Divergent narratives: community perceptions of illegal bushmeat hunting dynamics in the Serengeti-Mara ecosystem

Introduction

Global biodiversity is declining at an unprecedented rate, with illegal hunting identified as a primary driver of wildlife population loss, species depletion, and disruption of ecological processes across protected ecosystems (IPBES et al., 2019; Ripple et al., 2016). In East African savannas, illegal bushmeat hunting continues to threaten wildlife populations despite decades of investment in protected area expansion, anti-poaching operations, and community-based conservation (CBC) interventions (Lindsey et al., 2013; Rentsch and Packer, 2015). The persistence of illegal bushmeat hunting reflects complex socioeconomic realities, including poverty, food insecurity, population growth, limited livelihood alternatives, market demand, cultural traditions, and uneven integration of local communities into conservation governance (Nuno et al., 2013; Snyder et al., 2019; Zyambo et al., 2024; Zisadza et al., 2025). Effective conservation must therefore protect wildlife while simultaneously addressing the social conditions that shape people’s willingness to cooperate with conservation institutions.

Law enforcement remains a central response to illegal bushmeat hunting. Patrols, surveillance, arrests, fines, and other sanctions can decrease illegal bushmeat hunting activity by restricting access to protected habitats and raising the perceived consequences to infractions (Arias, 2015; Kahler and Gore, 2012; Holmern et al., 2007). However, an enforcement-only approach may produce fear-based compliance without necessarily creating trust or acceptance of the underlying governance. Such approaches may be viewed as externally imposed and unfair (Dowie, 2011). Research on protected area governance, therefore, distinguishes deterrence-based compliance from legitimacy-based cooperation, in which rules are more likely to be supported when institutions are viewed as fair, trustworthy, respectful, and responsive to local priorities (St John et al., 2010; Stern, 2008).

To complement enforcement, conservation institutions have consequently adopted CBC approaches centered on conservation outreach programs (COPs) (Mbise and Sosiya, 2023; Nilsson et al., 2016). COPs may include infrastructure development, educational support, livelihood assistance, and human-wildlife conflict mitigation measures, including electric fencing, livestock enclosures, rapid response services, and consolation schemes for wildlife-related losses (Kegamba et al., 2022, Kegamba et al., 2023). The intended purpose is to reduce livelihood vulnerability, increase tolerance toward wildlife, and decrease dependence on illegal resource extraction. As such, the implicit theory of change is that if communities receive economic and social benefits from conservation, they will reduce illegal hunting pressure and voluntarily support protected areas creating win-win outcomes for both biodiversity conservation and human well-being (Kegamba et al., 2023; Koricha and Adem, 2024; Salafsky and Wollenberg, 2000).

As one of the world’s most important wildlife regions, the Serengeti-Mara Ecosystem (SME) hosts numerous conservation interventions intended to protect its globally significant biodiversity and ecological processes. The region supports exceptional biodiversity, including the world’s largest terrestrial mammal migration (Snyder et al., 2019). Beyond its ecological significance, the ecosystem contributes substantially to Tanzania’s economy through tourism. At the national level, Tanzania’s travel and tourism sector generated TZS 18.6 trillion in 2023, representing 9.5% of economic output and supporting more than 1.4 million jobs (World Travel & Tourism Council, 2024). Despite the ecological and economic importance of this landscape, illegal bushmeat hunting persists across the ecosystem (Holmern et al., 2007; Rentsch and Packer, 2015). Numerous institutions implement community-oriented conservation programs in different ways, with substantial variation in their geographic coverage, intensity, and approach. This includes protected areas and community-managed conservation areas governed by several institutions (e.g., Tanzania National Parks (TANAPA), the Grumeti Fund (GF), the Tanzania Wildlife Management Authority (TAWA), and Wildlife Management Areas committees). Beyond governance, these entities provide numerous community-facing initiatives including social-service provision, livelihood support, and employment opportunities (Kegamba et al., 2022, Kegamba et al., 2023).

Evidence regarding the effectiveness of COPs in addressing illegal bushmeat hunting remains mixed and context-dependent (Roe et al., 2015). While benefit-sharing initiatives may strengthen community acceptance of conservation and improve relationships with conservation institutions, their influence depends on the perceived accessibility, continuity, fairness, and responsiveness of the benefits provided (Kideghesho et al., 2007; Kegamba et al., 2023, Kegamba et al., 2024). Such community perceptions also differ according to proximity to protected areas, dependence on wildlife resources, exposure to enforcement, access to conservation benefits, residence history, education, cultural context, and the distribution of conservation costs and benefits (Bitanyi et al., 2012; Afriyie et al., 2021; Kegamba et al., 2024). These differences are important because conservation-related behavior is shaped not only by material incentives and formal sanctions but also by knowledge, social norms, perceived barriers, institutional trust, and the meanings people attach to wildlife and place (Clayton et al., 2013; Bosone et al., 2022; Clayton and Myers, 2015). Moreover, COPs do not operate in a vacuum. These programs are implemented alongside stressors such as fear of law enforcement, multidimensional poverty, alternative employment constraints, human-wildlife conflict, and ever-shifting social norms that shape how people think of illegal hunting (Arias, 2015; Baral and Heinen, 2020; Imbo et al., 2026; Loibooki et al., 2002; Zyambo et al., 2024). Even within the same conservation landscape, villages may receive substantially different types of benefits, each with distinct eligibility criteria, household coverage, and program duration. As such, these families may ascribe conflicting meanings to the conservation outcomes because their interpretations are filtered through differing experiences the same pool of institutions. This complicates our ability to understand the cognitive and behavioral mechanisms through which COPs truly affect illegal bushmeat hunting.

