Segmenting Wildlife Value Orientations to Mitigate Human–Wildlife Conflict for Ecotourism Development in Protected Areas

ABSTRACT

Ecotourism is increasingly being adopted to improve the livelihoods of local communities and fulfill conservation goals. Many critical actors, including but not limited to governments and conservation organizations, believe that for ecotourism to be successful, local communities must appreciate and support wildlife conservation activities. Without evidence that communities’ value systems are oriented in a manner that aligns with the authorities’ sustainable development and wildlife conservation ideologies, strategies for how to achieve the most benefits for the most people will be based on anecdotes and assumptions. This study investigated residents’ value orientation towards wildlife in the context of Bardia National Park (BNP), Nepal. We surveyed residents from eight buffer zone communities surrounding BNP with a human–wildlife conflict (n = 871). Our findings revealed four distinct clusters of resident WVOs, creating challenges for BNP ecotourism development. Socio-demographic and resource management factors also play critical roles in shaping the sustainable development of BNP ecotourism.

KEYWORDS:

• Ecotourism
• wildlife value orientation
• protected area
• human–wildlife conflict
• conservation
• Nepal

Introduction

Ecotourism is a popular and universally adopted segment of the tourism industry (Fennell, 2020). The term “ecotourism” is often used alongside terms such as “protected area” (PA) tourism or “nature-based tourism” and touted to be the fastest-growing segment of the tourism industry (Das & Chatterjee, 2015). Ecotourism’s impacts on conservation, economic development, and societal wellbeing vary in degrees of success and failure (Cobbinah et al., 2017; Coria & Calfucura, 2012; Wondirad, 2019). For better or worse, ecotourism’s contributions toward achieving the sustainable development goals have embedded this segment of tourism into economies and the human experience (Honey, 2008; Nahuelhual et al., 2013).

The popularity of ecotourism as a local economic development and wildlife conservation strategy (Bell, 1987) has brought an increasing number of humans into proximity with wildlife. With increased wildlife conservation efforts, communities engaged in PA ecotourism may be subject to elevated, novel, and intense interactions with livestock predation or attacks by wildlife (Distefano, 2005; Gusset et al., 2009). Human–wildlife conflict (HWC) is of global concern and is defined as negative interactions between humans and wildlife or between humans concerning wildlife and is exacerbated around PAs (Peterson et al., 2010). For sustainable development strategies to achieve both human development and biodiversity conservation targets, the frequency of and ability to mitigate or solve HWCs resulting from ecotourism is critical.

Balancing the needs of humans and the wellbeing of wildlife can pose unique challenges to PA tourism development because these are inherently contradictory goals. For instance, previous research demonstrates that persistent HWC and the inability of authorities to address it can lead to unacceptable costs and conflict, and, ultimately, dissatisfaction or dissent among locals (Barua et al., 2013; Bello et al., 2017; Snyman, 2014; Walpole & Thouless, 2005). Moreover, negative outcomes, such as lost economic opportunities (Hackel, 1999), may influence locals’ interactions with tourists and the ecotourism development process. Researchers also suggest that top-down approaches, characteristic of PA governance and ecotourism development, can lead to disharmony within and dissent among communities, as authorities have a history of prioritizing wildlife conservation over local wellbeing (e.g. Mir et al., 2015; Morais et al., 2018). Structural prejudices aside, bankrolling actors hold fast to the idea that they can balance human and wildlife needs by ensuring a more equitable distribution of ecotourism benefits and enacting deliberative approaches to integrate local values, beliefs, and attitudes into PA decision-making and solving HWCs (Morais et al., 2015; Snyman & Bricker, 2019; Spiteri & Nepal, 2008).

