Forest and Communities

Deforestation, Conservation, and Socio-ecological

Relations in the Mau Forest, Kenya

Stefania Albertazzi
Valerio Bini
Guido Trivellini
Stefania Albertazzi, Valerio Bini, Guido Trivellini

FOREST AND
COMMUNITIES
Deforestation, Conservation,
and Socio-ecological Relations
in the Mau Forest, Kenya

FONDAZIONE UNIMI_PRESS 1
Forest and Communities: Deforestation, Conservation and Socio-ecological Relations in the Mau Forest,
Kenya / Stefania Albertazzi, Valerio Bini, Guido Trivellini. Milano: Milano University
Press, 2023.
ISBN 979-12-80325-78-5 (print)
ISBN 979-12-80325-80-8 (PDF)
ISBN 979-12-80325 71-6 (EPUB)
DOI 10.54103/milanoup.89

This volume, and Milano University Press publications in general, unless otherwise
specified, are submitted to an external refereeing process under the responsibility of
the Milano University Press Editorial Board. The works published are evaluated and
approved by the Editorial Board of the publishing house, and must be compliant with
the Peer review policy, the Open Access, Copyright and Licensing policy and the Pub-
lication Ethics and Complaint policy as reflected in MilanoUP publishing guidelines
(Linee Guida per pubblicare su MilanoUP).
The present work is released under Creative Commons Attribution 4.0 - CC-BY-SA,
the full text of which is available at the URL: https://creativecommons.org/licenses/
by-sa/4.0

This and other volumes of Milano University Press are available in open access at:
https://libri.unimi.it/index.php/milanoup

© The Authors for the text 2023

© Milano University Press for this edition
Published by Milano University Press
Via Festa del Perdono 7 – 20122 Milano
Web Site: https://milanoup.unimi.it
e-mail: [email protected]
The print edition of this volume can be ordered from all physical and online book-
stores, and is distributed by Ledizioni (www.ledizioni.it)

General note on figures and tables

Unless otherwise indicated, figures and tables have been prepared by the author(s) of
the chapter in question. The Editor is committed to striving to satisfy all copyright re-
quirements concerning graphics, images and tables for which it has not been possible
to identify the type of licence used.
Table of contents

Index of Figures 7
Index of Tables 9
1. Introduction  11
1.1 Theory and methods  12

Part I
The socio-environmental context
2 Research area 17
2.1 The Mau Forest: geographical and historical background 17
2.2 The three forest sections: Koibatek, Kiptunga, and Ndoinet 20
2.3 A brief history of Mau Forest settlement schemes  25
3 Deforestation in the Mau complex: a remote sensing analysis 29
3.1 General overview 29
3.2 Methodology 30
3.3 The case of the Southwest Mau Forest sector 33
4 Wildlife 43
4.1 Koibatek Forest  44
4.2 Kiptunga Forest 50
4.3 Ndoinet Forest 57

Part 2
The natures of Mau Forest
5 Silviculture 67
5.1 History 67
5.2 The nature of silviculture 69
5.3 The narrative of sustainable forestry 71
5.4 Silviculture: what future for monofunctional forests? 73
6 Settlement schemes and the forest as commons 75
6.1 The new geography of settlement schemes  75
6.2 Settlement schemes’ metabolism 77
6.3 Grazing and grassland 79
7 Forest conservation 83
7.1 Fortress conservation and plantation agriculture  83
7.2 Community conservation  88
 7.3 Nature conservation with direct political purposes  90
8 Indigenous agro-forest systems 95
8.1 An indigenous nature 96
8.2 Beekeeping 98
8.3 Participatory mapping of ecosystem services 101
8.4 Ecotourism 103
9 Conclusion 105
References 111
Archival documents 121
Index of Figures

