Green Grab by Bricolage – The Institutional Workings of Community Conservancies in Kenya

Across Kenya’s arid and semi-arid lands, vast rangelands are being transformed into community conservancies – common property arrangements managed for transhumance pastoralism and biodiversity conservation. The Northern Rangelands Trust (NRT) has spearheaded this transformation, promoting community conservancies as a model that conserves biodiversity while developing resilience, improving livelihoods, and promoting security among diverse pastoralist groups in Kenya. Building on recent critical engagement with the NRT model, this article reframes community conservancies as green grabs. In doing so, it makes two overarching contributions to wider debates. The first contribution complicates stereotypes about ‘grabbers’ and ‘grabbees’ and unsettles crude distinctions between political reactions to green grabs, social phenomena commonly portrayed as enacted from above and reacted to from below. Using the concept of bricolage , we show how actors at multiple scales with multiple identities participate – consciously and unconsciously – in reshaping institutional arrangements for managing communal lands and natural resources to align with conservation. The second contribution reveals how power works through emergent hybrid institutions, producing undesired and unintended outcomes. With this in mind, the article concludes that green grab by bricolage produces contradictory spaces animated by a seemingly adaptive, innovative, and progressive agenda, but constrained by historical patterns of access, accumulation, and domination.

communities to manage communal lands and natural resources for pastoralism (Enns 2017).
Like other Community-Based Natural Resource Management (CBNRM) arrangements, community conservancies in Kenya's ASALs are meant to give indigenous peoples or subsistence producers more control over how communal lands and natural resources are managed for conservation, and control over revenues derived from market-based approaches to conservation. 'The central tenet of this scheme is that, in contrast to colonial-style "fortress conservation", there should be no conflict between the economic survival of communities and foraging needs of wildlife; rather, they should complement each other' (Wolmer and Ashley 2003: 32). The use of economic incentives (e.g. jointly-owned tourism ventures, payment for ecosystem services, or new sources of employment in the conservation industry) to condition rural communities to the environmentally friendly behaviours espoused by NRT reflects the market-based or neoliberal nature of CBNRM (Dressler and Büscher 2008;Fletcher 2010;Roth and Dressler 2012). In short, market-based approaches to conservation 'promise to reconcile the longstanding tension between livelihood and conservation' (Roth and Dressler 2012: 363). Despite such lofty ideas and aspirations, critical researchers have problematised the contradicting outcomes associated with CBNRM in practice, with some seeing CBNRM as a form of green grabbing (see Green and Adams 2014). Given the vast literature documenting these contradicting outcomes, we avoid rehearsing debate about the good and bad of CBNRM in this article.
Instead, we contribute to recent debate about the differentiated impacts and variegated political reactions to land grabs (Hall et al. 2015), focusing specifically on green grabs or land deals made for conservation purposes (Fairhead et al. 2012). Critics of land grabs have recently come under fire for over-relying on stereotypes of 'grabbers' and 'grabbees' and for romanticising resistance by indigenous peoples or subsistence producers while downplaying a more complex and broader range of political reactions, including demands to be incorporated into land grabs on adverse terms (Borras and Franco 2013;Hall et al. 2015;Mamonova 2015). Such charges have sparked calls for a return to the question of agency in relation to land grabs, with the goal of documenting and better understanding political reactions that go beyond resistance (Hall et al. 2015). In contributing to recent debate about land grabs, this article shows how a critical institutionalist perspective offers nuanced insights into subtle ways that green grabs work in practice and into diffuse sources of power in green grabs.
More specifically, this article adds further context and complexity to linear conceptions of power prevalent in wider debate on peoples' experiences with and reactions to green grabs. Departing from the framing of land grabs as social phenomena enacted from above and reacted to from below (Hall et al. 2015), we show how green grabs occur subtly through the refurbishment and rearrangement of institutions governing communal lands and natural resources to align with the ideas and aspirations of CBNRM. Our analysis is informed by the concept of institutional bricolage, which refers to processes by which groups of individuals consciously and unconsciously participate in reshaping institutional arrangements for managing communal lands and natural resources based on whatever tools or resources are readily accessible to them -recognising that those resources and tools that are available depend on the life chances and lifeworlds of bricoleurs (Cleaver and De Koning 2015). Acknowledging the diverse roles that pastoralists and conservationists have played as bricoleurs across Kenya's ASALs, we consider how and why seemingly adaptive, creative, and progressive institutional arrangements for managing communal lands and natural resources result in green grabs.
Three assumptions about the interplay between institutions and power illuminate our analysis: 1) institutions are fashioned from tools and resources that are geographically and historically contingent and embedded in everyday life; 2) individual actions (e.g. adaptive responses, creative innovations, and everyday practices) and social structures (e.g. dominant discourses and policies, patterns of land distribution and tenure, and patriarchal forms of social organisation) interact to produce outcomes that are enabling for some while constraining for others; 3) semiotics derived from a variety of sources confer authority and legitimacy on institutions and the power relationships they embody (Cleaver 2012). With this in mind, institutional arrangements formed through bricolage are entwined with micro-level power relationships in everyday life and historical patterns of access, accumulation, and domination in society at large. As such, highly respected, wealthier, and/or well-connected people are capable of drawing on discursive and material resources from different domains of action to shape institutional arrangements in their favour. Although everyone has the potential to act in ways that transform or maintain the status quo, such action occurs both consciously and non-consciously. It is in the nonconscious realm where hegemonic discourses and material relationships inform peoples' perceptions of what is just or unjust, and so shape the actions pursued in response. These assumptions about institutions and power help to explain why relatively poor or marginalised groups do not always resist unjust forms of social organisation or why they might pursue incorporation into new, but equally problematic, relationships. Ultimately, these assumptions inform our conclusion that community conservancies in Kenya's ASALs represent contradictory institutional spaces animated by the promise of human adaptation but constrained by historically unequal and oppressive forms of social organisation. This article proceeds by bringing literature on green grabs (and CBNRM more broadly) into conversation with that on critical institutionalism. This section is meant to situate our analysis and discussion in relation to existing literature and to highlight our contribution to recent, relevant debate. Additional background information is then provided on the case study and research methods informing this article, before we begin to analyse specific institutional arrangements for managing communal lands and natural resources in community conservancies across Kenya's ASALs. This section begins by providing additional contextual information about NRT and proceeds with an analysis of key institutions governing revenue distribution and use, rangeland access and use, and conservation enterprises in community conservancies -focusing on the outcomes bricoleurs associate with these arrangements in practice. In discussing our analysis, we return to the concept of bricolage to expound on the institutional workings of power in community conservancies. The concluding section summarises the article's main contributions and reflects on the implications of green grab by bricolage for wider debate.