Researchers have demonstrated the difficulty of balancing rural livelihoods with PA conservation in Nepal. Sharma (1990) outlined the tenets of a long-term strategy to balance subsistence livelihoods in the Chitwan National Park region, emphasizing the need to see PAs as not merely a geographic space but also a cultural one, made alive by the customary rights and activities occurring within that space. Nepal and Weber (1995) were also early researchers to problematize park–people dynamics, expressing the need for authorities to respect existing and sustainable nature–culture relations and espouse participatory, financial, and educational strategies to ameliorate value conflicts. The Ministry of Forests and Soil Conservation (1999) later adopted the buffer zone concept to help locals to regain aspects of their livelihoods disrupted by PA creation. Researchers have declared that buffers zones and ecotourism now underpin top-down policies to link the disjointed aims of human wellbeing and wildlife conservation (Nyaupane & Poudel, 2011). Nonetheless, the paradigm is not foolproof. Spiteri and Nepal (2008) examined incentive-based programs in Chitwan National Park and concluded that the costs and benefits accrued by the residents differed, in that specific communities in proximity to touristic areas were receiving more benefits. KC et al. (2018) similarly noted the role of geospatial patterns in describing asymmetrical distribution of ecotourism’s costs, benefits, and varying levels of success. In summary, the schism between costs and benefits may be framed in terms of the value conflict introduced by ecotourism development policies (Lai & Nepal, 2006). Though wildlife is paramount to PA tourism in Nepal, researchers have yet to fully map residents’ values orientation towards wildlife, which may further elucidate another important dimension for addressing the aforementioned schism.

Wildlife value orientation (WVO) studies are a powerful theoretical approach to identify patterns in human–wildlife relationships and predict human behavior towards wildlife in certain situations (Cerri et al., 2017; King & Nair, 2013). The conceptual approach is situated within the cognitive hierarchy and formalizes the research of Steinhoff (1980) and others who have explored human value orientations toward wildlife as well as the occurrence of clustering values (e.g. Stern et al., 1995). Bright et al. (2000) noted that WVOs can help managers understand the ways in which humans relate to wildlife, evaluate policy, estimate demands for wildlife-related activities, and communicate with target populations about wildlife policy and recreation opportunities. Most WVO studies have focused on understanding how residents and managers relate to wildlife in North America (Manfredo et al., 2020), but studies have been expanding into novel, cross-cultural contexts and across stakeholder groups (e.g. Gamborg et al., 2019; Zainal Abidin & Jacobs, 2016; Zinn & Shen, 2007).

Research suggests that eliciting WVOs in the context of PA ecotourism can help decision-makers to better understand how tourism communities relate and react to wildlife (e.g. Manohar et al., 2012; Needham, 2010; Serenari et al., 2015). Understanding resident orientations towards wildlife in an ecotourism context is important because if WVOs are not oriented in a direction that aligns with wildlife conservation goals, the resultant dissent can yield failure to achieve desired targets (Bright et al., 2000; Teel et al., 2010; Teel & Manfredo, 2010). Certain orientations towards wildlife among communities can create either favorable or undesirable contexts or outcomes for ecotourism development (Manohar et al., 2012). A review of the literature revealed that there is a paucity of research investigating the linkages between ecotourism, WVOs, and HWC or how they diverge and align among different groups of people (e.g. Manohar et al., 2012). Our examination of the literature revealed a research focus on two of the three domains, though the third was still relevant (e.g. Freeman et al., 2021), or focuses on using tourist WVOs to characterize interactions with wildlife (e.g. Harvey Lemelin & Smale, 2007).

We aim to address this need by conducting a case study in Nepal, where ecotourism activities are rising while HWC remains an ongoing problem (LeClerq et al., 2019). Specifically, we investigate the factors influencing the WVOs of residents living near Bardia National Park (BNP), a cornerstone of the Nepal PA network and one of the largest and most popular parks in the country, but also dealing with rising HWC. The success of ecotourism policy hinges upon the mitigation of HWC. This study aims to (i) identify distinct clusters of residents based on their value orientation towards wildlife and how their value orientation differs among the clusters, (ii) examine the differences among the clusters for value orientation towards wildlife and resource management factors, and (iii) examine the differences among clusters regarding socio-demographic variables.

Wildlife value orientations

Grounded in social psychology, WVOs capture individual and collective wildlife-related experiences (Fulton et al., 1996). WVOs have become a popular conceptual and methodological approach for understanding and managing human–wildlife interactions and bolstering the inclusion of wildlife stakeholder expectations (Clark et al., 2017). WVOs do not reflect utility theory and the material value of wildlife. Instead, because they are situated at the bottom of the cognitive hierarchy, they reflect human thought. Values drive behavior, reflect desirable ends states and worldview, and anchor lower-order cognitions such as attitudes and beliefs (Rokeach, 1973). As is common with values, they change gradually but tend to remain stable (Ford-Thompson et al., 2015) and are influenced by contextual, multi-level factors (Manfredo et al., 2020).

