Stakeholder preferences for pangolin conservation interventions in south- east Nigeria

1. The overexploitation of biological resources severely threatens many species, requiring urgent and effective conservation interventions.


| INTRODUC TI ON
The overexploitation of species drives biodiversity loss (IPBES Secretariat, 2019;Maxwell et al., 2016). Hunting for subsistence and commercial purposes, for example, is responsible for declines in tropical bird and mammal populations (Benítez-López et al., 2017), negatively impacting the diet, customs and livelihoods of millions of people (Brashares et al., 2014). Addressing unsustainable species harvesting is challenging , often requiring interdisciplinary approaches to understand the social, political, ecological and economic factors (e.g. site-level governance) that shape conservation outcomes (Brashares et al., 2014;Reed et al., 2016).
Outcomes such as improved governance depend on an in-depth understanding of stakeholders' perspectives (Bennett, 2016), which can be used to formulate, implement and evaluate conservation policies and interventions (Young et al., 2013). Considered a 'fundamental prerequisite for the achievement of sustainable development' (UNCED, 1992; paragraph 23.2), stakeholder involvement in decision-making promotes legitimacy and trust in decision-making processes (Renn, 2006). Stakeholder involvement also increases the knowledge pool from which decisions are made (Svarstad et al., 2006), possibly informing more inclusive interventions. For example, awareness-raising, law enforcement, social sanctions, payments for ecosystem services and the application of new technologies are a suite of interventions that can help achieve desired conservation outcomes (Finer et al., 2018;Moore et al., 2018;Sachedina & Nelson, 2012;Thomas-Walters et al., 2022). However, these interventions must be applied appropriately to ensure public and stakeholder support and desirability (Reed, 2008). An essential component of intervention design is, thus, determining entry points (consensus) and areas of contention on the viability and adoption of different approaches (Carmenta et al., 2017). Fostering stakeholder participation is especially crucial when addressing complex, multilayered conservation challenges such as illegal wildlife trade (or wildlife trafficking) involving various actors, many of whom operate discretely.
Pangolins (Pholidota: Manidae) are scaly small-to medium-sized African and Asian mammals that are negatively impacted by wildlife trafficking (IUCN, 2021). All pangolins are listed as either Vulnerable, Endangered or Critically Endangered on the Red List of Threatened Species (IUCN, 2021), with African pangolins increasingly trafficked to Asia to supply demand for their scales used in traditional medicine (Emogor et al., 2021;Wang et al., 2020). The white-bellied pangolin  (Bassey et al., 2010;Fa et al., 2006;Soewu et al., 2020). Nigeria also plays a pivotal hub role in this Africa-Asia pangolin trafficking, with scales from possibly millions of pangolins trafficked each year from or through the country (Emogor et al., 2021).
Given the multi-layered nature of pangolin exploitation, effective and sustainable interventions require an in-depth understanding of the perceptions of diverse actors associated with pangolin exploitation and protection (Bennett, 2016). Previous studies have investigated perceptions towards pangolins in south-west Nigeria (Abdulazeez et al., 2020;Soewu & Adekanola, 2011). However, this region is culturally dissimilar to the south-eastern part of the country (Anedo, 2019). Additionally, stakeholders' preferences for interventions to curb pangolin decline have not been previously researched across pangolin range states. To address these gaps, we asked two research questions: (a) how do the stakeholders (hereafter stakeholder groups) that are involved in pangolin exploitation and protection perceive pangolins? and (b) how does support for interventions to reduce pangolin decline differ across stakeholder groups?
To investigate these questions, we used Q-methodology (hereafter Q; Brown, 1980Brown, , 1993, which is useful for understanding varying viewpoints among stakeholders (Zabala et al., 2018). Q facilitates the elicitation of sensitive information, as participants are not required to explicitly state their actions, perceptions or perspectives (McKeown & Thomas, 2013). Thus, the method is ideal for investigating sensitive topics in conservation (Carmenta et al., 2017). We present the first application of Q to an illegal wildlife trade scenario, perspectives underpins the need for a combination of targeted interventions at the site level to engage different stakeholders while highlighting the potential challenges to collaborative decision-making for species threatened by illegal wildlife trade.

Policy implications.
Our results stress the importance of targeted and contextspecific conservation interventions.