CBNRM, GREEN GRABS, AND CRITICAL INSTITUTIONALISM CBNRM and institutions
CBNRM rose to prominence in Africa during the mid-1990s, although its roots trace back to colonial efforts to govern communal lands and natural resources indirectly through leadership structures already in place among rural communities (Roe et al. 2009). Today CBNRM is based on developing 'arrangements for the decentralised sustainable utilisation of wild resources' (Wolmer and Ashley 2003: 31) by integrating the logics, markets, and values of conservation with existing natural resource management practices (Jones 2006;Igoe and Brockington 2007;Roth and Dressler 2012). In departing from coercive and exclusionary conservation models (e.g. fortress conservation) to pursue decentralised and more socially inclusive arrangements, CBNRM prioritises the adaptation of customary or informal institutions that embody in-depth knowledge of local ecosystems (Nelson and Agrawal 2008).
The logic of this approach is that designing institutions tailored to existing arrangements for natural resource management promotes good governance and economic viability in the conservation industry (Booth 2012). From this perspective, CBNRM reveals the hybrid nature of governance: whether intentional or not, formal rules and practices invariably combine with customary norms and informal practices to produce institutions that reconcile the policy imperatives of donors and governments with the needs and realities of rural communities (Cleaver et al. 2013). As critical researchers in Africa have demonstrated, the priorities and practices of conservation have long been at odds with those of rural communities that depend on access to communal lands and natural resources for their subsistence and survival (see Neumann 1998). As an alternative to fortress conservation, CBNRM avoids forcefully excluding rural communities from protected areas or coercing them into modern livelihoods or formal institutional arrangements. Instead, it seeks to embed new conservation rules and practices into existing leadership structures, decision-making processes, and everyday interactions, and vice versa (Cleaver et al. 2013). In this regard, CBNRM relies on technologies of government that work to create certain types of subjects (Agrawal 2005;Fletcher 2010Fletcher , 2017) that contribute to the conservation of communal lands and natural resources through the daily reenactment of behaviours, ideas, and values that are commensurable with the conservation industry (Agrawal 2005).
Many conservation policy-makers, practitioners, and researchers believe that getting institutional design right will lead to outcomes that are more democratic, inclusive, and economically viable than those associated with other conservation models (see NRT 2015b). However, a plethora of literature reveals just how challenging or unlikely it is for CBNRM to achieve such outcomes in practice (Brockington 2004;Jones 2006;Igoe and Brockington 2007;Roth and Dressler 2012;Martin et al. 2013). For example, mechanisms for devolving authority over wildlife to local-level actors often fail to ensure democratic and transparent public decisionmaking processes when transposed into diverse contexts and heterogenous communities (Agrawal and Gibson 1999;West 2006;Wright 2017). Moreover, the pursuit of economically viable conservation programmes tends to involve the commodification and privatisation/marketisation of communal lands and natural resources -processes that institutionalise new forms of exclusion, relationships of authority, and mechanisms for valuing natural resources that contradict the very ideas and aspirations of CBNRM (Songorwa 1999;Alexander and McGregor 2000;Igoe and Croucher 2007;Ojeda 2012;Wright 2017). For such reasons, many CBNRM arrangements bear the hallmark of green grabs: 'the appropriation of land and resources for environmental ends' (Fairhead et al. 2012: 238). Indeed, this article builds on recent critical engagement with NRT policies and programmes by reframing community conservancies as green grabs (see Greiner 2012;Mbaria and Ogada 2016;German et al. 2017;Bersaglio 2017).

CBNRM and green grabbing
As a type of land grab justified in the name of conservation, green grab implies the transfer of control over communal lands and natural resources 'from the poor (or everyone including the poor) into the hands of the powerful' (Fairhead et al. 2012: 238), generally speaking. Beyond gaining access to communal lands and natural resources, green grabbers (e.g. multi-national companies; transnational conservation organisations; national and international investors; and local elites like landowners, public stakeholders, and customary leaders) reinvest revenues generated by market-based approaches to conservation into programmes and ventures with the effect of intensifying their control over land, resources, and capital in rural society. It has been shown that related processes of commodifying and privatising/marketising communal lands and natural resources invariably privilege the interests of a minority of the population, especially already powerful actors (Igoe and Brockington 2007).
In this regard, green grabs are akin to primitive accumulation or accumulation by dispossession (Kelly 2011;Benjaminsen and Bryceson 2012;Green and Adams 2015). Marx (1976) originally used primitive accumulation to describe the appropriation and enclosure of common property in Great Britain, in which subsistence producers were divorced from the means of production through the institutionalisation and violent reinforcement of private property. This process transformed subsistence producers into wage labourers and the social means of production into capitalist relationships (Thompson 1975). Yet primitive accumulation is not a one-time thing; it is a basic ontological condition of capitalism (Luxemburg 1951;De Angelis 2001;Glassman 2006). The evolution and expansion of capitalism requires ongoing forms of primitive accumulation to forge new spaces (literal and figurative) for capital accumulation (Prudham 2013), processes Harvey (2003) reframed as accumulation by dispossession. As a form of green grab, CBNRM entails far more than changes as to how land is held, by who, and for what purposes. Such arrangements '"pull resource users into new market economies" and at the same time relieve local people of their land and/or resources while forcing them to become dependent on an industry (often ecotourism) over which they have no control and from which they often earn very little money' Büscher 2008: 454-455, as cited in Kelly 2011: 686).
The negative effects of green grabs around the world have been well-documented by critical researchers. In carrying out this work, however, the same researchers have been charged with proliferating stereotypical representations of 'grabbers' and 'grabbees'. These charges target land grabbing literature broadly but are just as applicable to literature concerned with inter-related processes of green grab. The types of representation in question depict indigenous peoples or subsistence producers as universally bound to a peasant way of life, inherently opposed to green grabs, and unwilling to adapt to capitalist markets and relationships (Mamonova 2015). The romanticisation of resistance to green grabs, in particular, has been criticised as a false or overly simplistic generalisation of what happens when green grabs hit the ground in diverse geographical and historical contexts (Alsonso-Fradejas 2015; Hall et al. 2015;Mamonova 2015). In response, increasingly more critical researchers are committed to focusing on 'the differentiated impacts and variegated political reactions to land deals', ranging from acquiescence to demands for incorporation to everyday or outright resistance (Hall et al. 2015).
