Landscape Research

ISSN: 0142-6397 (Print) 1469-9710 (Online) Journal homepage: https://www.tandfonline.com/loi/clar20

Contested mappings in a dynamic space: emerging

socio-spatial relationships in the context of REDD+.
A case from the Democratic Republic of Congo

Catherine Windey & Gert Van Hecken

To cite this article: Catherine Windey & Gert Van Hecken (2019): Contested mappings in a
dynamic space: emerging socio-spatial relationships in the context of REDD+. A case from the
Democratic Republic of Congo, Landscape Research, DOI: 10.1080/01426397.2019.1691983

To link to this article: https://doi.org/10.1080/01426397.2019.1691983

Published online: 02 Dec 2019.

Submit your article to this journal

Article views: 116

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=clar20
LANDSCAPE RESEARCH
https://doi.org/10.1080/01426397.2019.1691983

Contested mappings in a dynamic space: emerging socio-spatial

relationships in the context of REDD+. A case from the Democratic
Republic of Congo
Catherine Windey and Gert Van Hecken
Institute of Development Policy (IOB), University of Antwerp, Antwerp, Belgium

ABSTRACT KEYWORDS
This paper adopts a processual understanding of mapping to empirically Mapping; geospatial
explore the workings of satellite-based forest visualisations and maps knowledge; REDD+;
within the REDD+ process in DR Congo. Our analysis approaches maps socio-spatial relations;
socio-environmental justice;
as ongoing contingent practices and highlights the recursive interplay
performative assemblage;
between maps and the socio-natural world. We first show how REDD+ emergence
(mapping) assemblages enact a uniform portrayal of community-induced
threats to nature, which in turn legitimises a monoculture of abstract
space often to the detriment of communities’ authority over land and
their particular socio-ecological relationships. However, these mapping
attempts at reordering forest landscapes are locally met with and
reshaped by ever-emergent socio-spatial practices, ways of seeing and
appropriating landscape. Complexity, fluidity, and ambiguity are indeed
rendered absent by these seemingly immutable and complete represen-
tations although they are essential for understanding struggles over
resources. We conclude that adopting a processual understanding of
maps opens up ways of enacting socio-environmental justice.

1. Introduction
In 2009 the Democratic Republic of Congo (DRC) initiated the Reducing Emissions from Deforestation
and Forest Degradation ‘Plus’ (REDD+) global environmental policy process. Despite the very limited
data on forest cover change then available, the country’s annual gross deforestation was estimated at
0.23% for 2000–2010; substantially lower than the rates of Malaysia, Indonesia or Brazil (DRC-MECNT,
FCPF, & UN-REDD, 2015; Ickowitz, Slayback, Asanzi, & Nasi, 2015). One of the first major goals of
the REDD+ readiness phase was to gather scientific data to inform a ‘national consensus’ on the main
drivers of deforestation. These data would serve as input for building the country’s REDD+ strategy and
monitoring actions (UN-REDD & DRC-MECNT, 2012). The expert-based ‘national consensus’ was
grounded almost exclusively in the use of Landsat remote sensing satellite maps and geospatial
expertise, despite a range of critiques on the quality, validity and reliability of the data yielded
in four REDD+-sponsored studies (see Ickowitz et al., 2015; Moonen et al., 2016) and despite
the absence of local voices and on-the-ground research as denounced by several international activist
NGOs1 (e.g. Rainforest Foundation UK, interview 23/04/2018). The ‘national consensus’ posits shifting
cultivation and population growth as the main direct and underlying drivers of forest loss, with
industrial logging and mining activities having only limited impacts, though they are recognised
as potential future threats to forests (DRC-MECNT, FCPF, & UN-REDD, 2015).

CONTACT Catherine Windey [email protected]; [email protected]

This article has been republished with minor changes. These changes do not impact the academic content of the article.
© 2019 Landscape Research Group Ltd
2 C. WINDEY AND G. VAN HECKEN

On the basis of this consensus, the REDD+ strategy in DRC has largely evolved towards what is
labelled as jurisdictional or integrated landscape approach2 in policy discourses (McCall, 2016). The
landscape approach broadens the climate-, forest-, carbon- and performance-based initial emphasis
of REDD+ towards the integration of multiple stakeholders and multiple productive land-use types—
such as agriculture and mining—with a larger range of environmental and developmental goals at
a wider landscape scale. While the landscape approach is called ‘innovative’, some regard it as yet
another discursive commodity to ensure a continuous flow of finances to programmes that leave
previous conservation and development approach challenges largely untouched (Lund, Sungusia,
Mabele, & Scheba, 2017; McCall, 2016). As Clay (2016) argues, landscape approaches reproduce
uneven power relationships by promoting a particular vision of conservation coexisting with
productive and extractive activities. The use of zone-based, static land use models either disregards
or completely depoliticises local social-environmental dynamics; in part due to an increasing need
for expert knowledge and to the overly geospatialised interpretations of these dynamics (Clay, 2016;
McCall, 2016). The selective use of data, and the oversimplification of the messy world by employing
‘neutral’ scientific language, risks further exclusion of the poor (Ferguson, 1994; Li, 2007). With
technological advancements in satellite imaging and remote sensing tools ‘never-before-seen’ geo-
visualisations of forests have become more ubiquitous for studying deforestation and for policy
design and monitoring—making particular ‘realities’ visible while hiding others.
Building on insights from political ecology, Science and Technology Studies and critical carto-
graphy, this paper empirically looks at the role, use and contestation of geospatial environmental
visualisations and maps in producing ‘integrated’ green economic landscapes under REDD+ strate-
gies in DRC. We adopt a processual understanding of mapping (Kitchin & Dodge, 2007) to under-
stand how maps within DRC’s REDD+ assemblage reify particular socio-spatial relations and
identities while simultaneously being destabilised and recrafted in diverse spaces and times. After
discussing our conceptual, epistemological and methodological approach (Section 2), we then
analyse how a putative reality that blames local communities for deforestation is framed through
an assemblage of geospatial imageries, maps and discursive practices (Section 3). Subsequently we
examine how such a framing legitimises so-called inclusive REDD+ programmes that largely privilege
a monoculture (Santos, 2004) of abstract spaces managed through privately-held concessions (Clay,
2016) to the detriment of communities’ authority over land and over their specific socio-spatial
identities. Section 4 turns to the concrete implementation of such a rational spatialisation that tries
to mould a socio-ecological system to a given representation of what it should look like, through
geospatial and participatory mapping technologies (Pickering, 2013). Specifically, we analyse how
these mapping attempts are continuously reshaped by local dynamic socio-spatial practices, ways of
seeing and of appropriating landscape somehow rendered absent on the maps (Law, 2004). We
conclude that the continuous reliance on modernist technocratic framings and epistemologies and
their purely representational way of understanding the world denies both complexity and perfor-
mativity; thus always failing to account for emergence and multiple ways of being in and knowing the
world—often resulting in the perpetuation of socio-environmental injustices.