WVOs can be viewed as positive and negative (Cerri et al., 2017), and parsed into four distinct categories to segment and analyze the subjects’ values towards wildlife. The WVO literature suggests that mutualism and domination are dominant WVOs, but the concept also includes pluralist and distanced orientations. Recent research has further broken down WVOs into high/low categories (Gamborg & Jensen, 2016). Survey items measuring mutualism prioritize living in harmony with wildlife and elevating wildlife existence and wellbeing to a level parallel to that of humans. Meanwhile, domination-oriented subjects tend to relate to wildlife in a manner that prioritizes human needs, using wildlife for human purposes, and hands-on (e.g. lethal management) strategies to moderate human–wildlife interactions (Jacobs et al., 2014). Pluralism reflects a mix of mutualist and domination orientations, while a distanced orientation reflects limited engagement with either (Teel & Manfredo, 2010; Vaske et al., 2011).

Studies of WVOs have revealed critical insights as societies change and move away from rural livelihoods and utilitarian values and towards urbanization, deliberative institutional arrangements in wildlife conservation, and their implications for human wellbeing and biodiversity conservation (Manfredo et al., 2020; Teel & Manfredo, 2010). For instance, studies demonstrate the value of integrating various knowledge and spatial considerations into decision-making (Bright et al., 2000; Laverty et al., 2019), ways to improve communications about interactions with wildlife (Miller et al., 2018), policy preferences regarding species management (Serenari et al., 2015; Sijtsma et al., 2012), and the role of household dynamics in shaping WVOs (Clark et al., 2017; Zinn et al., 2002).

Research has revealed important predictors of reaction and behaviors within explorations of the WVO–HWC nexus, particularly in subsistence communities and working land contexts. Policy preferences are also popular variables of interest because they have a direct influence on outcomes. Researchers tend to measure support for or acceptability of management actions as a function of WVOs and how those dynamics impact the likelihood of living peaceably with wildlife (Bruskotter et al., 2017; Knackmuhs et al., 2019; Whittaker et al., 2006). Cerri et al. (2017) revealed that select species may be the target of individuals with a domination orientation, while a mutualist orientation can apply to a wide range of animals rather than just a single species. Capacity studies are also important to WVO research. These studies may examine knowledge, political or religious affiliations, and related variables (Zainal Abidin & Jacobs, 2016). Bright et al. (2000) noted that the knowledge capacity of segments of the public is critical to raise awareness about and support wildlife conservation and interactions. Demographics are also an important variable of interest in this context (Vaske et al., 2011) and a focal area of many WVO studies (e.g. Zinn & Pierce, 2002 (gender); Clark et al., 2017 (age); Manfredo et al., 2020 (place of residency, rural–urban)).

Methodology

Study context

In Nepal, the history of PA development and the conservation of wildlife and natural resources started in 1973 with the establishment of the Chitwan National Park. Since then, Nepal has increased its number of PAs and upped conservation priorities in the country, with about 25% of the surface area of the country allocated to PAs (Department of National Park and Wildlife Conservation (DNPWC), 2019). A total of 165,304 tourists visited the Nepal PAs in 2005–2006, and the number increased by 327.18% to 706,148 in 2018–2019 (DNPWC, 2017, 2019). Even though PAs were increasingly becoming a primary conservation tool for Nepal, the country only recognized the livelihood dependency on natural resources of communities surrounding PAs in the mid-1990s. Locals were regularly subject to HWC. As part of a comprehensive state strategy to mitigate park–people conflict, authorities established buffer zones to facilitate natural resource-dependent livelihoods. Gradually, ecotourism initiatives were introduced to incentivize communities living in the buffer zone and to tap into the growing ecotourism market. These ecotourism projects are promoted in PA settings to support local communities and improve their economic and social wellbeing.

Ecotourism prospects are growing faster in the BNP than the average growth rates in most PAs in Nepal. In the periods of 2005–2006 and 2018–2019, tourist visitation to the BNP increased by 1,661.69% (DNPWC, 2017, 2019). While the BNP is one of the largest PAs in the Terai region of Nepal, covering an area of 968 km² (Figure 1), over the years the BNP has undergone a series of shifts in name, size, and conservation regulations (Thapa Karki, 2013). Even though BNP was established in 1976, the buffer zone area of 507 km² was added only in 1996 to support the livelihoods of resource-dependent communities (DNPWC, 2019). Local communities rely on subsistence agriculture and forest resources such as timber, firewood, and fodder. Subsistence agriculture is only able to support 42% of households of the local communities (Bhattarai & Fischer, 2014). There are 19 buffer zone communities in the BNP (Figure 1). Thapa Karki (2013) reported that 30–50% of the revenue generated from the park is invested in the community development of the buffer zone areas. However, the distribution of revenue is often considered unfair (Allendorf et al., 2007; Thing et al., 2017). Recently, park authorities and WWF Nepal have been promoting community-based homestays to support rural livelihoods (KC, 2021).

