Geographies of fear – The everyday (geo)politics of ‘green’ violence and militarization in the intended transboundary Virunga Conservation Area

Despite decades of critique concerning violent conservation practices, ‘green’ violence and militarization for the protection of biodiversity are on the rise. This paper engages with the question of how and why these violent and militaristic dynamics unfold in relation to Transboundary Protected Areas in particular, which, paradoxically, are widely promoted as an instrument of peacebuilding. Based on long-term ethnographic fieldwork within the intended transboundary Virunga Conservation Area between the Democratic Republic of Congo, Rwanda and Uganda, I introduce a novel analysis of the instrumental role of fear into the conservation debate. Inspired by frameworks from emotional geographies and critical geopolitics, I examine fear discourses that are perceived, reproduced and enacted in everyday realities of individuals living within this violent geography. My analysis focuses on the narratives of park rangers who are trained and deployed as parastatal conservation actors by their respective governments. Rangers’ perceptions reflect a discursively constructed ‘geography of fear’ informing their understanding and implementation of violent and militaristic measures as ‘necessity’ for the protection of national and park borders. Their fear narratives further point to governmental interests in political sovereignty and territorial integrity undermining transboundary collaboration efforts. Ultimately, I argue that the analyses of fear can advance conservation studies by generating a deeper understanding of the persistence of ‘green’ violence and militarization. On that account, I advocate to include the voices of those who feel and perform these problematic practices in their everyday lives to create alternative, non-violent pathways in the future.


Introduction
Although critical scholars have contested violent practices undertaken in the name of conservation for decades (Neumann, 1998;Peluso and Watts, 2001;Shiva, 1990), an alarming increase in "green militarization" (Lunstrum, 2014) -the adaption of militaristic methods, tactics and equipment for protection of flora and fauna -has been widely documented in recent years (Adams, 2019;Bocajero and Ojeda, 2016;Duffy, 2014). Expanding on this concept, Büscher and Ramutsindela (2015) outline a continuum of forms of 'green' violence including but not limited to militaristic measures deployed in the specific context of Transboundary Protected Areas (TBPA). TBPAs are conservation sites that straddle the political borders of at least two countries (IUCN 2020). Promoted as 'peace parks' in Sub-Sahara Africa, they have been praised as contributing to encourage collaboration between (formerly) conflicting states, leading to economic development, sustainable conservation and ultimately regional peace (Ali, 2007;Barquet et al., 2014;Hanks, 2003). While increasing violence and militarization in and around TBPAs contradict this asserted association with peacebuilding, few studies have addressed these problematic developments in the specific context of 'peace parks' (a notable exception being Büscher and Ramutsindela, 2015). This article thus pursues a deeper understanding of how and why 'green' militarization and violence unfold within these spaces and thus undermine the cooperation and peace intended to manifest between countries who entered a transnational TBPA agreement?
To fill this research gap, this paper explores the case of the intended transboundary Virunga Conservation Area, 1 composed by Virunga National Park (PNVi) in the Democratic Republic of the Congo (DRC), Volcanoes National Park (VNP) in Rwanda, and Mgahinga Gorilla National Park (MGNP) in Uganda. Since TBPAs are border-crossing geographical spaces characterized by complex political dynamics and a multitude of interests between state and non-state actors, they require transnational agreements that are often informed by international development, conservation and peacebuilding trends (see e.g. UNEP, 2009). They are thus inherently geopolitical.
Within the Virungas, efforts to institutionalise cooperation among the DRC, Rwanda and Uganda on the TBPA model by the intergovernmental organisation Greater Virunga Transboundary Collaboration (GVTC) are undermined by segregation and militarization of the three National Parks as well as their respective rangers. Stories of frequent violent and lethal skirmishes between different armed groups within the Virungas and across state boundaries also contradict the regional peace aspirations expressed in the TBPA agreements. Further, militaristic measures undertaken in the name of conservation are jurisdictionally and discursively 'normalized' by the national governments of both Rwanda and Uganda, yet are reiterated and performed through the local and felt in the everyday experiences of conservation staff and parkadjacent communities alike. In order to shed light on this link between (trans)national dynamics and local realities within the context of TBPAs, I focus on the experiences and perceptions of of park rangers from the Rwanda Development Board (RDB) at VNP and the Uganda Wildlife Authority (UWA) at MGNP.
To develop my analysis of these dynamics, I offer a novel conceptual approach to exploring 'green' conservation violence inspired by frameworks from emotional geographies and critical geopolitics focused on the analysis of fear, understood as a discursively constructed experience. Scholars from emotional geographies emphasize the circular relationship between emotions and landscapes that shapes both social relations and political processes (Anderson andSmith, 2001, Davidson andMilligan, 2004;Tuan, 1979). Meanwhile, the subjective experience of emotions that informs individual behaviour in relation to certain spaces has been explored in the critical geopolitics literature. Critical geopolitics challenge conventional geopolitical scholarship that tends to reduce political struggles to contestation between dominant nationstates on the world stage. In contrast, work in critical geopolitics emphasizes analysis of personal, emotional experiences that can illuminate the reciprocal interrelation between (geo)political state discourses and their enactment by individuals in their everyday lives (Hyndman, 2007;Pain and Smith, 2008). Placing emphasis on individual emotions, some scholars specifically highlight the instrumentalization of fear by governmental discourses wherein geopolitical threat narratives are promoted to legitimize state-sanctioned violence in the name of national security (Cowen and Gilbert, 2008;Huysmans, 2006). Inspired by this approach, I propose to ground the analysis of green violence within transnational conservation efforts on the local scale in order to illuminate how fear is experienced and negotiated by actors charged with carrying out state-mandated conservation directives.
My aim is to examine how militarized and violent dynamics reveal the entanglement between conservation and (geo-)politics by analysing the emotional repercussions of these developments in park rangers' narratives about their day-to-day work. Trained and deployed as parastatal conservation actors by their respective governments, rangers' perceptions revealed patterns of fear that inform their enactment of and ostensible compliance with these dynamics. Based on rangers' narratives depicting the transboundary forest as a 'geography of fear', I argue that their fears function to legitimize 'green' violence, militarization and segregation of the National Parks, thereby serving underlying (geo-) political interests of sovereignty and integrity for the respective governments. Analysing fear discourses within conservation practices can thus aid in deconstructing the interlinkage among 'green' violence, militarization and concealed (geo-)political interests in TBPA development. Moreover, previous studies concerning the perceptions of fear have yielded deeper understanding of individuals' and groups' willingness to commit violence that have helped to inform better targeted interventions (see e.g. Agnew, 2001). Thus, understanding rangers' fears in the context of conservation enforcement can help to explain why different forms of 'green' violence persist and hence inspire pathways for pursuit of non-violent alternatives.
After a brief description of ethnographic methods used in this research, I explain how integration and synthesis of literature from critical geopolitics and emotional geography on fear can advance the debate on 'green' violence and militarization in the context of (transboundary) conservation. I then outline the historical context of the Virunga region before turning to the analysis of three dominant fear narratives and their underlying governmental interests that inform rangers' everyday realities. Based on this discussion, I argue that rangers' perception of a 'geography of fear' ultimately serves to legitimize militarization of the Virungas in the name of conservation, thereby undermining the transboundary collaboration efforts. In conclusion, I reiterate how the analysis of fear discourses can benefit critical conservation studies by highlighting the concealed geopolitics in everyday conservation work on the ground.