2. Mappings, power and emergence: a processual understanding

Political ecological scholarship on forest governance (e.g. Fairhead & Leach, 1996; Forsyth & Walker,
2008; McElwee, 2016; Robbins, 2001) has provided valuable insights into the ways in which Western
techno-scientific knowledge often lead to simplistic and largely ‘misread’ narratives of deforestation
that help constrain the livelihoods of local communities while availing space for more powerful capital
interests. Complex society-environment relationships are distilled into seemingly apolitical abstract
categories and reconfigured within a global environmental space and a unitary political-economic
rationale. Igoe (2013) specifically emphasises how scientific and aesthetic visual articulations of nature
have contributed to its abstraction, unitarisation and formation as a monolithic eco-functional object of
intervention, optimisation and commodification, while concealing and marginalising opposition and
 LANDSCAPE RESEARCH 3

alternatives to this vision. Similarly, Forsyth and Sikor (2013) show that the use of remote sensing data
for defining and measuring global REDD+ objectives might encourage both large-scale reductions in
deforestation rates and industrial selective logging rather than shifting cultivation and improvement in
forest quality/biodiversity that could locally be more beneficial.
Critical cartography (e.g. Harley, 1989; Pickles, 2004; Wood & Fels, 2008) has long analysed the
relationship between mapping and politics, the kinds of knowledge that maps produce and with
what effects, arguing that they do not simply represent the world but also produce it. Maps are
propositions about the world and are hence always political and ideological artefacts, embedded
in regimes of claims that construct meaning and ultimately behaviour. While this scholarship
makes indispensable contributions to our understanding of the power-laden character of geos-
patial environmental knowledge production, Kitchin and Dodge (2007) argue that it still tends
to consider maps as immutable and ontologically secure inscriptions. What has been less
explored is an epistemological approach to assess ‘lines of becoming’ and the multiple lives of
mappings, tracing out how they emerge time and again ‘in the entangled meshwork [. . .] of their
creation, use and unfolding of everyday life and space’ (Kitchin, Gleeson, & Dodge, 2013, p. 3;
Pickering, 2013). From this perspective, maps should be understood as a set of unfolding
practices to solve relational problems (e.g. spatial distribution of deforestation drivers) con-
stantly in the making, contingent, relational and ontologically unstable; they are always map-
pings. This necessitates a significant shift ‘from ontology (what things are) to ontogenetic (how
things become)’ (Kitchin et al., 2013, p. 15).
The ontogenetic nature of maps invites us to see them as assemblages that are imbricated
within other and/or larger assemblages, such as REDD+ in our case. From a Deleuzian and post-
representational perspective, the assemblage approach emphasises the constitution of
a contingent provisional unity between heterogeneous human and non-human components
that include material, discursive and social artefacts (Anderson & McFarlane, 2011). It is the
assembling processes of these components, in specific time and place, that are performative,
that is, they contribute to enact pre-existing social and material contexts (Yeow & Faraj, 2014).
Assemblages can hence produce different ontologies (e.g. counter-maps) and often follow unin-
tended routes as they are contingent on the fostering of new relations and the entrance of new
elements (Anderson & McFarlane, 2011).
Our starting point then is the contextually contingent (re)crafting of maps using specific data,
within DRC’s REDD+ assemblage of people and discursive regimes. Since a single coherent story of
deforestation is required for such a policy exercise, a single reality will necessarily emerge. We
concur with Wood and Fels (2008) that key here is the articulation of the map and the paramap,
that is, the verbal and nonverbal discourses that surround the map to position it and through
which argument and authority are conveyed. The paramap, in turn, consists of the perimap and the
epimap. The former refers to the physical map: title, legend, colours, graphs. The latter refers to the
discourse surrounding the map, that shapes reading and perception of the map. Although these
maps are given an appearance of immutability and predictability—their knowledge and message
fixed, and their emergent aspects edited out—they are in fact continuously recrafted, interpreted
and translated in many unpredictable ways every time they are engaged with (Kitchin & Dodge,
2007; Pickering, 2013). As Section 4 illustrates, maps are constantly remade and re-inscribed by
people’s everyday socio-spatial practices and bodily actions, as well as by the emergence of
previously invisible elements.
Our adopted approach thus refocuses (critical) cartographic research as ‘sciences of practices, not
[(un)truthful] representations’ (Kitchin & Dodge, 2007). This conceptual entry-point gives a more
active and political role to cartography. Most notably it reveals how maps emerge, within wider
contingent knowledge and discursive fields and forms of praxis, to enact some realities and solve
relational problems and to analyse the recursive interplay between maps and the socio-natural
world. The power of maps is relational rather than fixed and complete. Thus we aim to develop
a better understanding of the workings and work of mappings of Congolese forests and their people
4 C. WINDEY AND G. VAN HECKEN

from their genesis and unfolding within the REDD+ assemblage to some of their ‘multiple lives’ in the
world. Such insight highlights the sociality of maps, that is, how mappings differently affect various
actors and social relations such that technoscience and crucial issues of environmental justice are
and should be woven together.

