Hostname: page-component-7d8f8d645b-2q4x6 Total loading time: 0 Render date: 2023-05-29T20:25:27.543Z Has data issue: false Feature Flags: { "useRatesEcommerce": true } hasContentIssue false

Drivers of prohibited natural resource collection in Chitwan National Park, Nepal

Published online by Cambridge University Press:  07 April 2022

Scott T Yabiku*
Penn State University, 306 Oswald Tower, Penn State University, University Park, PA16802, USA
Abigail Sullivan
Boston University Earth & Environment, 685 Commonwealth Avenue, Boston, MA02215, USA
Abigail M York
School of Human Evolution and Social Change, Arizona State University, Tempe, AZ 85287, USA
Qunshan Zhao
Urban Big Data Centre 7-302, 7 Lilybank Gardens, University of Glasgow, Glasgow, G12 8RZ, UK
Jennifer E Glick
Penn State University, 601 Oswald Tower, Penn State University, University Park, PA16802, USA
Sharon J Hall
School of Life Sciences, Arizona State University, Tempe, AZ85287, USA
Dirgha J Ghimire
Population Studies Center, University of Michigan, 426 Thompson St, Ann Arbor, MI48106, USA
Li An
San Diego State University, 5500 Campanile Drive, San Diego, CA 92182-4493, USA
Author for Correspondence: Dr Scott T Yabiku, Email:
Rights & Permissions[Opens in a new window]


Protected areas (PAs) are critical for achieving conservation, economic and development goals, but the factors that lead households to engage in prohibited resource collection in PAs are not well understood. We examine collection behaviours in community forests and the protected Chitwan National Park in Chitwan, Nepal. Our approach incorporates household and ecological data, including structured interviews, spatially explicit data on collection behaviours measured with computer tablets and a systematic field survey of invasive species. We pair our data with a framework that considers factors related to a household’s demand for resources, barriers to prohibited resource collection, barriers to legal resource collection and alternatives to resource collection. The analysis identifies key drivers of prohibited collection, including sociodemographic variables and perceptions of an invasive plant (Mikania micrantha). The social-ecological systems approach reveals that household perceptions of the presence of M. micrantha were more strongly associated with resource collection decisions than the actual ecologically measured presence of the plant. We explore the policy implications of our findings for PAs and propose that employing a social-ecological systems approach leads to conservation policy and scientific insights that are not possible to achieve with social or ecological approaches alone.

Research Paper
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (, which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
© The Author(s), 2022. Published by Cambridge University Press on behalf of Foundation for Environmental Conservation


Protected areas (PAs) have been set aside for specific management goals (Geldmann et al. Reference Geldmann, Manica, Burgess, Coad and Balmford2019). Globally, the total area designated protected has increased dramatically, from negligible in the early twentieth century to c. 22.5 million km2 of land and 28.1 million km2 of sea in 2021, although PAs are often chosen based more on political considerations than scientific reasoning (Watson et al. Reference Watson, Dudley, Segan and Hockings2014, UNEP-WCMC & IUCN 2021). PAs now strive to balance social, natural and economic goals (Gutiérrez et al. Reference Gutiérrez, Hilborn and Defeo2011). Although the success of many PAs has been lauded, some PAs have been considered failures (Wilkie et al. Reference Wilkie, Carpenter and Zhang2001, Edgar et al. Reference Edgar, Stuart-Smith, Willis, Kininmonth, Baker, Banks and Barrett2014). These failures are frequently driven by inattention to the sociopolitical context in the initial design stages (Chuenpagdee et al. Reference Chuenpagdee, Pascual-Fernández, Szeliánszky, Alegret, Fraga and Jentoft2013, Watson et al. Reference Watson, Darling, Venter, Maron, Walston and Possingham2016). One important gap in understanding why PAs fail is unpacking the motivations of people who violate rules meant to protect these areas.

There are challenges to studying why some people abide by rules within PAs while others disregard them. First, it is difficult to obtain direct reports on rule violations from violators because violations are often either legally prohibited or socially unacceptable. Prior work has typically relied on indirect measures, although some studies have engaged in direct questioning or observation of violators (e.g., Yonariza & Webb Reference Yonariza2007, Bergseth et al. Reference Bergseth, Williamson, Russ, Sutton and Cinner2017). Second, even if rule violations are measured, information on the violations are rarely spatially explicit (Hofer et al. Reference Hofer, Campbell, East and Huish2000, Knapp et al. Reference Knapp, Peace and Bechtel2017). Without understanding the spatial distribution of rule violations, it is difficult to maximize the efficiency of enforcement resources, including the distribution and intensity of guard or patrol efforts (Hofer et al. Reference Hofer, Campbell, East and Huish2000). Third, many studies of rule violators often lack representative samples of the human populations that use local resources (e.g., Mann Reference Mann1995, Ward et al. Reference Ward, Stringer and Holmes2018). While non-representative studies contribute unique and rich descriptions, they may incompletely describe prohibited collection behaviours.