K E Y W O R D S
behaviour change interventions, community-based conservation, conservation governance, illegal wildlife trade, Pholidota, poaching providing crucial insights for informing future site-level conservation interventions and shaping governance mechanisms to better protect pangolins. Given Nigeria's primacy in pangolin trafficking (Emogor et al., 2021;Omifolaji et al., 2020), such interventions can considerably curb poaching and global trade.

| Study location
The primary study location was Cross River National Park in southeast Nigeria (5.5805°N, 8.7481°E; CRNP; Figure 1), where approximately 90% of participants lived or worked. We also recruited participants (border control officers) in the Abuja and Lagos metropolises. Cross River National Park is a biodiversity hotspot (Myers et al., 2000) of predominantly lowland and submontane forests.
It is divided into the southern Oban Division (~3000 km 2 ) and the northern Okwangwo Division (~920 km 2 ). Two of the three pangolin species in Nigeria occur in the park (P. tricuspis and P. tetradactyla).

| Q method
Q is a semi-quantitative method where participants from diverse stakeholder groups independently rank the same set of items (written statements or pictures) on a continuum (Zabala et al., 2018). The final set of items, also called the Q-set, represents subjective opinions about a topic, which then becomes the 'Q-sort' when ranked by participants. The Q-set is subjective because its composition depends on the researcher, who is expected to present all possible and relevant opinions, perspectives and perceptions on the topic of interest (the concourse). Common approaches to developing a concourse include a literature review and expert consultation. Q can also incorporate post-sort interviews to support Q-sort interpretation. The method facilitates in-depth evaluation of the highly subjective perception-assessment domain, helping to reduce bias and ensuring systematic analysis of viewpoints that form broader opinions (Watts & Stenner, 2012). Given Q typically uses purposive sampling, findings from the method cannot be generalised to the wider population-the method mainly seeks to elicit the variety of viewpoints held within the population, irrespective of their dominance.

| Stakeholder selection
We defined stakeholders as individuals, groups of people or formal organisations who influence and are influenced by actions or policies related to pangolin conservation and use, consequently qualifying as relevant contributors in decision-making processes (Freeman, 1984;Sterling et al., 2017). We targeted respondents based on a combination of stratified random, purposive and snowball sampling (Biernacki & Waldorf, 1981). We recruited a multi-layered stakeholder composition, representing those who know about pangolins (i.e. can describe a pangolin), including their primary threat of overexploitation, and are knowledgeable about their conservation.
We assessed respondents' knowledge about pangolins by first asking them to describe a pangolin verbally and sharing their thoughts about the conservation status of pangolins. All respondents we approached provided satisfactory morphological descriptions and linked hunting or poaching to pangolin survival.
We interviewed a total of 49 respondents (female = 9; male = 40) across seven stakeholder groups (seven per group), namely (i) hunter; (ii) small-scale farmer; (iii) civil servant; (iv) conservation organisation employee (management personnel of CRNP and Wildlife Conservation Society; WCS); (v) wild meat trader; (vi) CRNP ranger and (vii) wildlife unit of Nigeria Customs Service (NCS; see Table 1 for a description of stakeholder groups, location of participants from each group and rationale for group selection). We selected the stakeholder groups a priori to ensure contrasting representation of actors, that is, actors directly involved in pangolin conservation (ranger, WCS and NCS) and those involved in their exploitation (hunters and wild meat traders). We purposefully selected Lagos, Abuja and Calabar based on NCS and WCS staff presence. We then selected communities using a stratified random sampling technique, with each CRNP division split into four geographic quadrants. To account for the disproportionate sizes of the divisions, we randomly selected four communities in Okwangwo and nine in Oban. It was not always possible to cover all stakeholder groups in each community, but we attempted to ensure a comparable number across quadrants for the following groups: civil servants, farmers, hunters and wild meat traders. We received ethics approval from the Psychology Research Ethics Committee of the University of Cambridge (application number: PRE.2020.001), which mandated participant anonymity and free, prior and informed consent (we received written consent).