Figure 1. Bardia National Park and buffer zone communities (source: BNP, 2020).

Data collection

The study design was approved by the University of North Texas Institutional Review Board (IRB) (Study# IRB-19-187). The research team also acquired a research permit from the DNPWC and subsequently from the BNP authority in Nepal. Based on consultations with BNP officials, who know the issues and region best, we selected eight of the 19 buffer zone communities for data collection based on the number of reported HWC cases and compensation requests submitted to park authorities (Figure 1). We used a stratified random sampling technique (Paudyal et al., 2018) to ensure the equal representation of the communities. We allocated 110 questionnaires to each community, ensuring that at least 5% of households were represented in the study (resulting in a total of 871 usable survey questionnaires): Patabhar (n = 110), Geruwa (n = 110), Bindra (n = 105), Asaregaudi (n = 107), Shivapur Ekikrit (n = 110), Shreeramnager (n = 109), Thakurbaba (n = 110), and Suryapatuwa (n = 110) (Table 1).

Table 1. Selected buffer zone user committees and their demographic information (source: DNPWC, 2016).

S.N.	User committees	HH	Population			Number of user groups
			Male	Female	Total	Female	Male	Mixed	Total
1	Sriramnagar	1740	5674	5920	11594	14	14	2	30
2	Shivapur Ekikrit	1219	4414	4582	8996	12	10	3	25
3	Thakurbaba	1338	4094	4320	8414	12	12	7	31
4	Suryapatuwa	420	1386	1424	2810	4	4	1	9
5	Aseregaudi	1067	3553	3571	7124	9	9	3	21
6	Bindra	1025	3488	3584	7072	1	1	19	21
7	Geruwa	1029	3577	3733	7310	10	10	3	23
8	Patabhar	1981	7257	7160	14417	2	2	18	22

We surveyed the heads of households first, then recruited another household adult who was at least 18 years of age when the head of the household was not available during the month of July in 2019. The first author, a native of Nepal, and Nepalese field assistants collected the data while the field assistants were provided training before the data collection to familiarize them with the questionnaire and its content. The research team filled out the questionnaire on behalf of the participants. The importance of the study and potential practical implications resulting from the study were briefly explained prior to administering the survey, and the respondents were not compensated for their participation. We maintained flexible hours (early morning to evening) to increase our chances of meeting participants during what was the monsoon season as well as a period of high agricultural activity.

Measurement items and data analysis

We adapted the survey instrument items from Cerri et al. (2017) to measure the WVOs. We contextualized and situated wildlife resource management statements within the primary domains of interest in the PA literature. The items were not mutually exclusive and some overlap between domains was expected. We situated the survey items under the areas of knowledge, tourism benefits, conservation motivation, enforcement policies, and governance. We assessed legal awareness (Macura et al., 2011; Studsrød & Wegge, 1995; Thing et al., 2017) and potential for under-reporting (Gillingham & Lee, 2003; Pant et al., 2016). PA tourism is imbued with potential costs and benefits to local livelihoods. We explored these outcomes for communities, families, and individuals (Allendorf et al., 2007; LeClerq et al., 2019; Thapa Karki, 2013) and their preference for strategies to overcome existing HWCs—specifically, compensation (Bhatta et al., 2010; Karanth & Nepal, 2012), incentives (Shova & Hubacek, 2011), and deliberative arrangements (Thing et al., 2017). We also explored internal motivators known to influence views and behavior—specifically, identity as a resource steward (Bennett et al., 2012) and moral orientations regarding the degree to which locals agree they should shoulder the burden of conservation (Seeland, 2000), overall support for wildlife conservation (Allendorf, 2007), and implementing a penal code (Jana, 2007; Kahler & Gore, 2012). We also explored the role communities should have in decision-making. We used a five-point Likert scale ranging from (1) strongly disagree to (5) strongly agree for all the WVO measurement items.