PAs are complex social-ecological systems in which multiple pathways and mechanisms affect each other (Cumming & Allen Reference Cumming and Allen2017, Picone et al. Reference Picone, Buonocore, Claudet, Chemello, Russo and Franzese2020). Due to disciplinary separation, social and ecological processes were in the past often studied separately or with minimum consideration of their interaction with one another. With human–environment systems (Turner et al. Reference Turner, Matson, McCarthy, Corell, Christensen and Eckley2003) or coupled human and natural systems (Liu et al. Reference Liu, Dietz, Carpenter, Alberti, Folke and Moran2007), the social-ecological systems (SESs) framework assumes that such systems are integrated and complex, with humans and natural subsystems interacting with one another (Dietz et al. Reference Dietz, Ostrom and Stern2003). Studies of prohibited natural resource collection typically focus on a single part of the system and on data relating to ranger patrol observations (Critchlow et al. Reference Critchlow, Plumptre, Driciru, Rwetsiba, Stokes and Tumwesigye2015), respondent self-reports (Nuno et al. Reference Nuno, Bunnefeld, Naiman and Milner-Gulland2013), spatial mapping (Faulkner et al. Reference Faulkner, Stevens, Romañach, Lindsey and Le Comber2018) or stakeholder perspectives (Kahler et al. Reference Kahler, Roloff and Gore2013). We argue that multiple data perspectives are needed to better understand prohibited collection in complex social-ecological systems. Note that many studies use the term ‘illegal’ when describing prohibited natural resource collection activities, but we specifically use ‘prohibited’ as opposed to ‘illegal’ due to the negative and inappropriate implication that community members collecting in PAs are criminals.

The objectives of this research were to: (1) collect human data on the household drivers of prohibited collection and perceptions of Mikania micrantha in community forests; (2) collect institutional data from community forests managers on the perceived spatial distribution of M. micrantha; (3) collect ecological data on the spatial distribution of M. micrantha; and (4) examine the impacts of these human, institutional and ecological factors on prohibited collection. Each of our data sources benefits from: (1) direct measurement, (2) spatial precision and (3) representative sampling. We focus on M. micrantha because it has been called the ‘most serious weed in the tropical and subtropical areas of Nepal’ (Baral & Adhikari Reference Baral, Adhikari, Ellison, Sankaran and Murphy2017), and our fieldwork with residents, forest managers and non-governmental organization (NGO) officials indicated it to be the most concerning invasive species in the region. Although our setting is the area surrounding Chitwan National Park (CNP) in Nepal and the prohibited resources are fuelwood and fodder within park boundaries, we believe our comprehensive approach to prohibited collection is applicable to many complex social-ecological systems.


Research setting

The western Chitwan Valley of Nepal lies at 120–815 m elevation in the subtropical Terai region of Nepal, characterized by elongated valleys created by faults within the foothills of the actively uplifting Himalaya plateau (Lehmkuhl Reference Lehmkuhl1994). Much of the area was deforested and settled in the 1950s and 1960s, and today the 250-km2 western Chitwan Valley is home to over 200 000 individuals (Central Bureau of Statistics – Nepal 2016). While the northern region is intensively cultivated, the southern and western regions are dominated by CNP, a 932-km2 UNESCO World Heritage Site established in 1973 and internationally recognized for its largely intact primary forests, grasslands and habitats for endangered and vulnerable, charismatic mammal species. CNP is home to the vulnerable one-horned rhinoceros (Rhinoceros unicornis) and endangered Bengal tiger (Panthera tigris tigris) (Spiteri & Nepal Reference Spiteri and Nepal2008). The Forest Act 1993 created the existing community forest (CF) system – a type of decentralized, locally governed forest management – to prevent habitat degradation and to sustain the quality of life for growing numbers of buffer zone (the area surrounding CNP) residents (Nagendra Reference Nagendra2002). This arrangement allows for residents to legally collect resources from CF areas while CNP is rendered off-limits (Jones Reference Jones2007).

Areas near CFs and CNP are experiencing rapid human population growth and are threatened by three of the world’s 100 worst invasive exotic species: M. micrantha, Chromolaena odorata and Lantana camara (ISSG 2000). M. micrantha is particularly problematic, spreading quickly and degrading resources in these important landscapes (some CFs have lost up to 70% of their forest area to M. micrantha) (Clark Reference Clark2020). M. micrantha, or ‘mile-a-minute weed’, was discovered in the Chitwan region in the early 1990s and is thought to have spread during a large flooding event (Paudyal Reference Paudyal2007). M. micrantha is fire-adapted (Swamy & Ramakrishnan Reference Swamy and Ramakrishnan1988), and it is thought to proliferate in high-nutrient, high-light and high-moisture conditions (Swamy & Ramakrishnan Reference Swamy and Ramakrishnan1988, Zhang et al. Reference Zhang, Ye, Cao and Feng2004). Together, human population growth and increasing numbers of invasive species may lead to overburdened CFs and increase the appeal of prohibited collection in CNP.

Data collection

To understand prohibited collection, we collected data across multiple domains: human, ecological and institutional. In each domain, we ensured that our data were direct, spatially explicit and representative. Human subjects research was approved by the Institutional Review Board at Penn State University (Study #00005177).