Methods
This paper draws on twelve months of ethnographic fieldwork in 2018 conducted in and around the three National Parks that are envisioned to constitute the transboundary Virunga Conservation Area. Taking a participatory research approach, I spent a significant amount of life and work time with park rangers from RDB in Rwanda and from UWA in Uganda, partaking in tourist guiding, gorilla monitoring, and antipoaching patrols as well as living with rangers in different patrol outposts (permanent posts away from the park headquarters). To complement conversations and observation conducted during this time, I held semi-structured individual and focus group interviews with local community members, park rangers and their superiors holding positions as park wardens and in park management. Due to security restrictions, interaction with rangers in the DRC was limited to a granted stay at the Rumangabo headquarters within PNVi and interviews in Goma, the regional capital of North Kivu. Triangulation of data from these various sources was reinforced by interviewing transboundary conservation actors from GVTC and different regional (I)NGOs in all three countries, as well as by reviewing secondary literature from previous academic research, policy documents, national and international media releases and reports from international organisations.
As the perception and expression of emotions is subject to a specific socialized linguistic framing (Derrida, 2001), my focus on the expression of 'fear' is a subjective interpretation of local experiences. By 'fear,' I thus refer to narrated phrases in the commonly spoken language Kinyarwanda, 2 where to "be feared" (gutinywa) and "be fearful" (gingimiranya / ubwoba) hold similar contextual meanings and overlap with the same expressions in both English and French. In line with van Maanen's (2011: 2) emphasis of the interpretive act of fieldwork, my ethnographic methods aim to make sense of narratives constructed of recounted memories, stories, rumours and anecdotes of ranger's social world. Therefore, it is less important whether the content of these stories is 'true' or 'false' or merely invented for my presence as a researcher and a w3 (white, Western woman). Yet these limitations as well as the anonymization of all informants demand specific acknowledgement in a region of high surveillance shaped by historical and ongoing inter-and intrastate warfare where suspicion and mistrust are basic codes of conduct for survival.

Grounding on fear: 'green' violence, militarization and transnational conservation efforts
The creation of a transboundary protected area, defined as "a clearly defined geographical space that consists of protected areas that are ecologically connected across one or more international boundaries" (IUCN, 2020), entails substantial political negotiation between the involved governments and (international) non-governmental conservation actors and organisations. While conservation scholars have critically examined the underlying political interests of the multitude of stakeholders involved in the production of TBPAs (see e.g. Büscher, 2013;Fairhead, 2005), the instrumental role of fear informing these interests has not been taken into account so far.
Yet fear and its functionalities have been examined by a plethora of disciplines ranging from evolutionary biology, psychology, criminology, security and military studies, traditional and critical geopolitics to wider social sciences for centuries. The emotion of fear is generally defined as "a distressing emotion aroused by impending danger, evil or pain, whether the threat is real or imagined". 3 In situations of perceived danger, the reptilian complex within the human brain is understood to release hormones for fight-or-flight responses that cognitively justify the use of violence for protective purposes and survival. Hence, fear can be an emotional response to violence, while violence can be a physical response to fear, in defence or as a pre-emptive attack against a perceived threat.
Evolutionary psychologists have identified a spatial dimension in triggering fear. It was found that animals show 'spatial patterns of risk perception' that ultimately structure their ecosystem into go-and no-goareas referred to as their 'landscape of fear' (Gaynor et al., 2019). Since similar physiological responses to certain spaces or places have also been found in humans, Tuan (1979) adapted the concept of 'landscapes of fear' for analysis of human behaviour. Based on Lefebvre's argument that emotions and spaces interact reciprocally (ibid. 1974), he laid the foundation for emotional geographies in the social sciences that "refers both to psychological states and to tangible environments" (Tuan, 1979: 6). Building on this, Tulumello (2015) proposed to critically analyse the politics of fear behind the production of spaces revealing governmental processes of violence, power and controla dynamic that can be similarly observed in the production of TBPAs.
In order to shed light on the politicized interrelation between fear and these constructed geographical spaces, I also draw on research in critical geopolitics. Scholars from this discipline have focused attention on the geopolitical interests behind the instrumentalization of fear that is used to legitimize armament, violence and military interventions by states against a declared threat (Dalby, 1994;Girard, 1979;Hyndman, 2007). These threats are often informed by 'Othering' practices aiming to create a national identity distinguished from constructions of an external danger developed through fear discourses (Mountz, 2009). Research concerning the neurophysiological functionality of fear suggests that this perception of the dangerous 'Other' lowers the threshold to commit or support acts of violence that has been widely used within military tactics since records of war began (see e.g Homer's 'Ilias', ca. 8th century BCE). Fear has thus been identified as 'a tool of war that works through the minds' (Watson, 1978: 534). A prominent example can be found in 'war on terror' discourses under the U.S. Bush administration from 2001 onwards, showing how fear-inducing labels such as 'terrorists' were created and disseminated as propaganda to evoke popular support for the military invasion of Iraq (see e.g. Gregory and Pred, 2007;Robin, 2004;Sparke, 2007).
Similar to these geopolitical war tactics, critical conservation scholars have scrutinized how governments justify the increasingly violent and militaristic practices employed for the protection of flora and fauna through geopolitical discourses by linking conservation with the 'war on terror' (Duffy, 2016). The themed issue on 'green security' previously published in this journal highlights how those kinds of "globalised fear frameworks" discursively connect national security with protected areas framed as in need of exceptional violent and military protection (Ybarra and Kelly, 2016: 171). In here study of militarisation within the Maya forest in Guatemala, for instance, Ybarra demonstrates how "successful security strategies use imaginaries of existential threats to justify violence" (ibid. 2016: 195). Thus, violence can be enabled by threat narratives that criminalise people who enter protected areas unauthorized, for example by branding 'poachers-asterrorists' (Pennaz et al., 2017) or 'rogues for elimination' (Fairhead et al., 2012). I refer to the concept of 'green' violence in the context of TBPAs entailing an amplitude of material and non-material characteristics, including spatial enclosure and militarization of the protected areas as well as the implementation of discursive and physical forms of violence by conservation actors (Büscher and Ramutsindela, 2015: 10).
Yet thus far, this line of analysis has not focused on the use and consequences of fear as a tactic in 'green' conservation violence. In order to generate a deeper understanding of the paradox behind increasing violent and militaristic practices in the context of supposedly peacegenerating TBPAs, this paper thus seeks to illuminate the interrelation between (geo-)political interests and the implementation of 'green' violence by analysing individual fear narratives of park rangers in their day-to-day conservation work. Training attention on the local level allows one to account for the specific emotional experiences embedded in particular cultural, historical contexts and present conditions (Berger and Luckmann, 1966;Harré et al., 1986;Wouters, 1989). Since the expression of personal emotions can reflect underlying motivations for people's behaviour (Bourdieu and Loïc, 1992), my focus on individual, emotional experiences and narratives of park rangers may help to explain their ostensible commitment to and/or compliance with the enactment of different forms of 'green' violence. Previous studies from critical geopolitics suggest that fear narratives in the everyday can reflect how governmental discourses construct social realities pointing to the political interests behind the production of violence (Pain and Smith, 2008: 3). Hence, analysis of park rangers' fear narratives can offer a bridge between violence and militarization in the name of conservation on the ground, on the one hand, and its implications for national policies and practices in relation to the transboundary conservation efforts on the transnational level, on the other. To apply this analytical framework to the Virungas, I begin with a brief presentation of the historical context of the conservation sites in the DRC, Rwanda and Uganda.