We performed a two-step cluster analysis to group the residents with similarities (Pesonen & Tuohino, 2017). Cluster analysis measures distances and homogeneity among groups and suggests the ideal number of clusters (Kaufman & Rousseeuw, 2005). The mutualist group holds a high priority towards social affiliation and caring beliefs compared to appropriate use and hunting beliefs (Teel & Manfredo, 2010). Due to their high inclination towards mutualism, they are likely to support ecotourism development because it is consistent with their beliefs. A disinterested cluster does not have a penchant for either mutualistic or domination WVOs. As they do not hold either a mutualism or a domination orientation, this indicates a low level of interest in wildlife and wildlife issues (Teel & Manfredo, 2010). Usually, a cluster with a high score on every WVO is defined as “pluralist”, as they echo both domination and mutualism aspects (Teel & Manfredo, 2010). In our study, this group represents a balance, except that their hunting belief is the lowest among all the clusters. Their hunting belief is consistent with the current conservation policies (i.e. strict regulation without any privilege for wildlife consumption), but they also support the appropriate use belief. Therefore, instead of pluralist, they are referred to as conscious. The antagonistic group prioritizes human needs (i.e. the use of wildlife). Usually, a cluster with a high domination score and a low mutualism score is referred to as “traditionalist”. This group may lack affection for wildlife (Jacobs et al., 2014). Thus, this cluster is referred to as antagonistic as opposed to traditionalist because they will favor hunting.

We examined the reliability and validity of the measurements through factor analysis and confirmatory factor analysis using AMOS 26.0. We retained items exceeding a Cronbach’s alpha score of at least 0.70 for each construct and the goodness-of-fit indices exceeded the recommended values (CMIN/DF < 2.00, GFI > .90, CFI > .90, and RMSEA > .50) (Arbuckle, 2019). We employed two-step cluster analysis with hierarchical and K-means cluster analyses using SPSS version 24 to identify the optimal number of clusters. We verified the suggested number of clusters from the analysis with a dendrogram structure (Kaufman & Rousseeuw, 2005). Further, we performed one-way ANOVA and chi-square analyses to examine the differences among clusters Table 2.

Table 2. Reliability test of wildlife value orientations.

Value orientation belief dimension	Factor loading	Cronbach’s alpha
Appropriate use
It is acceptable for people to kill wildlife if they think it poses a threat to their life.	0.74	0.71
It is acceptable for people to kill wildlife if they think it poses a threat to their property.	0.79
It is acceptable to use fish and wildlife in research even if it may harm or kill some animals.	0.82
Fish and wildlife are on earth primarily for people to use.	0.64
Hunting
We should strive for world where there is an abundance of fish and wildlife for hunting and fishing.	0.81	0.72
Hunting is cruel and inhumane to the animals.*	0.63
Hunting does not respect the lives of animals.*	0.72
People who want to hunt should be provided the opportunity to do so.	0.73
Social affiliation
We should strive for a world where humans and fish and wildlife can live side by side without fear.	0.71	0.72
I view all living things as part of one big family.	0.82
Animals should have rights similar to the rights of humans.	0.75
Wildlife are like my family and I want to protect them.	0.71
Caring
I care about animals as much as I do other people.	0.80	0.87
It would be more rewarding to me to help animals rather than people.	0.91
I take great comfort in the relationships I have with animals.	0.82
I feel a strong emotional bond with animals.	0.76

Note: * indicates items used for reverse coding.

Results

We collected 871 valid survey questionnaires from households within the buffer zone. Among the respondents, 54% were males and 46% were females. A majority (50.1%) of the respondents were between 30 and 50 years of age, and more than half (55.5%) of the respondents reported primary education as their highest education level (Table 3).

Table 3. Demographic information.

Variable	Frequency (n)	Percent (%)
Gender
Male	470	54.0
Female	401	46.0
Age
18–29	227	26.1
30–50	440	50.1
51–65	156	17.9
Over 65	48	5.5
Education
Primary	483	55.5
Lower Secondary	156	17.9
Secondary/SLC	147	16.9
10 + 2/Intermediate	66	7.6
Bachelor’s Degree	17	2.0
Master’s Degree or Higher	2	0.2
Landholding Size (Kattha)
No Land	10	1.1
0.1–10	463	53.2
10.1–20	247	28.4
20.1–60	140	16.1
60.1–100	6	0.7
> 100	5	0.6
Annual Income (Nepalese Rupees)
0–49,999	241	27.7
50,000–99,999	237	27.2
100,000–149,999	148	17.0
150,000–199,999	79	9.1
200,000 and above	166	19.1
Total Damage (Nepalese Rupees)
0–19,999	265	30.4
20,000–39,999	239	27.4
40,000–59,999	138	15.8
60,000–79,999	118	13.5
80,000 and above	111	12.7
Household (HH) Members
1–2	71	8.2
3–4	292	33.5
5–6	304	34.9
7–8	111	12.7
9 and above	93	10.7