Human data. Household data were collected in autumn 2014. Social surveys included questions about the household, including structure, ethnicity, number of animals, area of land farmed, whether the household buys fodder and fuelwood and experiences with invasive species. To assess the spatial aspect of collection behaviours, a knowledgeable household member used an interactive tablet app that contained touch-enabled satellite imagery of the Chitwan area. The respondent circled areas where household members had collected different types of forest resources in the past year. Using tablets to collect spatial data in this population has been shown to be accurate (Yabiku et al. Reference Yabiku, Glick, Wentz, Ghimire and Zhao2017). The tablet app did not show CF or park boundaries (which could have influenced willingness to indicate collection in the CNP), and we did not ask about a prohibited activity (e.g., ‘Does your household collect fuelwood in the park?’), which could raise social desirability bias (Krumpal Reference Krumpal2013). In addition, using the tablet app provides greater spatial accuracy than responses to survey questions (Yabiku et al. Reference Yabiku, Glick, Wentz, Ghimire and Zhao2017).

Although the household survey interviewed 1235 households, our analysis uses 1036 households that resided in areas eligible for membership in one of 21 local CFs; we refer to these areas as ‘CF catchment areas’. Of the 21 CFs, 17 are within buffer zone CFs (BZCF) under the jurisdiction of CNP (Department of National Parks and Wildlife Conservation). The remaining four CFs are under the jurisdiction of the District Forest Office (Department of Forest). Although all share governing principles, the BZCFs receive additional financial resources from CNP for forest management, compensation for damage by wildlife and community development activities. Figure 1 shows the CF areas in relation to their proximity to CNP.

Fig. 1. Western Chitwan District and Chitwan National Park. Sources: Open Street Map, Department of National Parks and Wildlife Conservation, Ministry of Forest, Nepal, and primary data collection.

Households were selected through a three-stage process: first, wards (administrative units) in catchment areas were randomly sampled proportionate to population size in the 2011 Census; second, chosen wards were divided into equal sub-wards of c. 400 households each and sub-wards were randomly sampled; third, chosen sub-wards were enumerated to create a household sampling frame and households were randomly sampled. The household response rate for the face-to-face survey interview was 98%.

Ecological data

The ecological data were collected from 2013 to 2015 using a systematic vegetation survey throughout the CFs that surround the household settlements: 11 forests were surveyed in 2013 (August–November), 8 in 2014 (September–November) and 2 in 2015 (September–October). In the CFs, parallel transects were drawn every 200 m. Along each transect, one plot (composed of two 5 m × 5 m areas) was sampled every 50 m. Variables collected included canopy cover, evidence of disturbance, presence as well as percentage cover of the invasive species M. micrantha, C. odorata and L. camara and percentage cover of the three most dominant herbs and tree species. In total, 2219 plots were surveyed.

Institutional data

We measured both self-governance and market institutions. A representative from each of the 21 CF management committees was interviewed in 2014. Surveys measured management techniques, perceived challenges and experiences with invasive species. In addition, the representatives used a tablet to indicate areas of invasive species presence in their CF. Market organizations, governmental organizations and NGOs were enumerated in terms of the presence of organizations such as shops, schools, health clinics, restaurants and factories. Project staff walked all roads in each sub-ward and used a custom-designed tablet app with integrated GPS to enumerate the locations of all organizations. In the 21 catchment areas, 2335 such organizations were identified.

Data analysis

Dependent variable

Prohibited collection in CNP over the past year was directly collected from a household respondent. If the respondent drew a collection area for fuelwood or fodder (the most commonly collected resources) that overlapped with CNP boundaries, the household was coded 1 on prohibited collection and 0 otherwise.

Covariates of prohibited collection

We expected that the likelihood that a household engaged in prohibited collection would vary according to the following factors derived from our observations and existing literature (Shova & Hubacek Reference Shova and Hubacek2011): (1) demand for resources; (2) barriers to prohibited resource collection; (3) barriers to legal resource collection; and (4) alternatives to resource collection.

The demand for forest resources (fodder and firewood) was hypothesized to be positively associated with prohibited collection. Variables representing the demand for forest resources included the number of household members, number of dairy animals and size of farmland. Barriers to prohibited resource collection were hypothesized to reduce prohibited collection. This variable was measured according to the distance from the household to the nearest CNP boundary; greater distance made collection in CNP less convenient. Barriers to legal resource collection were expected to increase prohibited collection. Barriers included minority ethnicity (who are sometimes discouraged or excluded from CF membership), being in a catchment area with a small CF and being a member of a CF with a higher presence of the invasive species M. micrantha. Reflecting the multiple dimensions through which invasive species may increase prohibited collection, we measured M. micrantha in three ways: household perceptions of M. micrantha spread; the percentage of the CF covered by M. micrantha as indicated in the management committee survey; and the percentage of a CF’s plots with any M. micrantha (i.e., of all surveyed plots in a CF, we calculated the percentage of surveyed plots in each CF with any M. micrantha presence). Alternatives to resource collection were hypothesized to lower the likelihood of prohibited collection. Households with more alternatives were those with greater incomes, that report purchases (in addition to or instead of collection) of fuelwood and animal feed and that are in areas with a high density of market (non-governmental) organizations where alternatives to resource collection were readily available. Table 1 shows the data sources of the variables used in the analysis and provides details of how they were coded. Although we did not include the household members’ occupations, many households retained some agricultural activity even when a member had a non-farm occupation. Thus, even without an occupation variable, we believe a household’s dependency on natural resources was captured by the more proximate indicators (e.g., land area, dairy animals).