Following a return to the question of agency in broader land grabbing literature (Borras and Franco 2013;Hall et al. 2015), this article considers green grabs from a critical institutionalist perspective. Our approach is partially motivated by recent analyses that grapple with the pursuit of incorporation into green grabs or even acts of green grabbing on the part of indigenous peoples or subsistence producers stereotypically assumed to be diametrically opposed to or hapless victims of green grabs (see Gardner 2016;Astuti and McGregor 2017;Bersaglio 2017). We also build on a recent article informed by bricolage that analyses changing common property arrangements in a community conservancy elsewhere in Laikipia (German et al. 2017). In building on these works, our use of bricolage challenges some false stereotypes about grabbers and grabbees and, more specifically, unsettles crude distinctions between acquiescence, incorporation, and resistance prevalent in wider debate. From a critical institutionalist perspective, the institutional arrangements intertwined with green grabs are the product of complex and sometimes contradictory interactions between actors at multiple scales with multiple identities (e.g. pastoralists and conservationists); they are not simply enacted from above and reacted to from below (Hall et al. 2015). Thus, a return to the question of agency necessitates a return to the question of structure and, ultimately, analyses of green grabs informed by a critical perspective of institutions as political and power-laden processes. So, after saying more about the case study and methods behind this article, the rest of this article considers how green grab by bricolage 'creates room for manoeuvre and new possibilities for some people but simultaneously reproduces and reinforces social inequalities for others' (Cleaver 2012).

Case study overview
Laikipia County is 869,000 ha large and located in northcentral Kenya. The county encompasses a vast plateau west of Mt. Kenya that once served as a prime source of grazing land for transhumance pastoralists across the ASALs. Prior to British colonisation in the late 1800s, Il Aikipiak and other Maasais dominated much of the plateau. However, as part of Britain's efforts to transform Kenya's highlands into a 'white reserve', thousands of pastoralists were moved from the Laikipia Plateau to the Southern Maasai Reserve near present-day Tanzania (Hughes 2006). Such efforts eventually led to the formalisation of a racialised property regime, in which 'whites' were permitted to hold land individually in the so-called white highlands whereas 'blacks' were permitted to hold land collectively in 'native reserves' (Kanyinga 2009).
Current land use and tenure arrangements in Laikipia have been shaped by the legacies of this racialised property regime; especially in the county's conservation industry. Some estimates suggest that as much as 40 per cent of the county (or 382,400 ha) is owned by a white minority of the population, divided among 48 large-scale properties (Letai 2011(Letai , 2015. Forty-six of these properties are used for conservation and tourism and are either owned by settlers or have since been sold-off to international organisations like Fauna and Flora International or The Nature Conservancy (LWF 2012). Beyond the 46 properties that comprise most of the land used for wildlife conservation and tourism in Laikipia, 11 group ranches -i.e. 'privately titled collective rangelands used for communal livestock production' (Nelson 2012: 3) -owned by Maasai or Samburu pastoralists contribute 71,200 ha (or 7 per cent of Laikipia) to the conservation industry in the form of community conservancies (Letai 2011). In this article, we are primarily concerned with how Laikipia's three community conservancies, called Il Ng'wesi, Lekurruki, and Naibunga, fit into the county's conservation landscape. 1 Laikipia's network of private and community conservancies adds up to about 365,000 ha of contiguous space for conservation (LWF 2012). This network provides habitat for a wildlife population larger than that of Amboseli, Nairobi, Tsavo East, and Tsavo West National Parks combined (Kinnaird and O'brien 2012). In addition to containing a large population of wildlife, the county boasts high levels of biodiversity -including more than 95 species of mammals, 540 species of birds, and 87 species of amphibians and reptiles (LWF 2012). Laikipia is also famous for its endangered and critically-endangered megafauna. This includes half of Kenya's black rhinoceros population and the country's second largest elephant population, as well as less well-known species such as the reticulated giraffe (Giraffa camelopardalis reticulata), Grevy's zebra (Equus grevyi), and Laikipia hartebeest (a hybrid of Alcelaphus buselaphus cokei and lelwel). With the exception of black rhinoceroses, all community conservancies in Laikipia provide habitat for these and other endangered species.

Research methods
This article relies on qualitative data collected by the first author on this article during 11 months of fieldwork in Kenya between October 2014 and June 2015, as well as during followup trips in May 2016 and April and July 2017. The fieldwork involved ethnographic methods in case study research on community conservancies across Kenya's ASALs, including participation in tourism activities, observation of natural resource management practices, key informant interviews, informal interviews, and document analysis. The languages used during the research were primarily English, Kiswahili, and Maa.
In total, the first author spent over two months participating in tourism activities and observing natural resource management practices in community conservancies. Between 2014 and 2017, over 160 interviews were carried out with some 100 informants and interlocutors. Participants included individuals from pastoralist households in community conservancies (representing different ages, genders, and identity groups, although most identified as Maasai), administrative authorities in community conservancies, pastoralist civil society groups, biodiversity conservation organisations, and county governments, as well as representatives from broader populations of conservationists, landowners, and tourists in the region. The first author also spent seven days in Kenya's National Archives collecting data that added historical context to the fieldwork.
Most of the data informing this article derives from fieldwork activities that focused on Il Ng'wesi Conservancy, located on the north-eastern border of Laikipia. Il Ng'wesi Conservancy is 9,433 ha large and has a population of 1,448 people, most of whom identify as Maasai -the conservancy's name refers to a sub-section of Maasai who sometimes call themselves Il Ng'wesi or Laikipiak in reference to one of the original dominant Maasai groups on the plateau (Sobania 1993). Although Il Ng'wesi Conservancy was a focal point of fieldwork, general insights into community conservancies across Kenya's ASALs came from research activities pertaining to community conservancies in the nearby counties of Isiolo, Marsabit, and Samburu. In what follows, we analyse key institutional arrangements for managing communal lands and natural resources in NRT conservancies; focusing on institutions governing revenue distribution and use, rangeland access and use, and conservation enterprises.