The history of an idea -A transboundary Virunga conservation area?
The intended transboundary protected area straddling the Virunga volcanoes appears to be an exemplary case for the intensification of militarisation in conservation enforcement, which is popularly legitimized on two main grounds. First, the contiguous forest constitutes the last habitat of endangered mountain gorillas (besides a small, separated forest parcel in Uganda's Bwindi Impenetrable National Park). It is thus deemed a biodiversity hotspot by conservationists as well as a 'unique selling proposition' for one of the most expensive wildlife tourism attractions in the world (Trogisch and Fletcher 2020). Secondly, the TBPA is situated in the Great Lakes Regioninfamously labelled the site of "Africa's World War" (Prunier, 2011) -which is historically scarred by brutal colonial legacies and long-standing intra-and interstate warfare.
The historical entanglements among the DRC, Rwanda and Uganda point to the decisive geopolitical role of the Virungas in these armed conflicts and enable a deeper understanding of fear discourses that accelerate dynamics of segregation and militarisation among the three National Parks today. Long before colonialization, transnational interactions across the Virunga volcano chain have been characterized by territorial combats as well as intermarriages, settling, trading, labour and migration mobilities (Hochleithner, 2017;Nzabandora, 2006). The common language of Kinyarwanda spoken by locals settled around the volcanoes testifies to the century-old mobility and interconnection among communities in the region. In the colonial power struggles between 1890 and 1916, however, Germany, Belgium and Great Britain drew the borders for the Democratic Republic of the Congo, Rwanda and Uganda, respectively, by connecting the peaks of the Virunga volcano chain with a ruler on the map (Mamdani, 1996, Mathys, 2017Newbury 1998).
As a result of the Western 'discovery' of the mountain gorilla in colonial expeditions in 1902, international conservationists became another powerful usurping actor in contestations over the area. Manifested in the gazettement of the forest around the Virunga volcanoes as part of Albert National Park in 1925, Belgium colonial rulers established one of the first National Parks on the African continent encompassing in total roughly 8000 km 2 . After independence of the DR Congo in 1960 and Rwanda and Uganda in 1962, the forest was managed by the separate state entities and fluctuated in size due to fights between clearance for commercial and subsistence activities and conservation enforcement. The strict protection of mountain gorillas became prevalent due to international fame and funding garnered through Dian Fossey's research from 1967 until 1985 that subsequently led to an influx of international tourists.
Thus, for the past 100 years, colonial powers and international conservation organisations have staged the borderland forest surrounding the Virunga volcanoes as a 'space of exception' for mountain gorilla conservation requiring eviction of local inhabitants and military protection of the enclosed area (de Bont, 2017;Marijnen, 2018). As part of the colonial legacy, international threat discourses portending the looming destruction of forest and great apes by local people informed 'green' militarization practices based on harsh conservation laws (Ashaba, 2020). Their enforcement is managed by the three nations' respective protected area authorities: the Institut Congolais pour la Conservation de la Nature (ICCN), the Rwanda Development Board (RDB), 4 and the Uganda Wildlife Authority (UWA).
In contrast to the parastatal management of RDB and UWA, ICCN entered a public-private partnership with the British NGO Virunga Foundation in 2008 which changed the organisational structure, financing and management of Virunga National Park in the DRC significantly (Marijnen, 2017). As a result, this article focuses mainly on park rangers of RDB and UWA. Similar in their hierarchical organisation, both institutions are subordinated to the government, headed by one chief park warden and four to five wardens as well as administrative park management staff who supervise, instruct and manage the rangers in the domains of tourism, security, community relations and finance. While wardens in Uganda live at the headquarters of MGNP situated at the main entrance to the park, VNP wardens have residences in the next city or the capital Kigali.
Park rangers living at the headquarters or barracks at the outposts of VNP and MGNP carry out three main tasks on a daily base: monitoring of wildlife, patrols as deterrence and in response or in search of illegal park activities, and tourist guiding. In Uganda, UWA mainly deploys rangers to MGNP who not originate from the park-adjacent communities or the region to reduce social ties and clientelism. In contrast, RDB rangers at VNP mostly come from the surrounding communities and visit their families regularly. While there is a significant difference in terms of gender distribution within RDB and UWA, 5 I anonymize all quotes to avoid a gendered interpretation of the presented fear narratives as these differences did not appear as an important aspect of the present analysis.
In 1991, the International Gorilla Conservation Programme (IGCP), a transnational conservation NGO founded by the African Wildlife Foundation (AWF), Fauna and Flora International (FFI) and the World Wide Fund for Nature (WWF), initiated joint ranger patrols as well as monitoring and training programmes for conservation staff from ICCN, RDB (then ORTPN) and UWA 'to ensure effective management of the shared ecosystem' (IGCP, 2010a). These first efforts to establish a transboundary protected area (TBPA) for the conservation of mountain gorillas were in line with international development and conservation organisations promoting the triple-win effects of TBPAs intended to bring economic development, conservation collaboration and ultimately regional peace (UNEP, 2009). In this effort, funded by the Government of the Netherlands, the three national protected area authorities signed the first Memorandum of Understanding to transform the informal arrangement between rangers into an intergovernmental institution titled Greater Virunga Transboundary Collaboration (GVTC) in 2004. In the official treaty two years later, the respective ministries of the DRC, Rwanda and Uganda agreed to commit to "building trust, understanding and cooperation among wildlife authorities, nongovernmental organisations, users and other stakeholders to achieve sustainable conservation and thereby contribute to peace" (GVTC, 2006(GVTC, [2014: 6).
However, these formal attempts of establishing the Virungas as a TBPA in the model of a 'peace park' have proven difficult to operationalize, being opposed by nationalist attitudes, mutual hostile perceptions and accusations of the 'other' side by people living and working with(in) the different National Parks. This mental segregation was expressed in stories about periodic border skirmishes between different armed groups and the military from the DRC, Rwanda and Uganda carried out within or through the forest, yet official statements and media reports would only sporadically confirm those incidents, partially due to strict censorship by the government of Rwanda and Uganda (see e.g. BBC, 2018;International Crisis Group, 2020, Reuters, 2018. Political killings of mountain gorillas (UNESCO, 2007;UWA, 2020), tourists (Maekawa et al., 2013) as well as frequent targeted killings of park rangers from ICCN (Virunga, 2020) also contravene the 'official' peace intentions in the ministerial agreements on the national scale. Moreover, the transboundary conservation collaboration between the three governments and protected area authorities have decreased drastically, underscored by the diminishing support for GVTC requests, propositions and meetings. National solo efforts from the unexpected price increase for gorilla tourism by Rwanda in 2018, increasing border controls and closures and the militarization of the three National Parks undermine the agreements of the GVTC treaty and undermine any political authority of the trilateral organisation. One decisive example for this segregation is the unannounced increase of mountain gorilla tracking fees by the Rwandan government in 2018, which had tremendous impact on tourism dynamics and concomitant conservation challenges for the DRC and Uganda (see Trogisch and Fletcher 2020). After termination of Dutch funding in 2018, GVTC had to face liquidation as the three states have not acted upon their official lip services to secure the financing (as of February 2021). The 'transboundary Virunga Conservation Area' thus remains a label without spatial or practical manifestations.
Spending time with park rangers in their everyday life throughout 2018 shifted my attention to the lived realities of these segregation processes on the ground. The last joint patrol along the borders of the DRC between RDB and UWA with ICCN was recorded in 2003; meanwhile RDB and UWA continued along the Rwandan-Ugandan frontier until 2016. While patrols along and across borders together were frequently remembered by rangers from ICCN, RDB and UWA as friendly meetings for sharing information, monitoring data and lunches, the perception of their neighbours was subsequently coloured by suspicion, hostility and fear (personal interviews, 2018). These fear-induced narratives of rangers illustrate the mental and spatial segregation between the three states in the local conservation context that highlight the stark contrast with the intergovernmental 'peace park' rhetoric in the GVTC agreements.
Moreover, rangers enact and become the core of 'green' militarization as their training and daily conservation work expanded significantly to para-military and intelligence tasks. While scholars have critically analysed the discursive techniques of legitimizing militarization of Virunga 4 The protected area authority of Rwanda has been renamed from Office Rwandais du Tourisme et des Parcs Nationaux (ORTPN) to RDB as part of the change in official language from French to English in 2009. 5 In 2018, VNP engaged 135 rangers of which only six were women who carried out administrative and less strenuous guiding tasks. In contrast, around 30% of all rangers at MGNP were woman, including one of the four wardens who carries out the same tasks as her male counterparts.
National Park (Marijnen and Verweijen, 2016) and its interrelation with ongoing violent conflicts in the Eastern DRC (Vikanza, 2011), my major focus is on narratives of rangers from Uganda and Rwanda and the specific role the perception of 'the Congo' plays in the segregation, militarization and violence. To generate a better understanding of these dynamics, in the following section I examine how three state-cultivated fear narratives construct RDB and UWA rangers' understanding and performance in their everyday reality.