Table 1. Variables and data sources used in the analysis of prohibited resource collection in Chitwan National Park (CNP), Nepal.

CF = community forest.

Although ethnic groups are diverse, for analysis we adopted five categories from the literature (Axinn & Yabiku Reference Axinn and Yabiku2001): (1) Brahmin/Chhetri (historically the most advantaged); (2) Hill Janajati (moved to Chitwan from the hill regions; Tibeto-Burmese origin groups includes Tamang, Gurung, Magar, Rai and others, as well as Gurkhas); (3) Dalit (historically the most disadvantaged); (4) Newar (historically merchants and engaged in commerce and commercial business); and (5) Terai Janajati (indigenous to Nepal’s plain areas and include Tharu, Kumal, Bote and others).

Statistical modelling

Because our dependent variable is dichotomous, we used logistic regression to predict whether a household engaged in prohibited resource collection. To avoid skew, we applied a log transformation to several continuous predictors that had large ranges: park distance, CF size (area) and density of market (non-governmental) organizations. The sampling of households was clustered by CF catchment area. To accommodate for this, we included a random intercept for catchment area in our model. In random-effects logistic regression models, the coefficient can be interpreted as the relationship between the independent variable and the log odds of prohibited resource collection. A positive coefficient means a variable increases the log odds of prohibited collection; a negative coefficient decreases the log odds of prohibited collection. The models were estimated in R version 4.0.4 (R Core Team 2021) using the package lme4 version 1.1-26 (Bates et al. Reference Bates, Mächler, Bolker and Walker2015).


Of primary interest is the dependent variable: the percentage of households that reported prohibited resource collection in CNP, which is c. 11% (Table 2). This small percentage suggests that collection in CNP was an uncommon behaviour, in contrast to a behaviour that might be against the rules but was frequently performed by many community members.

Table 2. Descriptive statistics of households (n = 1036) in Chitwan National Park (CNP), Nepal. Percentages are given for categorical variables; means and standard deviations are given for continuous variables.

There was diversity in the factors that could increase or decrease a household’s tendency to collect resources in CNP (Table 2). Some households had substantial farm holdings (up to 88 kattha; c. 3 ha), while others had no land. On average, households were c. 2.25 km from the closest CNP boundary, although this distance varied from 50 m to nearly 10 km. Perceived M. micrantha invasion in a household’s assigned CF (based on the catchment area) varied across households. Approximately 20% of households thought M. micrantha was decreasing or not changing and c. 20% thought it was increasing slowly, but the largest group, at 50%, thought it was increasing rapidly; c. 10% answered ‘did not know’ to this question. When asked to circle the areas in which they believed M. micrantha was present, the households’ CF managers indicated areas that, on average, it encompassed 47% of their forest. As measured in our 2219 ecological plot surveys, M. micrantha was present, on average, in 14% of plots in the CF of which a household was a member. The managers and plot survey reports differed greatly.

We present our model with all predictors (Table 3), but we also tested predictors in sets so as to determine whether the results in our final model could have been affected by collinearity between predictors; the results were no different than if sets of predictors were estimated separately.

Table 3. Results of a random intercept logistic regression model (n = 1036) of variables associated with household prohibited resource collection in Chitwan National Park, Nepal.

a 95% confidence interval excludes zero.

b Reference group for dummy variables is ‘Household perceived M. micrantha is decreasing or not changing’.

c Reference group for dummy variables is ‘Caste: Brahmin’.

The amount of land farmed, number of dairy animals (cows, buffaloes, sheep and goats) owned and household size were not associated with the log odds of prohibited collection (Table 3). The logged distance from the respondent household to CNP was significantly associated with prohibited resource collection in the hypothesized direction. In other words, household distance from CNP represented a barrier to prohibited resource collection and so the log odds of prohibited resource collection were lower.

With respect to barriers to legal collection in the CFs, household perception of M. micrantha invasion in a CF was significantly associated with the log odds of prohibited collection. Compared to households that said M. micrantha was decreasing or staying the same (the reference group), households that thought M. micrantha was rapidly increasing were significantly more likely to engage in prohibited resource collection. The frequency of M. micrantha in plots in the household’s assigned CF, as measured in our ecological surveys or in the CF management committee tablet surveys, was not significantly associated with prohibited collection (Table 3). The logged area of the household’s CF was not associated with prohibited resource collection. Regarding alternatives to forest resources, households that bought more fodder and firewood were significantly less likely to engage in prohibited collection, but the density of market organizations and NGOs was not related to prohibited collection. The caste of the household, which was included as a control variable, was significantly associated with collection in CNP. Compared to Brahmin (the reference group), Hill Janajati, Dalit and Terai Janajati demonstrated significantly greater log odds of prohibited collection.


A systems approach to studying prohibited natural resource collection

We detailed a social-ecological systems approach, analysing prohibited natural resource collection in CNP by integrating social, institutional and ecological data. This approach gave us a holistic understanding of prohibited natural resource collection, and we believe incorporating social, institutional and ecological data provides unique insights into the drivers of prohibited collection behaviours. An understanding of the system, including its social and ecological characteristics, may be necessary to inform successful management policies for PAs (Carter et al. Reference Carter, López-Bao, Bruskotter, Gore, Chapron and Johnson2017).

