement South Africa: Lessons learned. Tour Econ 2010, 16:1037–1064. 2006, 27:39–50. 74. Ninan KN: Non Timber Forest Products and Biodiversity Conservation: 48. Lunstrum E: Mozambique, neoliberal land reform, and the Limpopo a Study of Tribals in a Protected Area in India. In Conserving and Valuing National Park. Geogr Rev 2008, 98:339–355. Ecosystem Services and Biodiversity: Economic, Institutional and Social 49. Almudi T, Berkes F: Barriers to empowerment: Fighting eviction for Challenges. Edited by Ninan KN. London: Earthscan; 2012:99–112. conservation in a southern Brazilian protected area. Local Environment 75. ETFRN NEWS 39/40: Globalisation, localisation and tropical forest 2010, 15:217–232. management. http://www.etfrn.org/etfrn/newsletter/pdf/etfrnnews3940.pdf. 50. Davis A: ‘ Ha! What is the benefit of living next to the park?’ Factors 76. Phamtrong N, Swan S: An investigation into the drivers of forest limiting in-migration next to Tarangire National Park, Tanzania. encroachment in Bidoup Nui Ba National Park, Vietnam. Vietnam: Report to Conserv Soc 2011, 9:25–34. WWF Greater Mekong - Vietnam Country Programme; 2009. 51. Hoole A, Berkes F: Breaking down fences: Recoupling social–ecological 77. Korhonen K, Rahkonen O, Hemminki E: Implications of integrated nature systems for biodiversity conservation in Namibia. Geoforum 2010, 41:304–317. conservation for human reproductive health: a case study from 52. Hartter J: Attitudes of rural communities toward wetlands and forest Ranomafana National Park, Madagascar. Dev South Afr 2004, 21:603–621. fragments around Kibale National Park, Uganda. Hum Dimens Wildl 2009, 78. Mishra C: Socioeconomic transition and wildlife conservation in the Indian 14:433–447. Trans-Himalaya. Journal of the Bombay Natural History Society 2000, 97:25–32. 53. Bedunah DJ, Schmidt SM: Pastoralism and protected area management in 79. Sheppard DJ, Moehrenschlager A, McPherson JM, Mason JJ, Pattenden J, Mongolia’s Gobi Gurvansaikhan National Park. Dev Chang 2004, 35:167–191. Khandker S, Koolwal G, Biggs B, King L, Basu S: Ten years of adaptive 54. Ogra MV: Human–wildlife conflict and gender in protected area community-governed conservation: evaluating biodiversity protection borderlands: a case study of costs, perceptions, and vulnerabilities from and poverty alleviation in a West African hippopotamus reserve. Uttarakhand (Uttaranchal), India. Geoforum 2008, 39:1408–1422. Environ Conserv 2010, 37:270–282. 55. Stronza A, Gordillo J: Community views of ecotourism. Ann Tour Res 2008, 80. Shrestha RK, Alavalapati JRR: Linking conservation and development: An 35:448–468. analysis of local people’s attitude towards Koshi Tappu wildlife reserve, 56. Yasuda A: The Impacts of Sport Hunting on the Livelihoods of Local Nepal. Environ Dev Sustain 2006, 8:69–84. People: A Case Study of Bénoué National Park, Cameroon. Soc Nat Resour 81. Bonaiuto M, Carrus G, Martorella H, Bonnes M: Local identity processes and 2011, 24:860–869. environmental attitudes in land use changes: The case of natural 57. Nyahongo J, Holmern T, Kaltenborn BP, Roskaft E: Spatial and temporal protected areas. J Econ Psychol 2002, 23:631–653. variation in meat and fish consumption among people in the western 82. Jim C, Xu SSW: Stifled stakeholders and subdued participation: Serengeti, Tanzania: the importance of migratory herbivores. interpreting local responses toward Shimentai Nature Reserve in South Oryx 2009, 43:258. China. Environ Manage 2002, 30:327–341. 58. Sarker A, Røskaft E: Human attitudes towards the conservation of 83. Gubbi S, Linkie M, Leader-Williams N: Evaluating the legacy of an protected areas: a case study from four protected areas in Bangladesh. integrated conservation and development project around a tiger reserve Oryx 2011, 45:391–400. in India. Environ Conserv 2008, 35:331. Pullin et al. Environmental Evidence 2013, 2:19 Page 41 of 41 http://www.environmentalevidencejournal.org/content/2/1/19 84. Infield M, Namara A: Community attitudes and behaviour towards conservation: an assessment of a community conservation programme around Lake Mburo National Park, Uganda. Oryx 2001, 35:48–60. 85. Sekhar NU: Crop and livestock depredation caused by wild animals in protected areas: the case of Sariska Tiger Reserve, Rajasthan, India. Environ Conserv 1998, 25:160–171. 