Fig. 2.1 - The Mau Forest 21

Fig. 2.2 - Mau Forest settlement schemes 26
Fig. 3.1 - Example of land cover classification, 2019 32
Fig. 3.2 - The Southwest Mau in 1995 34
Fig. 3.3 - The Southwest Mau in 2000 35
Fig. 3.4 - The Southwest Mau in 2003 35
Fig. 3.5 - The Southwest Mau in 2008 36
Fig. 3.6 - The Southwest Mau in 2014 37
Fig. 3.7 - The Southwest Mau in 2019 38
Fig. 3.8 - Variations of land cover in the Southwest Mau, 1986–2014 39
Fig. 4.1 - Koibatek Forest map resulting from participatory workshop 45
Fig. 4.2 - Mammals ranked by average detectability index in Koibatek Forest 47
Fig. 4.3 - Map of wildlife traces recorded during the forest walks in Koibatek
Forest and Kiptuget Forest 48
Fig. 4.4 - GIS map of bird species as indicated by community members 49
Fig. 4.5 - Map of the Kiptunga Forest 50
Fig. 4.6 - Average participatory detectability indexes for mammals, birds, and
ecosystem services in Kiptunga Forest 52
Fig. 4.7 - Camera traps positioned in Kiptunga Forest 54
Fig. 4.8 - Overlay of mammals’ ranges and birds’ ranges as identified by the
local community in Kiptunga Forest 56
Fig. 4. 9 - Ndoinet Forest map modified by the community 57
Fig. 4.10 - Mammals’ detectability index in Ndoinet Forest 60
Fig. 4.11 - Yellow-backed duiker (red) and bongo (blue) ranges in Ndoinet
Forest 61
Fig. 4.12 - Map of mammal and bird species in Ndoinet Forest 63
Fig. II. 1 - Distribution of the four socio-ecologies in the research area 66
Fig. 5.1 - Agriculture in a PELIS area, Koibatek Forest 70
Fig. 5.2 - Planted forest managed by Timsales in Kiptunga 73
Fig. 6.1 - Tea buffer zone (left) and farms close to the Ndoinet boundary 77
Fig. 6.2 - Livestock fence in Ndoinet transition Forest 80
Fig. 7.1 - Electric fence in Maasai Mau Forest sectors (Narok County) 88
Fig. 7.2 - Tea buffer zone along the Ndoinet Forest boundary 92
Fig. 8.1 - Cave used for rituals in Kiptunga Forest 98
Fig. 8.2 - Honey harvesting in Kiptunga Forest 101
Fig. 8.3 - Hotspots of ecosystem services in Kiptunga 102
Fig. 9.1 - Example of forest fragmentation and identification of contact points
for ecological corridor drawing in the East Mau Forest 108
Index of Tables

Table 3.1 - Land use change by class area, 1986–2014 40

Table 4.1 - Mammal species in Koibatek Forest according to the community
informants 46
Table 4.2 - Mammals’ detectability index in Kiptunga Forest 51
Table 4.3 - Camera trap shots and indirect signs of wildlife in Kiptunga Forest 54
Table 4.4 - Forest birds used as an indicator of biodiversity in Kiptunga Forest 55
Table 4.5 - Mammals present in Ndoinet Forest 59
Table 4.6 - Birds present in Ndoinet Forest 62
Table 8.1 - Plants gathered as food and medicines in Kiptunga Forest 97
1. Introduction

This book explores the relationship between forests and local communities
in the case of the Mau Forest protected area in Kenya by explaining the de-
forestation process, the conservation initiatives, and the different forms of in-
teraction between people and forest ecosystems. The Mau complex (380,000
hectares [ha]) is one of the most important forests in Kenya and East Africa. It
is rich in plant and animal biodiversity, is home to species of high conservative
value, and is claimed as ancestral land by the Ogiek indigenous people.
In the decade of the 2000s, the deforestation process occurring in the Mau
Forest due to a government redistribution of forest land for political reasons
received international attention and led to interest on the part of the United
Nations Environment Programme (UNEP) itself (headquartered in Nairobi)
and some conservationist non-governmental organizations (NGOs) (e.g., East
African Wildlife Society, Kenya Forest Working Group, World Wide Fund for
Nature, and the International Union for Conservation of Nature [IUCN]). This
process caused a successive phase of nature conservation that has become more
structured since 2012, when the Government of Kenya granted Water Tower
status to the Mau Forest, thus emphasizing the forest’s fundamental role in the
country’s hydrographic network. Loss of forest land and the subsequent con-
servation interventions have resulted in what is now called the “Mau Forest”.
This book analyzes three forest sections belonging to the broader Mau com-
plex—Ndoinet (and the Southwest Mau Forest sector in which this section
is located); Kiptunga (in the East Mau sector), and Koibatek (in the Mount
Londiani sector in the northern part of the forest). The book consists of two
parts. The first focuses on the characteristics of the forests, their biodiversity,
and the deforestation process. The second addresses more specifically the rela-
tionship between nature and society.
Part 1 comprises four chapters. In Chapter 2, we introduce the study area
in geographical and historical terms. In Chapter 3, we describe the deforest-
ation process that affected the Southwest Mau sector, the core of the Mau
Forest Complex, using an analysis of satellite images. Finally, in Chapter 4 the
cross-checked results of a participatory wildlife assessment and fieldwork in the
forests are presented, highlighting the different degrees of biodiversity charac-
terizing the different parts of the three forests.
Part 2 describes the relationship between the forest and the local communi-
ties through four social and ecological arrangements we have called socio-ecol-
ogies (Moore, 2003). Chapter 5 delves into the topic of forestry, investigating
the specific nature produced by the plantations of exotic trees (pine, eucalyptus,
and cypress) within the Mau Forest. Chapter 6 elaborates on the role of the
12 Forest and Communities