Methodological approach
The empirical material informing this analysis draws on five months of ethnographic field research
carried out by the first author between May 2015 and June 2018 in DRC, combined with a desk study
of REDD+ policy documents. The field research included 64 semi-structured expert interviews and also
informal conversations with international and national forest experts, technical advisers and consul-
tants from REDD+ lead implementing organisations, government agencies, representatives of inter-
national and national environmental NGOs and civil society organisations plus private actors engaged
in REDD+ discussions. Complementary qualitative fieldwork was undertaken in the intervention areas
of four REDD+ programmes in the hinterland of Kisangani: 96 semi-structured and informal interviews
focused on understanding local processes and meanings of land tenure and resource use. In order to
assess socio-spatial relations the fieldwork also included five low-tech participatory mapping exercises
combined with focus group discussions. Far from using maps from a positivist/realist epistemological
stance that emphasises the ontological security of maps, we used mapping as a contingent, relational
discursive process that can produce different ontologies. Maps in this view are ‘merely a reflection of
land use [and tenure] at a particular time under a particular set of circumstances’ (Roth, 2009, p. 222)
and a product of specific negotiations. We also asked four land users to draw individual sketch maps—
hand-drawn cognitive maps of their land use spaces. Although participatory (counter-)maps were
seen by the communities as a powerful way to claim land rights, both of our low-tech mapping
exercises also revealed their limitations in understanding a complex and dynamic tempo-spatiality. We
thus complemented the counter-maps with four transect walks guided by the same four above land
users, allowing us to highlight movement and flexibility.

3. Performing the green economic abstract space in DRC

DRC’s REDD+ assemblage has given primacy to remote sensing satellite mapping—largely seen from
a positivist perspective as detached global knowledge par excellence—as an entryway to solve the
problem of deforestation. Simultaneously, local and socio-political understandings of socio-environment
relationships have largely been relegated. In recent years, the American-based initiative ‘Global Forest
Watch’ (GFW) has been highly reinvested by REDD+ and forestry practitioners in DRC to monitor forests.
This ‘near-real time’ online forest monitoring system initiated by the 40+ partner World Resource Institute
(WRI) is often praised as ‘the’ technical feat finally offering reliable, objective forest data. The assemblage
of GFW partners—whether users or funders is never made clear by WRI—comprises conservation non-
profits, major environmental organisations supporting and employing ecosystem/landscape approaches,
and multilateral cooperation agencies funding REDD+ programmes as well as blue-chip private corpora-
tions which include two of the world’s largest agribusiness industries/palm oil buyers. As WRI’s CEO
asserted, thanks to GFW ‘[f]rom now on, the bad guys cannot hide and the good guys will be recognised
for their stewardship’ (WRI, 2014, para.2). However, GFW does not mention who the ‘bad’ and ‘good’ guys
are; this is left to each user, as GFW presents itself as a neutral actor, supporting sustainable management
of human-environment relationships.

A seemingly neutral and disconnected representation

Using an assemblage of GFW’s analysis of thousands of geospatial images, spatial demographic
datasets and administrative data, (para)maps were created by DRC’s REDD+ policy makers to simplify
perceptive judgements and act as reference objects to assign blame and spatially plan REDD+ actions
 LANDSCAPE RESEARCH 5

(Ehrenstein, 2014; Wood & Fels, 2008). One map-paramap assemblage can be found on the website of
the main REDD+ investment programme Central African Forest Initiative (CAFI)3. It includes a section
with 29 frequently asked questions (FAQs) about ‘CAFI and the forestry sector in DRC’ (CAFI, 2017a).
Among the 29 FAQs, one question/subpage claims that it seeks to provide ‘proven’, evidence-based
facts on the main drivers of deforestation and land degradation. The subpage starts with a short and
straightforward text—the epimap—stating and emphasising in bold script that the expansion of
subsistence activities (slash-and-burn agriculture, and fuelwood collection and harvesting) is the
main cause of deforestation and forest degradation and is hence closely correlated to the spatial
distribution of population. The text also highlights—as the other 28 FAQs uniformly do—that ‘contrary
to popular belief, [industrial] forest exploitation is not systematically a driver of deforestation and
degradation’ (CAFI, 2017b, para. 2, own translation). As such, the FAQs seem to largely defend and
legitimise the highly contested4 Sustainable Forest Management Programme (PGDF) that supports
logging companies in developing their forest management plans.
Then comes the map (Figure 1) and its perimap. They appear on all of the three FAQs’ subpages
dealing with deforestation drivers. The map displays four carefully chosen human and non-human
elements: forest cover loss, population distribution, roads and ‘forest’ concessions. The map draws
on three spatial data sources: GFW (forests), WorldPop (population) and the DRC’s Common
Geographic Reference System (roads and administrative limits). The map highlights forest cover
loss for the period 2000–2014 in a bright red colour and superimposes it on population density
explicitly suggesting a direct causal link. Logging concessions—framed as ‘forest concessions’—are
represented in non-threatening green, the traditional colour of trees and forests, suggesting the
absence of deforestation in those spaces. The (para)map design, with its particular use of map
colours and bold characters in the text, distracts the reader from the forest fragmentation that
appears within these concessions. Similarly, the (para)map only refers to undefined forest cover loss

Figure 1. Main map on the CAFI REDD+ investment’s programme website (CAFI, 2017b).
6 C. WINDEY AND G. VAN HECKEN

and silences different categories of forests and deforestation. It also does not refer to forest
degradation—an explicit REDD+ concern regarding forest quality—nor to biodiversity or any alter-
native local definitions of forest (use).
Yet, if we focus on cover loss, some GIS scholars—among whom some contribute to GFW—increas-
ingly argue that in DRC, ‘the majority of tree cover loss [. . .] is accounted for by shifting cultivation onto
previously farmed lands rather than new deforestation’ (de Araujo Barbosa, Maschler, Bonfils, & Molinario,
2018, para., p. 5), hence showing different interpretations of land use change. In the same way, the generic
definition of forest in DRC’s REDD+ official documents is an adaptation from the definition provided by
FAO and the Clean Development Mechanism, which can include (industrial) tree crop plantations (UN-
REDD & DRC-MECNT, 2012). ‘Forest’ in that discourse is understood as a unified category of carbon stock,
as is also made clear by another map (Figure 2)—the only one in the REDD+ national investment plan
(CAFI, 2015). In Figure 2, remote sensing maps are transformed into a rough cartographic representation
that performs the forest as scarce carbon stocks, that is, simultaneously as a threatened and a possessable
resource (Wood & Fels, 2008), while rendering all other elements invisible.
Returning to Figure 1, population—just like forest loss—is not qualified. While the (para)map
refers both to ‘population density’ and ‘spatial distribution of population’ other parts of the website,
as well as the REDD+ studies on deforestation drivers and strategy documents, use the term
‘population growth’. These concepts have very different meanings and implications. The REDD+

Figure 2. ‘Democratic Republic of Congo: a threatened carbon stock’ (CAFI, 2015).