Conservation interventions are often evaluated through visible outcomes such as reduced offtake of animals or fewer documented hunting events, while the emic narratives behind those outcomes receive less attention. Yet, human-dimensions research shows that the behaviors underlying these quantitative indicators of success are shaped not only by economic benefits, but also shaped by broader psychosocial and institutional factors (Bosone et al., 2022; Bott et al., 2003; Clayton et al., 2013; Clayton and Myers, 2015; St John et al., 2010; Vining and Ebreo, 2002). The uneven exposure to COPs across the SME creates an opportunity to compare the experiences of communities who live under similar ecological conditions yet inhabit different conservation outreach contexts. The same shift in illegal bushmeat hunting activity may carry different meanings across communities depending on who receives what benefits, and where. In this context, analytical lens such as Narrative Policy Framework (NPF) provides a structured way of examining how people identify relevant actors, attribute causes and consequences, and designate appropriate responses within a given policy problem (Schlaufer et al., 2022; Shanahan et al., 2018). This approach is rarely applied in relation to illegal bushmeat hunting, despite advocacy for approaches that better balance these complex perspectives within conservation narratives (Zhou et al., 2022).

In order to assess variation in how beneficiary and non-beneficiary communities perceive and explain illegal bushmeat hunting dynamics in the SME, we used a mixed-methods approach that combined household surveys, focus group discussions (FGDs) and key informant interviews (KIIs). Specifically, we investigate: 1) how communities with different levels of exposure to COPs perceive changes in bushmeat availability across 2018-2024; 2) what factors beneficiary and non-beneficiary communities attribute to those changes; 3) whether reported species associated with illegal bushmeat hunting, species richness, and conservation-status categories differ between groups; and 4) how beneficiary and non-beneficiary communities described hunting location selection, areas avoided by illegal hunters, and reasons for avoidance. Taken together, these findings contribute to a broader understanding of the human dimensions of conservation, particularly how conservation interventions are interpreted, experienced and legitimized at the community level.

Materials and methods

Study design

We employed a mixed-methods cross-sectional design to compare how COP beneficiary and non-beneficiary communities perceived and interpreted changes in illegal bushmeat hunting, the species associated with this activity, and the locations selected or avoided by illegal hunters. Data were collected during a single fieldwork period which was started from late January to early July 2025. Respondents were asked retrospectively to describe perceived changes in illegal bushmeat hunting across 2018-2024. The study was therefore cross-sectional because participants were interviewed once, rather than being followed repeatedly throughout the seven-year reference period. The design combined household surveys administered using a semi-structured questionnaire with FGDs, and KIIs. The semi-structured questionnaire included closed-ended questions that supported statistical comparisons between beneficiary and non-beneficiary communities and open-ended questions that allowed respondents to explain their perceptions in greater detail. FGDs and KIIs were used to provide additional contextual information and to clarify patterns emerging from the household survey.

We selected 12 villages and stratified them into two groups according to their exposure to conservation outreach programs (COPs), rather than their general access to ecosystem services. Beneficiary villages (n=6) were those receiving direct and sustained COP support from conservation institutions (TANAPA, GF, TAWA, or WMA), including infrastructure (schools, health facilities, water systems), livelihood support (agriculture, livestock programs), education scholarships, and human-wildlife conflict mitigation measures. Non-beneficiary villages did not receive these specific direct COP interventions during the study period and served as comparison group. Therefore, the terms “beneficiary” and “non-beneficiary” refers only to direct COP exposure in this study and do not imply that non-beneficiary communities received no indirect ecological, social or economic benefits from the SME.

Village selection followed a multistage procedure that combined stratification, random ward selection, and purposive village selection. Internal records obtained from the community-relations departments of the relevant conservation institutions were used to identify beneficiary villages distributed across 22 wards. QGIS was used during the first selection stage because most of the available administrative-boundary shapefiles were delineated only to ward level. All eligible beneficiary wards were mapped in a single spatial layer, and six wards were randomly selected using the random selection function under the QGIS Vector Research Tools. The same procedure was applied to the eligible non-beneficiary wards. Following ward selection, records from the respective conservation institutions and ward offices were reviewed to determine the number of villages per ward and confirm their exposure status to COPs. One village was then selected from each ward, guided by its proximity to the protected-area boundary, its COP exposure status, and the need to ensure broad geographical representation across the protected-area periphery. Finally, these villages were matched on key socioeconomic characteristics to minimize confounding (Table 1).

Table 1. Demographic characteristics of participants in household surveys of beneficiary and non-beneficiary communities within the Serengeti-Mara ecosystem.

n (%); Mean (SD).

Fisher’s exact test; Welch two sample t-test.Values are presented as n (%) for categorical variables and mean (SD) for continuous variables. Bold p-values indicate statistically significant differences between beneficiary and non-beneficiary respondents at p < 0.05. Fisher's exact test was used for categorical variables, and Welch's two-sample t-test was used for continuous variables.

Study area

Our study was conducted in 12 villages across Serengeti, Bunda, and Tarime districts, adjacent to protected areas within the SME in northern Tanzania (Figure 1). The area lies between approximately 1.4° - 2.2° S and 33.8° - 34.8° E, east of Lake Victoria and northwest of the Ngorongoro highlands. The villages are adjacent to Serengeti National Park (SNP), Ikorongo-Grumeti Game Reserve (IGGR), Ikona Wildlife Management Area (IWMA), and surrounding buffer zones. Although the broader Serengeti-Mara conservation landscape covers approximately 25,000km², this study focused only on villages located on the Tanzanian side of the landscape.

The SME contains grasslands, wooded grasslands, savanna woodlands, shrub lands, riverine forests, wetlands, and riparian systems, which collectively support diverse wildlife communities and large-scale ecological processes (Kija et al., 2020; Mligo, 2015). The landscape is globally recognized for the annual migration of wildebeest (Connochaetes taurinus), accompanied by zebra (Equus quagga) and gazelles (Eudorcas thomsonii and Nanger granti) (Kegamba and Siga, 2025; Sinclair et al., 2013; Snyder et al., 2019; TAWIRI, 2023). It also supports large herbivores, including elephant (Loxodonta africana) and black rhinoceros (Diceros bicornis), and large carnivores, including lion (Panthera leo), African wild dog (Lycaon pictus), cheetah (Acinonyx jubatus), and leopard (Panthera pardus). The ecosystem has a bimodal rainfall regime, with short rains occurring from November to December and long rains from March to May (Kiwango, 2024; Ogutu et al., 2024). Rainfall variability influences wildlife distribution and seasonal movements.