grasslands within the protected area in the life of local communities located
around the forest. The conservation of the forest in its various forms is de-
scribed in Chapter 7, underlying the economic and political interests at stake.
Finally, in Chapter 8 we discuss the renewed relationship between some indig-
enous Ogiek groups and the forest, based on the inclusion of traditional ele-
ments in new national and international networks. In conclusion, we highlight
the salient points of the research and propose seven strategies for conserving
the forest and enhancing a sustainable relationship between nature and society.

1.1 Theory and methods

The main objective of this book is to present the results of field research in
three specific sections of the Mau Forest conducted by combining geographical
and biological expertise; therefore, our work does not primarily have a theoret-
ical purpose. However, the research does fit into a specific theoretical frame-
work defined by the social construction of nature theory (Castree, 2001), which
can be summarized in the idea that “nature is defined, delimited, and even phys-
ically reconstituted by different societies” (Castree, 2001, p. 3). More specifical-
ly, the book draws on two perspectives, the political ecology approach (Bryant,
2001) and that of the literature on the “production of nature” (Smith, 1984). In
different ways and from complementary points of view, these two perspectives
allow us to reconstruct the process by which nature, in this case that of the Mau
Forest, co-evolves (Harvey, 2010) with the different social groups that interact
with it. From this perspective, the “nature” with which society is confronted
is always a product of previous actions. This is not the place to engage in an
in-depth debate on the concept of “first” and “second” nature (Smith, 1984;
Loftus, 2012); however, it is important to point out that what is now character-
ized as “nature”– in our case the Mau Forest – is anything but a primal datum
and is instead the product of prior social-environmental configurations.
First, certainly, the delimitation of a protected area called the “Mau Forest”
and its contemporary narrative as a “nature reserve” are cultural products of
a certain society that has experienced a specific history. In particular, some au-
thors (Peluso and Vandergeest, 2001) have traced the genealogy of the category
of “forest”, particularly in tropical areas, relating it to colonial history. In the
perspective we present, however, the “production of nature” is not limited to
the discursive level and invests the materiality of the forest; the different “for-
ests” found within the Mau reserve with their plant and animal species are the
product of past and present social choices.
The research that structures the book was carried out in the last 5 years
(2017–2022), although the biodiversity assessment in Kiptunga started in 2013.
The investigation is based on a mix of qualitative and quantitative method-
ology. The documentary analysis (gray literature) first addressed the technical
 Introduction 13

reports on deforestation and conservation processes (e.g., the Mau Task Force,
UNEP, the Ministry of the Environment, and the Kenya Forest Service [KFS]).
Furthermore, we consulted several archives at the National Archive of Kenya
in Nairobi, the Archive of the Rift Valley Province in Nakuru, and the Daily
Nation Archive in Nairobi. We also consulted the library of the KFS at the
Nairobi Headquarters, together with the colonial and/or post-colonial maps
and documents that in certain cases were available at forest stations. The
National Archive in Nairobi contains records of the colonial and post-colonial
period in the form of letters exchanged between administrators (at the provin-
cial and district levels and with the capital city), reports, and maps related to the
forest status of Mau. The Rift Valley Province Archive (RVPA) in Nakuru con-
tains documents from the post-colonial period that take the form of minutes or
reports related to eviction operations, Settlement Committee meetings, letters
between administrators, and letters from village communities. Furthermore, it
is possible to consult documents pertaining to the forestry, land, and adminis-
trative spheres in general. At the Archive of the Daily Nation, one of Kenya’s
leading newspapers, dozens of press articles relating to the Mau Forest, to
disputes about the resettlement program, to the illegal appropriations by the
country’s political class, and to the Ogiek issue were collected for the period
ranging from the 1970s to the second half of the 2000s (after that date, articles
from the Kenyan press can be found online). The KFS library at the Nairobi
Headquarters was also consulted, and numerous valuable reports relating to
the work of a British conservation project called the Kenya Indigenous Forest
Conservation Programme (KIFCON) in the 1990s were found. At the same
time, secondary data were collected from several offices dealing with forest
management, nature conservation, land use policies, and water management.
In the first phase of the research, we carried out a remote sensing analysis
to investigate land cover changes using Landsat satellite images (1995, 2000,
2003, 2008, and 2014) and Sentinel-2 satellite images (2019; see chapter 3 for
methodological details). During the fieldwork, we conducted semi-structured
interviews with 105 members of local communities around the Ndoinet Forest
section and with about 15 members of local communities close to Kiptunga
Forest. Furthermore, 30 semi-structured interviews were conducted with
Administration (Ward Administrators, Chiefs, various officers), the KFS and
Kenya Water Tower Agency (Forest Managers, Chief Ecosystem Conservator,
local rangers, Regional Coordinator), NGOs defending nature conservation
or indigenous rights (Rhino Ark, National Alliance of Community Forest
Associations, Kenya Forests Working Group, Ogiek People Development
Program), various researchers (from the Kenya Forestry Research Institute,
the Center for International Forestry Research, or the Integrated Forestry
Consultancy and Management Services), companies (Finlays Kenya Ltd.), and
community forest associations (CFAs) (Koibatek CFA, Ndoinet CFA, and
14 Forest and Communities