| Q statement selection
We used written statements for our study, separated into two the- (c) contemporary opinions about pangolins, such as their rumoured involvement in the origin and spread of COVID-19; (d) conservation efforts to curb direct and known threats to wild animals, which could be adopted for pangolins in CRNP and across Nigeria and (e) undocumented but probable drivers of threats to pangolins, such as their use as pets.
To select statements for each Q-set, we compiled two concourses using four primary sources: (a) scientific and grey literature;  dence.com/); (c) knowledge of ongoing management strategies in CRNP and (d) preliminary findings from a scoping study we conducted with 10 participants from three stakeholder groups in May 2020. We then selected 37 and 25 short and simplified statements for Belief and Intervention, respectively, from the concourse (concourse curation protocol and selected statements are presented in Appendix S1). Following Watts and Stenner (2012), we prioritised statements directly relevant to the study objectives and location F I G U R E 1 The three study locations, namely, Lagos, Abuja and Cross River, are highlighted in the insert (top left). Study sites in Cross River (yellow circles; n = 14) comprised the capital city, Calabar, and 13 communities (nine in Oban and four in Okwangwo) surrounding the Cross River National Park. Respondents in Abuja and Lagos were Nigeria Customs Service staff members.
when choosing the Q-set. For example, we excluded 'Plant trees to improve pangolin habitats' as habitat loss is not the primary threat to pangolins-at least in CRNP. We also avoided double-loading of statements (i.e. ensuring statements addressed distinct issues).

| Q statement administration
We conducted in-person Q-sorts following the multiple-participant design (see Watts & Stenner, 2012) from July to September 2020 using cards bearing written statements and a steep nine-point distribution Q response grid (+4 to −4) for both Q-sets (Appendix S2).
We chose a steep response grid to reduce the complexity during sorting, as not all our participants were pangolin experts or conservationists (Watts & Stenner, 2012). With zero as the reference, +4 represented 'most unimportant or most inappropriate', while −4 represented 'most unimportant or most inappropriate' (for Belief). For Intervention, +4 denoted 'most preferred', with −4 signifying 'most unpreferred'. We chose a steep response grid ( Figure 2) to reduce the complexity in sorting as not all our participants were pangolin experts or conservationists (Brown, 1980;Watts & Stenner, 2012).
During sorting, participants assigned a score to each statement by placing cards containing the statements within a cell. We arranged the cells below the +4 to −4 continuum, and because the grid followed an inverted normal distribution, there were limited spaces around the edges, which represented the highest consensus or disagreement ( Figure 2). Belief statements were sorted based on whether participants thought the statements were important or true (scale of importance or truth), while Intervention statements were sorted based on preference, with a 10-20-min break between both sessions. Participants sorted Intervention statements based on preference and not importance or effectiveness to account for the knowledge bias across stakeholder groups about interventions that could be effective for pangolin conservation. Before sorting, we TA B L E 1 Description of selected stakeholder groups, location of participants for each group and rationale for group selection. Note that the locations we recruited park rangers represent their places of duty at the time of this study, not their residence.

Location (number of participants) Rationale for inclusion
Civil servant State and local government staff, including schoolteachers Oban (1); Akor (1); Etara (1); Ekukunela (1); Alesi (1); Bamba (1); Anape (1) Civil servants represent people with a stable income source and are integral to the social structure where pangolin poaching and trade occur. Their inclusion is important for a neutral perspective, given they are not directly involved in pangolin conservation or exploitation

Conservation organisation
The management staff of CRNP and Wildlife Conservation Society, both actively involved in conservation interventions in CRNP Calabar (4); Butatong (2); Akamkpa (1) These organisations are currently working to safeguard pangolin populations and will be actively involved in implementing recommended pangolin conservation interventions. Some participants in this group are also trained scientists (ecologists)

Nigeria Customs Service
A Nigerian parastatal under the Ministry of Finance that is responsible for regulating illegal importation and exportation of goods Abuja (4); Lagos (3) NCS is critical to detecting and intercepting illegal pangolin shipments entering or leaving Nigeria Ranger CRNP staff whose duties include conducting periodic anti-poaching patrols in the park Butatong (2); Aking (3); Okwangwo (2) Rangers are critical to current efforts to reduce pangolin decline. They apprehend poachers found in the park, possibly deterring illegal hunting