Communities surrounding the protected areas are ethnically diverse and include Kuria, Ikizu, Ikoma, Jita, Natta, Sukuma, and other groups. They rely heavily on subsistence agriculture and livestock keeping, although small-scale trade and natural resources use also contribute to household income. Their proximity to protected areas exposes them to conservation benefits, access restrictions, enforcement, and human-wildlife conflicts.

The conservation landscape includes areas with different legal designations, management arrangements and permitted forms of wildlife use. TANAPA manages SNP, primarily for ecosystem protection and non-consumptive tourism and is not open to licensed tourist or resident hunting. TAWA manages IGGR, with GF providing conservation and community-program support in partnership with government and local stakeholders. TAWA identifies IGGR as hunting concessions, although photographic tourism is currently the predominant tourism activity. IWMA operates through a community-based management arrangement represented by an authorized association and participating villages. Within game reserves and Wildlife Management Areas, consumptive wildlife use may occur only when legally authorized; these areas do not permit unrestricted public hunting (United Republic of Tanzania, 2018, United Republic of Tanzania, 2015, United Republic of Tanzania, 2022). Tanzanian wildlife legislation requires a person or operator undertaking legal hunting to obtain the prescribed license or prior written authorization and to comply with approved species, quotas, locations, periods, and methods. The law does not authorize hunting in national parks in contravention of the National Parks Act and prohibits unauthorized hunting and the use of devices such as traps and snares in regulated wildlife areas (United Republic of Tanzania, 2022). In this study, illegal bushmeat hunting refers to the unauthorized capture or killing of wildlife for meat, including hunting without the required permission or license, outside an approved area, period, or quota, or through prohibited methods. This definition distinguishes the activities examined in the study from legally regulated tourist or resident hunting.

Legal wildlife use can generate revenue for conservation management and neighboring communities when institutions regulate and distribute the benefits effectively. Previous research in Tanzania has documented the use of revenue from photographic tourism, regulated hunting, concession fees, and other wildlife-related activities to support schools, health facilities, livelihood programs, local employment, WMA administration, and wildlife protection (Kegamba et al., 2022). These benefits vary among conservation areas and depend on their governance, revenue-sharing arrangements, and relationships with surrounding communities.

Funding and implementation arrangements for COPs also differed among institutions and programs rather than operating through a single common funding mechanism. TANAPA financed its Support for Community Initiated Projects (SCIPs) primarily through revenues generated by national parks, while participating communities were required to contribute 30% of the approved project budget. TAWA usually implements community projects in collaboration with private operators, NGOs, and development agencies. WMA generate revenue from photographic tourism, regulated hunting and concessions, and research fees to support community development, livelihoods, and conservation administration (Kegamba et al., 2022). In addition, the Grumeti Fund, a nonprofit organization that partners with government entities and local communities, finances its conservation and community programs primarily through donor funding and strategic alliances with institutions. This institutional variation creates differences in how communities experience conservation outreach, wildlife use restrictions, benefit sharing, and relationships with conservation authorities, providing the context for comparing beneficiary and non-beneficiary communities in this study.

Data collection

We used a mixed-methods approach comprising household survey, FGDs, and KIIs. This combination enabled triangulation of data from various sources (Bernard, 2018; Foya et al., 2023; Friant et al., 2020). We collected data in Kiswahili using Kobo Collect, a mobile data collection platform enabling real-time validation and quality control. The open-ended survey responses, FGDs, and KIIs were audio-recorded after obtaining the verbal consent from participants, transcribed verbatim, and translated into English for analysis. The household survey questionnaire, FGD, and KIIs guides, the operational definitions and coding rules for variables used in the quantitative and qualitative analyses are provided in Supplementary Tables 1, 2.

Household surveys

We conducted a household survey in 12 villages, including 6 beneficiaries and 6 non-beneficiary villages. In each village, eligible households were identified using an updated household register obtained from the respective village authority. An eligible respondent was a household head or another senior household member aged 18 years or above, lived in the village for at least 8 years and had sufficient knowledge of illegal bushmeat hunting practices, wildlife use, and conservation dynamics. Within each village 30 households were interviewed (Nyangoko et al., 2022; Loibooki et al., 2002), the interview was conducted with an eligible adult who was available, willing to participate, and able to provide informed responses to the survey questions.

When a selected household declined participation or no eligible respondent was available, another eligible household was approached as a replacement. To avoid concentrating interviews in a single part of the village, the research team approached households from different geographical areas and directions within each village. This procedure was intended to capture perspectives from across the village rather than from one neighborhood alone. The final household survey sample comprised 360 respondents, with equal representation from beneficiary and non-beneficiary villages: 180 respondents in each group. A semi-structured questionnaire was administered face-to-face to one respondent from each participating household. Household surveys are commonly used in bushmeat and conservation research to examine household characteristics, wildlife use patterns, conservation exposure, and local perceptions of resource use change (Foya et al., 2023; Friant et al., 2020).

Focus group discussions

We conducted a total of 12 FGDs, with one discussion held in each study village. Each discussion included 8 participants (Bernard, 2018; Chauka and Nyangoko, 2023), giving a total of 96 participants. We selected participants to reflect variation in gender, age, occupation, beneficiary status, and experience with local conservation and illegal bushmeat hunting dynamics.

Key informant interviews

We interviewed 24 KIIs, with two information-rich participants selected from each village. Key informants included village leaders, village game scouts, and respected elders with long-term knowledge of conservation interventions, benefit distribution, community conservation relationships, and illegal bushmeat hunting dynamics. KIIs generated specialized contextual and governance information from information-rich participants, complementing household survey and FGD data (Supplementary Table 1) (Bernard, 2018; Chauka and Nyangoko, 2023; Nyangoko et al., 2022).

