Situating ethno-territorial claims: dynamics of land exclusion in the Guarayos Forest Reserve, Bolivia

ABSTRACT Indigenous people have effectively employed ethno-territorial belonging as a rationale to claim land. However, as sympathetic scholar-activists, we run the risk of assuming that such rationale necessarily empowers the dispossessed. I show that this is not always the case by examining past and contemporary land disputes in the Guarayos Forest Reserve in northeast Bolivia. I thus argue for the need to situate ethno-territorial claims in specific conjunctures and in full recognition of the exclusion's double edge as observed by Hall, Hirsch and Li (2011). Doing so, in turn, calls for greater attention to class differentiation within competing ethnic groups.


Introduction
On the 31st of March 2017, a violent clash between two groups of Guarayo 'indigenous' people made it to national news in Bolivia. These groups represented competing factions within the Central de Organizaciones de Pueblos Nativos Guarayos (COPNAG). Founded in 1987, COPNAG was created as an umbrella organisation to represent the interests of the Guarayo people and, in particular, to lead their struggle for territorial recognition. That unfortunate morning, in which dozens were injured, the group presided by Eladio Uraezathen COPNAG's Presidentwas attending a meeting organised by the state's agency in charge of overseeing land and forest management in the country. Officials were meant to socializar 1 the scope and implications of a draft decree that sought to modify the restrictions of a Forest Reserve located at the heart of the Guarayos province. By changing the land use prescription, the proposal largely aimed to legitimise 'illegal' peasants' settlements. Not an hour had gone by when the meeting was abruptly interrupted by the other group. Led by COPNAG's then vice-president, Daniel Yaquirena, this faction significantly outnumbered its opponent and it was not long until it had effectively occupied the coliseum where the socializacion was taking place.
In the following days, Yaquirena and Uraeza exchanged accusations of 'land trafficking' in the press, each claiming to be the true steward of the Guarayos' territory against avasalladores 2 (Rojas 2017). However, having retained power for almost a decade, Uraeza's credibility was already very low amongst the Guarayo people. In contrast, Yaquirena inspired hope of a positive change in the leadership. Taking advantage of the momentum, he called a 'big assembly' in order to decide the future of COPNAG. In a tense environment, Yaquirena decisively defeated his opponent, thus becoming the new President of the Guarayos indigenous organisation. The new leadership emphatically rejected the state's proposal to modify the Forest Reserve's restrictions. It demanded for the negotiations to start anew, but this time from an unwavering premise: within the Guarayos province, nothing can be done without the active consent of COPNAG as the legitimate representative of the Guarayo people, the only rightful and ancestral owners of the territory. As I will show, such usage of ethno-territorial belonging was not new in the province, nor was it necessarily intended to advance the interests of the majority of the Guarayo people.
As a rationale to claim land, ethno-territorial belonging has of course been employed the world over and throughout history. It has nonetheless acquired different forms and meanings in specific places and conjunctures. The association between ethno-territoriality and indigenous identity is of particular relevance here. It is a relationship that has been propelled by the surge of 'identity politics' since the 1990s (Hale 1997). A significant milestone in struggles for 'indigenous territories' is the drafting of the 169 Indigenous and Tribal Peoples Convention of the International Labour Organization in 1989. Article 13 of the Convention explicitly calls for the inclusion of the concept of 'territories' when considering indigenous rights to land (ILO 1989). In parallel, the 1990s was also a time when global environmentalist discourses advanced a notion of 'indigeneity' as intimately linked to environmental conservation. An essentialized indigenous 'way of life' was therefore presented as compatible and harmonic with biodiversity preservation (Gadgil, Berkes, and Folke 1993;Bowen 2000).
Nevertheless, indigenous groups are not simple receptors of external discourses. Images of indigeneity are always negotiated not least because their essentialism fails to concur with actually existing conditions (Conklin and Graham 1995;Postero 2013). Groups react differently depending on specific circumstances. Some indigenous groups may strategically position themselves within the 'tribal slot' (Li 2000) in order to form alliances with the actors that produce and accept such a simplified frame. However, images of indigeneity should not be viewed only as an imposition from the outside; 'locals' may have well participated in their co-production whenever these push ahead their own agendas and interests (Conklin and Graham 1995;Li 2000). These indigenous representations may indeed bring opportunities, but they also entail tensions and contradictions. For example, Valdivia's (2005) study in the Ecuadorian Amazon shows how alliances between indigenous groups and environmentalist organisations can work well and endure or else clash and burst when the latter's expectations regarding 'authentic indigeneity' are not met.
In Bolivia, external agents have also played a prominent role in mediating processes of ethno-territorial claims. For instance, Reyes-García et al. (2014) have showed that both the spatial and institutional configuration of the Tsimane' territory were heavily influenced by conservation NGOs, logging companies, and state bodies. The interests of these external actors have consistently prevailed over the Tsimane's claim to land and natural resources, regardless of the neoliberal or 'productivist' direction of state policy. Asymmetries of power were also documented with regards to other lowland indigenous peoples. Bebbington (2013) found that Guaraní and Weenhayek populations in the Chaco region have been facing a series of 'procedural inequalities' that effectively prevent them from defending their territorial claims vis-à-vis hydrocarbon companies favoured by the state. Most notoriously, the TIPNIS conflict 3 exposed the tensions and contradictions between the state's extractivist agenda and the consolidation of indigenous rights even when these are constitutionally recognised. Fabricant and Postero (2015) have situated this conflict on a longer history of extractive industries and racialised relations in Bolivia's lowlands whereas Canessa (2014) emphasised that the TIPNIS case laid bare competing discourses of indigeneity in the country. He argued that an ecumenical discourse of indigeneity as a basis for state-building clashed with that of the lowland indigenous groups who demanded respect for cultural difference and protection from the state. It is in this latter sense that indigeneity has been connected with the notion of 'territory'.
Indeed, as argued by Tania Li (2010, 385), 'an important feature of indigeneity in most definitions is the permanent attachment of a group of people to a fixed area of land in a way that marks them as culturally distinct'. This attachment is usually conceived in communal or collective terms. Consequently, the consolidation of collectively held territories has become emblematic of indigenous struggles. Collective landholding is understandably seen as a foundational pre-requisite of hunting and gathering systems. The creation of the Original Communal Lands (TCOs) in lowland Bolivia has followed this rationale (Valencia García and Égido Zurita 2010; Anthias and Radcliffe 2015). Yet, whether or not indigenous livelihoods continue to be based on hunting and gathering systems is something that needs to be demonstrated rather than assumed from an ahistorical essentialist perspective.