Kiptunga CFA). More than 20 forest walks with community members and KFS
rangers were undertaken in the three forests1.
In addition, an in-depth 2-month work study on the participatory mapping
of biodiversity and ecosystem services in Koibatek Forest, Ndoinet Forest, and
Kiptunga Forest was conducted in 2017 together with members of local com-
munities (for more details on the methodology, see chapter 4). The process of
the participatory mapping of wildlife and ecosystem services was always tested
by transect walks in the forest aimed at detecting direct and indirect signs of
wildlife and, in the case of Kiptunga, by deploying camera traps on the ground2.

This work would not have been possible without the cooperation of many
people who accompanied us in the research. In particular, the Authors would
like to thank the members of the CFAs of Kiptunga, Koibatek and Ndoinet,
Samuele Tini, Duke Morema, the nonprofit organization Drones for Earth
(former Drone Adventures), Marco Cortesi, Enrica Soria, Anthony De Bortoli,
the staff of the NGO Necofa and the staff of the Kenya Forest Service.

1 For qualitative data processing, we used SPSS Statistics 22 and ATLAS.ti 7 software, and for
spatial analysis we used Google Earth, ERDAS IMAGE 2015, and QGIS software.
2 A transect walk is a method used in biological research for the survey of wildlife, particularly
to estimate population density. It is based on the definition of a sample area and of perpen-
dicular lines at a certain distance (e.g., 1 km) that are walked by researchers who count the
number of animals directly sighted or indirectly detected based on their signs (nests, dung)
in the case of poor visibility due to vegetation (Plumptre, 2000). This method differs from
more qualitative participatory-oriented walks where researchers move in an area (e.g., a rural
village, a forest) together with local community members with the aim of collecting data or
sketching a map in an exercise interspersed with observations, conversations, and discussions
about the social or ecological characteristics of the place (Chambers, 1994).
 Part I
The socio-environmental context

The first part of this book aims at introducing the social and environmental
framework of the Mau Forest, with specific reference to the three forest sec-
tions that constitute the research area—Koibatek, Kiptunga, and Ndoinet. The
selection of these three sections relates to the origin of the research project
that was initially part of an international cooperation initiative developed by the
NGOs Network for Ecofarming in Africa (Necofa) and Mani Tese, with the
support of private and public donors (particularly Cariplo Foundation and the
Italian Agency for Development Cooperation [AICS])1.
The forests were identified by the NGOs as strategic hotspots to fight
against deforestation and to develop forms of community conservation that
were the core of their development approach. The three areas in fact show sig-
nificant differences in both their social and environmental structures (presented
in chapter 2) but share two decisive elements—they lie at the core of the forest,
where important rivers find their sources, and they were subject to significant
controversies and conflicts, particularly at the end of the 1990s when large por-
tions of the forest were converted to agricultural use.
The reconstruction of these conflicts is the main object of Chapter 3, which
focuses on the Ndoinet section, probably the area that suffered the most com-
plex evolution in the Mau Forest, alternating between informal occupation, for-
mal settlements, evictions, and new forms of coexistence between the forest
and the local communities. Understanding the succession of these past events
is extremely important to read the present situation because it has defined what
kind of nature characterizes the forest today and the specific form of interac-
tion that the local communities have with the forest, the subject that is at the
core of the present research. Satellite images combined with data from local
and national archives (see chapter 1 for sources and methods) revealed an east–
west movement of the forest boundary that first shifted westward far beyond
the present limit and then receded to its present position during the first decade
of the century.
Chapter 4 presents the results of three participatory biodiversity assessments
developed in the research area to understand the size and characteristics of
the wildlife population, its distribution within the forest sections, and thus the