INSTITUTIONAL ARRANGEMENTS IN COMMUNITY CONSERVANCIES
Since being established in 2004, NRT has become one of the most influential conservation organisations in Kenya. It is now common to hear NRT referred to as the new government in the north -meaning the ASALs. 2 In the mid-1990s, Ian Craig -whose long-time settler family owned what is now a UNESCO World Heritage Site, Lewa Wildlife Conservancy (LWC) -began trying to persuade elders among pastoralist communities north of his family's land to implement wildlife conservation and tourism initiatives on their lands. After much persuasion, 3 Il Ng'wesi became one of the first group ranches to collaborate with LWC to implement CBNRM. Soon after, Namunyak in Samburu County entered a similar partnership with LWC. Today there are 27 community conservancies across the ASALs, under the umbrella organisation of NRT (NRT 2015a).
The NRT was established to coordinate and support the growing number of community conservancies emerging through collaborative efforts between LWC and pastoralist communities across the ASALs (NRT 2013). The organisation was given a mandate to develop conservancy institutions in group ranches and other communal property settings, monitor the performance of community conservancies and provide advice on how to manage conservancy affairs, and raise funds for community conservancies while providing quality assurance to donors from around the world (NRT 2015b). This mandate was meant to 'develop resilient community conservancies which transform people's lives, secure peace and conserve natural resources' (NRT 2015b: 8). Indeed, NRT sees itself as more than just a conservation organisation; it understands community conservancies as spaces that foster peace and stability in insecure parts of Kenya's ASALs and that catalyse sustainable development among pastoralists through the formalisation of customary institutional arrangements for natural resource management (NRT 2012).
Specifically, NRT defines a community conservancy as a 'community-owned and community-run institution that aims to improve biodiversity conservation, land management and the livelihoods of its constituents over a defined area of land traditionally owned, or used, by that constituent community' (2015b: 10). Departing from the approach to establishing conservancies outlined in Kenya's new Wildlife Conservation and Management Act -which encourages the acquisition of new land, or transformation of existing land, to create conservation areas -NRT integrates existing arrangements for managing communal lands and natural resources with conservancy institutions (NRT 2015b). 'In communities where traditional institutions (for governing resources) are strong and intact, conservancies must recognise these as the highest decision-making body and either formally or informally integrate them into the conservancy institution' (NRT 2015b: 12). According to NRT, integration occurs best in settings where the Boards of Directors for community conservancies mirror existing leadership structures: 'This streamlines community governance structures and ensures formal, administrative or customary institutions are represented on Conservancy boards' (NRT 2015b: 12).
To implement and fund community conservancies, NRT partners with international organisations such as USAID, The Nature Conservancy, DANIDA, and Agence Français de Développment (NRT 2017). NRT's close ties to LWC mean that the organisation has access to a range of extremely wealthy patrons, including members of the British Royal Family. In addition to partnering with these rather conventional supporters of CBNRM in Africa, NRT has partnered with less conventional conservation supporters. For example, the organisation recently signed an agreement with Tullow Oil to establish community conservancies around the Anglo-Irish company's sites of oil exploration and production in Turkana and West Pokot Counties. Given that pastoralists and conservationists have frequently found themselves at odds with each other throughout history, community conservancies implemented on land belonging to Borana, Gabbra, Maasai, Pokot, Rendille, Samburu, and Turkana pastoralists also represent unconventional alliances of sorts. In short, NRT's success in expanding the scale and scope of its influence has been achieved through alliances with conventional and non-conventional conservation actors. The pastoralist communities that have partnered with NRT have, in turn, deployed equal innovation -and perhaps more good faith -by pursuing unconventional alliances with the organisation.
The following sections outline key institutional arrangements for managing natural resources in community conservancies. They highlight specific institutional arrangements designed to manage the distribution and use of revenues generated by community conservancies, how communal lands and natural resources are accessed and used in community conservancies, and the livelihoods of people living in community conservancies. This discussion recounts the main advantages and disadvantages of such arrangements, based on the experiences and perspectives of pastoralists living and working in community conservancies, conservancy leadership, and pastoralist civil society organisations working in or with community conservancies. The vast majority of these individuals identify as descendants of Il Aikipiak Maasais. The point of this discussion is to provide the reader with insights into the institutional arrangements that have been produced through the integration of conservation logics, markets, and values with existing mechanisms for managing communal lands and natural resources for pastoralism.

Revenue distribution and use
The NRT recognises that the promise of institutional integration alone does relatively little to convince pastoralist communities to buy into the logic of CBNRM and to enter partnerships with the organisation. According to NRT, this requires the design and implementation of programmes that create 'leverage for conservation' (NRT 2015b: 8).
Leverage and support from communities for wildlife conservation comes from improved community attitudes towards conservation as a result of benefits people gain from conservancies, some of which are revenue and employment from conservancy enterprises. Conservancy enterprises must therefore have explicit links to the conservancy itself in order to create support for its conservation goals (NRT 2015b: 27). In other words, NRT uses the promise of direct and indirect benefits to leverage pastoralist communities into partnerships with the organisation. Direct benefits include revenue generating opportunities for individuals through conservation enterprises, such as wildlife tourism and access to livestock markets and micro-finance (NRT 2015b: 35). To start, though, this section focuses on the indirect benefits that donor partnerships and wildlife tourism ventures offer in community conservancies through revenue generation and distribution for community development initiatives.
While trying to ensure that pastoralist communities retain autonomy and ownership over communal lands and natural resources and derive benefits from conservation, NRT transforms existing community-level governance bodies -including customary councils of elders -into Boards of Directors. Boards consist of 12 elected representatives from pastoralist communities, county governments, Kenya Wildlife Service (KWS), private investors, and NRT. Each board member is elected by pastoralist communities for a three-year term and is eligible to run for two consecutive terms. Boards of Directors are responsible for ensuring that revenues generated from wildlife conservation and tourism activities are distributed throughout respective communities according to Benefit Distribution Plans. These plans outline if and how revenues are used to benefit communities by funding education and healthcare initiatives, infrastructure development, or sanitation and water programmes, for example. Under NRT, 40 per cent of revenues generated by community conservancies are reinvested in conservancy operations and the remaining 60 per cent allocated for community development. On an average, NRT provides 80 per cent of the total income generated by conservancies annually.
NRT's logic of using opportunities for community development to leverage pastoralist communities into CBNRM arrangements has proven to be effective. Prior to the implementation of community conservancies, very few people from pastoralist communities across Kenya's ASALs received a formal education. As a crude illustration of this, even until 2013, only 13.9 per cent of people in Laikipia had obtained a secondary education (KNBS-SID 2013). To access health services, moreover, most people had to travel long distances on bicycle or by foot. Even then, accessible services were limited and of poor quality. A definitive lack of good quality roads across Kenya's ASALs also made the outcomes of medical emergencies grave -including childbirth -and complicated simple feats like transporting livestock to markets. Lastly, limited access to education and health services was compounded by a definitive lack of infrastructure in pastoralist communities. In this regard, CBNRM has enabled some people in some communities to have their citizenship rights fulfilled through partnerships with NRT in a context where local and national governments had largely neglected to fulfill such rights (Funder and Marani 2015).