Inhabiting the Amazonian rainforests of northeast Bolivia, the Guarayo people were once nomad communities of hunters and gathers. By the end of the nineteenth century, however, their socio-political and economic organisation changed radically. Franciscan missionaries had effectively 'reduced' them to permanent settlements establishing a feudal-like labour regime. This system remained in place until the secularisation of the missions in 1938-39 when power was transferred to a local state authority called 'the delegate'. In practice, however, a similar system of labour control was installed. The state administrator was entitled to use Guarayos' labour in the productive activities he deemed necessary, to arrange and/or inspect their labour contracts with others, and even to collect their salaries in order to oversee their 'fair distribution'. 4 While the system was abolished in 1947, its enduring legacy was the formation of a new local elite of civil administrators who first turned into the local merchants and gradually became ranchers as they came to control large tracks of land (García Jordán 1998, 2007Pereira 2013).
By the 1970s, as will be discussed below, the region became a major site for timber extraction and the Guarayos' economic and political prospects partially improved due to the repercussions of the 1953 Agrarian Reform (Pacheco 1998;Vallejos 1998). However, it is in the last decades that the once isolated and scarcely populated region 5 has been radically transformed by the expansion of agrarian capital. In fact, it has arguably become the most dynamic and conflictive indigenous frontier in Bolivia (Vadillo 2009). Well-connected by the main road Santa Cruz-Trinidad, the dynamism of the Guarayo province has attracted the most diverse group of people, from landless peasants of distant arid regions in the country's highlands to big soy farmers coming from neighbouring countries. Marked ethnic and class differences are discernible amongst this highly complex constellation of actors, although their competing interests and contradictory agendas are much more difficult to untangle. At its core, disputes in the province are about who gets to benefit from land and who is excluded. 6 In this article, I examine past and contemporary land disputes in the Guarayos region of eastern lowland Bolivia. Specifically, my analysis is focused on the dynamics of land exclusion derived from the restrictions imposed by the Guarayos Forest Reserve that was created back in 1969. I pay particular attention to the people who have been excluded and the changing usage of ethno-territorial claims. I track the Forest Reserve's exclusionary effectsand the reactions to thesein two different conjunctures. The first one is marked by the expansion of timber exploitation during the 1970s under the auspices of Hugo Banzer's military ditactorship. The second one corresponds to the period of rapid frontier expansion and settlement supported by the MAS government of Evo Morales, particularly from 2013 onwards.
By comparing the two conjunctures, my analysis shows that the very same land use restrictions can lead to very different ethno-territorial claims. It reveals that the Guarayo people have articulated distinct ethnic identities in tune with particular constellations of power. I therefore argue for the need to clearly situate ethno-territorial claims in particular conjunctures. The case analysed here also demonstrates the salience of recognising exclusion's 'double edge' (Hall, Hirsch, and Li 2011) when examining land disputes. Doing so helps to put in sharp focus the key question of who benefits and who loses. To adequately pinpoint the who of exclusion, my empirical evidence stresses the need to pay greater attention to class differentiation within competing ethnic groups.
In the next section, I outline the theoretical framework guiding the analysis which is based on the 'powers of exclusion' as put forward by Derek Hall, Philip Hirsch and Tania Li (2011). I then turn to analyse the case of the Guarayos Forest Reserve during the aforementioned conjunctures based on year-long fieldwork conducted in Bolivia between September 2017 and August 2018. I spent ten months living in this region and the remaining two visiting archives and key informants in the cities of La Paz and Santa Cruz de la Sierra. Data for the analysis include 94 semi-structured interviews with leaders, state officials and people representing different classes which were triangulated with participatory mapping and observation. In the concluding section, I wrap up the argument by discussing the article's main contentions in the light of the case presented.

Theoretical orientation
Allusions to land conflicts are ubiquitous in studies of the Guarayos Province. Yet, the actual dynamics at work have not been at the centre of most analysis. Available literature has been chiefly focused on forest governance (e.g. Vallejos 1998;Martínez 2002), the historical trajectory of the Guarayo people (e.g. García Jordán 1998; Pereira 2013) and the consolidation of their indigenous rights (e.g. Acuña 2006;Martínez 2010). An exception is the book chapter by Vadillo (2009) that features a brief but rather informative case study of land conflicts in the southern part of the province. As a former Director of the National Institute of Agrarian Reform (INRA), Alcides Vadillo was acutely aware of the particularly volatile and conflictive situation in the province where, to use his words, '[land conflicts] have become permanent and systemic, any difference is motif of clashes and confrontation ' (2009, 181). While his conclusions reiterate the rather common 'solution' of strengthening property rights, he makes an intuitive but important point: 'When it comes to land access, recognising the [property] right of someone means denying that of other' (Vadillo 2009, 139).
A similar observation underpins Derek Hall, Philip Hirsch and Tania Li's (2011) approach to land exclusion. Given that 'all land use and access requires exclusion of some kind', exclusion is truly inevitable, an intrinsic element of land relations (2011,4). Conceptualising exclusion in this manner highlights the fact that it simultaneously creates security for some and insecurity for others. Exclusion is therefore a double edge sword. Whose claim should prevail is essentially a question of political perspective with different ideologies justifying their own type of exclusion as positive. From the tradition of critical agrarian studies, taking sides in situations of massive land dispossession by speculative agribusiness capital is fairly easy. It gets trickier though when contradictory land claims are made by groups considered part of the democratic agrarian transformation. Typically, ethno-territorial claims clash with others made in the name of landlessness or livelihood entitlement. Indeed, as Hall, Hirsch, and Li (2011) point out, exclusion's double edge necessarily entails troubling dilemmas difficult to solve. In Bolivia, tensions between lowland 'indigenous' peoples and highland 'peasant' migrants present such dilemmas, as I note in the concluding section of the paper.