Subsistence hunter
People who hunt wild animals for local consumption and income when sold to consumers or vendors in local wild meat markets. They mostly use unsophisticated firearms and the income generated is usually for household maintenance Oban (1); Akor (1); Ntebachot (1); Etara (1); Bamba (2); Anape (1) Poaching is the main threat to pangolins. Thus, hunters are notable contributors to pangolin exploitation

Small-scale farmer
People who engage in agriculture mainly for subsistence Oban (2); Akor (1); Etara (1); Alesi (1); Bamba (1); Anape (1) Pangolin hunting does not require special skill or equipment and farmers kill pangolins they encounter in modified landscapes such as plantations Wild meat trader People who buy wild meat from hunters living in or around CRNP and sell them in rural or urban communities surrounding the park Akor (1); Ntebachot (1); Ekong (1); Etara (1); Alesi (1); Bamba (1); Anape (1) Wild meat traders are notable players in the illegal pangolin trade (a primary driver of pangolin decline) asked participants to group the statements into 'agree', 'disagree' and 'neutral' to enhance their familiarity with the statements. After sorting, we asked follow-up questions on their rationale for their rankings, focusing on the statements on the extremes of the grid.
We read the statements to participants that could not read them.

| Data analysis
We analysed both sets of Q-sorts separately in two broad steps, following Brown (1980) and Zabala (2014). The first involved reducing the Q-sorts to factors (i.e. main viewpoints or perspectives). We used principal component analysis and varimax rotation (Brown, 1993), which involves identifying commonality among Qsorts and grouping highly correlated Q-sorts across participants as distinct factors (see Brown, 1980;Watts & Stenner, 2012). The second step comprised three parts. The first part centred on flagging the Q-sorts that significantly loaded each factor. The Q-sorts that significantly load a factor are highly correlated with the factor, meaning that the statements that make up those Q-sorts are important contributors to defining and understanding the factor. We automatically flagged the Q-sorts that significantly loaded the factors and then manually inspected them for confounds (i.e. where a Q-sort has a similar correlation with more than one factor). In the second part, we estimated the scores of statements for each factor (i.e. z-scores and factor scores; metrices used to assess levels of participants' agreement with the statements). The z-scores are weighted averages of the scores assigned to a particular statement by the flagged Q-sorts. Their comparison highlights the similarity of the factors towards a given statement. On the other hand, the factor scores are the z-scores transformed to the scale of the sorting grid. Lastly, we compared factors to identify distinguishing and consensus statements. A statement was considered as distinguishing if the absolute difference between the z-scores of the statement for each pair of factors was statistically significant (at 0.05 p-level).
Conversely, a consensus statement showed no significant differences between any pair of factors. We conducted our analysis using qmethod package in R-version 4.1.0 (R Core Team, 2022; Zabala, 2014). Our Methods are summarised in Figure 3, with analytical details presented in Appendix S3.