Interlocutors in Il Ng'wesi and other community conservancies described access to education and health services, and to improved infrastructure, as the most important thing CBNRM has done for pastoralists -as 'the number one benefit of conservation' 4 -along with access to improved security services. While expressing gratitude to donors and tourists for helping pastoralist communities generate revenues for community development, many informants and interlocutors are simultaneously discontent with the fact that CBNRM has left them dependent on foreign actors for their wellbeing. 5 For pastoralist rights advocates, especially, this paternalistic dependency is seen as a form of neo-colonialism -as 'Colonialism 2.0', as one pastoralist rights activist stated. 6 This notion of dependency stems largely from the fact that wildlife tourism has failed to live up to its promise as a way for pastoralists to generate and control their own revenues for community development.
Although Il Ng'wesi enjoys 'pride of ownership' over its ecolodge (NRT 2015b), a competitive and fickle tourism market, coupled with tourists' latent prejudices against seeing livestock in conservation areas, have undermined its promise as a tool for liberating the Il Ng'wesi community from poverty and under-development. For example, an outbreak of Ebola as far away as Guinea might cause tourists booked at the eco-lodge to cancel reservations for refunds. This happened during the tourism high season in 2015, causing the Board of Directors to temporarily close the eco-lodge due to lack of business. The eco-lodge was closed again for part of 2016 and, amid violent conflicts between migrant pastoralists and conservancy personnel in Il Ng'wesi and Laikipia, the eco-lodge was closed for part of the 2017 high season as well due to safety concerns. 7 In addition to the various external factors that make international tourism an unreliable source of income, tourists are influenced by prejudices that might deter them from going on safari in community conservancies in the first place. Safari guides working in community conservancies explained to the first author on this article that it is common for tourists to lose their tempers at the sight of domestic livestock in conservancies: 'I didn't come all this way to look at cows!', a safari guide recalled one angry tourist shouting angrily during a game drive. 8 In another instance, a tourist at at a private conservancy elsewhere in Laikipia exclaimed to the first author, 'You went to Il Ng'wesi? There's nothing but a lot of dik-dik up there!' 9 With such factors in mind, wildlife tourism has proven to be less profitable than other conservation enterprises in community conservancies. In 2014, for example, wildlife tourism injected $468,000 into community conservancies (NRT 2015a). This is less than the $789,295 brought into the conservancies via cattle and handicraft sales combined, which amounts to $684,188 and $105,107 respectively (NRT 2015a). For such reasons, creating institutional mechanisms for Boards of Directors to control the distribution and use of revenues generated from wildlife tourism may do relatively little to ensure the autonomy of pastoralist communities, leaving them dependent on external interventions and donor support to deliver the promises used to leverage communities into partnerships with NRT in the first place.

Rangeland access and use
Improving the condition of communal lands and natural resources among pastoralist communities and securing their access to such resources is another key point of leverage that the NRT uses to incentivise pastoralists to implement community conservancies on their land (NRT 2015b). As indicated, the changing landscape of pastoralism across Kenya's ASALs has forced individuals and groups to adapt their livelihoods -for example, by pursuing new arrangements for accessing and using communal lands and natural resources or new opportunities for livelihood diversification. In exploiting this opportunity to create leverage for conservation among pastoralists, NRT alters existing approaches to rangeland management and introduces new mechanisms for securing communal lands and natural resources when partnering with pastoralist communities.
For example, NRT facilitates training and other logistical support for wildlife rangers in community conservancies to secure space for conservation in the ASALs. Wildlife rangers are paramilitary forces comprised of both permanent officers stationed in community conservancies and mobile units specialised in deterring wildlife offences, such as illegal hunting. Typically, wildlife rangers are also members of the Kenya Police Reserve (KPR) in their respective communities and, therefore, tend to already be residents in the community conservancies where they are employed. Wildlife rangers are paramilitary forces deployed to enforce conservancy boundaries and regulations, mediating access to natural resources for pastoralists during times of dearth and plenty. However, wildlife rangers also provide general security services to people living in community conservancies -acting as first responders to cattle raids, livestock thefts, and other emergency situations. In other words, community conservancies provide pastoralists access to new mechanisms and procedures that enable to them forcefully defend their territory.
To carry out their conservation mandate and provide security services in community conservancies, wildlife rangers are granted access to military intelligence, technology, and weapons. In partnership with KWS and KPR, wildlife rangers also receive special training in combatting wildlife crimes, collecting intelligence, and enforcing conservancy regulations. It is worth noting that Kenya's new WCMA increases the scope and punishment of wildlife offences, which wildlife rangers have the authority to enforce in collaboration with the national government. In securing space for conservation across Kenya's ASALs, and acting as security agents in under-serviced areas, wildlife rangers also play a vital role in managing how communal lands and natural resources are accessed and used.
Specifically, wildlife rangers monitor access to natural resources among community members and regulate natural resource uses allowed or disallowed by Grazing Coordinators and Grazing Committees that might already exist at the community level -group ranches tend to already have such individuals and institutions in place before community conservancies are implemented. The NRT's current approach to managing communal lands and natural resources for shared use is known as 'holistic rangeland management'. Holistic rangeland management is said to combine customary grazing practices with modern techniques (LWF 2017). In simple terms, it involves grazing livestock in tightly clustered groups in designated grazing blocks in community conservancies: 'when huddled close to one another the animals act as a "bulldozer" breaking the ground and allowing for water and nutrient flow, whilst at the same time they implant seeds and add fertilizer' (LWF 2017). Grazing blocks are rotated to distribute nutrients from livestock dung and other beneficial aspects of their behaviour throughout community conservancies, enabling livestock to play a role in regenerating plants for shared use. While the practices of holistic rangeland management reflect centuries of customary adaptation by pastoralists in the ASALs, new discourses, logics, and technologies are said to add a modern element to this centuries-old approach to managing communal lands and natural resources.