Data analysis

Quantitative analysis

We analyzed quantitative data using R statistical software (version 4.4.3). Before analysis, we checked the dataset for completeness, consistency, and coding errors, then standardized variable names and response categories. We constructed derived variables to support comparison between beneficiary and non-beneficiary communities (Supplementary Table 2). Beneficiary status was coded according to village level exposure to COPs. We grouped age into three categories (<36, 36-55, >55 years), residence duration into (<10, 10-29, 30+ years), bushmeat availability categories across 2018-2024 period, and coded the direction of change as decreased, stable, increased. We calculated species richness as the number of distinct wildlife species each respondent reported as associated with illegal bushmeat hunting. We also grouped reported species into conservation status categories using the IUCN Red List of Threatened Species. Species listed as Critically Endangered, Endangered or Vulnerable were grouped as threatened/high-value species; Near Threatened species were grouped as medium conservation concern, and Least Concern species were grouped as common. We coded open-ended responses on factors attributed to changes in bushmeat availability and reasons for hunting location avoidance into analytical categories for comparison across beneficiary status. Full variable definitions, coding rules, response formats, and analytical uses are provided in Supplementary Table 2.

We generated descriptive statistics to summarize respondent characteristics, COP awareness, species associated with illegal bushmeat hunting, and locations reportedly avoided by illegal hunters. Furthermore, we used a chi-square test (χ²) for categorical variables (e.g., perceived changes in availability, factor attributions, and species targeted patterns); Fisher’s exact tests were used when expected cell counts were <5 (to avoid chi-square approximation errors); and the Welch’s two-samples t-tests were used for continuous variables (e.g., age, residence time, and species richness). Statistical significance was set at α = 0.05. When chi-square or Fisher’s tests revealed significant group differences, we reported effect sizes as percentage-point differences.

Qualitative analysis

We analyzed qualitative data from open-ended household survey responses, FGDs, and KII transcripts using a thematic analysis approach supported by QualCoder version 3.8.2. We imported qualitative data into QualCoder to organize transcripts, apply codes, write analytic memos, and retrieve coded excerpts for comparison across beneficiary and non-beneficiary communities. Codes captured specific meanings within participant accounts, such as fear of punishment, increased ranger patrols, conservation education, employment opportunities, electric fencing, lack of projects, and others, which were then systematically grouped into broader analytical categories. These categories included the perceived benefits of COPs, explanations for changes in bushmeat availability, narratives of enforcement, education, economic support, spatial adaptations in hunting practices, and recommendations to improve conservation effectiveness. Codes were iteratively reviewed and refined to ensure internal consistency within categories and clear distinctions between them. Through iterative interpretation, these categories informed broader themes indicating patterns of beliefs regarding conservation interventions, enforcement, legitimacy, and behavioral adaptation. Coding combined deductive, inductive, and selected in vivo approaches (Braun and Clarke, 2006). Deductive codes were informed by the study objectives and conservation behavior literature, particularly concepts related to conservation exposure, enforcement risk, perceived benefits, livelihood constraints, institutional trust, and conservation legitimacy (Bosone et al., 2022; Clayton et al., 2013; St John et al., 2010). Inductive codes were developed through repeated reading, using participants’ own words to capture emergent trends around fear, enforcement, conservation benefits, and absence of projects.

We used thematic analysis to identify patterned explanations of bushmeat availability, species associated with illegal bushmeat hunting, locations avoided by illegal hunters, and perceived effects of COPs. The results section presents coded patterns and interpretive themes, while the discussion explains their conceptual significance for conservation legitimacy, enforcement-based compliance, and community experiences of conservation interventions. We selected representative quotations from coded excerpts to illustrate analytical claims and demonstrate how interpretations were grounded in participant accounts. Qualitative findings formed the primary basis for analysis, while quantitative results were used to provide complementary comparative evidence and strengthen observed patterns. After the themes were developed, we drew on the Narrative Policy Framework to consider how beneficiary and non-beneficiary accounts identified relevant actors, attributed causes and consequences, and interpreted enforcement and conservation outreach. We did not code the full set of NPF elements; Supplementary Table 3 provides the final qualitative codebook, including codes, definitions, illustrative excerpts, analytical categories, and interpretive themes.

Results

Demographic characteristics

The household survey included 360 respondents equally distributed between beneficiary (N = 180) and non-beneficiary (N = 180) communities (Table 1). The sample comprised 60.00% men and 40.00% women. Nearly half of the respondents (47.22%) were aged 36-55 years, while 36.11% were >55 years. Educational attainment was generally low, with 79.17% having attained only primary education and 2.78% having attained tertiary education. Farming was the dominant livelihood (76.67%), followed by small-scale entrepreneurship and livestock keeping. Residence time was long, with 68.06% having lived in the area for >30 years, indicating substantial familiarity with local hunting dynamics over time. Beneficiaries and non-beneficiaries did not differ significantly in age, gender, education level, or occupation (p > 0.05), but beneficiaries had significantly longer residence times than non-beneficiaries (p < 0.001).

Implemented conservation outreach programs

Beneficiary respondents identified four broad categories of COPs implemented in their villages: infrastructure and social-service support (87.22%), livelihood and economic support (42.78%), conservation education and capacity building (31.11%), and human-wildlife conflict response and animal-health interventions (28.89%) (Figure 2). Infrastructure and social-service support included schools, health facilities, water systems, and other village development projects. Livelihood and economic interventions included agricultural support, livestock projects, small-business development, beekeeping, savings groups, and employment opportunities. Conservation education focused on raising awareness of wildlife protection, protected-area values, scholarships and the long-term benefits of conservation. Human-wildlife conflicts response included electric fencing, building strong livestock enclosures, village response efforts, consolation support, and animal-health support.

Household, FGD, and KII evidence converged on the same broad intervention types (Table 2). Published research from the Greater Serengeti Ecosystem similarly identifies social-service provision, livelihood support, and employment as principal benefit-sharing mechanisms. The literature comparison provides regional corroboration rather than administrative verification that every intervention was delivered in each sampled village.

Table 2. Triangulation of conservation outreach programs reported by beneficiary household respondents, focus group discussions and key informants’ interviews, compared with published evidence from the Greater Serengeti Ecosystem.

Bold values indicate the percentage of beneficiary household respondents who reported each conservation outreach program category.