1 The research was partly conducted within the framework of two international cooperation
projects: “Local economies, biodiversity protection and responsible tourism development”
(funded by Cariplo Foundation and Regione Lombardia, 2014–2015) and “IMARISHA!
Rural energy for the fight against climate change and for the protection of the environment
(Kenya)” (funded by the Italian Agency for Development Cooperation, 2017–2019).
16 Forest and Communities

existence of strategic hotspots for biodiversity conservation. The issue of wild-

life is central both because biodiversity is a specific forest value and because
the future of forests, particularly in Africa, is closely linked to the possibility of
establishing a sustainable relationship between human and non-human animals.
The reason for choosing a participatory research methodology lies not only in
the effectiveness of the method but also in the desire to identify conservation
pathways that are not imposed from above on local people and that, on the
contrary, see the community at the center of the process.
The three context chapters show the irresolvable entanglement between the
natural and social components that is at the root of the present study. To deal
with such complexity, it was necessary not only to mobilize different discipli-
nary expertise but also to establish a continuous dialogue between the authors,
two geographers, and a biologist to cross the barriers that still exist between
the natural and human sciences in terms of approaches, methodologies, and
language. In this way—“at the appointment of giving and receiving” in the
words of Léopold Sédar Senghor—we wanted to force the limits of individ-
ual disciplines, illuminating their blind spots and trying to produce original
interpretations.
2 Research area

2.1 The Mau Forest: geographical and historical

background
The Mau Forest lies in the Kenyan Rift Valley and is the largest tropical
montane forest in East Africa. The forest complex consists of 16 main sectors
and six non-contiguous satellite areas, covering approximately 380,000 ha. All
the sectors, with the exception of Maasai Mau in the southern part of the forest
complex, are classified as state forest reserves. The area has been gazetted since
1932 and is now managed by the KFS.
The forest complex extends along a 150 km north–south axis in the counties
of Elgeyo Marakwet, Uasin Gishu, Baringo, Nandi, Kericho, Nakuru, Bomet,
and Narok at an altitude between 1200 m and 3000 m. The landscape is typical
of the Rift Valley, with steep slopes alternating with plateaus and reliefs with
softer shapes. The fertile soils of volcanic origin and the humid climate (1,000
mm–2,000 mm of rain per year) make this area particularly suitable for agri-
culture and livestock, favoring human settlement. The Mau complex is in fact
located in one of the most densely populated areas of Kenya, a country where
82% of the territory is classified as arid and semi-arid (less than 700 mm of rain
per year; UNEP, 2009).
The forest harbors the sources of 12 important rivers of western Kenya that
feed lakes on which tens of millions of people depend—Lake Victoria, Lake
Nakuru, Lake Turkana, Lake Baringo, and Lake Natron. These rivers contribute
greatly to agricultural activities, energy production, and the tourism sector.
In terms of commercial agriculture, the lands around the forest are home
to extensive tea plantations that date back to the colonial period and are now
owned by a number of multinational companies, namely Unilever1 and Finlays,
large Kenyan families (including the family of former President Daniel Arap
Moi), and a number of producers affiliated with Kenya Tea Development
Agency Holdings Limited (KTDA Ltd.), the historic state-owned company pri-
vatized in 2000.
The forest’s ecosystem services also provide vital support to small-scale
farming. The region is the ancestral land of the Ogiek people (approximately
52,000 people in the last census; KNBS, 2019), originally a hunter–gatherer

1 In July 2022, Unilever announced it had completed the selling of its Kenyan tea plantation
to the CVC Capital Partners Fund VIII for € 4.5 billion (https://www.unilever.com/news/
press-and-media/press-releases/2022/unilever-announces-completion-of-the-sale-of-its-
tea-business-ekaterra/).
18 Forest and Communities