Next to access to social services and infrastructure, interlocutors in community conservancies cite security as the most important benefit that CBNRM has brought to their communities. At the same time, however, the proliferation of paramilitary personnel and military grade weapons across Kenya's ASALs is understood by many informants to be aggravating tensions over communal lands and natural resources and provoking conflicts with deadly outcomes. For example, in June 2015, wildlife rangers at Il Ng'wesi exchanged fire with Samburu pastoralists that had entered the conservancy to graze and water their livestock. The conflict left six people dead, including two Il Ng'wesi members. During a similar incident in 2017, tourists had to be evacuated from the eco-lodge in a helicopter supplied by LWC. 10 Such conflicts are not new across Kenya's ASALs, but CBNRM does provide a new justification for the use of lethal violence to defend territory. One NRT document reports that there were 71 incidents of cattle rustling in its conservancies, as well as 17 human deaths due to insecurity in 2014 (NRT 2015a). Interestingly, both statistics increased from 2013 and correlate with a significant decrease in the number of elephants killed illegally in community conservancies.
Beyond the violence associated with securing communal lands and natural resources for conservation across Kenya's ASALs, the rangeland management practices enforced with the assistance of wildlife rangers have led to new forms of exclusion within community conservancies. Accordingly, it is relatively common to observe -or hear about -pastoralists in community conservancies disregarding conservancy bylaws and watering and grazing their livestock in wildlife areas. 11 Moreover, when asked about holistic rangeland management, interlocutors often reply with statements that slander or taunt NRT: 'Ha! Holistic what? We have been doing that for centuries'. 12 Holistic rangeland management is widely perceived as a co-optation of customary institutions that legitimates the exclusion of pastoralists from communal lands and natural resources that are supposed to be owned and managed by them under NRT. For such reasons, pastoralist rights activists outside community conservancies typically refer to community conservancies as a form of 'land grab' (which we in turn call green grab), to hybrid institutions as 'colonialism', and to NRT as 'Lords of Impunity'. 13 Despite NRT's insistence that community conservancies are spaces that foster peace and stability in insecure parts of the ASALs, the organisation's approach to managing rangeland access and use is perceived as intensifying conflict over communal lands and natural resources and introducing new forms of exclusion into pastoralist communities.

Conservation enterprises
In creating leverage for conservation through income generating or enhancing opportunities, NRT focuses on developing wildlife tourism in community conservancies and on enhancing access to markets and micro-finance for existing livelihoods. To begin with, NRT's approach to wildlife tourism mainly involves facilitating investment in safari camps and lodges through a range of ownership, management, and benefit distribution arrangements. Il Ng'wesi, for example, is owned and operated by Il Ng'wesi members -members that range from local politicians to individuals with diverse livelihood portfolios to strictly pastoralist households comprised of men and women. The NRT suggests that this ownership model affords communities the greatest degree of autonomy over decisions related to wildlife tourism and the revenues it generates. However, tourism ventures might also be community-owned but privately managed, such as Tassia in Lekurruki. In some cases, lodges are both owned and operated by private companies. The NRT sees itself as playing a mediating role of an 'honest broker' in its efforts to foster community-private sector partnerships in community conservancies while ensuring transparency and consensus in negotiations (NRT 2015b). As the advantages and disadvantages of wildlife tourism have already been discussed in relation to the management of revenue distribution and use in community conservancies, this section focuses on the management of conservation enterprises designed to generate income and livelihood opportunities for pastoralists through CBNRM.
Beyond mediating investments in wildlife tourism, NRT facilitates access to cattle markets for pastoralists in community conservancies through a for-profit division of the organisation called NRT-Trading (NRT-T). As many community conservancies have yet to develop wildlife tourism ventures -and because wildlife tourism has not proven to be all that successful in generating revenues in community conservancies -NRT's cattle marketing initiative has become one of the more beneficial aspects of CBNRM for communities. Through this programme, NRT-T purchases cattle from pastoralists in community conservancies, 'fattens them' on settler-owned ranches and conservancies, 14 and then sells them for slaughter about 18 months after purchase. NRT's cattle marketing initiative is also praised for offering 'more than fair' prices for cattle. 15 For example, a cow that might typically be sold for $221 at the local market might be bought by NRT-T for $288. 16 Through this exchange, the seller is required to pay a levy to their respective conservancy, as is NRT-T. This levy provides conservancies with a small, additional source of income. The income generated by NRT-T through its cattle sales are reinvested back into the organisation.
Another key way that NRT is investing in conservation enterprises is by facilitating 'financial literacy' training, micro-credit, and access to markets for beadwork and other handicrafts (NRT 2015a(NRT , 2015b. These interrelated initiatives primarily target women, whereas employment in tourism and cattle sales target men. Initially, NRT provided financial and business training to women in community conservancies. Upon completing the training, participants became eligible to apply for small loans. According to NRT, 737 women received training in financial literacy, 253 women received training in business development, and 98 women were ultimately able to access $17,280 in micro-credit in 2014 (NRT 2015a). On average, each woman received $176 in micro-credit, but loans were only granted to women in 3 community conservancies (NRT 2015a). Since 2011, funding for micro-credit has declined, as NRT transitions towards mediating access to handicraft markets through NRT-T.
Although small loans were meant to help women kick-start handicraft businesses that sold merchandise to tourists visiting community conservancies, NRT's most recent programme provides women access to handicraft markets through NRT-T and its affiliated organisation, Beadworks. This programme is like NRT's cattle marketing initiative: women purchase raw materials at cost from NRT-T, which they use to make beaded decorations and jewelry. Participants in the programme then sell their goods back to NRT-T via monthly markets, where they get paid, receive feedback on the quality of their goods, and in turn purchase materials for new merchandise. NRT carries out quality control, packaging, and markets the goods to clients in Kenya and around the world (NRT 2015b).
Individuals who have been granted access to new markets and valuations for cultural products (i.e. cattle, beads, etc.) speak favourably about NRT-T's efforts to boost their livelihoods. Yet they are quick to downplay the impacts such programmes have on their lives, let alone in wider society. For example, access to cattle markets through NRT-T is partly determined by the performance of community conservancies.
In other words, NRT-T selects who it purchases cattle from based on the conservation and governance performance of the community conservancies in which potential beneficiaries reside. Additionally, each conservancy is allocated a limited number of cattle that they can sell to NRT-T and the organisation does not purchase cattle on a regular basis. One pastoralist from Il Ng'wesi explained that, 'if NRT buys a cow from you today, you might not see them again for one or one and a half years'. 17 Moreover, NRT-T will only purchase cattle from pastoralists, even though sheep and goats are prolific in community conservancies.