The analysis revealed that four clusters were ideal for the current study. One-way ANOVA and chi-square analyses used to examine the differences among clusters were found to be statistically significant at a .01 alpha level in relation to the participants’ WVOs, resource management factors, and demographic characteristics. As a result, the data were grouped into four clusters based on the participants’ WVOs and labeled based on their characteristics: cluster 1 (n = 258), cluster 2 (n = 269), cluster 3 (n = 225), and cluster 4 (n = 119). Dolnicar et al. (2014) indicates that the threshold of a 70 sample size for each cluster is ideal to find the maximum differences among clusters and achieve valid results.

Less than one third (29.6%) of the participants were classified into cluster 1 (n = 258) (Table 4). Cluster 1 showed above-average mean scores for social affiliation and caring beliefs compared to other clusters identified in the study. The mean score for the appropriate use belief was below the mid-point (3) of the scale, and the score was the lowest compared to the other clusters. Additionally, the score for the hunting belief was below the mid-point (3) of the scale. Therefore, cluster 1 was labeled as “mutualist” after these characteristics. Cluster 2 accounted for 30.9% of the sample, representing the largest share among the clusters. The group showed below the mid-point (3) of the scale for hunting belief and was closer to the neutral rating for the other value orientation beliefs, thus the group was labeled as “disinterested”. Cluster 3 represented over one fourth (25.8%) of the sample and showed generally positive and higher than average scores in all the value orientation beliefs. However, the group had the lowest mean score for hunting belief. Therefore, this cluster was categorized as “conscious”, as their value orientation beliefs recognized the appropriate use belief along with the social affiliation and caring beliefs, but did not support the hunting belief. Cluster 4 had the smallest share of the sample (13.7%) among the clusters, but this group’s appropriate use belief was the strongest among the clusters. This group also showed the lowest social affiliation and caring beliefs among the clusters. The mean score for the hunting belief was the highest among the cluster, even though the mean score was close to neutral. Overall, residents’ strong support for the appropriate use belief and neutrality for the hunting belief while maintaining disagreement for social affiliation and caring beliefs indicate their conflicting relationships with wildlife. Thus, this cluster was characterized as “antagonistic”.

Table 4. Wildlife value orientation clustering.

Value orientations	Mutualist n = 258	Disinterested n = 269	Conscious n = 225	Antagonistic n = 119
Appropriate use	2.67	2.80	4.03	4.27
Hunting	2.35	1.72	1.45	2.88
Social affiliation	4.61	3.71	4.67	2.86
Caring	4.39	2.60	3.63	1.63

Note: Bolded score indicates that the mean for this cluster is significantly higher than that for all other clusters, while an underlined score indicates a mean that is significantly lower than that for all the other clusters. The significance for all statements is p <0.01.

To investigate the factors that characterized each cluster, their knowledge of wildlife rules and regulations, perceived benefits of tourism, conservation motivation, and opinions regarding enforcement policies and conservation governance were analyzed (Table 5). Compared to the other clusters, clusters 1 (mutualist) and 3 (conscious) appeared to show more positive ratings on all the items shown in the table. Cluster 1 (mutualist) had the highest mean score for knowledge of wildlife rules and regulations and opinions on enforcement policies, whereas cluster 3 (conscious) placed the highest rating on conservation motivation and conservation governance. Furthermore, cluster 2 (disinterested) showed relatively moderate mean scores on most of the items, but perceived enforcement policies were rated lowest compared to the other clusters. Cluster 4 (antagonistic) placed the lowest rating on tourism benefits, conservation motivation, and conservation governance among the clusters.

Table 5. Resource management factors clustering.