Wildlife species reported as associated with illegal bushmeat hunting

Respondents identified wildlife species that they associated with illegal bushmeat hunting or the illegal bushmeat trade during the study period, with the most frequently mentioned species being abundant migratory and savanna ungulates. The five most frequently mentioned species were: wildebeest (78.3%), zebra (51.7%), impala (Aepyceros melampus; 43.3%), buffalo (Syncerus caffer; 31.9%), and topi (Damaliscus lunatus; 14.7%). Other frequently mentioned species included eland (Tragelaphus oryx), warthog (Phacochoerus africanus), giraffe (Giraffa camelopardalis), Grant’s gazelle (Nanger granti), and hippopotamus (Hippopotamus amphibius) (Table 3). These percentages represent the proportions of respondents who mentioned each species; they do not measure verified hunting incidents, wildlife offtake, or respondents’ personal involvement in illegal bushmeat hunting.

Table 3. Top 10 wildlife species most frequently reported as associated with illegal bushmeat hunting or sold as illegal bushmeat in communities surrounding the Serengeti-Mara ecosystem.

Species richness, defined as the number of distinct wildlife species reported by each respondent in relation to illegal bushmeat hunting, ranged from 0 to 20 species, with an overall (x̄ =2.7, SD = 1.9). Mean reported species richness was significantly higher among beneficiaries (x̄ = 3.03, SD = 1.85) than among non-beneficiaries (x̄ = 2.29, SD = 1.95; t = 3.73, df = 357, p = 0.001). This indicates that beneficiary respondents reported a broader range of species associated with illegal bushmeat hunting or the illegal bushmeat trade than non-beneficiary respondents (Figure 3).

To assess conservation relevance, reported species were classified according to the IUCN Red List of Threatened Species. Species listed as Critically Endangered, Endangered, or Vulnerable were grouped as threatened/high-value species (elephant, rhino, giraffe); Near Threatened species were grouped as medium conservation concern species (buffalo, hippopotamus, eland); and Least Concern species were grouped as common (wildebeest, zebra, impala, topi, warthog, hare, dik-dik, and duiker). Species-level classifications and coding decisions are provided in Supplementary Table 2. Overall, 8.90% of respondents reported at least one threatened species, 37.50% reported medium-concern species, and 90.00% reported common species in relation to illegal bushmeat hunting. Reported targeting of threatened/high conservation concern species did not differ significantly between beneficiary and non-beneficiary (χ² = 2.78, df = 1, p > 0.05). These percentages describe the conservation-status categories mentioned by respondents; they do not measure the prevalence, frequency, or intensity of illegal bushmeat hunting.

Perceived changes in bushmeat availability across 2018-2024

Across both beneficiary and non-beneficiary communities, 53.61% of the respondents described bushmeat as not available or never encountered during the study period. A non-beneficiary respondent summarized this pattern by stating, “Nowadays, bushmeat is no longer seen as before; it has become very rare”. Only a small proportion (3.89%) reported high availability, defined as daily or several times per week (Figure 4). The statistical comparison revealed no significant difference between beneficiaries and non-beneficiaries in their perceptions of change in bushmeat availability (χ² = 3.52, df = 2, p > 0.05).

Community explanations for changes in bushmeat availability

Respondents attributed the perceived changes across the study period to multiple factors. Statistical analyses revealed significant differences between groups regarding the factors attributed to these changes, particularly concerning protection and law enforcement, conservation education, and tangible benefits. Protection and law enforcement were cited by 76.1% of non-beneficiaries and 56.7% of beneficiaries (χ² = 14.39, df = 1, p <0.001). Conservation education was reported by 36.7% of beneficiaries and 7.8% of non-beneficiaries (χ² = 41.80, df = 1, p < 0.001), while tangible benefits and livelihood alternatives were cited by 27.8% and 2.8%, respectively (χ² = 41.55, df = 1, p < 0.001). Attributions regarding social interventions did not differ significantly between groups (χ² = 0.03, df = 1, p > 0.05). These findings indicate differences in respondent explanations associated with direct exposure to COPs, but they do not establish causal effects of the programs (Figure 5).

Across both beneficiary and non-beneficiary communities, respondents associated declining bushmeat availability with intensified protection of wildlife and conservation areas, including increased ranger patrols, arrests, fines, surveillance, and stricter penalties for illegal bushmeat hunting. However, the meanings attached to these changes differed between groups. Non-beneficiary communities more often explained reduced bushmeat availability through fear, punishment, and restricted access to protected habitats, whereas beneficiary communities more often linked illegal bushmeat hunting decline to conservation education, livelihood support, community benefits, employment opportunities, and changing social values toward wildlife.

Thematic analysis of open-ended household responses, FGDs, and KIIs identified four main themes: conservation understood through coercion and compliance; conservation education, benefits, and livelihood security fostering conservation legitimacy; shifting social norms and livelihood aspirations away from illegal bushmeat hunting; and enforcement driving adaptation rather than cessation of illegal bushmeat hunting. These themes explain how communities interpreted changes in bushmeat availability and illegal bushmeat hunting practices in relation to conservation interventions, wildlife protection, and habitat access.

Conservation understood through coercion and compliance

The theme of understanding conservation through coercion and compliance was dominant in both groups, but was particularly prominent among non-beneficiary participants. Respondents overwhelmingly explained declining bushmeat availability through intensified law enforcement and coercive conservation control. They attributed the decline in bushmeat availability to increased patrols, strengthened ranger presence, surveillance technologies, arrests, imprisonment, and strict legal penalties for illegal bushmeat hunting. In these accounts, reduced bushmeat availability was therefore understood as the outcome of increased risk of detection and restricted access to protected habitat.

Non-beneficiary respondents frequently described conservation as a system of surveillance and coercive regulation that discourages illegal bushmeat hunting through fear of legal consequences. Respondents repeatedly mentioned statements including: “security has improved”, “hunters fear being arrested”, and “if caught, you are imprisoned for 30 years”. These narratives appeared frequently across responses. FGD participant from a non-beneficiary village explained:

“Surveillance and law enforcement have been significantly strengthened, with many illegal hunters arrested and others required to pay substantial fines, and the government has increased its emphasis on enforcing strict laws”.

Similarly, another non-beneficiary village household participant stated:

“Since the government strengthened patrols and increased the number of rangers, bushmeat has become difficult to obtain because illegal hunters are afraid of being caught.”