Women express similar critiques of NRT-T's handicraft programme. Because the flow of tourists to community conservancies is inconsistent and unpredictable, producing goods for the tourist market has not proven to be profitable. Additionally, demand for handicrafts does not match the number of women participating in NRT programmes. To illustrate this point, an informant belonging to one women's group in Il Ng'wesi explained that, should the conservancy receive an order for 150 belts, each women's group in the community might only end up making about 20 belts. A typical group may have 25-30 members, all of whom share in the profit of the goods they produce and sell. 18 'We can make them, we just can't succeed. We don't have the market', she said. 19 An informant from a women's rights organisation added that it is problematic to confine women's contributions to conservation to handicraft-making, 'as if that's all they can do'. 20 Comparatively, women remain under-represented in leadership positions and decision-making processes within community conservancies. 21 Such insights suggest that NRT's efforts to integrate informal, customary livelihoods with CBNRM through conservation enterprises has failed to convince pastoralists that their bargain with NRT is going to deliver the results they were initially promised.
Before concluding, the next section returns to the concept of bricolage to expound on how power operates in and through the types of community conservancy institutions discussed above to produce undesired or unintended outcomes. In doing so, it further clarifies how institutions mediate the interface between individual behavior and society to produce outcomes that contradict the very ideas and aspirations of CBNRM (Cleaver 2002). Focusing on the institutional interface of agency and structure, pastoralists need not be understood as either hapless victims acted on from above or autonomous actors reacting from below, but as political agents embedded in contradictory rules and practices of green grabs.

THE INSTITUTIONAL WORKINGS OF POWER IN COMMUNITY CONSERVANCIES
To begin, a bricolage perspective demands consideration for processes through which some pastoralist communities have pursued unconventional partnerships with NRT while simultaneously adapting CBNRM arrangements to better serve their desires, interests, and needs. Through this lens, community conservancies are outcomes of pragmatic political manoeuvres on the part of pastoralists and conservationists alike -manoeuvres that correspond with the reconfiguration of ideologies of nature, however slight. Through this manoeuvring and reconfiguration, old institutional arrangements are modified and new ones are invented; institutional mechanisms of diverse origins are recycled and repurposed; and adapted institutions are imbued with meaning and authority through the everyday challenges they respond to and the actions in which they are embedded (Cleaver and De Koning 2015). The hybrid arrangements resulting from bricolage are also embedded in diverse cultures and forms of social organisation that underpin rural society. From NRT's perspective, for community conservancies to function effectively in practice, conservation logics, markets, and values need to become rooted in the established rules and practices that organise everyday life among pastoralist communities. Accordingly, NRT emphasises working through existing institutional arrangements for managing natural resources rather than entirely replacing them -here it is worth emphasising that existing institutional arrangements have likely already been adapted, borrowed, and hybridised, such as arrangements for managing group ranches (German et al. 2017). This is evident in the emphasis that NRT places on creating leverage for conservation among pastoralist communities by developing appropriate conservation enterprises. For example, by designing institutions that enable livestock and handicraft producers in community conservancies to benefit from conservation, however marginally, NRT seeks to blend customary livelihood practices with market-based approaches to conservation and vice versa. In this example, pastoralists can be understood as shaping new institutional arrangements in community conservancies unconsciously through the everyday normative actions that underpin their livelihoods. Holistic rangeland management is another example of how pastoralists have shaped institutional arrangements in community conservancies and have participated in adapting conservancy institutions to meet the needs of livestock and wildlife. As NRT explains (2015b: 12), community conservancies 'build on these traditional institutions and their governance structures and land management practices are a mix of formal and customary'. From this perspective, bricolage is an adaptive and dynamic process; but it also has limitations.
As existing leadership structures and institutional arrangements are reworked in community conservancies to produce outcomes like holistic rangeland management, bricoleurs engage in adaptive, innovative, and pragmatic courses of action. Yet they are also confined to act within the literal and figurative constraints imposed on them by the environment, society, and related power relationships. As Levi-Strauss (1962) famously remarked, a bricoleur might make a lampshade out of an umbrella stand but the same umbrella stand cannot be made into a space shuttle. Not only would a bricoleur face physical limitations in trying to turn an umbrella into a space shuttle, they might also be subjected to societal pressures or sanctions that dissuade them from pursuing such a foolish project in the first place. In the case of community conservancies, it is instructive to reflect on the exercise of agency by groups of women. Whilst women can apparently form associations and benefit from bead-making enterprises through NRT-T, it is probable that such agency might not extend to them becoming members of Board of Directors -certainly not a dominant social category among Boards of Directors. Moreover, by engaging with community conservancies through 'traditional' women's livelihood activities, which are less income generating and less central to conservation goals than those typically performed by men (e.g. cattle herding), they are arguably made complicit in reinforcing their own subordination through participation in community conservancy institutions. In this regard, multiple existing institutions and diffuse sources of power place tangible limits on processes of bricolage and the range of possible outcomes associated with it (Sehring 2009;Petursson and Vedeld 2015).
The resulting patchwork of institutions that ends up being produced through bricolage serves multiple functions (Cleaver and De Koning 2015). For example, new institutional arrangements for managing access and use of natural resources in community conservancies also serve as emergency response mechanisms and as mechanisms for defending communityowned land from trespassers. Additionally, although NRT-T's livestock marketing programme serves as a new mechanism by which pastoralists can access cattle markets, the imposition and collection of an exchange levy on cattle purchases in community conservancies generates revenue for conservation activities -at least in principle. Moreover, by making access to cattle markets contingent on the conservation performance of communities, the livestock marketing programme serves as a mechanism that incentivises or places pressure on pastoralists to uphold their respective community's new conservation mandate. In such ways, multi-functional institutional arrangements formed through bricolage are adaptive responses to particular social environments, but also serve to embed and extend the reach of the powerful across different domains of action.
Additionally, a bricolage lens suggests that multifunctional institutional arrangements need to be seen to fit with established rules and practices if they are to be accepted and ascribed meaning and authority. This is exemplified by the function of local Boards of Directors in community conservancies, which are designed to mirror existing leadership structures and decision-making processes among different pastoralist communities. The subtle process of achieving fit between old and new institutional arrangements -of 'naturalisation' (Douglas 1987) -is facilitated by calls on imagined or real traditions; new symbols, discourses, and power relationships; and cultural, normative, or cosmological notions of right and wrong (Cleaver and De Koning 2015: 5). Such calls reflect in one claim made by a safari guide at Il Ng'wesi lodge, who explained that 'conservancies allow us to milk cattle with one hand and wildlife with the other'. 22 The use of 'milk' in this statement serves to legitimise Il Ng'wesi's pursuit of CBNRM, symbolising the supposed centrality of pastoralism in community conservancies by linking a fundamental subsistence practice (i.e. milking cattle) with the life-sustaining potential of conservation.