Statements	Mutualist n = 258	Disinterested n = 269	Conscious n = 225	Antagonistic n = 119
Knowledge	3.77	3.45	3.34	3.52
I have heard of the parks and Wildlife Conservation (NPWC) Act. I am aware of current Wildlife laws. It is common of my community to report wildlife damage to park authorities.
Tourism Benefits	2.73	2.10	2.89	1.89
Wildlife tourism has been beneficial to my community. My family has benefitted from wildlife tourism. Wildlife tourism is sufficient to compensate for any loss of our customary livelihoods.
Conservation Motivation I would support a household financial incentivization program to protect wildlife. My family considers ourselves wildlife conservationists. Wildlife conservation has been beneficial to my household.	3.77	3.04	4.23	2.38
Enforcement Policies	4.30	3.53	3.84	3.71
The park hires community members to enforce park laws. Those who kill problem wildlife should be penalized. To protect wildlife, rangers should punish suspected wildlife law breakers.
Conservation Governance	4.05	3.81	4.28	3.27
The public should be involved in wildlife decision-making Locals should bear some of the cost of wildlife conservation. My community should be involved in wildlife management.

Note: Bolded score indicates that the mean for this cluster is significantly higher than that for all other clusters, while an underlined score indicates a mean that is significantly lower than that for all the other clusters. The significance for all statements is p <0.01.

Additional demographic characteristics for the clusters are shown in Table 6. Cluster 1 (mutualist) appears to be young, male, educated, and reported the highest annual income and landholding size compared to other clusters. Cluster 2 (disinterested) can be described as older and less educated compared to the other groups. In particular, this group reported the lowest landholding size and least damage from wildlife, as well as the highest satisfaction with compensation. Cluster 3 (conscious) is characterized as being female-dominant and having the least annual income and the highest damage from wildlife compared to the other clusters. The group’s satisfaction mean score with compensation was the lowest when compared to the other groups. Cluster 4 (antagonistic) can be described as a female-dominant and less educated group compared to the other clusters. The group had a relatively low annual income and satisfaction with compensation and high damage from wildlife. The results from one-way ANOVA and chi-square analyses revealed that the differences among clusters in terms of variables shown in the tables were statistically significant at the .01 alpha level.

Table 6. Socio-demographic characteristics clustering.

Characteristics	Mutualist n = 258	Disinterested n = 269	Conscious n = 225	Antagonistic n = 119
Age (younger than 30)	37.5%	21.1%	21.3%	26.1%
Male	68%	50%	45%	45%
Education (Primary)	49.4%	62.2%	50%	60.2%
Annual Income	NRs.164,129.55	NRs.121,223.11	NRs.76,179.25	NRs.77,297.30
Landholding Size (<10.1)	40.2%	61.5%	58.7%	58.8%
Total Damage	NRs.42,992.73	NRs.36,432.22	NRs.47,104.89	NRs.46,272.27
Happiness with Compensation^a	2.74	2.92	1.56	1.97

Note: Bolded score indicates that the mean for this cluster is significantly higher than that for all other clusters, while an underlined score indicates a mean that is significantly lower than that for all the other clusters. The significance for all statements is p <0.01.

^a The item is coded on five-point scales ranging from 1 (very unhappy) to 5 (very happy).

Discussion and conclusions

Ecotourism development in the BNP will likely succeed if younger generations are supportive of the government’s aim to achieve a participatory but authoritative approach. Our results revealed that mutualists tended to be younger, male, and held more natural and financial capital. They were also well informed about wildlife laws and regulations and supportive of enforcement policies. Youths were reported to be actively involved in conservation activities conducted by the BNP officials (KC, 2021). Hence, operating under a dynamic of government-led deliberative PA ecotourism development, youths would have the opportunity to see the value of a democratic but highly regulated ecotourism milieu to minimize negative impacts (Weaver, 2000), and recognize and appreciate the tradeoffs that come with embracing a sustainable PA ecotourism program. Likewise, officials can use these results to recognize the gaps in their community engagement strategies. Large portions of our sample will have different levels of understanding, capital, engagement, motivations, and caring when it comes to BNP ecotourism.