Several respondents also described the expansion of surveillance systems and community informant networks as contributing to illegal bushmeat hunting decline. In some villages, respondents reported that modern technologies and collaboration between conservation authorities and local informers reduced illegal hunters’ ability to remain undetected in protected areas.

Although enforcement narratives dominated non-beneficiary responses, some respondents also connected illegal bushmeat hunting persistence to limited livelihood opportunities. One household respondent from non-beneficiary village stated:

“Illegal bushmeat hunting continues because there are no projects helping people obtain income.”

Another household respondent similarly explained:

“We do not have projects to support us, which is why some youth continue illegal bushmeat hunting.”

These responses indicated that enforcement constrained illegal bushmeat hunting, but the absence of livelihood support remained an important explanation for why illegal bushmeat hunting persisted.

Beneficiary respondents also emphasized enforcement, showing that outreach did not replace the perceived need for wildlife protection. However, they more frequently embedded enforcement within broader explanations involving education and community benefits. One household participant from a beneficiary village explained the combined role of enforcement, education, and conservation benefits:

“Protection has been strengthened, and people fear being arrested with bushmeat; people have also become educated and have seen the benefits brought by conservation projects, so they decided to stop illegal bushmeat hunting”.

Conservation education, benefits, and livelihood security foster conservation legitimacy

Beneficiary accounts linked conservation education, visible development benefits, livelihood support, employment, and human-wildlife conflict response to perceived reductions in illegal bushmeat hunting. Participants described these interventions as increasing knowledge of the value of wildlife, creating alternatives to illegal bushmeat hunting, and making conservation benefits visible at the household and community levels. Some also linked livestock and crop protection measures to greater livelihood security and tolerance of wildlife. FGD participant from a beneficiary village linked conservation education to a changed understanding of wildlife values:

“The conservation education taught us why protecting wildlife is important. We now understand that animals bring tourists, and tourism brings development to our villages.”

Also another FGD participant similarly explained,

“People received education about the importance of wildlife, and now we understand that conservation benefits the whole community, not a few individuals”.

Participants also described visible conservation benefits, including schools, health facilities, scholarships, livestock support, agricultural assistance, water infrastructure, village development projects, and revenue-related community support. These benefits were presented as evidence that wildlife protection and habitat conservation could produce tangible gains for local communities. One household respondent from a beneficiary village explained,

“Because of the projects brought to the village, people now see the value of wildlife and have reduced illegal bushmeat hunting activities.”

Another KII respondent from the beneficiary village linked livelihood support directly to reduced dependence on illegal bushmeat hunting:

“We received livestock support from the Grumeti Fund. This reduced our need to engage in illegal bushmeat hunting because we have other ways to get food and income.”.

Human-wildlife conflict response also formed part of how beneficiary communities described the value of conservation outreach. Respondents particularly mentioned electric fencing and building strong livestock enclosures as reducing crop damage, improving farming reliability, and strengthening household food security. One key informant from a beneficiary village explained,

“Since the electric fence was installed, wildlife no longer destroys our farms, so people can farm successfully and obtain crops, which have helped to reduce illegal bushmeat hunting.”

Another household respondent from beneficiary village similarly noted:

“The fence helped us protect crops and livestock, and this reduced the need to engage in illegal bushmeat hunting.”.

Employment opportunities further shaped beneficiary explanations of reduced hunting. Participants referred to youth and former hunters being employed as Village Game Scouts (VGS), security personnel, tourism workers, casual laborers, or conservation staff. One key informant from beneficiary village stated,

“Some former hunters were employed as village game scouts, while others started farming and business activities. This helped reduce illegal bushmeat hunting.”

Another household respondent explained:

“Youth now have employment opportunities and no longer depend on illegal bushmeat hunting for survival.”.

Shifting social norms and livelihood aspirations away from illegal bushmeat hunting

Participants across beneficiary and non-beneficiary communities described broader social and generational changes that reduced the social desirability of illegal bushmeat hunting. In these accounts, illegal bushmeat hunting was increasingly portrayed as risky, labor-intensive, socially undesirable, and less compatible with current livelihood aspirations. Respondents associated this shift with education, farming, small businesses, mining, motorcycle transport, religious teaching, and changing expectations among younger generations. One non-beneficiary village leader explained:

“People have become educated, and today’s youth do not like illegal bushmeat hunting jobs because they have other opportunities and education.”

Another key informant from a beneficiary village stated:

“In the past, young people only hunted, but now many attend school or engage in farming and business.”

Other respondents explained that experienced illegal hunters had aged, migrated to mining areas, or shifted into farming, motorcycle transport, or conservation-related employment. These accounts indicate that reduced bushmeat availability was not explained only through law enforcement. Participants also described a wider social transition in which illegal bushmeat hunting was becoming less desirable as a livelihood pathway and less aligned with changing community aspirations.

Enforcement drives adaptation rather than cessation of illegal bushmeat hunting

Participants described strengthened enforcement as reshaping where illegal hunters operated, rather than eliminating illegal bushmeat hunting. Across beneficiary and non-beneficiary communities, respondents reported that illegal hunters increasingly avoided areas associated with high enforcement risk, including protected areas, reserve boundaries, WMA zones, ranger patrol areas, fenced locations, and places where community reporting was perceived to be strong. Overall, 85.60% of respondents reported awareness of specific areas that illegal hunters currently avoid, and this pattern did not differ significantly between beneficiary and non-beneficiary respondents (χ² = 1.17, df = 1, p > 0.05).

Respondents linked the locations avoided by illegal hunters to increased ranger patrols, arrests, fines, imprisonment, surveillance, electric fencing, and community reporting. One household respondent from a beneficiary village explained,

“Illegal hunters now avoid protected areas because security has become stronger.”

A village leader from a non-beneficiary village linked avoidance directly to patrols and punishment:

“People fear going into protected areas because patrols have increased and the punishment is severe.”

Another respondent from the beneficiary village emphasized the role of community surveillance:

“People fear illegal bushmeat hunting activity now because even community members can report them to the authorities.”.