Expounding on different forms of bricolage, and the variable actions of bricoleurs, De Koning and Cleaver (2012) identify three processes that occur when customary and informal institutions encounter new and formal ones in settings like community conservancies: aggregation, alteration, and articulation. Aggregation refers to the creative combination of introduced institutions with existing institutional mechanisms and normative actions, leading to hybrid institutional arrangements. Holistic rangeland management is an example of aggregation, as it allows pastoralists in community conservancies to graze and water cattle in wildlife areas -albeit in a closely regulated manner -while contributing positively to local ecosystems from NRT's perspective. Alteration refers to the ongoing adjusting and adapting of institutions to make them better fit existing livelihood practices or notions of identity. Examples of alteration in community conservancies include mechanisms that allow wildlife rangers to act as first responders in emergencies, or as paramilitary forces that defend territory, rather than simply as wildlife officers. Finally, articulation refers to the assertion of culture, identity, or tradition in resistance to new hybrid institutional arrangements. Examples of articulation in community conservancies often include acts of everyday resistance -such as breaking into wildlife areas at night to graze and water livestock or slandering NRT. More broadly, everyday resistance refers to 'the ordinary weapons of relatively powerless groups: foot dragging, dissimulation, false compliance, pilfering, feigned ignorance, slander, arson, sabotage, and so on' (Scott 1985: 29). Consideration of bricolage practices associated with naturalisation, and of articulation specifically, leads to further reflection on the limitations and possibilities of human agency in settings like CBNRM.
Although bricoleurs participate in shaping institutional arrangements through conscious and unconscious actions, previous sections indicate that bricolage remains an authoritative process that is shaped by historical patterns of access and accumulation and by corresponding power relationships among bricoleurs in present-day Kenya. Working visibly and invisibly through taken-for-granted social orders (Lukes 2005), power 'shapes processes of bricolage through public decision-making, in everyday social interactions, and through broader social-structural factors' (Cleaver et al. 2013: 175). For example, a widespread lack of familiarity with the economic principles of conservation among pastoralists means that people are inherently disadvantaged in public decisionmaking processes that impact their lives in community conservancies -even though NRT prioritises the design of democratic and transparent institutions that are said to protect the autonomy and ownership pastoralist communities have over communal lands and natural resources. For the sake of illustration, a member of Il Ng'wesi leadership described how the Board of Directors failed to adequately inform the community about the implications of value deprecation during one AGM, when trying to achieve a consensus on the fate of the conservancy's Land Cruisers: 'We tried to explain the idea of value depreciation to the community during a recent AGM. They simply could not understand it. They asked, "Does the Land Cruiser still work? Do the wheels still spin? Then how can it depreciate?"' 23 Indeed, consideration for everyday social interactions in community conservancies and for broader social structures reveals how power works in and through hybrid institutions to reproduce and reinforce inequality; even in the context of institutional arrangements considered to be adaptive, innovative, and progressive. For example, foreign tourists' cultural biases against seeing livestock in wildlife areas while on safari dissuades some from spending time and money in community conservancies and, as such, serves as one bench mark against which NRT rewards or punishes partner communities. In broader terms, the legacies of a racialised property regime contribute to the fact that settler landowners have the privilege and power to devise mechanisms for leveraging pastoralist communities into conservation in the first place -a point that should not be understated in Laikipia (see Mbaria and Ogada 2016;McIntosh 2016;Bersaglio 2017). Although power works subtly through public-decision making processes and everyday social interactions in community conservancies, as well as through broader social structures, its effects place real and observable constraints on the opportunities available to pastoralist communities to adapt, negotiate, and resist hybrid institutional arrangements for managing communal lands and natural resources (Giddens 1984;Rabe and Saunders 2013;Ojha et al. 2016).

CONCLUSION
Building on recent critical engagement with the expansion of community conservancies across Kenya's ASALs, this article reframes the refurbishment and rearrangement of pastoralist institutions governing communal lands and natural resources to align with the ideas and aspirations of CBNRM as a form of green grab. Specifically, our analysis and discussion reveal the subtle ways that green grabs occur through processes of bricolage: the conscious and unconscious reshaping of institutional arrangements for managing communal lands and natural resources by bricoleurs, who make use of whatever resources and tools happen to be at their disposal (Cleaver 2012). In the case of Il Ng'wesi and other community conservancies across Kenya's ASALs, NRT policies and programmes represent a toolkit containing many of the resources available to transhumance pastoralists in the region. By relaying contextual insights offered by informants and interlocutors in Kenya, our empirical contribution offers glimpses into some of the undesired and unintended outcomes of community conservancies. Doing so also provides evidence of how green grabs work in and through established rules and practices governing communal lands and natural resources, as well as how the outcomes of such phenomena reflect historical patterns of access and accumulation in Laikipia specifically -a stronghold of settler society in Kenya (see Mbaria and Ogada 2016;McIntosh 2016;Bersaglio 2017).
The article's main theoretical contribution derives from its engagement with the question of agency in wider debate on land grabs and, by implication, green grabs (see Borras and Franco 2013;Hall et al. 2015). We suggest that a return to the question of agency should occur in tandem with nuanced consideration for the structures that shape individual behaviours and, in turn, limit the range of outcomes possible through CBNRM. In framing our analysis, we bring green grabbing literature into conversation with that on critical institutionalism to show how institutions mediate complex and sometimes contradictory interactions between actors at multiple scales with multiple identities to produce green grabs. Specifically, we depart from linear conceptions of power that frame green grabs as enacted from above and reacted to from below in pursuit of a more diffuse understanding of power that ultimately complicates false stereotypes of grabbers and grabbees and unsettles crude distinctions between acquiescence, incorporation, and resistance (see Gardner 2016;Astuti and McGregor 2017;Bersaglio 2017). Our analysis and discussion begin to show how green grabs occur through bricolage, which involves interrelated processes of aggregation, alteration, and articulation among diverse individuals and groups of bricoleurs. In this regard, green grab by bricolage produces contradictory institutional spaces animated by the promise of adaptation, collaboration, and ingenuity, but constrained by historically unequal patterns of access, accumulation, and domination between bricoleurs.