ethnic group that is now settled around the forest. Alongside this original pop-
ulation, other agro-pastoral tribes belonging to the broader Kalenjin ethnic
group have settled in the area over the past few decades, with government sup-
port, thus increasing the pressure on the forest (see sub-chapter 2.3).
Moreover, it should be highlighted that the rivers that originate in the forest
make an important contribution to the hydroelectric sector, which produces
30% of the national energy. Finally, tourism, which generates more than 10%
of the national GDP, also benefits from the ecosystem services produced by
the Mau Forest. In particular, two famous tourist destinations, the Maasai Mara
National Reserve and the Lake Nakuru National Park, depend on rivers the
sources of which are located in the forest (GoK and UNEP, 2008).
For all these reasons, the forest plays a crucial role in the development of the
country, and the government has included the Mau Forest among the five most
important country’s “water towers”—the Mau Forest Complex, the Aberdare
Mountains, Mount Kenya, Mount Elgon, and the Cherangani Hills—that have
to be protected with special actions. However, beyond this strategic role of
“water tower”, the Mau Forest ecosystem has extraordinary value in terms of
vegetational and animal biodiversity.
The vegetation inside the forest changes with altitude. It is in fact progres-
sively denser as the altitude increases up to 2,500 m above sea level (asl), when
bamboo begins to prevail. In the higher areas, it is characterized by grasslands.
In the areas of dense forest that are mainly located between 2,000 m asl and
2,500 m asl, there are some important species of indigenous vegetation, such
as Albizia gummifera, Prunus africana, Olea capensis, Ficus thonningii, and Podocarpus
latifolius. Besides the native vegetation, there is also an important presence of
allochthonous species that was introduced in colonial times to foster forestry
activities. The vegetational patterns produce very different ecosystems: on the
tree plantations, there is a drastic decline in biodiversity compared to areas of
indigenous forest, where there are still many animal species of high naturalistic
value. In particular, these areas are home to some species considered at risk of
extinction, such as the African golden cat (Caracal aurata, IUCN, Vulnerable)
and the yellow-backed duiker (Cephalophus silvicultor, IUCN, Near Threatened).
The story of the Mau Forest is strictly connected with the historical and po-
litical dynamics. During the colonial period, the Mau complex was surrounded
by British settlements in the plateau of the Rift Valley in what were renamed
the “White Highlands”. The preference of the British colonizers for this area
was due to the favorable climate and the presence of fertile lands, a dense water
network, and a forest rich in tree species (Morgan, 1963). The status of “for-
est reserve” was granted to the different forest sectors between the 1930s and
the 1950s to protect the timber that the colonial government intended to be
a strategic asset for both the Kenya–Uganda railway and for the development
of the logging sector. These objectives are also behind the creation of the first
 Research area 19

tree plantations with exotic species in the first decades of the 20th century
(Ofcansky, 1984).
At the same time, the gazettement of Mau as a natural protected area re-
sulted in the eviction of the Ogiek indigenous groups who lived in the forest.
Those operations ensured the dispossession of Mau Forest and the sedentari-
zation of these semi-nomadic groups, thus ensuring cheap labor for the col-
onizers’ farms (Cavanagh, 2017). With independence, the forest was declared
the Central Forest Reserve (1964). However, the organization changed little,
and colonial companies such as Timsales—now administered by Kenyan man-
agement—continued to control large portions of the forest for silviculture in
association with the Forest Department.
The situation changed dramatically in the 1990s when a conservation pro-
ject called KIFCON (1991–1994), funded by the UK government, envisaged
the conversion of part of the protected area to settlement zones to definitely
sedentarize the Ogiek people living irregularly in the forest (KIFCON, n.d.).
Following the KIFCON recommendations but subverting the original purpose,
the government of President Moi (1978–2002, Kenya African National Union
Party [KANU]) set up an ambitious settlement scheme initiative (1994–2001)
that produced decisive socio-territorial transformations in the region. First, the
protected area lost more than 60,000 ha, with the highest percentages in the
Eastern (-54%) and Southwestern (-27%) sectors, at that time converted to
settlement and agricultural use for the population (RVPA, FOR. 13/5/1 Vol.
1). Second, the initiative, originally aimed at the Ogiek community, attracted
thousands of people coming from the other Rift Valley Province districts, such
as Bomet, Kericho, and Baringo. As will be explained in sub-chapters 2.3 and
3.1, forest land was used by the Kenyan government to build its political sup-
port in a delicate phase of the country’s transition to democracy (Boone, 2012;
Morjaria, 2012).
In 2001, the government issued a notice for the degazettement of the 60,000
ha of forest, thus sanctioning the land cover change that occurred in the previ-
ous decade. This process was heavily contested inside the country (Mau Forest
Task Force, 2009), and worldwide environmental NGOs protested openly
against what was seen as a strong deforestation act.
Even after the degazettement, Mau settlement schemes continued to attract
migration flows that led to the irregular occupation of large parts of the forest.
Subsequently, particularly in the period 2000–2008, several evictions took place
(Daily Nation, 2002; Daily Nation, 2002b; interviews from 2019 and 2020),
followed by further encroachments. In 2008, the alarm raised by the UNEP
and the government of Kenya (GoK and UNEP, 2008) marked the start of a
forest conservation phase that became particularly effective in 2012 with the
granting of the “water tower” title to Mau, thus emphasizing the fundamental
20 Forest and Communities