A historically constructed 'geography of fear'
The first narrative circulated amongst the park rangers in Rwanda and Uganda presents the Virungas as a 'geography of fear' based on legacies of insecurity from the historical warfare inside the transboundary forest justifying strict conservation measures of enclosure and military protection. As Pain (2009) argues, 'geographies of fear' are made of personal perceptions of spaces regardless whether the triggered emotions rely on factual experiences of threat or on their imagination. In Rwanda and Uganda, rangers' narratives illustrate how their spatial perception of the Virunga volcanoes depict such a 'geography of fear' informed by factual and fictive memories of the Rwandan civil war in 1990 that culminated in the genocide in 1994.
While navigating through the parks for their daily patrols, rangers from MGNP and VNP showed 'spatial patterns of risk perception' that map a landscape into go-and no-go-zones (Gaynor et al., 2019). No-gozones pointed to spaces of 'death' and 'danger' 6 inside the forest that rangers on both sides recalled as memories of their 'personal experiences' 24 years after the end of the Rwandan war. These detected zones informed their paths in avoiding certain hill tops, valleys and natural volcanic caves that they referred to as rebel hide-outs or execution spaces as well as mass graves from the war (personal interviews, 2018). In interviews, some rangers specifically demarcated these dangerous sites in sketching maps to illustrate the topographical knowledge required for their day-to-day conservation tasks in the Virungas. These verbal and visual illustrations show how rangers picture the forest as a 'geography of fear' based on memorized and reiterated 'experiences' of violence and threats in the past.
Yet these narrated 'experiences' were seldomly consistent with the personal timeline of ranger's deployment or their age. 7 As Fujii (2010) pointed out, invented stories, rumours and silences are frequently part of local testimonies in political sensitive contexts and can indicate suspicion towards external actors, and researchers in particular. However, "their value might lie in the meaning with which the narrator endows the events or moments," reflecting "the speaker's state of mind, aspirations, and desires" (Fujii, 2010: 234). At first, I assumed that especially young rangers' aspirations behind telling these stories of violence and killings were to impress by depicting how they confront the fearsome dangers of their work. Over time, however, I found resemblance between rangers' 'fictive' fear narratives and dominant threat discourses circulated amongst their respective superiors in the park management and in the media. Other studies have found that governments gain popular support by drawing on historical legacies and previous events to reinforce narratives of insecurity that mobilize and instrumentalize previous fears within the population (Sparke, 2007: 341). In the following I explore such a historical, geopolitical dimensions behind fear discourses used to legitimize the enclosure and 'green' militarization of MGNP in Uganda and VNP in Rwanda from 1990 onwards.