 LANDSCAPE RESEARCH 7

strategy makes things clearer: the lack of family planning and education leads to uncontrolled
population growth that ‘poses huge problems in terms of environmental and natural resources
preservation, for forestry ecosystems among others’ (DRC-MECNT, FCPF, & UN-REDD, 2015, p. 77,
own translation). The strategy does not provide a further explanation for this causality. Similarly,
Figure 1 only tells us that human presence and forest cover loss are correlated. No political-economic
or historical contextualisation is provided regarding the resettlement of population around roads
during colonial times or conflicts forcing relocation, or about unequal access to resources, distribu-
tion of wealth or market access. In Section 4 we show that such information could lead to very
different understandings of deforestation drivers. However, REDD+ synthesis of deforestation drivers
in DRC acknowledges that defining underlying drivers is very complicated and direct observation of
these underlying drivers is impossible (UN-REDD & DRC-MECNT, 2012); meaning, in the document’s
vocabulary, that it is impossible to observe/represent them with geospatial technology from an
office in Kinshasa or from elsewhere in the world. ‘Population’ is the exception: it appears to be easily
represented (geo)spatially, and is largely depicted as a technically manageable matter of family
planning. Heterogeneous dynamics of resource use among a seemingly homogeneous ‘population’
remain hidden.
So, REDD+ (para)maps and documents simultaneously co-produce homogeneous categories of
forests and users that are driven by implicit norms (Forsyth & Sikor, 2013). Figure 3, for instance,
shows the REDD+ priority zones of the CAFI Programme Intégré Oriental that were determined on the
basis of ‘deforestation hotspots’, a.k.a. ‘zones with high population density’ (FONAREDD & UNDP, n.
d., p.13). In reality, densities largely differ from one zone to another: from more than 400 people/km2
in Kisangani to an average of 6 people/km2 in more remote regions. The legend points out that the
hotspots correspond to livelihood activities, that is, slash and burn agriculture (in red circles),
fuelwood production, artisanal logging and mining. Industrial activities in the area are not identified,
for example, in the Tshopo province, 16 logging concessions (occupying 21,5% of the province’s
territory) and 2 palm oil tree plantations; neither are protected areas (4% of the Tshopo Province;

Figure 3. Priority zones of the CAFI Programme Intégré Oriental, in 3 provinces (FONAREDD and UNDP, n.d).
8 C. WINDEY AND G. VAN HECKEN

31% of Bas-Uélé Province). However, as our fieldwork revealed, the presence of logging concessions
and roads attracts ‘migrants’ seeking livelihoods as well as many high-ranking political and military
elites who start cash-crop plantations or engage in artisanal logging. This suggests much more
heterogeneity than the category ‘population’ reveals. Also the omissions in the map hide the fact
that the presence of concessions or protected areas often forces increasing numbers of peasants to
encroach other forests for their livelihood activities, as explained by several of our informants.

A ‘successful’ discourse
The (para)maps’ propositional logic of a community-induced deforestation was continuously
restated—especially at central level in Kinshasa—during our interviews with national and inter-
national REDD+ consultants in DRC and with various other state and non-state actors involved in
or targeted by the programme. Our analysis of these interviews and of programme documents
shows a constant use of essentialising discursive binaries such as artisanal versus industrial,
artisanal versus formal, subsistence versus productive, slash-and-burn versus intensive, collective
versus individual. ‘We have never seen that the collective can manage anything correctly’, one of
the REDD+ national coordination’s main (international) Technical Officers told us [25/02/2016]. In
an attempt to explain the reliance in the hinterland of Kisangani on traditional subsistence
agriculture that causes deforestation rather than on cash crop/productive agriculture,
a Provincial Coordinator of REDD+ programmes told us: ‘here people are passive [. . .] they are
less dynamic, entrepreneurial than in other regions’5 [01/03/2016]. Our interviews with govern-
ment actors and with national civil society organisations (CSOs) in Kinshasa—whose represen-
tatives are typically urban elites—revealed a similar strong prejudice against those who are seen
as primitive peasants with age-old practices. While CSOs constantly challenge state institutions’
legitimacy to make decisions regarding forest policies and REDD+, the large majority of them
never question the mainstream narrative of deforestation precisely because it appears as
scientific and depoliticised, as neutral and normatively superior. Typically, CSOs recognise that
REDD+ design in DRC is not perfect but they do not question the foundations of the programme
and its continuation, unlike international environmental NGOs [President of the Congolese civil
society platform for REDD+, 28/05/2018]. In turn, as a high-ranking official at the Tshopo
Provincial Ministry of Environment told us [11/05/2018], for state actors, assigning logging or
agricultural concessions can represent much larger financial gains for ministers at central level
(rent-seeking behaviours) compared to supporting local livelihoods. In the areas around
Kisangani targeted by REDD+ programmes, the well-oiled discourse about communities’ ‘beha-
vioural problems’ was also repeated by local opinion leaders trained and informed via REDD+
communication and awareness plans. In contrast, the impact of industrial companies was rarely
mentioned except when the president of one of the (trained) local Agricultural Administration
Councils, asserted that ‘even if an agricultural plantation deforests, it stays in a limited space
while [local] population, they, are always on the move’ [21/03/2016]. ‘Subsistence agriculture,
that one, is insatiable’ [one REDD+ national coordination’s Technical Officer, 25/02/2016].
While these actors’ justifications for REDD+ orientation are different, these binaries can funda-
mentally be understood as embedded in a socio-tempo-spatial monoculture of productivity (Santos,
2004). This monoculture normatively privileges a linear, sequenced and bounded time-space asso-
ciated with the rationality of formal industrial and commercial activities. It corresponds to what is
seen as a productive time-space that is legitimised by the different actors we interviewed through its
economic significance for the country’s green growth. This uniform discourse simultaneously creates
the non-existence of periodic, discontinuous and unbounded time-spaces by describing informal
artisanal and subsistence activities as irrational. It is perceived as an incidental time-space in a double
sense: as backward and secondary for national growth, and as unplanned, disordered and uncon-
tained in relation to the natural environment. On this basis rights to land and forest resources and
legitimacy to deforest (or not) are distributed. Clearing trees appears acceptable as long as it is
 LANDSCAPE RESEARCH 9