For instance, a continued lack of support or exclusion from ecotourism or related conservation programs could further alienate those who are disinterested in or antagonistic towards either wildlife conservation or BNP tourism. This group reported less total damage, but they also held the lowest landholding size and were relatively unhappy with the compensation (unhappy or neutral), further suggesting that the existing compensation program has not been well-received (KC, 2021). Elevated damage from wildlife to communities dependent on subsistence agriculture perhaps led to this disassociation or resentment (Bhattarai & Fischer, 2014). A timely and transparent compensation program could potentially help mitigate HWC while minimizing the residents’ antagonistic and uncooperative attitudes (Bello et al., 2017). A buffer zone management plan stipulates the provision to return 30–50% of the income earned by PAs to community-development and income-generating activities for residents (Aryal et al., 2019; Ministry of Forests and Soil Conservation, 1999). A provision of benefit-sharing mechanisms allows avenues to garner support for ecotourism development (Hunt et al., 2015; Snyman & Bricker, 2019; Umar & Kapembwa, 2020). Our data suggest that critical actors such as state and park officials should consider designing and administering community engagement programs that encourage broader participation and involvement in BNP ecotourism initiatives. Our findings can inform the use of a wider lens to appeal to a range of WVOs and demographic profiles to better integrate locals who feel alienated from the larger development process (Paudyal et al., 2018). As such, these programs will need to effectively demonstrate that the costs do not outweigh the benefits to appeal to the wide range of WVOs (Hackel, 1999).

To appeal to the female-majority conscious cluster, ecotourism in the BNP region will need to provide an opportunity to earn income from the existence of wildlife while mitigating HWC or compensating this group for wildlife damage. Wildlife conservation motivation is high among this group, and they may be willing to make a livelihood adjustment to benefit wildlife conservation efforts but require support to do so. Residents within this group may be the most approachable, particularly by state officials whom they seem to respect. This group might be effective in critical roles to address HWC and ecotourism development issues because they appear to approach HWC with a level-headed approach and recognize that a balance between human and wildlife wellbeing is required for park tourism to work for everyone.

Ecotourism development and wildlife conservation will encounter resistance from antagonistic types. This group is a female-dominant group and had among the least formal education and income levels, high damage from wildlife, and low satisfaction with compensation. These residents may be motivated to compromise by dramatically reduced wildlife damage and increased material benefits attached to ecotourism development. These domination-types holding antagonistic and uncooperative attitudes (Bello et al., 2017) are likely to change their values and attitudes towards wildlife and conservation efforts with the provision of benefit-sharing mechanisms (Snyman & Bricker, 2019) and the effective implementation of community-oriented buffer management plan (Aryal et al., 2019; Thapa Karki, 2013). As changing human values is difficult (Rokeach, 1973) and they underpin support for wildlife management (St John et al., 2019) and ecotourism development (Fletcher, 2009), research should further investigate what policy arrangements are viable for as many locals as possible.

Before the introduction of the buffer zone in 1996 (Ministry of Forests and Soil Conservation, 1999), HWC and residents’ resentments of park authority could be characterized as severe. The establishment of the buffer zone and expansion of community forestry programs in surrounding buffer zone communities increased access to resources and addressed HWC to some extent (Thapa Karki, 2013). Moreover, due to the popularity of ecotourism in a few national parks and conservation areas such as Chitwan National Park, Mt. Everest National Park, and Annapurna Conservation Area, the benefits accrued from ecotourism development also helped mitigate the conflicts and support conservation goals (Nyaupane & Poudel, 2011). However, benefits have often been reaped by gateway communities and those with the most capital (KC et al., 2018; Spiteri & Nepal, 2008). Therefore, planning for ecotourism development ideally requires a holistic approach without any bias for geographical locations to achieve the equitable distribution of ecotourism benefits and mitigated HWC. Specifically, communities with high HWC incidents and wildlife damage and without the means to help themselves should be prioritized in the planning of PA ecotourism development.

HWC issues continue to dominate PAs around the world (Peterson et al., 2010). This study reiterates that WVOs studies are necessary to inform the tourism development process and, hopefully, promote more sustainable outcomes. Community development and educational programs may be made more effective based on the insight gained from empirical studies on WVOs. As ecotourism development is still intricately associated with PA conservation, understanding residents’ orientation towards wildlife, their perception of resource management aspects, and their socio-demographic factors create opportunities to implement effective strategies and ensure that the time, energy, and resources (e.g. financial, human) invested in conservation efforts produce positive outcomes. Ecotourism promises to foster sustainable destination development, which remains a questionable goal for the most part with limited success stories, while it has been argued that this goal is used as a marketing tactic (Wondirad, 2019). Every aspect that potentially compromises the success of PA ecotourism projects should be considered in the planning and development of ecotourism.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by University of North Texas–Division of Research & Innovation.