Enclosure of Mgahinga Gorilla National Park, Uganda
In the guise of exclusive conservation enforecment, the Ugandan government under Yoweri Museveni initiated the enclosure of Mgahinga forest to create an undisturbed, invisible and well-protected hide-out for the rebel army Rwandan Patriotic Front (RPF) to support their attack against the government of president Juvénal Habyarimana in Rwanda (UNHCHR and OHCHR, 2010;Epstein, 2017). Therefore, the gazettement of Mgahinga to a National Park category II mandated the eviction of 2400 people from their former homelands within the fenced area 8 briefly after the beginning of the Rwandan civil war on the 1st October 1990. With international funding from various conservation agencies, the park headquarters, three permanent patrol outposts and a girdling stone wall were erected on the forest frontier in the following years, officially declared to 'protect' adjacent residents from wildlife exiting the forest (Blomley, 2003). These material manifestations made Mgahinga forest literally a 'fortress' (Brockington, 2002) in the name of mountain gorilla conservation (Butynski and Kalina, 1993).
Yet testimonies of local residents are in line with official reports maintaining that the government of Uganda utilized this enclosure to shield their strategic and military support for the RPF from witnesses. A resident from Mgahinga emphasized that this timing of the National Park creation and the begin of the Rwandan civil war was no coincidence: "Everybody knew who sat inside the forest because the rebels from Rwanda told us how to hide from the bombs. We were so stupid, we did not know what bombs are, so we climbed on trees to see better. The government closed the park, so we could not see." As this quote exemplifies, many locals recall their first-hand experiences of the deployment of the Tutsi-dominated RPF led by today's president Paul Kagame and the warfare with Rwanda unfolding within the forest. From 1991 to the genocide in Rwanda in 1994, Mgahinga Gorilla National Park became a military base, refuge and battleground supported by Ugandan president Yoweri Museveni that forged an alliance between the two governments in subsequent years. While rangers' sense-making of the enclosure was aligned with the official conservation purpose, their perception of dangerous places inside the transboundary Virunga forest points to their mental spatialization of the area in reference to this historical warfare. Thus, memories and imaginaries of the Virunga forest as a 'geography of fear' were reproduced and strengthened how rangers understand and legitimize the enclosure and militarized enforcement of MGNP then as now.

Enclosure of Volcanoes National Park, Rwanda
The 'geography of fear' perceived by rangers from the Rwandan side of the Virungas illustrates how the historical enclosure and militarization of VNP is entangled simultaneously with international conservation demands and Rwanda's geopolitical interests. From 1990 onwards, VNP and its adjacent hinterland became the main theatre of war between the RPF invading from MGNP and the Rwandan military under Habyarimana. As the VNP headquarters and outposts were abandoned, the park was used as a resource base and battleground between the warring parties. Subsequently, the forest became an atrocious execution space and graveyard during the genocide in Rwanda in 1994. In its aftermath, VNP provided one of the main routes for refugees fleeing into neighbouring Zaire including the Habyarimana government and its military organizations, among them the interahamwe who is held responsible for perpetration of the genocide. 9 While younger rangers pointed out historical places of urupfu 6 In Rwanda, rangers referred to specific sites with the Kinyarwanda word 'urupfu' (death). In Uganda, rangers used the English terms 'death' and 'danger' interchangeably to demarcate a site they would avoid. 7 Some older rangers indeed presaged traumatizing personal experiences but would not elaborate on those.
["death"], park rangers who have been on duty during this period and survived did not speak about their experiences or shifted the conversation to Rwanda's present conservation successes. Similarly to those stories and silences, other studies on (pre-)genocide memories found that many Rwandans either 'choose collective amnesia' (Buckley-Zistel, 2006: 133) or align their stories to strong state-led discourses (Nymeth Brehm and Fox, 2017: 121). One of the most compelling of those discourses focuses on Rwanda's conservation achievements to create popular support for the gazettement and militarization of the National Park.
With an estimated funding of over six million USD by international conservation organisations, aid agencies and NGOs (Milner-Gulland and Mace, 1998: 304), the new Rwandan government re-established VNP under the pretext of reclaiming mountain gorilla habitat in 1996. Similar to MGNP, VNP became a strictly enclosed 'fortress' surrounded by partial stone wall constructions, trenches and nine permanent patrol outposts for rangers in addition to numerous military bases along and within the park. This strict enclosure assuaged the fears of primatologists accusing Rwanda of reducing the mountain gorilla population to the brink of extinction and hence resulted in unabated international investments. Yet, more importantly, VNP also served the new Tutsi-dominated Rwandan government under Kagame to erect a military bulwark against Zaire in fear of revenge from the expelled but largely intact former Hutu government, with related military parts of the interahamwe reinvigorating within and around Virunga National Park (PNVi). Meanwhile, the enclosure of VNP enabled the Kagame government to send armed groups through the safe, concealed passage of the Virunga forest into Eastern Zaire to carry out pre-emptive strikes against the expelled former regime. With military support from its Ugandan ally, these counterinsurgencies were intended to exterminate suspected Hutu living in refugee camps, surrounding villages and within PNVi. In response to Rwanda's interventions, Mobutu Sese Seko, then president of Zaire, mobilized ethnic hatred against Zairian residents considered to be Tutsi, defined as 'Kinyarwanda-speakers' or 'Rwandophones,' who were accused of supporting Kagame's Tutsi-dominated regime (Mathys, 2017: 469). In consecutive years of warfare, officially accounted for as the 'two Congo Wars' (1996-1997; 1998-2003), approximately two million people have been killed (UNHCHR and OHCHR, 2010). In this aftermath of the genocide in Rwanda, the mutual exclusion and killing orders along the Hutu-Tutsi dichotomy by the Rwandan and Zairian (later Congolese) governments resulted in wide-spread atrocities and new waves of refugees in both directions across the borders (Okosun and Kibiswa, 2013).
Besides this geostrategic utilisation of VNP's enclosure, park rangers were assigned a significant role in Kagame's counterinsurgencies that evidenced the blurring lines between conservation and military tasks. Under Mobutu's command to exclude the Tutsi-related population from Zaire, public resentment increased leading to intimidation, violence and killings of community members with an associated Tutsi heritage, including ICCN rangers. As a result of this first Congo war, 10 some rangers with this attributed Tutsi ethnicity fled in fear of their lives to Rwanda. During the 'second Congo war' from 1998 to 2003, the Rwandan government made use of these rangers' precarious situation and topographical knowledge of the Congolese side of the forest to recruit them to detect Hutu hide-outs within PNVi, in exchange for jobs, asylum, andin case of accused Hutu-affiliationamnesty (personal interviews, 2018). As an example representative of a number of testimonies, a former ICCN ranger described the command to lead the Rwandan army through PNVi into the Eastern Congo as follows: "Nobody knew when we would start to guide the soldiers, because nobody should know what we do and where we go. Not even I did know. After the genocide everybody was a spy. I just showed the places in the forest where they [Hutu refugees, authors' note]

hide. I put down the GPS coordinates and then they [Rwandan soldiers, authors' note] would -you know. There was nothing left."
The mutually incited fears along ethnic lines by the Rwandan and the Zairian/Congolese government served to exploit park rangers for military, cross-border interventions through the enclosed forest. As became evident in collaboration workshops organised by GVTC in 2018, these historical fears continue to inform feelings of mistrust, suspicion and hostility between Rwandan and Congolese rangers. Such fears offer one explanation for why park rangers increasingly oppose the former transboundary collaboration with their neighbouring conservation counterparts, compounding governmental withdrawal of support for the Virunga Conservation Area project in the model of a TBPA.