Discussion

Divergent narratives and conservation legitimacy

Across communities surrounding the SME, similar conservation outcomes may carry markedly different social meanings depending on communities’ exposure to and engagement with COPs. We found that beneficiary and non-beneficiary communities both reported perceived declines in bushmeat availability but explained those declines through divergent conservation narratives. Research on COPs has often focused on how interventions shape policy acceptance, including studies of conservation attitudes in western Serengeti (Kideghesho et al., 2007), the perceived acceptability of benefit-sharing mechanisms in the Greater Serengeti ecosystem (Kegamba et al., 2023), and benefit-sharing experiences across protected areas in Tanzania (Kegamba et al., 2022). However, our work emphasizes that how COPs shape communities’ emic lived experiences and conceptualizations of policy outcomes is also vital. Our results broadly support the idea that the construction of community narratives is a core intermediate step toward collective and, in this case, conservation-related action (Czarniawska, 2004).

Enforcement, compliance, and conservation legitimacy

Experiences with law enforcement remained prominent in both groups; however, non-beneficiaries’ narratives relied more heavily on enforcement pressure, fear of repercussions, surveillance, and restricted access to protected habitats to explain change. This stronger emphasis on prohibitions and punishment among non-beneficiary respondents supports the idea that conservation was frequently understood through coercion and compliance. Where negative experiences are more tangible than program benefits, people may adhere to laws without developing trust in or voluntary support for conservation institutions (St John et al., 2010). This can be counterproductive: in some contexts, increased fear of enforcement decreases willingness to disclose information on threats to wildlife (Sjöstedt et al., 2022). Enforcement-centered accounts frame compliance as an adaptive response to the perceived likelihood of detection and prosecution (Duffy, 2016; Holmern et al., 2007; Kahler and Gore, 2012; Nuno et al., 2013; Rentsch and Packer, 2015). In this way, factors often measured as drivers of conservation behavior, including knowledge of and appreciation for ecosystem services (Børresen et al., 2023), may be secondary motivators compared to negative perceptions of governance. This distinction has implications for the long-term sustainability and effectiveness of COPs because compliance is not equivalent to cooperation or commitment. Recent multi-country evidence indicates that willingness to comply with protected-area rules depends not only on the severity of sanctions, but also on whether those rules are administered transparently, consistently, and respectfully, with meaningful opportunities for communities to voice concerns (Ibbett et al., 2025). It is important to note that while beneficiary respondents also acknowledged patrols, arrests, and punishments, they more often situated these factors as one aspect of a broader conservation framework that also included education, livelihood alternatives, employment, and human-wildlife conflict mitigation as tangible community benefits. Such interventions are known to be most effective when tailored to needs that communities themselves identify as locally relevant (Kegamba et al., 2022, Kegamba et al., 2023; Kideghesho et al., 2007; Mbise and Sosiya, 2023; Roe et al., 2015; Salafsky and Wollenberg, 2000). Given this, a combined approach that entails consistently pairing enforcement with positive community interventions may produce more durable behavioral change than sanctions alone.

Prominent species in community narratives

The species most frequently associated with illegal bushmeat hunting were common and migratory ungulates, specifically wildebeest, zebra, impala, African buffalo, and topi. Previous research in the Serengeti ecosystem has likewise identified these species among those hunted, consumed, confiscated, or sold as bushmeat (Rentsch and Packer, 2015; Schilling et al., 2020; Kegamba and Siga, 2025). The emphasis on these species in narrative accounts likely reflects several factors, including these animals’ relative availability and accessibility, together with household demand for food and income, enforcement pressure, cultural preferences, dissatisfaction with conservation benefit sharing, and access to wildlife within protected areas (Bitanyi et al., 2012; Holmern et al., 2007; Loibooki et al., 2002; Nuno et al., 2013; Rentsch and Packer, 2015; Van Velden et al., 2020; Zisadza et al., 2025). These findings indicate that illegal bushmeat hunting in savanna systems is not only a threat to rare or charismatic species, but also a recurring pressure on common herbivores that support predator-prey interactions, migration dynamics, tourism value and habitat connectivity (Kegamba and Siga, 2025; Lindsey et al., 2013; Ripple et al., 2016). Sustained illegal offtake of even non-threatened wildlife can affect population dynamics, migration processes, predator-prey relationships, and ecosystem functioning (Rogan et al., 2017; Rentsch and Packer, 2015).

Beyond narratives that emphasized reliance on non-threatened species, even low-frequency reports of threatened, high-concern species remain important sources of information. Although this study relied on community reports rather than verified enforcement or wildlife-monitoring records and therefore cannot estimate actual offtake, reports associating elephants, rhinoceroses, and giraffes with illegal bushmeat hunting indicate community-level awareness of threats affecting species of high ecological, conservation, and tourism significance. In the SME where these megafaunal species underpin crucial ecosystem processes and bolster tourism economies, even reported knowledge of threatened species targeted warrants attention (Rentsch and Packer, 2015; Sinclair et al., 2013; TAWIRI, 2023). In such cases, community narratives can complement formal enforcement and wildlife-monitoring records by identifying risks that may otherwise remain undocumented (Roe and Booker, 2019). Such approaches have been valuable for species-specific conservation, especially where ecological information is limited (Nash et al., 2016). The absence of a significant difference between beneficiary and non-beneficiary communities does not demonstrate that COPs had no effect on illegal hunting of threatened species. Rather, knowledge of hunting involving threatened species was present across both groups, and likely requires broader-scale but community-tailored interventions.