planned and bounded in space and time, like continuous and linearly growing industrial exploita-
tion, something the same REDD+ national coordination’s Technical Officer referred to as ‘accom-
panying deforestation’ [25/02/2016].
The overall goal of DRC’s REDD+ programme is to ‘supervise’ and incentivise local rural communities
to settle their ‘harmful’ activities and maintain forest cover in order for DRC to be able to support
increasing private and international demand for commercial and industrial agricultural land, timber
and minerals (DRC-MECNT, FCPF, & UN-REDD, 2015). The involvement of the private sector and, in
particular industrial timber corporations, has gradually increased as the perspectives of REDD+ in DRC
entering the investment phase became clearer, allowing them to both define and benefit from the
REDD+ agenda (DRC-MECNT, FCPF, & UN-REDD, 2015; Mpoyi, Nyamwoga, Kabamba, & Assembe-
Mvondo, 2013). The REDD+ CAFI programme, as one of its stakeholders told us, is ‘really a rational
approach, to reinforce control’, it is a ‘big zoning and land use plan for development’ [19/04/2018]. In
other words, it is about spatialising what Igoe (2013, p. 43) conceptualises as an ‘eco-functional
nature [. . .] amenable to technological re-orderings that will optimise economy and ecology, or at
least accommodate putatively inevitable growth with minimal disruption to ecosystems and human
well-being’. In this logic, the model of the privately-held, sustainably managed and zone-based scaled
concession acts as a socio-spatial reference institution. The demarcation of land-use zones (e.g.
permanent production forests, conservation zones, subsistence agriculture zone, large-scale agricul-
tural land), advocated by the REDD+ spatial-based approach to landscape operates with the same
practices of industrial logging for delimitating territory and resources and negotiating the circum-
stances of extraction, production and social benefits’ distribution (Clay, 2016). REDD+ in this case can
be seen as a scaling up of the model of the forest management plan to a larger landscape. The approval
of a highly contested US$18 million PGDF programme in the REDD+ investment plan is a striking
example of this approach. CAFI justifies the programme by appealing to the definition of REDD+
activities adopted by the conference of parties in 2010 under the Framework Convention on Climate
Change, that indicates ‘sustainable management of forests’ as one possible activity (CAFI, 2017a). In
practice, PGDF mainly supports the logging sector to develop forest management plans to organise
their production more ‘rationally’, profitably and ‘ecologically’ while increasing control over community
and artisanal activities within the concession boundaries (FONAREDD, 2016).

4. Unfolding mappings of the hybrid eco-functional landscape

In the hinterland of Kisangani the case of the establishment of the forest management plan (FMP) of
a foreign company’s timber concession (ITC)6, whose entrance is situated just 20km from the city,
provides a good example of a codification and mapping of a smaller scale integrated landscape. It relies
heavily on geospatial mapping technologies but also on participatory mapping as ‘means of translating
the community’s land [use and] tenure system into spatial information, which will be technically
accepted and understood by [all landscape stakeholders]’ (Astuti & McGregor, 2015, p. 2285). While
the example raises important issues of ownership and authority over landscape, it also emphasises the
ontological insecurity of the dominant abstract representation of space and its seemingly immutable
and predictable categories of forests and population as we described in the previous section. In fact,
maps and space ultimately emerge in context and are constantly (re)made in a dialectical relation with
everyday socio-spatial practices and with people who engage with this type of dominant spatial
representation (Kitchin et al., 2013; Sletto, 2002), leaving planners with unexpected challenges and
outcomes.
Two elements of the FMP directly impacting communities are key. One concerns the mapping of
boundaries of ethnic subgroups’ lands on which a concession is installed so as to distribute the benefits
of the social responsibility contract, that is, the part of the FMP that establishes specific socio-economic
investments by the company for local communities. This is done on the basis of maps and decrees from
the colonial Belgian administration and by the use of GPS coordinates collected by a team composed of
a company’s representative, local communities’ representatives, provincial administration and a GIS
10 C. WINDEY AND G. VAN HECKEN

expert from the provincial forest administration. A second element concerns the spatial organisation
and delimitation of the concession into four broad use categories: industrial timber production area,
rural development zone (for community activities), conservation and protection zones. The demarca-
tion of these land-use zones is mainly premised on two socio-ecological categories that allow estima-
tion of community forest conversion rates: density of forest cover and existing road networks as
identified by satellite imagery. The size of the Rural Development Zone (RDZ) was estimated for a 25-
year period. However, in 2017, about two years after the plan’s approval, the RDZ’s boundaries were
already reached and activities involving tree clearing (agriculture, charcoal production or artisanal
logging) were taking place even within the limits of the production zone, that is, outside of the RDZ.
This completely ‘remapped’ the zone (Figure 4). As two consultants involved in the establishment of
the FMP told us with consternation, ‘everything is moth-eaten in this concession’ despite the commu-
nities having approved the limits of the RDZ in a participatory way [29/05/2018]. The astonishment
suggests a lack of understanding of (and/or willingness to engage with) socio-political dynamics
shaping access to land and the complex and multiple land use dynamics in this peri-urban area.
The participatory nature of the mapping of community land tenure and use boundaries was
intended to legitimise the FMP. Through the participatory process the FMP assigned bounded areas
of the concession to four ethnic subgroups with the idea that as customary land owners they would