Exclusion, moreover, is structured by and reflexive of power relations. It does not come about naturally or randomly. Therefore, unlike Vadillo's (2009), Hall, Hirsch, and Li's notion of exclusion is not limited to the concept of property as 'it refers not just to the presence or absence of rights but to the broader array of powers that prevent people from benefiting from land ' (2011, 8). Consequently, they propose to explore land exclusion as the result of a complex interaction between regulation, force, the market and legitimation, four 'powers of exclusion' that can be succinctly conceptualised as follows: Regulation, often but not exclusively associated with the state and legal instruments, sets the rules regarding access to land and conditions of use. Force excludes by violence or the threat of violence, and is brought to bear by both state and non-state actors. The market is a power of exclusion as it limits access through price and through the creation of incentives to lay more individualized claims to land. Legitimation establishes the moral basis for exclusive claims, and indeed for entrenching regulation, the market and force as politically and socially acceptable bases for exclusion. (2011,5) Although in practice these 'powers' are inextricably fused, their separation is a useful heuristic move to tease apart the central elements at work in a given exclusionary regime. Concrete investigation of their intersection, however, demands a focus on historically specific processes and actorsthat is a 'conjunctural approach' of the type developed with unusual sophistication by Antonio Gramsci (1971). Such analysis requires examination of 'multiple strands of interlinked historical trajectories and how they come together' as well as 'unpacking complex, mutually constitutive relations of conflict and cooperation over history' (Peluso 2011, 815). It is an approach that has been skilfully applied to land dynamics by scholars such as Tania Li (2014) and Nancy Peluso (1992Peluso ( , 2011. As noted already, I submit that competing land claims can only be understood and fully appreciated when these are clearly situated in specific conjunctures. For one thing, such an approach demands analysing competing groups as historically situated and socially determined, hence avoiding their essentialization. It highlights social differentiation along the lines of class, gender, race, among others; a point brought home by agrarian scholars working in the Marxist tradition (O'Laughlin 2009; Bernstein 2010; Lerche and Shah 2018). Competing groups should not be considered homogenous unitary social formations, but complicated contradictory assemblages of socially differentiated subjects.
Guided by the empirical evidence, I emphasise the importance of class differentiation to properly explain the who of exclusion. Ethnic divides figure prominently in analysis of land conflicts, particularly in indigenous frontiers. Yet the class differences within the competing ethnic groups are usually overlooked or downplayed. This limitation has been highlighted by prominent scholars in critical agrarian studies. For instance, Henry Bernstein (2014) has shown that treating 'indigenous communities' or 'peasants' as internally coherent social categories is highly problematic analytically. He argued that this often leads to bracketing together diverse and conflicting interests which in turn undermines the construction of viable political programmes. Similarly, Tania Li (2010, 396) has lamented the 'reluctance to examine microprocesses of dispossession' amongst indigenous people and small-scale farmers. She contended that failing to recognise processes of class differentiation within broad identity-based groups impoverishes the debate on agrarian capitalism. This issue has also been framed as the question of the inherent 'partial representation' of agrarian social movements by scholars such as Saturnino Jun Borras, Marc Edelman and Cristóbal .
Attention to class differences is indispensable to explore dynamics of land exclusion. This is so because land disputes invariably take place within a broader structural process of accumulation of capital and resources. Understood in a materialist sense, the class divide largely determines the resources and the specific strategy that a given group is able to mobilise to exclude others. In other words, the differentiated access to means of production conditions agency. In this regard, as noted by Akram-Lodhi (2012, 613), the 'regulatory, market, coercive and legitimation powers of exclusion have strong class dimensions and are strongly expressive of class relations'.
The origins of the Guarayos Forest Reserve: military governments, loggers and the Guarayo 'peasants'

Containing colonos' expansion
The Guarayos Forest Reserve was established in 1969 during the military government of President René Barrientos. An ambitious and charismatic air force officer, in 1964 Barrientos led a military coup d'état against incumbent Victor Paz Estenssoro, the leader of the Movimiento Nacionalista Revolucionario (MNR), a nationalist political party that had ruled the country since the 1952 National Revolution. Barrientos' coup marked the end of the emancipatory period known in Bolivian history as 'Revolutionary Nationalism' (1953)(1954)(1955)(1956)(1957)(1958)(1959)(1960)(1961)(1962)(1963)(1964) and inaugurated instead a succession of military dictatorships that lasted until the early 1980s. 7 In this period, a new strict top-down approach in development planning emerged of which the creation of the Guarayos Forest Reserve is an example. Imperious in tone, the supreme decree for its creation stated that 'colonos settlements of any kind are strictly forbidden as is cutting trees or forest clearance for agricultural purposes in the entire geographical extension thus delimited' (Decreto Supremo No. 8660 1969).
Despite the universalistic prose, this new regulation targeted a particular kind of 'colono'. To the south of the area, groups of immigrant peasants had been settling over the last decades and the restriction was meant to contain their further expansion. Most of these immigrants had come from the western highlands after the 1953 agrarian reform. Given labour shortages in the lowlands and the increasing scarcity of land for the growing population in the highlands, this demographic movement was supported by the MNR government as a win-win solution. Indeed, it was considered a necessary condition for the development of a modern capitalist agricultural sector in the lowlands (Flores 1953;Sandoval Arenas 2003).
By the late 1960s, however, Santa Cruz's landed elite was increasingly worried about the mid and long-term effects of this migration. Largely composed of Spanish descendants or mestizos, this class of hacendados controlled nearly 43% of the farmland in Santa Cruz by 1950. In contrast to their counterparts in the highlands, their powerful position was not significantly affected by the 1953 Agrarian Reform as most of them managed to convert their properties into 'agricultural enterprises' (Eckstein 1983;Kay and Urioste 2008). While they initially welcomed the availability of extra labour, they quickly became anxious about the fact that migrant peasants aspired to become independent small farmers and not just wage labourers. This hacendado class feared an eventual challenge to their hegemonic position. Consequently, they fostered a regionalist and racist narrative to portray immigrant peasants as a threat to 'order and peace' and did not hesitate to call on the state to intervene on their behalf to 'contain' their expansion.