| Perspectives about benefits, burdens and beliefs associated with pangolins
At least six statements shaped each of the five perspectives about the Belief associated with pangolins. Only statement 14 (S14; Figure 4) was highlighted as a consensus statement (see Appendix S4 for results of pairwise factor comparisons). However, S24, S25 and S11 showed marginal consensus across the perspectives-none of them distinguished any perspective, with their z-scores clustered ( Figure 4).
3.1.1 | Belief1: Interests in pangolin conservation Statements 1,4,6,9,10,15,18,19,26,31,32 and 37 shaped this perspective ( Figure 4). Q-sorts that significantly loaded the perspective were completed by the following: conservation organisation (n = 5), NCS (n = 7) and ranger (n = 1). The perspective suggests that pangolins' insectivorous feeding habits help to reduce ant and termite populations that would otherwise infest farmlands around CRNP (S1). The perspective further suggests that catching a pangolin F I G U R E 2 Example of a Q-sort on burdens, benefits and beliefs (Belief) associated with pangolins in south-east Nigeria. An employee of the Nigeria Customs Service completed this Q-sorts. The shaded top row represents the scale of the response grid. The numbers (from −4 to +4) were imprinted on the grid before sorting began. The grids below the top row represent spaces where participants were required to place statements based on the instructions associated with the Q-set (importance and truth, in this case). The numbers displayed represent Q-statements in the Belief Q-set (see Table 1 in Appendix S2).
is not a sign of good luck (S4) and that Christianity does not forbid the killing or selling of pangolins (S6), adding that pangolins are not perceived as ugly animals (S10). This perspective agrees with S9 (Most people in my community think pangolin meat is delicious) and S26 (Pangolin meat is considered a luxurious or royal delicacy in my community), also submitting that pangolin scales are used locally for medicine (S15). The perspective holds that pangolin poaching is not exclusive to local hunters but that people not living around pangolin habitats visit these places to hunt them (S18), including paying hunters in these areas to do so (S19): 'People come [into CRNP] from Cameroon to hunt pangolin' (Interview; conservation organisation).
The remaining statements that defined this perspective centred on trends in pangolin populations and participants' willingness to support pangolin conservation. The perspective suggests that pangolin populations in CRNP have reduced compared with populations 5 years ago (S31) and believes that current hunting levels will not substantially reduce pangolin populations (S32). The perspective also shows people's willingness to contribute towards the conservation of pangolins (S37).
3.1.2 | Belief2: Unplanned, opportunistic poaching Statements 2, 3, 4, 6, 10, 28 and 29 shaped this perspective. The Q-sorts that significantly loaded the perspective were completed by the following: civil servant (n = 1), farmer (n = 1), hunter (n = 2), ranger (n = 1) and wild meat trader (n = 2). This perspective believes that pangolins are zoonoses carriers (S2). The perspective does not perceive pangolins as a bad omen (S3) but rather believes that catching a pangolin brings good luck (S4). During the post-sort interviews, participants residing in communities adjacent to pangolin habitats (hereafter local communities) stressed that seeing a pangolin during a hunting trip is considered a good sign because of their rarity. They added that catching the mammal is considered lucky because, in most cases, the action does not require a cartridge as the animal is commonly captured by hand or killed with a machete: 'Whenever someone goes to the forest or farm, especially those without hunting weapon(s), their prayer is to see and pick pangolin on the road; it shows that the trip will be favourable' (Interview; farmer). 'Pangolins are very difficult to catch, it's likely to hunt the whole of the rainy season without catching one. It is not easy, except you hunt in the night and luckily you catch one' (Interview; hunter).
Like Belief1, this perspective holds that Christianity does not forbid the killing of pangolins (S6) and that pangolins are not ugly animals (S10). The perspective suggests that pangolins are held locally as pets (S28) and that their hunting is opportunistic (i.e. hunting trips do not target pangolins but attempt to capture them when encountered; S29).

F I G U R E 3
Summary of the Methods we used to (a) assess people's perceptions towards pangolins in south-east Nigeria (Belief) and (b) understand stakeholder groups' preferences for conservation interventions to reduce pangolin decline (Intervention). Depictions are from freeicons.io, flati con.com and Ellingsen et al. (2014).
TA B L E 2 Summary of the themes defining each perspective related to (a) benefits, burdens and beliefs associated with pangolins in south-east Nigeria (Belief) and (b) stakeholders' preferences for management interventions to curb pangolin decline in Nigeria (Intervention).

Contribution across stakeholder groups (number of respondents)
Gender composition

Belief1: Interests in pangolin conservation
• There is external influence in pangolin poaching Conservation organisations (5); NCS (7); ranger (1)  (2); farmer (1); hunter (1); wild meat trader (2) 2 4 • Mobilise local people in community-science initiatives • Support communities to make by-laws against poaching pangolins but not their use in traditional medicine • Do not develop action plans for pangolins TA B L E 2 (Continued) F I G U R E 4 Q-set representing benefits, burdens and beliefs (Beliefs) associated with pangolins in south-east Nigeria. The symbols and colours correspond to the five extracted perspectives (bottom left). The z-score on the x-axis shows the weighted average by each perspective to a given statement. The z-score differences between factors determine the degree of agreement between statements, with an increase in the levels of agreement towards the statements by the factors from the top to the bottom of the plot. Filled symbols indicate the distinctiveness of the statement to the corresponding factor (perspective). The statements are sorted in descending order using the standard deviation of their z-scores for all factors. Q-sorts that significantly loaded this perspective were completed by the following: civil servant (n = 2), hunter (n = 2), ranger (n = 1) and wild meat trader (n = 1). This perspective holds that it is taboo to kill (S7) and sell (S8) pangolins. It further suggests that people not part of local communities do not come to adjacent forests to hunt pangolins (S18) and that the current population of pangolins in the CRNP and surrounding forests far exceeds their population 5 years ago (S31). The perspective disagrees that people who catch pangolins are solely entitled to their parts (S22), adding that pangolin scales are not used in making jewellery (S27). The perspective also suggests displeasure towards community members who are found to have killed a pangolin (S23). The taboo associated with killing and selling pangolins and associated communal displeasure towards these actions were commonly expressed in Alesi during post-sort interviews. Members of this community believe that pangolins are sacred animals because during a communal clash several hundred years ago, pangolins helped their ancestors escape annihilation by stretching to form a bridge over a river so they could cross to safety: 'We believe that anyone who kills a pangolin will be cursed; scales like those on pangolins will grow over the person's skin' (Interview; wild meat trader).