'Green' militarization against 'the Other'
The second narrative serving to legitimize 'green' militarization of the National Parks and their rangers in Rwanda and Uganda demonstrates how the politics of fear unfold in 'Othering' discourses against an external enemy symbolized in a generalized image of 'the Congo'. As Mountz states, the geographical aspects of these 'Othering' processes "rely heavily on spatial metaphors to locate and place identities and difference in the landscape" (ibid. 2009: 332). Since the metaphor of the Virungas as a 'geography of fear' was fuelled by the historical war alliance assimilating Rwanda and Uganda in the aftermath of 1994, 11 narratives of rangers from RDB and UWA still located the source of their fears in 'the Congo'. In the course of my fieldwork in 2018, rangers were eager to share stories from loudly declared hostilities to tacit, subtle but no less angst-ridden anecdotes about the chaos, insecurity and the barbaric behaviours of Congolese people on 'the other' side of the park.
Yet most Rwandan and Ugandan rangers admitted they have never experienced an incident with a Congolese, had never stepped on Congolese soil or even spoken with a Congolese person. Their fear narratives point to a public 'Othering' discourse constructing a unifying national identity of park rangers as state agents distinguished from the DRC. 'Othering' functions as a discursive tactic to legitimize violence declared as measures for security against an external enemy (Mountz, 2009: 328). Militarization of the National Parks was hence rationalized as a security measure against the enemy labelled 'the Congo', a generalizing title to merge landscape, inhabitants, the government and armed groups situated in the DRC into a threat to national territory, wildlife and tourists. This 'Othering' of the DRC is substantiated by rangers' reiteration of two specific violent incidents presented in the media that impacted gorilla conservation and tourism efforts of Rwanda and Uganda.

Rwandan enmity against 'the Congo'
The incident in Rwanda refers to the deadly attacks by "several thousand interahamwe" against civilians in 1997 that disrupted the just re-opened tourism and conservation activities at VNP (East, 1999: 74). Since this incident, president Kagame continuously attributed crossborder attacks against Rwanda to the Democratic Forces for the Liberation of Rwanda (FDLR), one of the most dominant armed groups in the Eastern Congo that emerged from the interahamwe, remnants of the former Hutu government and its supporters (see e.g. The New Times 2018). Further alleging that the Congolese government prevents 10 The DRC was named Zaire under Mobutu Sese Seko in 1962. After the coup by the Alliance of Democratic Forces for the Liberation of Congo-Zaire (AFDL) aided by the Rwandan and Ugandan government overthrowing Mobutu, Zaire was renamed to the Democratic Republic of the Congo in 1998. 11 The relationship between Rwanda and Uganda deteriorated significantly since the severe warfare between Rwandan and Ugandan forces on Congolese territory during the Second Congo War (BBC, 2000) and political accusations and economic trading barriers in the past years (Reuters, 2019). extradition as well as supporting the militias for decades, governmental discourses established a clear link between threats to Rwanda's national security and 'the Congo' as a territorial and political entity. In addition, media sources incriminate the Congolese army, Les Forces armées de la République démocratique du Congo (FARDC), who are accused of transgressing the territorial boundaries and challenging Rwandan border integrity (BBC, 2018;Reuters, 2018;The New Times, 2012).
RDB rangers would consistently agree with this state-led fear discourse that frames 'the Congo' as the enemy of Rwanda by providing a sanctuary for the 'genocide perpetrators' and support for the 'Hutu rebels' within PNVi (focus group interview, 2018). This rhetoric underscores how fear discourses are linked to the imagined space of the Eastern DRC serving as an imperative for the militarization of VNP. While stories about violent attacks and killings between different armed groups on the Rwandan side of the forest frequently occurred during fieldwork in 2018, all questions about these skirmishes posed to RDB rangers on site were commonly dismissed with the words: "On ne sait pas…" ["We don't know"] or with vague reference to rumours. When asked about their fears of getting involved and physically harmed in these skirmishes during patrols along the Congolese border, rangers predominantly restated their general fear of meeting 'a Congolese' anywhere who would kill them for being a Rwandophone, and thusassumedly -a Tutsi. This prejudicial fear points to the historical legacy of ethnic mobilisation that fuelled the 'Othering' and alienation of rangers from ICCN and RDB, further undermining the envisioned transboundary collaboration. From this perspective, rangers expressed their firm belief in the necessity of their military education, armament and conjoined movement with the national army to protect tourist, gorillas and Rwanda's national borders against 'the Other', namely 'the Congo'.
At the same time, Rwandan rangers similarly voiced their feelings of security within Rwanda due to their countries' military capabilities. One ranger expressed the sense of protection in the day-to-day patrols with the phrase "chaque arbre, un soldat" ["Each tree, one soldier"]. The statement refers to the high density of soldiers from the Rwandan Defence Force (RDF) deployed within VNP and its surroundings that I witnessed first-hand on daily patrols. As soon as we were invisible from the public after passing trench and stone wall to enter the park, four to ten wellequipped soldiers with bulletproof vests and machine-guns would join the group of three rangers armed with rifles. These conjoined activities point to the double function of VNP's 'green' militarization that serves to merge the increasingly popular conservation paradigm of militarized environmental protection with Rwanda's security concerns by establishing a military bulwark against the DRC.