Drivers of prohibited natural resource collection in Chitwan National Park

None of the measures that we expected to increase the demand for forest resources were associated with prohibited collection, but measures of other factors were predictive in expected ways. Households that perceived M. micrantha as rapidly increasing and those that belonged to traditionally disadvantaged and marginalized ethnic groups were more likely to participate in prohibited resource collection. In addition, households that did not buy alternatives to resource collection and those closer to CNP also were more likely to take part in prohibited collection.

There are several potential explanations for the finding that a household’s perception of M. micrantha invasion was a predictor of prohibited collection, but M. micrantha invasion measured from other perspectives (our representative ecological surveys and the community forest committee tablet assessment) did not predict prohibited collection. Human behaviour is often motivated by perceptions and attitudes (Schlüter et al. Reference Schlüter, Baeza, Dressler, Frank, Groeneveld and Jager2017), and it is possible that perceptions of M. micrantha invasion that are misaligned with ecological measurements of M. micrantha invasion are influential in prohibited resource collection. In addition, people’s perceptions may be incomplete. Distrust between CF members, CF officials and CNP officials has been a barrier to gaining information regarding M. micrantha (Sullivan et al. Reference Sullivan, York, White, Hall and Yabiku2017). Furthermore, our ecological surveys measured entire CFs using systematic sampling. The areas of the CFs that households most frequently visit are often only a fraction of the entire CF. Similarly, CF officials’ perceptions of M. micrantha probably also were influenced by selective observation.

Our finding that households from traditionally marginalized ethnic groups are more likely to participate in prohibited collection in CNP is consistent with historical context, as Nepal’s government has a history of formally and informally marginalizing non-Brahmin ethnicities (Nightingale Reference Nightingale2011). Much of this exclusion results from injustices against Indigenous peoples and their dispossession from their native lands (Maharjan Reference Maharjan2017). PAs have been established in Nepal without the consent of the Indigenous peoples who have historically lived there (Stevens Reference Stevens2013). In some cases, Indigenous groups have continued to reside in PAs in protest and have established their own Indigenous Peoples and Community Conserved Territories and Areas (e.g., within Nepal’s four high Himalayan national parks) (Stevens Reference Stevens2013). In CNP and other Terai and western national parks, Indigenous peoples including the Tharu people were displaced as PAs were established or expanded (Basnet Reference Basnet2017). The consequences of this injustice persist. Households belonging to native ethnic groups often have higher poverty, fewer material resources and fewer formal social support structures, such as equal access to the legal collection of CF resources (Sullivan et al. Reference Sullivan, York, Yabiku, Sabetti and Castiglione2016).

Crafting successful policies to deter resource collection in protected areas

Policies need to carefully define and understand what poverty and marginalization mean in a given situation and link these to drivers of prohibited hunting, poaching or resource collection (Duffy et al. Reference Duffy, St John, Büscher and Brockington2016). Many conservation studies rely on economic definitions that emphasize income measures and ignore structural context, including cultural and ethnic relationships. Prohibited wildlife hunting and natural resource collection may be tied to prestige, customs and local cultural beliefs, going beyond reasons such as a lack of access (Montgomery Reference Montgomery2020). Beyond providing paid labour and other market-based remedies, the results of this study point to the need to measure multidimensional aspects of poverty and marginalization. Ultimately, understanding the local context (e.g., histories of discrimination) in relation to engaging in prohibited natural resource collection is an important policy foundation.

In Chitwan, some qualitative work and our quantitative models indicate that collecting prohibited resources from CNP may be related to a lack of adequate natural resources being obtainable via legal means (e.g., within the CF or through purchase). Conservation policies that address unequal legal access to natural resources may yield beneficial outcomes. Differences between specific policies may be less important to the success of the policy than having policies that are linked to local intrinsic motivations and social justice issues (Cetas & Yasué Reference Cetas and Yasué2017). A conservation policy in Chitwan that addresses local motivations for resource collection requires an understanding that access to CF resources can be unequal across multiple dimensions.

Conservation policies often displace social-environmental pressures to surrounding lands, which may not be legally protected by the formal policies of adjacent PAs (Dou et al. Reference Dou, da Silva, Yang and Liu2018). These spill-over effects complicate efforts to create conservation policies that address structural inequalities in natural resource access and livelihoods (Ma et al. Reference Ma, Zhang, Huang and Wen2020). In Chitwan, the formal rules of the community forestry programme have helped to conserve the forest land governed by those rules (Thing & Poudel Reference Thing and Poudel2017), but it is possible that those same rules have shifted resource collection to adjacent forests that are not governed by community forestry policy. To avoid the unintended consequence of degrading the land surrounding PAs, conservation policies must be holistic (Dou et al. Reference Dou, da Silva, Yang and Liu2018). Potential spill-over effects in Chitwan are connected to underlying issues of unequal access to natural resources. Policies that can address these effects will need to confront this inequality in order to be successful.