| Belief5: Pangolins for food and a delicacy
Statements 3, 5, 12, 26, 29 and 33. The Q-sorts that significantly loaded this perspective were completed by the following: conservation organisation (n = 1), civil servant (n = 2), ranger (n = 1) and wild meat trader (n = 2). The perspective holds that seeing a pangolin does not cause bad luck (S3) but it suggests that killing a pangolin can cancel out bad luck caused by seeing the animal (S5). This perspective also suggests that pangolins are hunted for their meat, which is consumed for food (S12), adding that pangolin meat constitutes a luxurious or royal delicacy (S26). In addition to suggesting that pangolin poaching is targeted (S29), the perspective holds that, on average, pangolins reproduce more frequently than African brush-tailed porcupines Atherurus africanus (S33): 'People hunt pangolin for food, and that is the major reason we hunt' (Interview; civil servant). 'Pangolins can reproduce four to five [young] whereas porcupines reproduce two to three' (Interview; farmer).

| Perspectives on pangolin conservation interventions
At least nine statements shaped the four perspectives on stakeholders' preferences for pangolin conservation interventions. None of the statements was labelled as consensus, although S11 showed marginal agreement across the perspectives, suggesting that the stakeholder groups agreed on their view about the usefulness of research for pangolin conservation (S11; Figure 5).
3.2.1 | Intervention1: Awareness and law enforcement over economic incentives Statements 2,3,4,5,6,9,10,12,14,15,16,18,19 and 23 defined this perspective. The Q-sorts that significantly loaded this perspective were completed by the following: civil servant (n = 1), conservation organisation (n = 5), farmer (n = 2), hunter (n = 1), NCS (n = 3) and ranger (n = 5). The perspective shows strong support for awarenessraising campaigns (S14) but not for interventions that require local communities to make by-laws to prohibit pangolin hunting (S3) and the use of their parts in traditional medicines (S2). It also strongly opposes interventions whose sole emphasis is to pay hunters to stop hunting pangolins (S04): 'Giving poachers money to stop killing pangolins will create more poachers because they will all need money and the numbers will increase…when the money stops coming, they will go back to poaching' (Interview; conservation organisation).
The perspective similarly opposes interventions that centre on recruiting people living around pangolin habitat to become whistleblowers, that is, informing law enforcement about violations (

| Intervention2
: Promote ecotourism and support local advocates Statements 2,3,4,8,9,10,14,15,16 and 23 shaped this perspective. The Q-sorts that significantly loaded this perspective were completed by the following: civil servant (n = 1), farmer (n = 2), hunter (n = 1), NCS (n = 2) and wild meat trader (n = 1). Like Intervention1, this perspective does not support community-centred laws prohibiting pangolin poaching (S3) and also does not support the use of pangolin parts in traditional medicines (S2). Furthermore, it does not support interventions centred on financially compensating hunters not to kill pangolins (S4). The perspective does not prefer the establishment of a harvesting quota for pangolins (S9), the involvement of wildlife experts in border inspection duties (S10), awareness-raising campaigns on the status of pangolins (S14) and increased enforcement on local food vendors (S16). This perspective strongly prefers communitybased interventions where people living in local communities are recruited as stipendiary advocates for pangolin protection (S8). The perspective also supports pangolin-related tourism (S23) and establishing rehabilitation centres for pangolins rescued from the illicit pangolin trade (S15).