Ugandan enmity against 'the Congo'
The incident fuelling 'Othering' amongst UWA rangers in MGNP refers to a supposedly Hutu-affiliated militia killing of eight tourists and four park rangers on the borderland between the DRC and MGNP's neighbouring Bwindi Impenetrable National Park (BINP) in Uganda in 1999. The militias declared the attack revenge for the ongoing military interventions into the DRC by conjoined forces of Rwanda and Uganda and their international support by theUS and UK in the context of the second Congo War (HRW, 1999). Since the incident led to the complete halt of tourism, Uganda's most important export sector, the government of Uganda interpreted the attack inside the National Park as a political tactic to harm the national economy (BBC, 1999). In response, discourses amongst park rangers and wardens linked threats to the safety, sovereignty and economic prosperity of Uganda with the 'insecurity' emanating from the Eastern Congo.
In focus group discussions, many UWA rangers reproduced this threat discourse, further referring to the national and international media presentation about the killing of tourists in BINP as 'the jungle massacre' and 'a horrific act of slaughter' (The Independent, 1999;BBC, 1999). Despite official investigations concluding that the armed group belonged to 'Rwandan Hutu rebels' who sought revenge for the Ugandan backing of the Tutsi-dominated Kagame government and the supporting governments of USA and UK (HRW, 1999), rangers recounted the killings as evidence for, as one condensed it, "the barbarity and lawlessness of the Congo". This depiction dehumanized Congolese people and generalised them into a vague, impersonal object -'the Congo' -that needs to separated and kept out of Uganda.
This 'Othering' discourse led to a mental segregation between Congolese and Ugandan rangers, on the one hand, while further manifesting the spatial segregation of the transboundary protected area, on the other. Underpinning their perception of the Virungas as a 'geography of fear', UWA rangers repeatedly pointed out the invisible borders to the DRC within the forest that demarcated no-go-zones for them. One UWA ranger explained why it was vital to identify these invisible borderlines within the forest: "We just move along here with GPS. That tells you: stop, now you are close to Congo. When you get there, they jail you.
[…] They mistake you for a rebel." Numerous fear-related stories about the consequences of accidently stepping on Congolese soil were shared between rangers and informed the movement of the patrols in MGNP on a daily basis, underlining the impossibility of joint conservation practices between the two states (see Fig. 1) Several insurgencies in other border-adjacent National Parks in Uganda led the Museveni government to discursively link conservation tasks with national security duties (Ashaba, 2020: 8), resulting in the introduction of 'under-cover' military personnel from the Uganda People's Defence Force (UPDF) into ranger staff since the 2000s. Their 'undercover' status refers to their being indistinguishable for tourists, as the soldiers live together with rangers and perform the same conservation tasks including tourist guidance, gorilla monitoring and patrols. While this military-conservation partnership in Uganda has historical roots informed by (post-)colonial legacies and the gradual militarization of society under the Museveni regime since 1986 (Ashaba 2020), fear discourses directed against 'the Congo' served UWA rangers at MGNP to make sense of their four to six month paramilitary training, the establishment of UWA's own intelligence unit and the state-mandated collaboration with the UPDF.
Due to their similar uniforms, rangers frequently raised their fears of being mistaken for the military when patrolling in proximity to the Congolese border, as one ranger expressed in the following way: "Those people [Congolese] don't ask, they flush you out [kill] before they look." Such fears further fuelled rangers' sense of 'need' for military training and UPDF support while the armies' mandate to enforce national security simultaneously mobilized rangers' sense of national responsibility for protecting park and borders against 'the Other'. This discursive framing of 'the Congo' as a threat simultaneously to national security and conservation evidences the functionality of fear in mobilizing ranger's physiological reaction of a fight-response. Thus, fear narratives circulating amongst rangers served to increase their acceptance of and commitment to violent responses as a form of defence of both self and nation. The construction of enmity against the DRC hence substantiated mental and spatial segregation between rangers from ICCN, RDB and UWA on the ground, enabled the military build-up of the parks and rendering transboundary collaboration among the park authorities impossible.

An insight from 'the Other' side
In the course of 2018, I observed that ICCN rangers have been largely absent from the trinational GVTC workshop that were mostly taking place in Rwanda. GVTC representatives as well as staff from RDB and UWA accused the PNVi management of being unwilling to cooperate since it receives international funding from a wide range of donors through its public-private partnership with the British NGO Virunga Foundation and is therefore not dependent on or interested in the benefits of transnational collaboration (see also Hsaio 2018). In contrast to these allegations, Congolese park rangers and park management personnel explained, in interviews at the PNVi headquarters in Rumangabo and their two offices in Goma, North Kivu, that they reject sitting together at one table with Rwandans or Ugandans for several reasons. Firstly, a common narrative amongst PNVi staff depicted the two neighbours as the cause of prevalent insecurity in Eastern Congo by perpetuating militia activities, illegal resource extractions and violence affecting the conservation efforts of PNVi. While accusations were also directed against the Congolese government for supporting and orchestrating (foreign) militia activities at the expense of conservation efforts, one managerial officer emphasized that primarily the absence of governance, law and order in North Kivu has been exploited by Rwanda in the past decade. ICCN rangers also expressed their grievance concerning the killing of more than 200 PNVi rangers in the past ten years seen as a consequence of this incited insecurity in the aftermath of the genocide in Rwanda.
Secondly, facing the threat of violent collisions with armed groups in their daily conservation duties, ICCN rangers and staff from the PNVi Air Wing blamed the governments of Rwanda and Uganda for financing and supporting a wide range of militias operating within PNVi, such as the Ugandan Allied Democratic Forces (ADF) and remnants of the Mouvement 23 rebellion. Despite the complex, constantly changing landscape of armed actors, international reports on the control areas of armed groups reinforce these statements about the militia presences (see e.g. International Alert 2010; Kivu Security Tracker 2021; OHCHR 2020).
Thirdly, staff from the PNVi management related that they hold the government of Rwanda indirectly responsible for the destruction of PNVi through illegal park activities practiced by the FDLR. Since the FDLR consist mostly of former Rwandan Hutu who established their stronghold after the genocide in Rwanda within the Congolese side of the forest, PNVi staff accused the Kagame regime of lacking political will for amnesty and repatriation of the militias. In addition, some members of the PNVi management expressed their resentment against the ongoing military interference by Rwanda in North Kivu (confirmed in reports by UNSC 2020; Kivu Security Tracker 2020), perceived as geostrategic interventions to maintain insecurity and chaos in Eastern Congo in order to facilitate the illegal exploitation and trafficking of valuable resources such as coltan, gold and timber. While these accusations have been supported by a number of international organisations, reports also point to the illicit exploitation networks of multinational companies in collaboration with the Congolese government and the Congolese army FARDC (International Alert 2010;Global Witness, 2018;Reuters, 2012;UNEP, MONUSCO and OSESG, 2015;UNSC 2020). Official exports alongside smuggling of minerals in concession with Congolese officials pass through and significantly contribute to the economy of Rwanda (tin, tantalum, cobalt) and Uganda (gold) (see e.g. Epstein, 2017;Vogel and Raeymaekers, 2016).
Overall, the perceptions of PNVi staff reinforce Ugandan and Rwandan blame narratives of 'the Congo' and shed light on the complex entanglements between state and non-state actors from the three countries in their geo-economic interests, the ongoing involvement of Uganda and Rwanda in the Eastern DRC and in prevalent violent conflict in the PNVi surroundings. In contrast to these formal and informal transboundary collaborations over resource extraction and supply chains, narratives of Congolese, Rwandan and Ugandan rangers underline persistent segregation in the ostensible conservation collaboration that render transboundary conservation policies and practices void of substance.

