Elephants, people and conservation: re-examining conservation interventions and human-elephant interactions through multispecies ethnography in the so

Background: In Tanzania and Kenya, past and present conservation endeavours have been focused on managing human-wildlife interactions (HWI) through exclusionary practices and delineating boundaries between spaces for people and wildlife (Barua, 2014). Such boundaries are reproduced through the establishment of protected areas of various form (wildlife corridors, community-based conservation areas etc.) that essentially alter the relationship that local people have with their environment and wildlife and represent 'nature' as separated from people (Khumalo & Yung, 2015). Often these are established excluding agricultural-based communities, livestock, and related livelihood activities, and entail legislative changes that regulate local communities' behaviours and cultural practices (Massé, 2016). Further, the socio-economic and political landscapes resulting from such policies and practices, and the institutions designed to manage them, drive conflicts between people and wildlife (HWC) (Anthony et al., 2010). Wildlife, particularly wide-ranging species, regularly cross institutional and physical (e.g. electric fences) boundaries coming into conflict with rural communities, threatening people's wellbeing and livelihoods (Gupta, 2013). Additionally, the creation of boundaries between people and wildlife contributes to negative perceptions of local communities towards charismatic megafauna, which are seen as having a political, economic or land-use advantage over local people (Lee & Graham, 2006). Conflict mitigation strategies have been mostly concerned with expanding protected areas, wildlife deterrents and fences, translocations of animals and relocation of people, selective culling, and monetary compensation for losses inflicted by wildlife to people (e.g. elephants trampling livestock or injuring/killing people) (Lee & Graham, 2006). All of these measures are based on the idea that wildlife does not belong in human settlements and on the continuation of a reproduction of wilderness (de Silva & Srinivasan, 2019; Boonman-Berson et al., 2016).

In recent years, along the southern Kenya and northern Tanzania borderland landscape, both countries have established a growing network of wildlife corridors and community-based protected areas to allow for wildlife movement between countries and national parks. While these conservation interventions are depicted as "win-win" solutions that generate benefits for local communities, studies show that these are unlikely to generate net economic gains for local people (e.g. Noe & Kangalawe, 2015; Benjaminsen & Svarstad, 2010; Igoe & Croucher, 2007). Instead, these often resulted in the loss of customary land rights and disempowerment of local communities, and consequently produced more opposition to conservation. Anger and frustration of local communities caused by these circumstances often translate into retaliatory killings of wildlife as forms of subordinate resistance towards the government's regulations (Benjaminsen & Svarstad, 2010; Holmes, 2007). As such, approaches that demarcate separate areas for nonhuman nature are not only problematic for their social impacts but also fundamentally flawed, as they essentially alter the longstanding relationship between people and wildlife and their ways of coexisting (de Silva & Srinivasan, 2019). Simultaneously, conservation interventions that aim to constrain wildlife to bound spaces proved inadequate, since wide-ranging species continues to trespass into human spaces facing the danger of being killed as an act of retaliation, in self-defence, or as "problem animals" by local wildlife management authorities (de Silva & Srinivasan, 2019; Benjaminsen & Svarstad, 2010).