| Intervention3
: Support communities to make anti-poaching by-laws Statements 2,3,9,10,13,14,19,21,23 and 25. The Q-sorts that significantly loaded this perspective were completed by the following: hunter (n = 2), ranger (n = 1) and wild meat trader (n = 2). This 'Government should focus on the source; if we are comfortable and stop killing pangolins, there will be no need to buy scanners for borders' (Interview; hunter).
3.2.4 | Intervention4: Wildlife officials needed at borders Statements 1,2,3,5,9,10,12,13,17,23 and 24 shaped this perspective. The Q-sorts that significantly loaded this perspective were completed by the following: civil servant (n = 2), farmer (n = 1), hunter (n = 1) and wild meat trader (n = 2). The perspective supports the involvement of wildlife experts in border inspections (S10) and the establishment of community laws to prohibit the use of pangolin parts in traditional medicine (S2) but not the creation of laws by local communities to prohibit pangolin hunting (S3). This perspective prefers the establishment of community science programs, where people in local communities are involved in monitoring pangolin populations (S24). The perspective also supports awareness-raising campaigns (S14), the establishment of a harvesting quota (S9), the provision of communal infrastructure to communities as a reward for upholding poaching regulations (S5) and pangolin tourism (S23). The perspective does not support creating an action plan for pangolins in Nigeria (S17 and S1) and training border enforcement personnel on techniques to detect pangolin trafficking (S13). It also does not support strengthening the penalties associated with pangolin poaching (S12).

| DISCUSS ION
We used Q to elicit perspectives towards pangolins and stakeholders' preferences for management interventions for pangolins in south-east Nigeria. We extracted five perspectives associated with Belief and four with Intervention, identifying consensus statements in both Q-sets. Our findings accounted for 58% and 44% of the variance related to Belief and Intervention, respectively. These results surpass the 35%-40% range recommended by Watts and Stenner (2012) and are similar to the results of other Q studies (Carmenta et al., 2017;Gannon et al., 2020). Despite attempts to recruit a comparable number of males and females, we struggled to recruit more females partly because of the snowball and purposive sampling methods we used.

| Leveraging consensus across stakeholders
Only one statement in each of the Q-sets stood out as consensus statement, meaning that all the stakeholder groups shared a common view of the statements (two other statements showed marginal consensus). The low consensus (four of 62 statements) possibly arose from the inherent dichotomy among stakeholders, that is, contributors to pangolin conservation and those who directly or indirectly threaten them, except for civil servants, who, in theory, represented a neutral group. For Belief, stakeholders reported that they were happy about the presence of pangolins in CRNP (S14), expressing their awareness of Nigerian laws prohibiting pangolin hunting (S24).
Stakeholders also agreed that pangolins do not pose a threat to crops

| The diversity and complexity of perceptions towards pangolins
We identified five dominant perspectives on Belief associated with pangolins. These perspectives differ substantially across topics, including the nature and primary driver of pangolin exploitation and the role of disease transmission in eating wild animals-a debate boosted by COVID-19. As captured distinctively in Belief1, the interest in contributing to pangolin conservation is associated with the perceived decline in wild pangolin populations ( Table 2), supporting the notion that knowledge or perceptions about species decline strongly affect policy responses (see Rodrigues et al., 2006).
However, Belief1 harbours a representation bias. It is formed only by the viewpoints of stakeholders directly or indirectly involved in pangolin conservation or protection (Table 2). Thus, despite the interest from those in the conservation sector, stakeholders who directly threaten pangolins did not show interest in protecting the species. Given the consensus on awareness of the laws prohibiting pangolin poaching, the apparent lack of interest might reflect the socioeconomic situation of members of the stakeholder groups living around pangolin habitat (e.g. wild meat traders and hunters), who may prioritise pangolin exploitation over the conservation of the animal, perhaps because of the opportunity cost of giving up killing and trading in pangolin.
Another dominant viewpoint of Belief1 is the likelihood of external influences in pangolin poaching. This viewpoint was also captured by Belief3 formed by respondents living or working in CRNP, strongly suggesting that people visit communities adjacent to CRNP to poach pangolins, including financially incentivising hunters to poach them.
Such external influences on pangolin poaching have been observed to varying degrees in other places, including China (Nash et al., 2016) and the Democratic Republic of the Congo (Malimbo et al., 2020).
We did not assess the extent of any external influence in our study site. However, Belief5, formed largely by viewpoints from stakeholder groups living and working directly around CRNP (  (Friant et al., 2015). Although there is no evidence linking pangolins to the coronavirus pandemic (Xiao et al., 2021), the impact of zoonotic diseases underpins the need for effective awareness-raising campaigns regarding the apparent dangers of consuming wild meat (Hilderink & de Winter, 2021). Furthermore, awareness-raising interventions can yield a greater impact when combined with other interventions (Smith et al., 2020).
An effective awareness-raising strategy could be to target children living in the landscape, which could facilitate biocentric reasoning towards pangolins from a young age (i.e. promote the ideology of the intrinsic value of pangolins that is separate from their worth as a commodity; Špur et al., 2020;Kahn & Kellert, 2002). This approach could also foster protective attitudes towards pangolins among adults (Damerell et al., 2013).  (Williams et al., 1999).