One possibility for addressing inequality in access to forest resources in Chitwan is creating a programme to build local capacity – in terms of both logistical resources and shifting social norms – so as to redistribute resources between neighbouring CFs to provide for low-income households, for recently established households and for those facing other hardships. Our social-ecological systems approach also revealed that it is critical for managers and stakeholders to consider both social and ecological data in order to fully understand people’s resource collection decisions. This is important because household perceptions of M. micrantha as rapidly increasing did not always align with the ecological data on M. micrantha. Local officials with broader knowledge of M. micrantha may be able to guide households in collecting forest resources in areas with less M. micrantha, but only if households have established trust with these officials (Shrestha et al. Reference Shrestha, Shrestha, Sharma, Thapa-Parajuli, Devkota and Siwakoti2019). Thus, strengthening relationships between local conservation officials (e.g., CNP and conservation NGO officials) and local households should be included as part of any policy to reduce prohibited resource collection in CNP. Policies that address the drivers of prohibited natural resource collection, such as supporting traditionally marginalized households and addressing perceptions of M. micrantha, are critical to reducing prohibited natural resource collection in PAs (Roe Reference Roe2015). Community members should be involved in the development of any programme designed to reduce prohibited resource collection (Nyaupane et al. Reference Nyaupane, Poudel and York2020) in order to increase the likelihood that social, political and other structural factors will be addressed.

Individual conservation programmes are not capable of overturning or reversing structural issues such as multidimensional poverty and historical mistrust between officials in power that often drive prohibited natural resource collection (Duffy et al. Reference Duffy, St John, Büscher and Brockington2016). Any policy to reduce prohibited natural resource collection in CNP, and indeed any PA, must attempt to address these underlying issues as opposed to solely targeting individual behaviours. This is why, based on our results, we propose establishing policies that build and strengthen the community-level capacity to elevate and support households that lack access to necessary natural resources. In our case, the underlying assumption is that households are engaging in prohibited resource collection in CNP because they need access to unequally distributed resources or resources that they perceive as having been invaded by M. micrantha. Thus, supporting relationship-strengthening among households and conservation officials and capacity-building for equitable resource access should reduce prohibited collection.



Financial support

This research was supported by National Science Foundation under Grant No. 1211498, Dynamics of Coupled Natural and Human Systems Program. Support was provided by the Population Research Institute at Pennsylvania State University, which is supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (P2CHD041025). Dr Qunshan Zhao has received UK ESRC’s ongoing support for the Urban Big Data Centre (UBDC) (ES/L011921/1 and ES/S007105/1).

Competing interests


Ethical standards

The authors assert that all procedures contributing to this work comply with applicable ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.