role played by the Mau Forest Complex in the water network of Kenya2. The
legal creation of this new category was accompanied by the establishment of
a dedicated organization (the KWTA) designed to manage forest assets of na-
tional importance. Since then, the Mau Forest Complex has not experienced
significant alteration of its cover.
Even if the encroachments on the Mau boundary in the research area are no
longer a main issue, the establishment of highly populated settlements around
the border of the Mau Forest puts strong pressure on the protected area, par-
ticularly regarding firewood collection and grazing. The ethnic composition of
the population in the settlements saw some changes as well, such as the relative
decrease of the Ogiek component and the parallel growth of other groups
linked to President Moi (Kipsigis, Tugen)3. This social transformation contrib-
uted to the establishment of a conflictual environment that still characterizes
the area.

2.2 The three forest sections: Koibatek, Kiptunga, and

Ndoinet
The research area covers three sections of the central part of the forest:
Koibatek (part of the Mount Londiani sector), Kiptunga (East Mau), and
Ndoinet (Southwest Mau). The areas around these forests host approximately
850,000 people, 720,000 in the areas of Kiptunga (Molo and Njoro sub-coun-
ties) and Ndoinet (Kuresoi North and Kuresoi South sub-counties) and 130,000
in the Koibatek sub-county (KNBS, 2019).

2 The “water tower” status is a legal entitlement established in 2012 by the corresponding
Kenya Water Tower Agency (KWTA), a state corporation that falls under the Ministry of
Environment. The granting of the title marks the recognition of a forest’s national impor-
tance in terms of water supply and conservation and the need for a more comprehensive ap-
proach in its management. In fact, the Agency is characterized by the adoption of a multiple
ecosystem services perspective (rather than a mere exclusive focus on forest cover or water
supply) and by being a coordinating body of national offices, local communities, and devel-
opment partners (see https://watertowers.go.ke/history/). In practical terms, the Agency’s
operations are unclear. In the case of Mau, the KWTA is very active in the Maasai Mau sec-
tor, where it coordinates nature conservation initiatives, perhaps because it is not an officially
gazetted forest. In the other Mau sectors, it is actually the Kenya Forest Service that is the
most active stakeholder.
3 The Kipsigis and Tugen, together with the other sub-tribes (Keiyo, Marakwet, Nandi, Pokot,
and Sabaot) comprise the macro ethnic group of Kalenjin people. They come from the Rift
Valley region, but their presence in the area adjacent to the Mau forest is due to migration
flows that occurred in the 1980s and 1990s, particularly in relation to the creation of govern-
ment settlement schemes.
 Research area 21

Fig. 2.1 - The Mau Forest (edited by the authors, GoK and UNEP, 2008)
22 Forest and Communities

Koibatek Forest
Koibatek Forest covers an area of about 9,000 ha and is located mainly in
Baringo County. The physical characteristics of the forest reflect those of the
entire Mau complex, with rainfall ranging from 800 mm in the lower areas to
1,200 mm in the higher areas and temperatures between 15 °C and 35 °C, mak-
ing it relatively cool from June to October and warmer from December to
March. In this forest is particularly evident the volcanic structure that char-
acterizes the whole area of the Mau complex; in fact, Koibatek occupies the
southern slopes of an extinct volcano (KFS, 2013).
From a hydrographical point of view, the forest contributes to the water
basin of the Molo River, the main tributary of Lake Baringo, the second largest
lake in the Kenyan Rift Valley and a center of great importance for biodiversity,
particularly for birdlife (Ramsar site since 2002).
Koibatek has been a protected area since 1932 and, unlike the other two sections,
has not lost much area over the decades (about 100 ha). However, the forest is sur-
rounded by a highly populated area that has conditioned its evolution. Of the three
forests, Koibatek is the one where the silvicultural component is greater, with 2,967
ha of plantations (about 30% of the total, mostly in the lowest part), while indig-
enous forest occupies about 3,300 ha (37%) and bamboo another 2,500 ha (30%).
The presence of tree plantations is directly linked to the Plantation Establishment
Livelihood Improvement Scheme (PELIS), a government program that allows the
local community to temporarily cultivate land in protected areas designated for
forestry (see sub-chapter 5.2), while assuring the maintenance of the plantation.
The main species used in forestry are cypress (Cupressus lusitanica, 1,580 ha) and
pine (Pinus patula, Pinus radiata, 700 ha), with minor areas devoted to eucalyptus
(Eucalyptus sp., 170 ha). The highest part of the forest is characterized by important
indigenous vegetation to which corresponds a richer fauna, with several species
classified as “threatened” by IUCN (e.g., buffalo, antelopes). There are also signifi-
cant cultural sites (caves) in this portion of the forest.