In creating the Guarayos Forest Reserve, Barrientos was partially answering the hacendados' demand for 'order' in the region. But his government also had its own interest in administering the region's forest riches. Reports of local merchants turning into prosperous loggers had already reached the ministries in La Paz. Not only was the region unusually well-endowed, but there was already a rudimentary road in place and international timber prices remained attractive. Therefore, controlling these forest resources became a priority for the government and the creation of the Reserve was legitimised in the name of a modern, efficient and orderly extraction, or as stated in the decree: 'the zone constitutes one of the economic areas better qualified in terms of forest resources in the Republic … its industrial exploitation ought to abide by scientific and administrative norms that preserve and contribute to its increase and profitability' (Decreto Supremo No. 8660 1969).

Reserved for loggers
However, the restrictions imposed by the Guarayos Forest Reserve only started to be enforced during the long and vicious military dictatorship of President Hugo Banzer . Coming from a hacendado family, Banzer largely used his power to advance the interests of the lowland traditional landed elite (Dunkerley 1984). During his government, this agrarian class came to control 3.8 million hectares including hundreds of huge properties above 10,000 hectares each (Soruco 2008). Nevertheless, Banzer did continue encouraging the migration of highland peasants as their labour was still required by a landed elite that had not significantly mechanised their fields. The hacendados' dilemma remained unsolved: bringing in more agricultural labourers was necessary, but also potentially dangerous. With Banzer in the presidency, however, they trusted the state could contain and manage peasant settlements or, as they usually put it, 'impose order'. They also took comfort in the fact that, in case of conflicts and tensions, the state will unequivocally defend their interests over those of peasants.
Banzer's approach is clearly reflected in the Guarayos Forest Reserve case. In 1974, his government promulgated two relevant decrees. The first one authorised the expansion of the peasant settlement of San Julian, located about 100 km to the south of the Reserve. This settlement had started in 1971 when the first 120 families were established thanks to a project organised by a civil organisation called 'United Churches Committee' and the recently created National Institute of Colonization. Branded as successful, a second phase of the project secured financial support from Germany and USAID and had the aim of settling another 5000 peasant families until 1980 (Hess 1979). To accommodate this expansion, Banzer granted an extra 752,833 hectares part of which overlapped with the Guarayos Forest Reserve, de facto reducing its size. The second decree enacted the first General Law of Forestry in the country. It is with this law that the Guarayos Forest Reserve was consolidated, and its restrictions became effective. The law created the Centro de Desarrollo Forestal (CDF) as the state office in charge of overseeing, regulating and executing forestry activities. The CDF was given the power to authorise timber exploitation both within state land and private properties, as this new law separated land rights from those over the forest. Although the law declared forest resources as public patrimony, these were effectively put into private hands. Indeed, concessionary contracts were exclusively allocated to private companies. Most of these companies were owned either by powerful loggers already operating in the region or by other members of the landed elite turned forestry 'entrepreneurs' due to their close connections with Banzer's administration (Pacheco 1998).
The Guarayos region was one of the first where the new policy was implemented. Wellendowed and with restrictions against 'colonos' already in place, the region was particularly attractive to loggers. Using political influences, twenty companies of well-connected families quickly obtained their concessionary contracts and came to control nearly the entire area of the Guarayos Forest Reserve. 8 As shown by Vallejos (1998), these logging companies carried out a disorderly and highly selective extraction. The search for timber was erratic, destructive and largely inefficient. It was not the sort of technically-sound rational exploitation that the creation of the Guarayos Forest Reserve was supposed to bring about. Its exclusionary effects, nevertheless, were very real. Indeed, loggers systematically relied on the restrictions set for agricultural activities to exclude peasants. They did not hesitate to intrude and extract timber from smallholders' plots without any consent, let alone compensation.
Guarayo 'peasants': exclusion and reaction Yet the excluded peasants were not migrants. The closest settlement of immigrant peasants, San Julián, was only beginning to consolidate and families from there would only go further north by the early 2000s. Although the Reserve's prohibitions were supposed to affect 'colono' settlements, those effectively excluded were Guarayo smallholders. Loggers and Guarayo people increasingly ended up in land conflicts as forest extraction intensified. With the promulgation of the General Forestry Law, loggers enjoyed complete state support for their claims and so approached Guarayo small farmers with arrogance and irreverence; as one of them recalled: 'Companies at that time simply advanced over the forest … Of course, [they claimed] everything was part of their concession, the entire Guarayos region. The concession might had been theirs, but the plots belonged to the Guarayos' (Fieldwork notes 28/04/2018).
The reverberations of the 1953 agrarian reform, however, had modified the situation of Guarayo smallholders. On the one hand, it enabled them to engage in independent agricultural production by effectively challenging the socio-political and economical system inherited from the Franciscan missions. More concretely, the agrarian reform provided the basis for people organising in sindicatos campesinos (peasant unions). Under the modernist project of the MNR, indigenous people were invited to think of themselves as 'peasants' and to abandon 'indian' or 'native' identities considered backward and antimodernist (Albó 2008;Postero 2013). While culturally homogenising, this legitimised self-organisation in the region as never before. Guarayo 'peasant' unions emerged in the 1960s and became stronger in the mid-1970s as a response to loggers' abuse. In as much as Guarayo smallholders began to articulate the identity of their interests in opposition to those of the loggers, they started to form a class (Thompson 1966). In this particular conjuncture, this emerging class employed ethno-territorial belonging as rationale to defend their land claims, as eloquently described by an old leader of the Guarayo 'peasants': There was this arrogant owner of the Santa Maria company. He was another one threatening people with his gun in order to exploit timber … . He had chased people with bullets, but we stopped him. I was a leader back then, the Agrarian Inspector for Peasant Justice, 9 and he came to my zone of San Gregorio where we had bunches of good standing timber. I told him: 'mister this is our timber our grandfathers kept this forest for us. This sector is called Guarayos for a reason. You can take the timber, but you need to pay for each log or else improve our lane, that is what the law says' and then I showed him the Law of Agrarian Reform. In the past we did not have that sort of document in our hands. Then I told him: 'What you have in your hands [the General Forestry Law], here it is, I also have it' And so I left him without words … . That day I confronted him with my [support] base. We had only few old rifles, but everyone was with their tools, carrying machetes and shovels. (Fieldwork notes 19/05/2018) This telling testimony unveils the different 'powers of exclusion' at work in this particular conjuncture. Force was violently exercised, particularly against Guarayo farmers, and the threat of force was central to the strategies of both sides. Interestingly, the testimony also points to the class character of state regulations. The General Forestry Law was for loggers what the Agrarian Reform Law for Guarayo 'peasants'. While both could find a particular piece of legislation to validate their claims over the land, only the Guarayos were able to articulate a discourse of ethno-territorial belonging to further legitimate their position: 'our grandfathers kept this forest for us. This sector is called Guarayos for a reason'. The latter became an important factor of unity amongst Guarayo smallholders who gradually built up an umbrella organisation representing all their 'peasant' unions, the Federación Especial de Trabajadores Campesinos de los Pueblos Guarayos (FEDTCG).