Axinn, WG, Yabiku, ST (2001) Social change, the social organization or families, and fertility limitation. American Journal of Sociology 106: 12191261.CrossRefGoogle Scholar
Baral, HS, Adhikari, B (2017) Mikania micrantha: its status and impact on people and wildlife in Nepal. In: Ellison, CA, Sankaran, KV, Murphy, ST (eds.), Invasive Alien Plants: Impacts on Development and Options for Management (pp. 5972). Wallingford, UK: CABI International.CrossRefGoogle Scholar
Basnet, BB (2017) Contested indigeneity in Africa and Nepal: challenges for access and rights to land and natural resources [www document]. URL CrossRefGoogle Scholar
Bates, D, Mächler, M, Bolker, B, Walker, S (2015) Fitting linear mixed-effects models using lme4 . Journal of Statistical Software 67: 148.CrossRefGoogle Scholar
Bergseth, BJ, Williamson, DH, Russ, GR, Sutton, SG, Cinner, JE (2017) A social-ecological approach to assessing and managing poaching by recreational fishers. Frontiers in Ecology and the Environment 15: 6773.CrossRefGoogle Scholar
Carter, NH, López-Bao, JV, Bruskotter, JT, Gore, M, Chapron, G, Johnson, A et al. (2017) A conceptual framework for understanding illegal killing of large carnivores. Ambio 46: 251264.CrossRefGoogle ScholarPubMed
Central Bureau of Statistics – Nepal (2016) Population distribution and composition [www document]. URL Google Scholar
Cetas, ER, Yasué, M (2017) A systematic review of motivational values and conservation success in and around protected areas. Conservation Biology 31: 203212.CrossRefGoogle ScholarPubMed
Chuenpagdee, R, Pascual-Fernández, JJ, Szeliánszky, E, Alegret, JL, Fraga, J, Jentoft, S (2013) Marine protected areas: re-thinking their inception. Marine Policy 39: 234240.CrossRefGoogle Scholar
Clark, MD (2020) Evaluating the social and ecological drivers of invasive plant species abundance in sub-tropical community forests of Nepal [www document]. URL Google Scholar
Critchlow, R, Plumptre, AJ, Driciru, M, Rwetsiba, A, Stokes, EJ, Tumwesigye, C et al. (2015) Spatiotemporal trends of illegal activities from ranger-collected data in a Ugandan national park. Conservation Biology 29: 14581470.CrossRefGoogle Scholar
Cumming, GS, Allen, CR (2017) Protected areas as social-ecological systems: perspectives from resilience and social-ecological systems theory. Ecological Applications 27: 17091717.CrossRefGoogle ScholarPubMed
Dietz, T, Ostrom, E, Stern, PC (2003) The struggle to govern the commons. Science 302: 19071912.CrossRefGoogle ScholarPubMed
Dou, Y, da Silva, RFB, Yang, H, Liu, J (2018) Spillover effect offsets the conservation effort in the Amazon. Journal of Geographical Sciences 28: 17151732.CrossRefGoogle Scholar
Duffy, R, St John, FAV, Büscher, B, Brockington, D (2016) Toward a new understanding of the links between poverty and illegal wildlife hunting. Conservation Biology 30: 1422.CrossRefGoogle Scholar
Edgar, GJ, Stuart-Smith, RD, Willis, TJ, Kininmonth, S, Baker, SC, Banks, S, Barrett, NS (2014) Global conservation outcomes depend on marine protected areas with five key features. Nature 506: 216220.CrossRefGoogle ScholarPubMed
Faulkner, SC, Stevens, MC, Romañach, SS, Lindsey, PA, Le Comber, SC (2018) A spatial approach to combatting wildlife crime. Conservation Biology 32: 685693.CrossRefGoogle ScholarPubMed
Geldmann, J, Manica, A, Burgess, ND, Coad, L, Balmford, A (2019) A global-level assessment of the effectiveness of protected areas at resisting anthropogenic pressures. Proceedings of the National Academy of Sciences of the United States of America 116: 2320923215.CrossRefGoogle ScholarPubMed
Gutiérrez, NL, Hilborn, R, Defeo, O (2011) Leadership, social capital and incentives promote successful fisheries. Nature 470: 386389.CrossRefGoogle ScholarPubMed
Hofer, H, Campbell, KL, East, ML, Huish, SA (2000) Modeling the spatial distribution of the economic costs and benefits of illegal game meat hunting in the Serengeti. Natural Resource Modeling 13: 151177.CrossRefGoogle Scholar
ISSG (2000) 100 of the World’s Worst Invasive Alien Species: A Selection from the Global Invasive Species Database. Auckland, New Zealand: Invasive Species Specialist Group.Google Scholar
Jones, S (2007) Tigers, trees, and Tharu: an analysis of community forestry in the buffer zone of the Royal Chitwan National Park, Nepal. Geoforum 38: 558575.CrossRefGoogle Scholar
Kahler, JS, Roloff, GJ, Gore, ML (2013) Poaching risks in community-based natural resource management. Conservation Biology 27: 177186.CrossRefGoogle ScholarPubMed
Knapp, EJ, Peace, N, Bechtel, L (2017) Poachers and poverty: assessing objective and subjective measures of poverty among illegal hunters outside Ruaha National Park, Tanzania. Conservation and Society 15: 2432.CrossRefGoogle Scholar
Krumpal, I (2013) Determinants of social desirability bias in sensitive surveys: a literature review. Quality & Quantity 47: 20252047.CrossRefGoogle Scholar
Lehmkuhl, JF (1994) A classification of subtropical riverine grassland and forest in Chitwan National Park, Nepal. Plant Ecology 111: 2943.CrossRefGoogle Scholar
Liu, J, Dietz, T, Carpenter, SR, Alberti, M, Folke, C, Moran, E et al. (2007) Complexity of coupled human and natural systems. Science 317: 15131516.CrossRefGoogle ScholarPubMed
Ma, B, Zhang, Y, Huang, Y, Wen, Y (2020) Socioeconomic and ecological direct and spillover effects of China’s giant panda nature reserves. Forest Policy and Economics 121: 102313.CrossRefGoogle Scholar
Maharjan, K (2017) Local communities’ responses towards land dispossession in Nepal: contradictions, perception and resistance. Journal of Land and Rural Studies 5: 3149.CrossRefGoogle Scholar
Mann, BQ (1995) Quantification of illicit fish harvesting in the Lake St Lucia Game Reserve, South Africa. Biological Conservation 74: 107113.CrossRefGoogle Scholar
Montgomery, RA (2020) Poaching is not one big thing. Trends in Ecology & Evolution 35: 472475.CrossRefGoogle Scholar