86. Fiallo EA, Jacobson SK: Local Communities and Protected Areas: Attitudes of Rural Residents Towards Conservation and Machalilla National Park, Ecuador. Environ Conserv 1995, 22:241–249. 87. Harada K: Attitudes of local people towards conservation and Gunung Halimun National Park in West Java, Indonesia. J For Res 2003, 8:271–282. 88. Hartter J, Goldman AC: Life on the Edge: Balancing Biodiversity, Conservation, and Sustaining Rural Livelihoods around Kibale National Park, Uganda. Focus Geogr 2009, 52:11–17. 89. Ite DUE: Community perceptions of the Cross River national park, Nigeria. Environ Conserv 1996, 23:351–357. 90. Cihar M, Stankova J: Attitudes of stakeholders towards the Podyji/Thaya River Basin National Park in the Czech Republic. J Environ Manage 2006, 81:273–285. 91. Alexander SE: Resident attitudes towards conservation and black howler monkeys in Belize: the Community Baboon Sanctuary. Environ Conserv 2000, 27:341–350. 92. Rugendyke B, Son NT: Conservation costs: Nature-based tourism as development at Cue Phuong national park, Vietnam. Asia Pacific Viewpoint 2005, 46:185–200. 93. Saayman M, Saayman A, Ferreira M: The socio-economic impact of the Karoo National Park. Koedoe 2009, 51:1–10. 94. Just RE, Hueth DL, Schmitz A: The Welfare Economics of Public Policy: A Practical Approach to Project and Policy Evaluation. Cheltenham, UK: Edward Elgar; 2004. 95. Shrestha RK, Alavalapati JRR, Seidl AF, Weber KE, Suselo TB: Estimating the local cost of protecting Koshi Tappu Wildlife Reserve, Nepal: a contingent valuation approach. Environ Dev Sustain 2007, 9:413–426. 96. Jemal H, Bogale A, Hagedorn K: Welfare loss for pastoralists due to wildlife protection areas: the case of Awash National Park, Ethiopia. Quarterly Journal of International Agriculture 2008, 47:217–238. 97. Kremen C, Niles J-O, Dalton MG, Daily GC, Ehrlich PR, Fay JP, Grewal D, Guillery RP: Economic incentives for rainforest conservation across scales. Science 2000, 288:1828–1832. 98. Han F, Yang Z, Wang H, Xu X: Estimating willingness to pay for environment conservation: a contingent valuation study of Kanas Nature Reserve, Xinjiang, China. Environ Monit Assess 2011, 180:451–459. 99. Maharana I, Rai S, Sharma E: Environmental economics of the Khangchendzonga National Park in the Sikkim Himalaya, India. GeoJournal 2000, 50:329–337. 100. Jones N, Iosifides T, Evangelinos KI, Florokapi I, Dimitrakopoulos PG: Investigating knowledge and perceptions of citizens of the National Park of Eastern Macedonia and Thrace, Greece. International Journal of Sustainable Development & World Ecology 2012, 19:25–33. 101. Asciuto A, Schimmenti E: Comparison between Willingness-to-Pay expressed by a panel of forestry experts and by a sample of non-expert respondents in a pilot survey conducted for a contingent valuation study. Mediterranean Journal of Economics, Agriculture and Environment 2005, 4:42–51. 102. Abbot JIO, Mace R: Managing protected woodlands: fuelwood collection and law enforcement in Lake Malawi National Park. Conserv Biol 2001, 13:418–421. 103. Kavanagh J, Oliver S, Lorenc T, Caird J, Tucker H, Harden A, Greaves A, Thomas J, Oakley A: School-based cognitive-behavioural interventions: A Submit your next manuscript to BioMed Central systematic review of effects and inequalities. Heal Sociol Rev 2009, 18:61–78. and take full advantage of: 104. Petticrew M, Tugwell P, Kristjansson E, Oliver S, Ueffing E, Welch V: Damned if you do, damned if you don’t: subgroup analysis and equity. J Epidemiol Community Health 2012, 66:95–98. • Convenient online submission 105. Reid R, Stone M, Whitely T: Economic value of wilderness protection and • Thorough peer review recreation in British Columbia. (Victoria BC): FRDA Working Paper; 1995. • No space constraints or color figure charges doi:10.1186/2047-2382-2-19 • Immediate publication on acceptance Cite this article as: Pullin et al.: Human well-being impacts of terrestrial • Inclusion in PubMed, CAS, Scopus and Google Scholar protected areas. Environmental Evidence 2013 2:19. • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit

Journal

Environmental EvidenceSpringer Journals

Published: Oct 28, 2013

There are no references for this article.