Loggers, on their side, sought to further reinforce their claims by appealing to the state. As members of Santa Cruz's landed elite, they had direct influence over Banzer' administration. Indeed, the state was quick to react, and a new decree was enacted. As if referring directly to the latter conflict, the official document 'denounced the clearance of parcels for agricultural and colonization purposes', by people who claimed 'protection from authorizations given by the Agrarian Reform' (Decreto Supremo No. 12268 1975). Through this new decree, the government annulled any competing documents within the Guarayos Forest Reserve, recognising only the concessionary contracts. In a rather blatant manner, the government made it clear that the zone was set aside for the benefit of private companies only.
However, Guarayo 'peasants' were already organised and in a much stronger position. Despite Banzer's latest decree, loggers could no longer extract timber at will and had to enter into agreement with the FEDTCG. For example, FEDTCG archives include a 1977 contract between the Federation and the representative of the company operating in the highly endowed area known as 'La Chonta'. The Federation was acting on behalf of a group of Guarayo 'peasants' residing in the proximities. The company agreed to pay a total of Bs. 110.000 and to build a road from the site of extraction to the main town in exchange for all the Mara trees (Swietenia macrophylla King) found in the peasants' plots. The money was to be administered by the Federation and entirely used in the construction of a secondary school. The latter was indeed built and was still functioning at the time of my fieldwork.

Old restrictions, new conditions
The boundaries of the Guarayos Forest Reserve have not been altered since 1974. Over its 1,087,124 hectares, the old restrictions to settlements and agricultural activities remain formally in place. Yet, over the last two decades, its landscape has changed radically. Timber extraction ceased to be the main economic activity as gold mining and agriculture have become increasingly important. Agriculture, in particular, has expanded rapidly in spite of the legal restrictions. According to official geo-referenced data, timber extraction is practiced in only 12% of the Reserve whereas around 26% of its forest has been cleared for agriculture (ABT and INRA 2018). Indeed, the region has become one of the country's most dynamic agricultural frontiers receiving an unprecedented influx of big and small farmers. Most of the newcomers came from soy-producing areas located to the south. Demographic data clearly reflects the latter. According to the latest census, the population in the Guarayos province has increased from 31,577 people in 2001 to 48,301 in 2012 with an annual growth rate of 3.10%the highest of all the rural provinces in the Santa Cruz department (INE 2001(INE , 2012. This new conjuncture was also shaped by important changes in Bolivia's land legislation during the 1990s. 10 Three reforms are of particular relevance here. First, Banzer's General Forestry Law was replaced by a new one that established stricter control over the activities of logging companies and introduced a universal fee of one dollar per hectare conceded by the state. Forest companies reacted to the new legislation either by leaving business or massively reducing the area under their control. In the Guarayos Forest Reserve, logging concessions shrank by approximately 70% and, as a result, large swathes of land returned to state control (Vallejos 1998).
Second, a new type of communal property called Tierras Comunitarias de Origen (TCO) was legally recognised. Simply put, TCOs are ethnic territories collectively titled in favour of indigenous organisations. The creation of TCOs was largely the result of the historical struggle of lowland indigenous people who in 1990 had staged an emblematic march for 'territory and dignity' of more than 1000 km towards the city of La Paz (Valencia García and Égido Zurita 2010). Yet, as argued by Fabricant and Postero (2015), neoliberal governments incorporated the TCO demand into what has been described as 'neoliberal multiculturalism'thus attaching particular essentialised images of indigeneity to it. In August 1996, COPNAG submitted a legal demand for a TCO over an area of 2,194,433 hectares, virtually all the jurisdiction of the Guarayos province. Based on a study of 'spatial needs', the state has titled an area of 1,352,152 hectares in favour of COPNAG (INRA 2018).
Third, the state began a process to 'regularise' land property rights throughout the country with a technical and legal procedure that became known as saneamiento. The term is perhaps best translated as 'healing the records', a figure of speech that aims to convey a reboot of land administration. The implications for areas declared as 'Forest Reserves' were as follows. The saneamiento process could only recognise as legaland thus eligible for land titlingproperties that fulfil two requirements: (1) being established before the promulgation of the new land law in October 1996; and (2) corresponding to indigenous and peasant communities or small proprietors. 11 Medium and large-scale properties should thus face size reductions. The resulting expropriated landcolloquially called recortes (cut offs)should return to state hands. Importantly, within 'Forest Reserves', all state land is automatically categorised as 'unavailable' for redistribution (Acuña 2006).
More recently, the land policy of Evo Morales' administration has also had an important if contradictory impact. During his first mandate (2006)(2007)(2008)(2009), the state decisively pursued the consolidation and expansion of TCOs throughout the country. In the case of the TCO Guarayos, for instance, seven of the ten collective land titles were issued by his administration. His initial policy also modified the process of saneamiento so as to maximise land expropriation of middle and large properties in favour of peasant and indigenous communities. Over the years, however, the MAS government has veered from its initial intentions. After years of confrontation with Santa Cruz's landed elite, both sides sought a rapprochement since around 2010. The new pact was formally staged in 2013 when Evo Morales was invited to inaugurate Santa Cruz's international business fair 'Expocruz'. This was highly symbolic because Morales had not been invited before, despite the fact that it was customary for the country's president to open the fair each year. Ever since, the promotion of agribusiness expansion has become the cornerstone of the MAS agrarian policy in the lowlands (McKay and Colque 2016). This put the government in a contradictory position when trying to mediate processes of agrarian change in indigenous frontiers as noted below.