Figure 4. Landsat 8 satellite map of a part of ITC’s concession, from 10/05/2017. The mapped RDZ was reshaped as it extended
outside of its foreseen boundaries (we highlighted some of these extensions with red circles).
 LANDSCAPE RESEARCH 11

regulate access to forest and have customary power and control over the RDZ. Similar to colonial
land administrators, the FMP assumed that people living together in one space are part of a same
homogeneous ethnic subgroup and are governed by one common chief, thereby ignoring the
heterogeneity within so-called ‘communities’ and different interests at local level. Indeed, the FMP
and the demarcation of the RDZ was signed by chiefs of the four ethnic subgroups. However, the
land ownership based on a first occupancy status by one of these ethnic subgroups (the Bevenzeke),
was strongly contested by other groups who were not recognised in the plan. This conflict was
fuelled by political deputies who had clear interests in enlarging their constituencies.
As expected, our numerous conversations with different groups never provided an answer to the
primary occupancy question, nor did the historical maps of the region. While the FMP seeks to fix
boundaries between different groups and to record permanent customary ownership, for a long
time such ownership and authority were fluid in this part of the concession. This was due to a historic
absence of road infrastructure that meant forests thrived because of a relative lack of competition for
resources. By definition, this was a non-prescriptive space characterised by fluid appropriation,
diversity and disorder. In this specific space—identified as vacant land during colonial times—
boundaries were continuously negotiated in time and space rather than being permanent and
static. Tenure and also land use were flexible rather than defined in a comprehensive way. The
ones who are now identified as ‘migrants’ by the Bevenzeke, created villages and even became
customary chiefs who could grant access to land and allocate (very open) user rights to newcomers.
Land use activities of the different ‘migrant’ groups were effectively unregulated as resources were
plentiful. They could (and still can in practice) freely cut trees for opening a diamond quarry, for
agriculture or for charcoal production. However, with the extension of ITC’s permanent logging road,
large plots of land have been easily acquired by urban elites or alienated forcefully by military or
political players for land speculation or for establishing cash crop plantations, charcoal production,
or artisanal logging. The urban elites’ impact on forests is in fact often much higher than local
communities’ who have less means to exploit extensive resources.
These heterogeneous communities resist the mapped production of a bounded ethnic space
through their everyday dynamic land appropriation and use thus defying the official representation
of space. Our participatory and sketch mapping exercises revealed the difficulty of delimiting fixed
categories of occupation and use. Rather, the discussions these exercises raised and the transect walks
we carried out highlighted a spatial epistemology based on movement (Roth, 2009). Socio-spatial
practices, characterised by periodicity and extension, often lead villagers to install small temporary
camps in the forest that are used for several weeks to carry out diverse agricultural and forest activities,
then abandoned for years and then re-used. The landscape, far from being separated into units of
agricultural lands and forests is envisaged in a much more holistic way: fallow swidden lands are
interrupted by denser secondary forests in the middle of which one can find a small field of plantain
banana or a stack of logs waiting to be transformed into charcoal. New villages and land use activities
settle along new logging roads, leaving some more remote areas (sometimes temporarily) unoccupied.
The sporadic maintenance of former logging roads leaves a fragmented forest structure with varying
levels of cover and degradation; these patterns mostly do not appear on maps, since small logging
roads are usually not detected by satellites. As such, various categories of forest cohabit through
continua, successions, cycles and change (Robbins, 2001). Additionally, often hidden from the main
road, powerful elites alienate large areas of land and resources, sometimes in the production zone of
the logging concession, to install cacao or palm oil plantations, or for artisanal logging.
This ‘dwelling space’ is thus in a constant state of becoming, imprinted by non-bounded socio-
spatial identities and practices, culture and power relations rather than ‘already formed, ready to be
occupied’ (Roth, 2009, p. 210). It highly contrasts and challenges the abstract space that zone-based
planning aims at producing through an assemblage of mapping and boundary-making practices. Our
case study illustrates the constant state of transformation such an assemblage is always embedded in.
While it aims at reordering a landscape that would bear the imprint of (desired) homogeneous
categories (new codes) of community and forest circulating within REDD+ communities, this attempt
12 C. WINDEY AND G. VAN HECKEN

is constantly contested and destabilised by local socio-cultural and political relationships. Remote
sensing and participatory GIS maps produced as part of the FMP were in this sense socially and
politically reshaped when they were engaged with by heterogeneous communities. The maps’ appar-
ent ontological security allowed the Bevenzeke to claim formal rights to land and to benefit from ITC’s
social responsibility contract while jeopardising the possibility of managing a changing reality in more
dynamic ways—as was previously the case—hence fuelling conflicts between different groups. At the
same time, the maps’ ontological assumptions were highly challenged as they revealed their failure to
truthfully capture communities’ tempo-spatiality.

5. Conclusion
Through the application of a processual and performative understanding of mapping we have, in
this paper, challenged modernist and purely representational conceptions of maps and socio-
environmental phenomena. We have argued that the usual way of governing the planet solely on
the basis of supposedly true and complete representations (Pickering, 2013) not only fails to account
for emergent and complex socio-ecological problems but also denies alternative ways of knowing
and being in the world. In the context of DRC’s REDD+ process we have shown that satellite-based
maps act as reference objects—secure representations within an epistemic regime of claims and
meanings that is deeply rooted in a monoculture of linear and bounded socio-tempo-spatiality.
While these maps appear as ‘natural arbiters’ for defining landscapes, in practice they reflect an
‘already settled dispute about the nature of [. . .] landscape[s]’ (Robbins, 2001, p. 175). As such, they
reveal particular embedded values about desirable productive, formalised and privatised landscapes
inherent in carbon forestry (Forsyth & Sikor, 2013). Dominant mappings in REDD+ assemblage
contribute to produce a unified spatial ontology (this is there) (Wood & Fels, 2008) of deforestation
that blames the livelihood practices of a homogeneous and apolitical category of ‘local population’.
Simultaneously the complexity of land use and power struggles over resources are obscured.
Through these processes, the REDD+ landscape approach continues to push Western techno-
scientific knowledge to arbitrarily create green economic and ‘rational’ socio-spatial identities and
practices despite a purported holistic way of looking at land use with special attention given to local,
native and marginalised groups.
As Section 4 above showed, environmental planners and cartographers are faced with constant
challenges and unpredicted outcomes as maps continue to be thought as ontologically secure and
purely technical rather than relational. Our argument illustrates that while satellite-based and partici-
patory maps in REDD+ governance are used for providing certainty and control over a seemingly
neutral environmental space, they are reappropriated in different ways for solving relational problems
and are thus contingent and never stable. Rather, these maps (as any map) enact some realities rather
than others and thus always contain uncertainties and traces of exclusion—whether there are
excluded communities, categories of forest, land use activities or spatial identities—that open them
to contestation, reinterpretation and assignment of new meanings. To paraphrase Pickering (2013), the
usual way of governing human–environment interactions that assumes a chain of causes and effects
too often ignores temporal and spatial emergence in this (too) lively world. ‘We can interfere performa-
tively with it, and it will respond, but there is no guarantee whatsoever that the response will be what
we expect’ (Pickering, 2013, p. 19).
To be clear, we are not advocating against any uses of geospatial knowledge in REDD+ environ-
mental governance. We rather argue for maps’ re-socialisation, that is accounting for their performa-
tivity and their partiality, as well as their provisional and emerging character. It is about imagining other
ways of approaching mapping that pays much more attention to how and why decisions are made, to
the diverse context-dependent ways maps can emerge, and how maps and the world are always co-
constituted. This type of mapping rather than enframing is much more open to negotiation and debate,
and to performative interaction with more elusive dimensions of existence. More importantly—as it is
impossible to disentangle the world from procedures to report on it, as the power of maps is relational
 LANDSCAPE RESEARCH 13