A scapegoat to justify 'green' violence
The third narrative unfolds in a scapegoating practice accusing 'the Congo' of 'poaching' activities within VNP and MGNP that allows rangers to rationalize the command for increasingly violent conservation measures against people entering the park illegally. Through scapegoating, a person or a group attributes blame for a problem to an external group in order to persecute and punish the supposed cause (Girard, 1979). Since the strict enclosure of MGNP and VNP, park authorities stated that the major problem in their conservation efforts is the increase in illegal park activities due to persistently high levels of poverty in park-adjacent communities who depend on basic resources from the park (Bush et al., 2010). These activities comprise the collection of water, bamboo or honey, logging as well as hunting small animals, an act that is referred to as 'poaching' in conservation law (CGIS/DFGI, 2018). However, studies have found that 'poaching' in VNP and MGNP is mostly carried out by unarmed local residents with self-built snares hunting for means of subsistence (Sabuhoro et al., 2020).
Nevertheless, Rwanda and Uganda expanded the law enforcement duties of rangers where punishments for 'poaching' range from public exposure and humiliation to imprisonment against people who illegally enter the National Park. While shoot-on-sight policies are not formally implemented in the conservation law and I found no evidence for such measures in Rwanda, Ugandan President Museveni declared them acceptable and imputable. During my fieldwork in 2018, I witnessed the killing of a hunter caught in MGNP and the concealment of the incident as well as the lack of criminal prosecution of the ranger responsible. Other forms of punishment I observed include the killing of hunters' dogs (justified as 'destruction of their poaching weapon'), confiscation or destruction of food resources, assault and public naming and shaming in front of family, children and the community. Carrying out these forms of 'green' violence against people violating conservation law is predicated on rangers' commitment to and/or compliance with national law enforcement policies and practices.
Given the permanent deployment of rangers in the headquarters or outposts in Uganda and Rwanda, many developed friendly relationships with local community members in spending their free time together. While some rangers preferred to stay in their barracks and avoiding contact with people from the park-adjacent communities, many others made a lot of effort to 'go slumming'. When taking me along, rangers explained thatin order to enjoy these social pleasures and be accepted by civiliansthey must not look like military and cannot talk about their work in public. They changed their uniforms to civilian clothes, going to the local bar or public-viewings for international football matches or actively participating in community activities, such as church meetings or playing team sports together. Nevertheless, most admitted that they suffer from their duty to violently punish local people who entered the park illegally and the social exclusion they experience as a consequence (personal interviews, Bugeshi, Muhabura, 2018).
As Mathys asserts, "Violence cannot be understood without understanding the discursive concepts that justify it for the actors" (2017: 467). Scapegoating appeared to help rangers resolve the personal cognitive dissonance consequent to enacting violent commands against community members they interact with privately. On anti-poaching patrols with rangers, I found that they utilized 'the Congo' as a scapegoat for 'poaching' activities that helped them to legitimize the need for these violent measures. Regardless of the reality that most law enforcement practices are carried out against people from their own communities, rangers' fear-induced stories about the barbarity of 'the Congo' served to conjure what Girard (1979) calls a 'surrogate victim' to distance their own communities from illegal park activities. Thus, rangers externalised the blame and could redirect the required aggression for violence onto this 'Other', impersonal 'Congo'. This scapegoating strategy further helped rangers to maintain their "perceived personal moral value by minimizing feelings of guilt over one's responsibility for a negative outcome" (Rothschild et al., 2012(Rothschild et al., : 1148, namely the act of punishing.
This discursive link between 'the Congo' and the environmental destruction of 'poaching' activities was further strengthened by narratives from park rangers' superiors justifying violent conservation measures. The chief park warden of MGNP exemplified how 'the Congo' is blamed for 'poaching' within the Ugandan side of the forest: "What makes the harder situation here is the transboundary nature of the park. Now we are not talking about poaching from the communities here, but we are talking about poaching from the insurgent forces from Congo. This poaching by these forces that move in and out 'cause they must survive […]. They are eating animals, they are cutting for shelter, they are cutting for wood, all those stories to survive." In addition, the security warden of VNP attributed poaching activities within the Rwandan part to the incapability of Congolese ranger patrols while simultaneously criticizing the transboundary nature of the forest that enables trespassing: "The political situation in the Congo affects the security of the whole region […] because for example the Mikeno sector [PNVi, DRC] is not well patrolled, which leads to increased transboundary poaching. Poachers find it very easy to cross the border from DRC side and get involved in poaching activities in areas along the border and many times cross borders to poach in Rwanda and Uganda." Since "scapegoating…enables a group to convert an anxiety into a fear, thus legitimising hostile utterances and actions against bodies 'outof-place'" (Haldrup et al., 2009: 42), park wardens' perceptions reinforce the fear of 'the Congo' amongst rangers, further legitimizing use of 'green' violence for conservation law enforcement.
This scapegoating practice of creating an external threat to the nation-state is a common military tactic, reducing potential barriers to carrying out violence by individuals (Watson, 1978). Rangers of VNP and MGNP are mobilized by this scapegoating that makes 'the Congo' the external threat for the protection of flora and fauna that ultimately legitimizes the use of violence in defence of the national park and borders. Consequentially, rangers' attitude to collaborate with their counterparts across the borders have deteriorated and undermined the institutionalised transboundary collaborative efforts made by GVTC. Ultimately, by denying GVTC any authority over their National Parks' management, the governments of Uganda and Rwanda retain a state monopoly on violence through their parastatal conservation authorities that contradict the transnational peacebuilding intentions of a TBPA.

Conclusion
This paper has examined how fear functions to legitimize modes of 'green' violence, militarization and segregation of the three National Parks that were intended to form the transboundary Virunga Conservation Area between the Democratic Republic of Congo, Rwanda and Uganda. Since state-cultivated discourses in the form of public rhetoric and media accounts inform everyday realities on the local scale, I explored three fear narratives common among park rangers in Uganda and Rwanda that depict perceptions of a 'geography of fear', practices of 'Othering,' and scapegoating, respectively. These interrelated narratives shed light on rangers' understanding, enactment and sense-making of their militaristic and violent conservation duties, mobilizing their sense of responsibility and necessity to protect both national park and borders. The resultant mental and spatial segregation between the three National Parks undermines attempts to foster transboundary collaboration to institutionalize the Greater Virunga Transboundary Collaboration. In this regard, I argue that analyzing the link between fear discourses and 'green' militarization practices on the local level illuminates the complex and concealed entanglements between conservation and (geo)political interests on the transnational scale. In the Virunga context, agents of the Ugandan and Rwandan states instrumentalize rangers' fears to justify the militarization of the borderland under the guise of mountain gorilla conservation to stage the transboundary forest as a military buffer zone and bulwark to 'the Congo' deemed to serve cross-border military interventions in the context of the genocide in Rwanda and its aftermath.
The study has also demonstrated that an interdisciplinary approach combining critical geopolitics and emotional geographies can advance debates on transboundary conservation dynamics by introducing the analysis of fear discourses and their geopolitical instrumentalization to generate a deeper understanding of why 'green' violence and militarisation unfold and persist in many parts of the world. Building on this analysis, I suggest that more nuanced research concerning the role of emotions beyond fear in contemporary conservation efforts could benefit the critical conservation literature in general. Therefore, I advocate grounding critical discussions of conservation practices in the emotional realities of those who feel and enact these practices in their everyday lives, in order to include their voices in efforts to create alternative and non-violent pathways for conservation in the future.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.