Nagendra, H (2002) Tenure and forest conditions: community forestry in the Nepal Terai. Environmental Conservation 29: 539539.CrossRefGoogle Scholar
Nightingale, AJ (2011) Bounding difference: Intersectionality and the material production of gender, caste, class and environment in Nepal. Geoforum 42: 153162.CrossRefGoogle Scholar
Nuno, ANA, Bunnefeld, N, Naiman, LC, Milner-Gulland, EJ (2013) A novel approach to assessing the prevalence and drivers of illegal bushmeat hunting in the Serengeti. Conservation Biology 27: 13551365.CrossRefGoogle ScholarPubMed
Nyaupane, GP, Poudel, S, York, A (2020) Governance of protected areas: an institutional analysis of conservation, community livelihood, and tourism outcomes. Journal of Sustainable Tourism 30: 120.CrossRefGoogle Scholar
Paudyal, A (2007) Buffer zone resources and community conservation: a case study of Piple buffer zone Village Development Committee, Chitwan National Park [www document]. URL Google Scholar
Picone, F, Buonocore, E, Claudet, J, Chemello, R, Russo, GF, Franzese, PP (2020) Marine protected areas overall success evaluation (MOSE): a novel integrated framework for assessing management performance and social-ecological benefits of MPAs. Ocean & Coastal Management 198: 105370.CrossRefGoogle Scholar
R Core Team (2021) R: a language and environment for statistical computing [www document]. URL Google Scholar
Roe, D (2015) Conservation, crime and communities: case studies of efforts to engage local communities in tackling illegal wildlife trade [www document]. URL Google Scholar
Schlüter, M, Baeza, A, Dressler, G, Frank, K, Groeneveld, J, Jager, W et al. (2017) A framework for mapping and comparing behavioural theories in models of social-ecological systems. Ecological Economics 131: 2135.CrossRefGoogle Scholar
Shova, T, Hubacek, K (2011) Drivers of illegal resource extraction: an analysis of Bardia National Park, Nepal. Journal of Environmental Management 92: 156164.CrossRefGoogle ScholarPubMed
Shrestha, BB, Shrestha, UB, Sharma, KP, Thapa-Parajuli, RB, Devkota, A, Siwakoti, M (2019) Community perception and prioritization of invasive alien plants in Chitwan–Annapurna Landscape, Nepal. Journal of Environmental Management 229: 3847.CrossRefGoogle ScholarPubMed
Spiteri, A, Nepal, SK (2008) Distributing conservation incentives in the buffer zone of Chitwan National Park, Nepal. Environmental Conservation 35: 7686.CrossRefGoogle Scholar
Stevens, S (2013) National parks and ICCAs in the High Himalayan region of Nepal: challenges and opportunities. Conservation and Society 11: 2945.CrossRefGoogle Scholar
Sullivan, A, York, A, White, D, Hall, S, Yabiku, S (2017) De jure versus de facto institutions: trust, information, and collective efforts to manage the invasive mile-a-minute weed (Mikania micrantha). International Journal of the Commons 11: 171199.CrossRefGoogle Scholar
Sullivan, A, York, A, Yabiku, S (2016) Moving toward sustainability: integrating generational equity into institutional analysis. In: Sabetti, F, Castiglione, D (eds.), Institutional Diversity in Self-Governing Societies (pp. 229259). Lanham, MD, USA: Lexington Books.Google Scholar
Swamy, PS, Ramakrishnan, PS (1988) Effect of fire on growth and allocation strategies of Mikania micrantha under early successional environments. Journal of Applied Ecology 25: 653658.CrossRefGoogle Scholar
Thing, SJ, Poudel, BS (2017) Buffer zone community forestry in Nepal: examining tenure and management outcomes. Journal of Forest and Livelihood 15: 5770.CrossRefGoogle Scholar
Turner, BL, Matson, PA, McCarthy, JJ, Corell, RW, Christensen, L, Eckley, N et al. (2003) Illustrating the coupled human–environment system for vulnerability analysis: three case studies. Proceedings of the National Academy of Sciences of the United States of America 100: 80808085.CrossRefGoogle ScholarPubMed
UNEP-WCMC, IUCN (2021) Protected Planet Report 2020 [www document]. URL Google Scholar
Ward, C, Stringer, LC, Holmes, G (2018) Protected area co-management and perceived livelihood impacts. Journal of Environmental Management 228: 112.CrossRefGoogle ScholarPubMed
Watson, JE, Darling, ES, Venter, O, Maron, M, Walston, J, Possingham, HP et al. (2016) Bolder science needed now for protected areas. Conservation Biology 30: 243248.CrossRefGoogle ScholarPubMed
Watson, JE, Dudley, N, Segan, DB, Hockings, M (2014) The performance and potential of protected areas. Nature 515: 6773.CrossRefGoogle ScholarPubMed
Wilkie, DS, Carpenter, JF, Zhang, Q (2001) The under-financing of protected areas in the Congo Basin: so many parks and so little willingness-to-pay. Biodiversity & Conservation 10: 691709.CrossRefGoogle Scholar
Yabiku, S, Glick, J, Wentz, E, Ghimire, D, Zhao, Q (2017) Comparing paper and tablet modes of retrospective activity space data collection. Survey Research Methods 11: 329344.Google ScholarPubMed
Yonariza, Webb EL (2007) Rural household participation in illegal timber felling in a protected area of West Sumatra, Indonesia. Environmental Conservation 34: 7382.CrossRefGoogle Scholar
Zhang, LY, Ye, WH, Cao, HL, Feng, HL (2004) Mikania micrantha H. B. K. in China – an overview. Weed Research 44: 4249.CrossRefGoogle Scholar
Figure 0

Fig. 1. Western Chitwan District and Chitwan National Park. Sources: Open Street Map, Department of National Parks and Wildlife Conservation, Ministry of Forest, Nepal, and primary data collection.

Figure 1

Table 1. Variables and data sources used in the analysis of prohibited resource collection in Chitwan National Park (CNP), Nepal.

Figure 2

Table 2. Descriptive statistics of households (n = 1036) in Chitwan National Park (CNP), Nepal. Percentages are given for categorical variables; means and standard deviations are given for continuous variables.

Figure 3

Table 3. Results of a random intercept logistic regression model (n = 1036) of variables associated with household prohibited resource collection in Chitwan National Park, Nepal.