rather than inevitable—such a cartography can bring in socio-environmental justice concerns. It indeed
opens up to enacting what was rendered invisible or marginalised by dominant representations of the
world, and to a willingness to account for what ‘surprises’ modernist science such as different socio-
spatial identities and ways of inhabiting the world.

Notes
1. International environmental NGOs (like Greenpeace, Rainforest Foundation UK (RFUK) or Global Witness) much
more than Congolese environmental civil society organisations are contesting these studies and the framing of
the REDD+ programme in DRC. While a detailed analysis of the heterogeneity of opinions among NGOs/CSOs
would exceed the scope of this paper, it is important to note that their views on the REDD+ process, forestry
reforms and industrial logging are sometimes divergent.
2. In REDD+ literature, both terms are used and refer to the same model for aligning multiple land-use types with
administrative jurisdictions and coordinate multiple goals, initiatives and stakeholders. In the remainder of the
paper we only refer to ‘landscape approach’.
3. CAFI and its related fund FONAREDD (National REDD+ Fund) is a partnership between DRC, multilateral and
bilateral donors, for the implementation of the REDD+ investment phase. It acts as a coordinating body of the
REDD+ process in DRC. Its website has become one of the main references regarding the REDD+ overall strategy
and programmes in DRC.
4. PGDF was and still is contested by some international activist ENGOs and some Congolese civil society
organisations.
5. Quotes from interviews have been translated from French as literally as possible by the authors.
6. For reasons of anonymity, we refer hereafter to the fictive acronym ITC (International Timber Company).

Acknowledgments
The field research informing this article was supported by the IOB Research Fund of the Institute of Development Policy
at the University of Antwerp and by mobility grants from the Research Foundation – Flanders (FWO). We are extremely
grateful to the many people in Kinshasa and Kisangani and its hinterland who welcomed us into their homes and
offices. We would also like to thank the special issue’s editors and two anonymous reviewers who provided very helpful
comments on earlier versions of this article. The responsibility for the positions expressed as well as any remaining
errors is exclusively ours.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work was supported by the Institute of Development Policy (University of Antwerp), Grant ID: IOB Research Fund;
and by the Research Foundation – Flanders (FWO), Grant ID: Grants for a long stay abroad (V412416N, K224216N,
V410818N).

Notes on contributors
Catherine Windey is a PhD Candidate and Academic Teaching Assistant at the Institute of Development Policy,
University of Antwerp. Her work lies at the crossroads of anthropology, critical geography, political ecology and
Science and Technology Studies, and focuses on global environmental politics, socio-spatial relationships, local under-
standings of environmental change and environmental justice. More specifically, her current research looks at material-
discursive practices in forest conservation mechanisms, that homogenise ecologies, societies and places within the
abstract space of (eco-)modernisation and at how they reconfigure and are resisted by everyday socio-spatial,
economic, cultural and political practices in relation to the natural environment. Over the last 5 years she has conducted
fieldwork in DR Congo and previously, in Cameroon. She is also the project leader of an academic cooperation project
(VLIR-UOS) with two Congolese universities that is dedicated to building knowledge and academic capacity in political
ecology of forest resource management and deforestation in DRC. She teaches classes in participatory and action
research methods and in the governance of natural resources.
14 C. WINDEY AND G. VAN HECKEN

Gert Van Hecken is Assistant Professor in International Cooperation and Development at the Institute of Development
Policy, University of Antwerp. His main expertise and research interests lie in the area of global and local environmental
governance and development, and more specifically in the local socio-political dynamics triggered by climate change
and development finance instruments, such as carbon and biodiversity markets and payments for ecosystem services.
For over fifteen years he has lived and worked in Nicaragua, both as a researcher on social-environmental dynamics in
rural communities and as a practitioner for a rural development NGO.

ORCID
Catherine Windey http://orcid.org/0000-0003-4027-8712
Gert Van Hecken http://orcid.org/0000-0001-8496-3194