Certifying 'unavailable' land
The saneamiento within the Guarayos Forest Reserve has been riddled with conflicts. By 2018, 66% of the land had been titled (i.e. 717,961 hectares) and the other third was pending. Of the titled area, 169,013 hectares were declared as 'unavailable' state land (INRA 2018). Land disputes were increasingly revolving around the control of this state land which was seen as vacant by competing actors despite the fact that it was illegible for redistribution. In relation to this issue, the government found itself in a contradictory position. While it tried to enforce the land use restrictions legally attached to the Forest Reserve, it simultaneously supported agribusiness expansion towards the region. In other words, it regulated to keep state land out of the market, but, at the same time, it provided the incentives for its commodification.
Nonetheless, most of this 'unavailable' state land had already been commodified. Contradicting romanticised images of indigeneity, it was COPNAG leaders who had put it on sale. Claiming to be 'the real owners of the territory', they devised a parallel system for land allocation in the province. This alternative system mimics the official one. New settlements are first authorised by a COPNAG Board Resolution. Each member of the new settlement is then issued a 'certification' thatjust like a land titledeclares the holder as the owner of the property and includes a detailed cadastral map. The second paragraph of these resolutions explains that authorising new settlements is part of COPNAG's policy to ensure that Guarayo people gradually occupy their own territory. However, in practice, settlements and certifications are only available to those who can pay for them. Those buying land from COPNAG have largely been non-Guarayo people, usually either capitalised ranchers/farmers seeking to expand their properties or self-identified 'peasants' of highland origin who arrived at the region in the hope of securing a small parcel of land. Paradoxically, COPNAG's land allocation system excludes the vast majority of the Guarayo population from accessing land. 12 In fact, in the mid-2000s, Guarayo people pushed their leaders to form a 'disciplinary tribunal' to investigate such land deals. The internal investigation revealed that nearly 175,000 hectares had been irregularly 'certified' by leaders in exchange for underhand payments. The accusations of 'land trafficking' by Guarayo leaders noted in the introduction refer to these corrupt practices.
Ethno-territorial belonging has been discursively used to legitimise COPNAG's system of 'certifications' and so to defy the state's right to administer land. COPNAG leaders claim sovereignty over land not only within the boundaries of the TCO but throughout the entire province, as Daniel Yaquirena explained: Our initial TCO demand was for two million hectares and only one million has been titled … . We might not be versed in laws, but our grandfathers knew the real limits [of the territory], they walked from one extreme to the other. Ancestrally, land titles did not matter. I am the president not only of titled areas, but of the entire Guarayos province. If we were to take the [Bolivian] state to an international court of indigenous people rights, we would win the demand. (Fieldwork notes 11/06/2018) The use of ancestral belonging as criterion for entitlement has proven to be powerful. With very limited material resources, COPNAG has become a recognised authority in the province. Almost invariably, new settlers seek COPNAG's consent from the very moment of arrival. They pay for the 'certifications' in the hope that these will serve as proof of recognition and inscription of rights by the indigenous organisation. Yet these 'certifications' do not yield property rights per se. They function more like 'permits' to occupy a given area. Although COPNAG's resolutions may authorise the 'definitive use of the land', there are no guarantees. In fact, as land becomes scarce and its prices go up, Guarayo leaders have started to evict the very people they themselves had 'certified' in the past. COPNAG found itself in a particularly strong position vis-à-vis other actors and the state because Guarayo leaders were able to further legitimise their claims by making connections between a conservationist image of indigeneity and the regulation of the Forest Reserve. They took advantage of inconsistent state agendas, poor maps, and fuzzy boundaries to argue that right was on their side. They were also capable of mobilising substantial force to enforce their system of land allocation.
Emblematic of this intricate conjuncture are the land disputes around the so-called 'recortes' (cut offs). As noted above, these are areas resulting from processes of land expropriation within the Forest Reserve which, given the established restrictions, are legally unavailable for redistribution. Recortes are particularly targeted by immigrant peasants to establish new settlements because these areas tend to be already cleared of forest and/or well-connected to roads. News of any freshly made recorte quickly circulate throughout the region, raising expectations and tensions. Unsurprisingly, this is source of great anxiety for the local landed elite who constantly complains that 'reverted or cut off land are invaded immediately by colonos' (CAO and FEGASACRUZ 2018). By 2018, there were a total of 101 recortes in the Guarayos Forest Reserve encompassing an area of 28,433 hectares (INRA 2018). The next section provides a case study of the conflicts in one of these areas.

Colla migrants vs indigenous locals? Land exclusion on the ground
The Comunidad 3 de Octubre is one of the 90 peasant settlements found in the Guarayos Forest Reserve (INRA 2018). Located 18 km from the province capital, this settlement consists of 35 immigrant peasant families. The average size of parcels is 33 hectares and only two families own more than one parcel. Small-scale cattle raising is the predominant activity and most families grow banana and mandioca for self-consumption. This land used to be part of a larger property owned by a big rancher of the traditional local elite. In 2009, the process of saneamiento reached the area and his estate was greatly reduced from 2300 to just 500 hectares, the maximum legal limit for small ranches. It is in the recorte thus createdof roughly 1800 hectaresthat these families of immigrant peasants settled.
The first to claim this land, however, were not the immigrant peasants but Guarayo people. In fact, the settlement was originally created as an 'indigenous community' by COPNAG's local branch. Although 24 Guarayo families figured as the beneficiaries, the settlement of those indigenous families did not materialise. Instead, only three of the alleged 'beneficiaries' took control of the land and quickly put it on sale. They divided the area into parcels of 50 hectares and offered each one for USD 3000. One of them had a position at COPNAG and the other two were her relatives. These people belonged to an emerging Guarayo elite that had used their leadership positions for personal gain. It is from this small group of people that immigrant peasants first bought the land. Standard purchase contracts were signed and COPNAG issued the corresponding resolution 'authorising the settlement and definitive use of the land' (COPNAG 2012, 1), for which the peasants were charged extra.