Spatial adaptation of illegal hunting to enforcement

Respondents were aware of areas currently avoided by hunters. If hunters adjust their movement and activity patterns in response to changing patrol effort, enforcement may reshape the spatial organization of illegal hunting without eliminating the activity (Dancer et al., 2022; Kuiper et al., 2025). This would be consistent with broader evidence that hunters shift from risky areas but continue hunting, a process consistent observations that patrol effort, detection probability, sanction, and landscape accessibility can influence the spatial distribution of illegal resource extraction (Hilborn et al., 2006; Xu et al., 2020). Broadly, respondents described that hunters avoid areas not only where enforcement was visible, but also where community reporting was perceived to be strong. Apparent reductions in illegal bushmeat hunting in highly monitored locations could coincide with perceived shifts toward remote or less frequently patrolled parts of protected areas, boundary and buffer zones, or surrounding village lands. Recent advances in conservation technology and adaptive patrol planning show that patrol records, ranger-based monitoring, camera traps, EarthRanger, and predictive risk models can improve the identification of priority areas to enable the dynamic allocation of enforcement effort (Xu et al., 2020). Our findings reinforce the view that illegal hunting is an adaptive cycle between hunters and law enforcement, resembling the “arms race” described elsewhere in conservation-governance research (Sjöstedt and Linell, 2021). Yet this adaptability is not always captured by static, patrol-centered management approaches. Integrating community reports with standardized patrol records, wildlife monitoring, and spatial modeling would help distinguish genuine declines in illegal bushmeat hunting from changes in location, timing, methods, or detectability.

Mitigating human-wildlife conflict as a mechanism for enhancing livelihood security and conservation support

Human-wildlife conflict response emerged as one pathway through which conservation outreach may link livelihood security with wildlife protection. Beneficiary respondents described electric fencing and related interventions as reducing crop damage, improving farming reliability, and reducing the perceived need to rely on illegal bushmeat hunting as a coping strategy. This finding is consistent with literature showing that local tolerance of wildlife and support for conservation are influenced by the costs and benefits communities experience from living near protected areas (Kegamba et al., 2022, Kegamba et al., 2023; Kideghesho et al., 2007; Roe et al., 2015). This finding should be interpreted cautiously because the human-wildlife conflict response was not reported equally across all villages and was not measured as an independent causal intervention. However, it suggests that conservation programs may gain social relevance when they reduce the livelihood costs of living near wildlife. For communities adjacent to protected habitats, interventions that reduce crop loss and support household security may strengthen tolerance toward wildlife and improve the social conditions needed for long-term conservation.

Towards a more equitable distribution of conservation benefits

The stark differences between beneficiary and non-beneficiary communities raise critical equity concerns: non-beneficiaries may face the same conservation restrictions (e.g., hunting bans, land-use limitations) as beneficiaries but receive minimal or no benefits. This could enable a ‘two-tier’ conservation system where some communities experience conservation as a partnership, while others experience it as externally imposed regulation. This inequity risks generating resentment, eroding conservation legitimacy among excluded communities, and potentially fueling non-compliance as an act of resistance (Franks and Schreckenberg, 2016; Oldekop et al., 2016). Ensuring equitable benefit distribution, either by expanding COPs to currently excluded villages or by transparently explaining inclusion criteria, is essential for both ethical and pragmatic reasons. Funding and staffing constraints may limit the reach of interventions, but community knowledge can help conservation institutions prioritize where and how limited resources are directed (Tourinho et al., 2023).

Study limitations and future directions

Several limitations should be considered when interpreting these findings. First, the cross-sectional design cannot establish neither causal nor quantifiable effects of COPs on illegal bushmeat hunting. The study did not independently measure wildlife offtake, patrol effort, arrest records, or changes in wildlife abundance and distribution. Second, the study relied on self-reported perceptions of a sensitive illegal activity and may therefore be affected by recall bias, social desirability bias, or reluctance to disclose information (Nuno et al., 2013; St John et al., 2010). Although the mixed-methods design, indirect questioning, and assurances of confidentiality may have reduced these risks, they could not eliminate them. Accordingly, the higher species richness reported by beneficiaries may reflect greater openness with researchers rather than actual differences in illegal hunting. Reported spatial shifts should be interpreted as perceived rather than confirmed displacement. Third, beneficiary status was assigned at the village level, and individual households may have differed in their actual engagement with conservation activities. Finally, the findings are context-specific and should not be generalized automatically to other protected-area landscapes, where perceptions of illegal hunting, benefit sharing, enforcement, and conservation legitimacy may differ (Afriyie et al., 2021; Bitanyi et al., 2012; Kegamba et al., 2023).

Future research could use fuzzy cognitive mapping (Gray et al., 2015; Nyaki et al., 2014) or participatory causal diagramming (Coletta et al., 2021) to compare how communities construct relationships among conservation outreach, enforcement, livelihoods, governance, and illegal hunting. In our study, we cannot clarify how non-beneficiaries’ awareness of others’ benefice influences their own narratives. Therefore, future comparisons could distinguish among beneficiaries, non-beneficiaries who are aware of COP benefits provided elsewhere, and non-beneficiaries with little or no awareness of those benefits. A combined COM-B and Theoretical Domains Framework approach could further structure comparisons of capability, opportunity, motivation, and related behavioral determinants across these groups (Michie et al., 2011; Atkins et al., 2017).

Conclusion

Beneficiary and non-beneficiary communities perceived broadly similar declines in bushmeat availability from 2018 to 2024 but explained those changes very differently. Non-beneficiary respondents emphasized their experiences with law-enforcement actors and policies, whereas beneficiary accounts situated enforcement within broader narratives of positive engagement with conservation interventions. Both groups demonstrated a willingness to share information on patterns of species utilization and activity distribution, and conservation institutions should create meaningful avenues through which to not only utilize this knowledge but also invite communities to act as co-owners of policy and surveillance. Overall, this paper shows not simply that COPs may influence attitudes toward conservation, but that they may also shape the narratives through which people interpret conservation outcomes and, in turn, their subsequent behavior. Our findings suggest that the unequal distribution of conservation programs across landscapes where conservation law is ubiquitous may generate narratives underpinned by distinct cognitive models, rather than merely different opinions.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. Generative AI was used to support language editing and improve clarity. The authors collected data, performed data analysis, wrote the manuscript, reviewed, verified, and approved all content, interpretations, citations, analyses, and conclusions. The authors take full responsibility for the preparation of the manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fcosc.2026.1901135/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: David Jennings, Defenders of Wildlife, United States

Reviewed by: Greyson Nyamoga, Sokoine University of Agriculture College of Forestry Wildlife and Tourism, Tanzania

Milosava Matejevic, University of Novi Sad, Serbia