| Law enforcement
These interventions may involve financially compensating communities, so they must be properly designed to ensure public compliance with the by-laws. Attempts should be made to assess the overall effectiveness of such interventions to guide future actions .

| Ecotourism
Pangolin-related ecotourism was preferred by all stakeholder groups except conservation organisations and rangers. Ecotourism can benefit conservation, including promoting local support for conservation (Diedrich, 2007;Lindsey et al., 2005;Stronza et al., 2019). Although the contributions of ecotourism to the survival of threatened species have been shown to outweigh associated costs in most cases (Buckley et al., 2016), such programs must be carefully developed to minimise adverse effects on biodiversity: reduction in species richness through increased trail use (Lei et al., 2022), degradation of habitat through infrastructure development and zoonotic disease transfer (Shannon et al., 2017). Possible pangolin-centred tourism must also strive to incorporate local communities in the design and implementation phases (Stronza, 2008).

| CON CLUS IONS
This study presents a novel application of Q in assessing people's perceptions of illegal wildlife trade issues. We have shown the diverse viewpoints that the different stakeholder groups hold towards pangolins and their conservation, providing insights into stakeholder support and the desirability of viable interventions to curb pangolin decline. Our application of the method to investigate the sensitive topic of pangolin poaching and trade facilitated the recruitment of certain respondents (hunters and wild meat traders), as they were not required to state their actions. Most participants from the groups above declined to partake in post-sort interviews, suggesting a possible challenge in data collection if we used conventional qualitative interviews. However, Q only seeks to highlight the prominence and diversity of opinions within groups, and the findings from the method cannot be generalised across the broader populations because of the non-random sampling.
Our study indicates that local demand for meat drives pangolin exploitation-a proposition requiring further empirical scrutiny. This new evidence is contrary to popular opinion that hunters specifically set out to kill pangolins, suggesting that pangolin scales from southeast Nigeria that may end up in international shipments are remnants of subsistence hunting for their meat. Our results underscore the need for targeted interventions compared with places where international demand may drive pangolin poaching. Implementing behaviour change interventions to prevent hunters from killing pangolins may, thus, be more cost-effective because the opportunity cost of not killing a pangolin is relatively lower compared with regions where poaching is driven by international demand for scales.
Our results on preferences for pangolin interventions suggest a spatial divergence: people living around CRNP largely supported community-centred initiatives (such as anti-poaching by-laws, community-science programs and payment for ecosystem services), while those living away from the park markedly favoured awareness and law enforcement interventions. This underpins the need to consider all relevant stakeholders' perceptions, values and perspectives when making conservation policies (Cook et al., 2013). Our results showcase the challenge of collaborative decision-making and highlight the importance of implementing complementary interventions to engage different stakeholders. We recommend further research on the perspectives of international actors in wildlife trafficking and the effectiveness of the above-preferred interventions.

ACK N O WLE D G E M ENTS
We thank all our study participants for their time and contributions.
We are grateful to the following organisations for contributing to this

CO N FLI C T O F I NTE R E S T S TATE M E NT
Charles A. Emogor is a non-stipendary fellow of the Wildlife Conservation Society that financially contributed to this work. Staff members of WCS were interviewed for this study.

DATA AVA I L A B I L I T Y S TAT E M E N T
We have uploaded our study data to the Mendeley data repository.