However, the rancher was not prepared to relinquish the land so easilynotwithstanding the results of the saneamiento. After all, in his eyes, the peasant settlement was illegal and illegitimate. Immigrant peasants were unwelcomed outsiders, collas 13 who should have no place in Santa Cruz. To chase away peasants, the rancher devised an intimidation strategy based on two fronts. A legal case was opened accusing peasants of cattle robbery and a series of brutal attacks were planned against the settlement. The first attack was led by the rancher himself who arrived at the settlement with a group of nearly one hundred people. He claimed to have the support of the 'real owners of the territory' (i.e. COPNAG leaders) and gave the peasants an ultimatum to either leave or face the consequences. As the gang left the settlement, they destroyed and burnt the communal house to prove they were serious. Peasants ignored the ultimatum and so the rancher prepared a second attack. This was led by one of the COPNAG's leaders that had sold land to peasants. Claiming to be the rightful owner of the land as a Guarayo woman, she engaged in a heated exchange of insults with the settlers before a physical clash between the factions took place. The settlers prevailed as they had received reinforcements from neighbouring settlements also in conflict with the same group of land dealers. Shortly afterwards, the outcome of the legal case was known. The allegations made by Suarez were rejected due to insufficient evidence. Defeated on both fronts, the rancher finally abandoned his attempts to evict peasants.
Taking advantage of the momentum, settlers decided to hit back. Three peasant settlements presented a formal joint action for 'land trafficking' against the Guarayo land dealers. An official commission from the INRA arrived at the area and found enough evidence supporting settlers' allegations and so recommended to initiate a legal case. While the report emphasised that the area in question constituted 'unavailable state land', peasants celebrated the visit and report as the final victory against the big rancher. They vowed to continue their struggle to lift the Reserve's restrictions and obtain land titles, but they were nevertheless relieved to have an official report confirming that neither the rancher nor the Guarayo land dealers have a legitimate claim on that land. It is interesting to note that peasants sought the state presence in the settlement regardless of their unlawful situation. Likewise, the state did not pursue any concrete action to evict them, although it ratified the disputed area as 'unavailable state land'. The latter reflects the contradictory position in which the government found itself trying to reconcile legal norms with the interests of its most loyal constituency in the region.
The calm in this peasant settlement did not last long. A new and more violent conflict took place by the end of June 2017. This time the dispute was not with a powerful rancher of the local elite, but with two brothers that had also arrived as migrant 'peasants' in search of land. They had settled in the neighbouring community of San Andres back in the early 2000s. In contrast to most immigrant peasants, these two people arrived with substantial capital to buy two parcels of land each and open shops of motorcycle parts in the main town. As their businesses flourished, they started to accumulate plots from their neighbours and gradually came to control most of the land in their settlement. Seeking further expansion, they devised a plot against settlers of 3 de Octubre and found powerful allies at COPNAG, including the then president Eladio Uraeza. On the 28th of June, the two brothers and Uraeza led a group of around fifty people into the settlement. Armed with rifles, chainsaws and gasoline, they started to destroy and burn residents' properties. Residents had little choice but to escape. As they ran away, Uraeza allegedly shouted 'run, go back to your towns fucking collas, this land belongs to us' (Fieldwork notes 12/06/2018). According to the official police investigation, twenty-two houses and eighteen motorcycles were burned and almost all cattle pens, corrals and tools were destroyed (Policía Boliviana 2017).
Five days later, the two brothers returned to settle a new group of people. The brothers had hired some people to guard their share of land, while others were newly arrived immigrant peasants thatjust like the residentshad bought the land from COPNAG leaders. Yet, having defended the land for more than five years, the residents were not planning to leave. Quite the contrary. When the new group reached the settlement, they found residents rebuilding their houses with the help of others from neighbouring peasant settlements. Violence quickly erupted and this time the band of the residents prevailed and took revenge by destroying all the motorcycles and supplies of their opponents. This conflict made it to regional news. The most important newspaper in Santa Cruz framed it as a dispute between 'locals' (the brothers' group) and highland 'migrants' (the residents), implying that the former were legitimately evicting the latter (El Deber 2017). In relation to the actual events, the story and this divide are fallacious but reflect the longstanding narrative against immigrant peasants.
After the violent clash, the dispute continued through other means. Both groups submitted legal actions against the other and staged demonstrations in the main town demanding justice. In contrast to the brothers, however, settlers of 3 de Octubre constituted an actual peasant union affiliated to the region's main peasants' organisation. Consequently, their demonstration was far more numerous as the organisation issued a formal call to peasants from all over the province. Having shown their strength, the 'Peaceful March against Arson, Land Trafficking and Discrimination', as it was called, enacted a resolution which opened: We have arrived at this warming town of noble people, searching better days for our families.
Here we have organised ourselves in peasant communities entering through the big door of the Guarayo organization COPNAG. However, we feel insulted by some indigenous leaders and ex-leaders who gave us with resolutions, certifications, and even notarized purchasing contracts, the parcels where we currently work inside state land … Nevertheless, our parcels are sold to other unwary people or greedy business men, and then they want to chase us away burning our houses and destroying everything we have worked for throughout this time. (CSUTC-AG 2017)

Conclusions
In the light of the case presented, in this concluding section I discuss the two central contentions of the paper: (1) the need to clearly situate ethno-territorial claims in particular conjunctures; and (2) the importance of paying greater attention to class differentiation within ethnic groups to accurately pinpoint the who of land exclusion.
Attention to context-specificity is certainly not in short supply within critical perspectives of development, particularly in the disciplines of anthropology and geography. Yet recognising what is 'distinctive' of a given place does not make for an integral analysis of specific conjunctures. This involves attention to the complex, contradictory and yet mutually constitutive relations of force and conflict unfolding in particular places and times. Conjunctures are the historical terrains in which disputes take place. As such, these largely shape and determine the meaning, form, and effects of competing land claims. By situating ethno-territorial claims in specific conjunctures, their actual form and effects become visible, concrete. Historicizing demystifies and allows approaching competing claims not as given discreet demands, but as manifestations of agency that are historically sedimented and produced.