References
Anderson, B., & McFarlane, C. (2011). Assemblage and geography. Area, 43(2), 124–127.
Astuti, R., & McGregor, A. (2015). Responding to the green economy: How REDD+ and the one map initiative are
transforming forest governance in Indonesia. Third World Quarterly, 36(12), 2273–2293.
CAFI. (2015). Plan d’investissement REDD+ (2015–2020). Retrieved from: http://www.cafi.org/content/dam/cafi/docs/drc-
documents/DRC_2015_Plan%20Investissement%20National%20REDD%20RDC%2024%20Novembre%202015.pdf
CAFI. (2017a). CAFI et le secteur forestier en RDC: Foire aux questions [Online]. Retrieved from: http://www.cafi.org/
content/cafi/fr/home/events/drc-forests—frequently-asked-questions.html Accessed 1 March 2018
CAFI. (2017b). CAFI et le secteur forestier en RDC: Foire aux Questions. FAQ 03: What contributes most to deforestation
and forest degradation in DRC? [Online]. Retrieved from http://www.cafi.org/content/cafi/fr/home/all-news/drc-
forests—frequently-asked-questions/what-contributes-most-to-deforestation-and-forest-degradation-in.html
Accessed 22 August 2018
Clay, N. (2016). Producing hybrid forests in the Congo Basin: A political ecology of the landscape approach to
conservation. Geoforum, 76(Supplement C), 130–141.
de Araujo Barbosa, C., Maschler, T., Bonfils, D., & Molinario, G. (2018, August 28). New map helps distinguish between
cyclical farming and deforestation in the Congo Basin [Blog post]. Retrieved from https://blog.globalforestwatch.org/
data/new-map-helps-distinguish-between-cyclical-farming-and-deforestation-in-the-congo-basin
DRC-MECNT, FCPF, & UN-REDD. (2015). Stratégie-cadre nationale REDD de la République Démocratique du Congo. Version
3. Retrieved from: https://www.forestcarbonpartnership.org/sites/fcp/files/2015/March/Strategie-cadre%20natio
nale%20REDD%20de%20la%20RDC.pdf
Ehrenstein, V. (2014). Géopolitique du carbone : L’action internationale pour le climat aux prises avec la déforestation
tropicale (Doctoral Dissertation). Ecole Nationale Supérieure des Mines de Paris, Paris.
Fairhead, J., & Leach, M. (1996). Misreading the African landscape. Society and ecology in a forest Savannah Mosaic.
Cambridge: Cambridge University Press.
Ferguson, J. (1994). The anti-politics machine: Development, depolitization and bureaucratic power in Lesotho.
Minneapolis: University of Minnesota Press.
FONAREDD. (2016). Note d’Idée de programme au Fonds National REDD+. AMI n°4 – programme de Gestion durable des
forêts. Retrieved from: http://www.fonaredd-rdc.org/wp-content/uploads/2017/06/NIF-FONAREDD-GESTION-FORET-
AFD-05072017.pdf
FONAREDD, & UNDP. (n.d.). Programme Intégré Oriental. Retrieved from: http://www.cafi.org/content/dam/cafi/docs/
drc-documents/DRC-Approved%20Programmes/DRC-PI-Orientale-UNDP/DRC-UNDP-PIREDD-Oriental-Prodoc.pdf
Forsyth, T., & Sikor, T. (2013). Forests, development and the globalisation of justice. The Geographical Journal, 179(2),
114–121.
Forsyth, T., & Walker, A. (2008). Forest guardians, forest destroyers: The politics of environmental knowledge in Nothern
Thailand. Seattle and London: University of Washington Press.
Harley, J. B. (1989). Deconstructing the map. Cartographica: the International Journal for Geographic Information and
Geovisualization, 26(2), 1–20.
Ickowitz, A., Slayback, D., Asanzi, P., & Nasi, R. (2015). Agriculture and deforestation in the Democratic Republic of the
Congo. A synthesis of the current state of knowledge (CIFOR Occasional Paper no.119). Bogor: CIFOR.
Igoe, J. (2013). Nature on the move II: Contemplation becomes speculation. New Proposals: Journal of Marxism and
Interdisciplinary Inquiry, 6(1–2), 37–49.
Kitchin, R., & Dodge, M. (2007). Rethinking maps. Progress in Human Geography, 31(3), 331–344.
Kitchin, R., Gleeson, J., & Dodge, M. (2013). Unfolding mapping practices: A new epistemology for cartography.
Transactions of the Institute of British Geographers, 38(3), 480–496.
Law, J. (2004). After method. Mess in social science research. Abingdon: Routledge.
 LANDSCAPE RESEARCH 15

Li, T. M. (2007). The will to improve. Governmentality, development, and the practice of politics. London: Duke University
Press.
Lund, J. F., Sungusia, E., Mabele, M. B., & Scheba, A. (2017). Promising change, delivering continuity: REDD+ as
conservation fad. World Development, 89, 124–139.
McCall, M. K. (2016). Beyond “Landscape” in REDD+: The imperative for “Territory”. World Development, 85(Supplement C),
8–72.
McElwee, P. (2016). Forests are gold: Trees, people, and environmental rule in Vietnam. Seattle: University of Washington
Press.
Moonen, P. C. J., Verbist, B., Schaepherders, J., Bwama Meyi, M., Van Rompaey, A., & Muys, B. (2016). Actor-based
identification of deforestation drivers paves the road to effective REDD+ in DR Congo. Land Use Policy, 58, 123–132.
Mpoyi, A. M., Nyamwoga, F. B., Kabamba, F. M., & Assembe-Mvondo, S. (2013). Le contexte de la REDD+ en République
Démocratique du Congo. Causes, agents et institutions (CIFOR Occasional Paper no.84). Bogor: CIFOR.
Pickering, A. (2013). Being in an environment: A performative perspective. Natures Sciences Sociétés, 21, 71–83.
Pickles, J. (2004). A history of spaces: Cartographic reason, mapping and the geo-coded world. London and New York:
Routledge.
Robbins, P. (2001). Fixed categories in a portable landscape: The causes and consequences of land-cover categorization.
Environment and Planning A: Economy and Space, 33(1), 161–179.
Roth, R. (2009). The challenges of mapping complex indigenous spatiality: From abstract space to dwelling space.
Cultural Geographies, 16(2), 207–227.
Santos, B. D. S. (2004). The WSF: Toward a counter-hegemonic globalization. In J. Sen, A. Anand, A. Escobar, &
P. Waterman (Eds.), World social forum: Challenging empires (pp. 235–245). New Delhi: The Viveka Foundation.
Sletto, B. (2002). Producing space(s), representing landscapes: Maps and resource conflicts in Trinidad. Cultural
Geographies, 9(4), 389–420.
UN-REDD & DRC-MECNT. (2012). Synthèse des études sur les causes de la déforestation et de la dégradation des forêts en
République Démocratique du Congo. Retrieved from: https://unredd.net/index.php?option=com_docman&view=
list&slug=studies-publications-and-other-resources-500&Itemid=134.
Wood, D., & Fels, J. (2008). The natures of maps. Chicago IL: Chicago University Press.
WRI. (2014, February 19). RELEASE: Global forest watch: Dynamic new platform to protect forests worldwide [Online
press release]. Retrieved from: https://www.wri.org/news/2014/02/release-global-forest-watch-dynamic-new-
platform-protect-forests-worldwide.
Yeow, A., & Faraj, S. (2014). Technology and Sociomaterial Performation. In B. Doolin, E. Lamprou, N. Mitev, & L. McLeod
(Eds.), Information systems and global assemblages. (Re)Configuring actors, artefacts, organizations. IS&O 2014. IFIP
advances in information and communication technology (Vol. 446, pp. 48–65). Berlin, Heidelberg: Springer.