The case of the Guarayos Forest Reserve neatly corroborates this. The very same land use restrictions imposed by the Reserve make for quite distinct land disputes in two different conjunctures. Ethno-territorial claims, in particular, differed both in their form and effects. In the 1970s, this rationale was used by Guarayo people to defend and negotiate their claim over the region's forest resources. It facilitated their organisation to counter powerful state-backed loggers. Ties of belonging, however, were not framed as a direct defiance of the state's authority. While Banzer's regulations were contested, ethno-territorial identity was used to further legitimise their sense of entitlement that had emerged due to the repercussions of the 1953 Agrarian Reform. That is why Guarayo smallholders articulated the 'peasant' identity advocated by the MNR's 'modernist' project. They were signalling their desire to become full citizens of the Bolivian state. Furthermore, at that time, Guarayos' ethno-territorial claim did not possess an environmentalist connotation. Their dispute with loggers was not framed on conservationist lines but on the politics of redistributionthey contested being excluded from the benefits of timber extraction.
In the second conjuncture, in contrast, Guarayo people articulated an identity as 'indigenous', particularly those affiliated to COPNAG. Their ethno-territorial claim was premised on a defiance to the state's legitimacy to define land uses. The very sovereignty of the state was explicitly challenged by Guarayo leaders who reclaimed jurisdiction over the entire province. To operationalise their control, they installed a non-transparent system of 'certifications'. The fact that such system mimics the official one signals not only the reproduction of a technical fetishism, but also a replacement of authority. The system was presented as an instrument for the re-occupation of the territory by its ancestral legitimate owners, but, in practice, land was allocated to the best bidder. Empirical evidence suggests that it is a system from which only a tiny Guarayo elite benefits. This renewed invocation of ethno-territoriality, moreover, tactically mobilises environmentalist images of indigeneity. The first time I met COPNAG's president, Daniel Yaquirena, he opened the interview positioning the Guarayo people as stewards of the environment with a 'natural conservationist tradition' (Fieldwork notes 25/04/2018). Such images, however, have lost traction over the years as accusations of 'land trafficking' have increased. Indeed, Leonardo Tamburinione of Bolivia's most prominent scholar-activists for indigenous rightstold me with frustration that they cannot work anymore 'with indigenous leaders who are only interested in putting the territory up for sale' (Fieldwork notes 27/06/2018). 14 Derek Hall, Philip Hirsch and Tania Li's (2011) framework proved to be a useful heuristic device to unpack land conflicts in the Guarayos Forest Reserve. Competing groups sought to exclude each other by mobilising 'powers of exclusion' of their own. Ethno-territorial belonging was the most powerful form of legitimation in both conjunctures, but not the only one. Loggers' concessions were legitimatised on the name of efficiency and orderly management, while immigrant peasants implicitly invoked nationhood when seeking state intervention in the conflicts. Force was another key power. It was exercised by all types of actors, both in its violent forms and as intimidation. Although there were violent clashes in the 1970s, it is in the more recent conjuncture that force became central to defend or seek access to land. Cases of arson and violent eviction were increasingly common since the late 2000s. Tellingly, when Uraeza lost the vote at COPNAG's assembly, Guarayo people themselves chanted 'fuera quema casas!' (out arson man!) (Fieldwork notes 12/05/2018).
The power of the market had a lot to do with the increasing conflict. Most obviously, the rising land prices were a key incentive. But it was also the power of the market that had pushed newcomers into the Guarayos region in the first place, although for different reasons. Most immigrant peasants arrived at this indigenous frontier to start anew after losing their land somewhere else, usually due to downward spirals of debt. In contrast, big farmers and better-off 'peasants' were responding to favourable commodity prices and arrived at the region with the aim of expanding their operations. Finally, the power of regulation was more consistently expressed in the restrictions of the Forest Reserve. Yet changes to the country's land legislation during the 1990s were radically transformative. They had immediate effects on the ground such as the creation of the so-called 'recortes'. But, more importantly, the new rules shaped the broader land relations throughout the lowlands by recreating an ethnic divide between lowland 'indigenous' peoples and highland 'peasant' migrants.
Indeed, as argued by Bottazzi and Rist (2012), the country's process of land registration reinforced such dichotomy by associating the categories 'indigenous' and 'peasant' with particular land tenure arrangements. This created 'a spatial, cultural and institutional segmentation that goes beyond land appropriation, also embracing access to resources, health care, education, and diverse public infrastructures ' (2012, 547). In the Apolo region, for example, Fontana (2014) demonstrated that the introduction of identitarian criteria for the allocation of land was the main factor enabling land conflicts. In this case, the 'indigenous vs peasants' ethnic divide emerged amongst people that used to belong to a single common organisation. It was a dissident group who decided to articulate an 'indigenous' identity as Leco people to advance an ethno-territorial claim to obtain a TCO title. Others have also framed land conflicts along the 'indigenous vs peasants' ethnic divide, noting the troubling dilemmas involved when different criteria of entitlement collide (e.g. Von Stosch 2014;Sanchez-Lopez 2015).
Such processes of ethnicization have undoubtedly reconfigured land relations in lowland Bolivia. Yet, as shown here, the 'indigenous vs peasants' dichotomy is insufficient to fully explain land conflicts on the ground. In dynamic indigenous frontiers like the Guarayos province, significant processes of 'accumulation from below' (Bernstein 1996) are intensifying class differentiation within ethnic groups. Clashes are not between homogenous ethnic groups, but between different classes with opposing interests. As noted by McNeish for the TIPNIS case, 'a common mistake of external observes is to assume clear relationships between class and ethnicity ' (2013, 238). Indeed, as I have demonstrated, land conflicts in the Guarayos Forest Reserve do not concur with the simplistic 'colla migrants vs indigenous locals' framing. Just like there is an emerging Guarayo bourgeoisie profiting from their leadership positions, immigrant 'peasants' are also increasingly differentiated along class lines. In fact, the latest conflict at the 3 de Octubre settlement is more accurately described as one between two petty capitalists and a group of resource-poor peasants. While they all arrived at the region as migrant 'peasants', the first two were able to accumulate and eventually went against their own 'countrymen'. Recognising exclusion's 'double edge' (Hall, Hirsch, and Li 2011) has the advantage of putting in sharp focus the who of exclusion. To adequately pinpoint the latter, however, I have argued that greater attention to class differentiation is needed. This is perhaps particularly the case in indigenous frontiers where ethnic divides tend to be overemphasised.

Notes on contributor
Enrique Castañón Ballivián is a PhD candidate at SOAS, University of London. His doctoral thesis explores the process of agrarian change brought about by agribusiness expansion in eastern lowland Bolivia with a focus on the changing labour dynamics and the intersection between class and ethnicity.