Kiptunga Forest
Kiptunga Forest now occupies an area of 10,300 ha within the eastern sector
(which has a total area of approximately 31,000 ha) within Nakuru County. The
entire East Mau has suffered significant losses of forest area, particularly in the
1990s, when 35,000 ha (54% of the eastern forest sector at that time) were con-
verted to settlement schemes marked by small-scale agriculture. The forest area
covers the ridge (2,900 m asl) that separates the Rift Valley watershed and the
drainage basin of Lake Victoria. The hydrographic network of the forest is cru-
cial because here are the sources of some of the most important watercourses
of the region—the Molo River, directed to Lake Baringo lake; the Njoro River,
which feeds Lake Nakuru; and the Mara River that, after crossing the lands of
the Maasai Mara and Serengeti parks, flows into Lake Victoria (KFS, 2015).
 Research area 23

The forest was put under protection in 1958 with the double intent of pro-
tecting the vegetation and ensuring the regular course of forestry. The coex-
istence of these two purposes can still be seen today; the forest hosts 7,300
ha of indigenous forest (70%) and 2,000 ha of tree plantations (19%)4. The
socio-environmental value of the indigenous forest is partly limited by its great
fragmentation; of the four sections that form the forest (Chebuin, Olengape,
Kiptunga, and Kiboyet), only one (Chebuin) is relatively uniform and hosts the
greatest animal biodiversity.
Kiptunga is the only section of the forest in which there are still Ogiek villages
within, but on the edge of, the protected area, allowing this population to maintain
the socio-environmental relationship with the forest that characterizes its culture,
including honey production, collecting medicinal herbs, and performing rituals.
This is also why in this section of the forest the claims of the Ogiek are particularly
strong and the latent conflicts most critical (Mkawale and Gachui, 2020).

Ndoinet Forest
The Southwest sector of Mau constitutes the largest sector of the forest
complex (approximately 60,000 ha) and is composed of four forest sections
the boundaries of which are administrative rather than ecological—Ndoinet
(about 20,000 ha), Itare, Maramara, and Kericho. The landscape of Ndoinet
is characterized by undulating reliefs and valleys that decline towards the west,
following various watercourses that are part of the Sondu–Miriu and Mara river
basins (Lake Victoria watershed). The elevation of the southwest sector varies
between 1,900 m and 2,500 m, and the area receives approximately 2,000 mm
of annual rainfall, with average annual temperatures in the range of 12 °C–21
°C (Konoin Kabara Community Forest Association-Itare, 2013; KFS, 2018).
Like other parts of the forest complex, the environmental context is charac-
terized by its volcanic origin, with nutrient-rich clay soils particularly suitable
for agriculture (Courtney Mustaphi et al., 2014). The vegetation follows an al-
titudinal zoning from west to east, with montane forest (at elevations below
2,300 m) gradually giving way to highland bamboo (Arundinaria alpina) mixed
with grasslands and finally to sclerophyllous montane vegetation near the es-
carpment ridge. Tree species characteristic of the lower montane forest include
Aningeria adolfi-friedericii and Strombosia scheffleri.
This area has suffered heavily from human intervention. With the 2001
degazettement, the Ndoinet Forest (at that time known as Tinet Forest5) lost

4 The other vegetation types are bamboo (4.3%), grasslands (3.4%), and bushlands (2.6%).
5 The part of the forest belonging to Southwest Mau and now occupied by government set-
tlement schemes was known in earlier decades as Tinet Forest (RVPA, FOR 13/5/1 Vol. 3).
Interestingly, the name is of Ogiek origin and refers to a tree important to the indigenous
group. Today, Tinet is the name of the southern part of the area, part of Kuresoi South
sub-county.
24 Forest and Communities

almost 25,000 ha, an area larger than that of the current forest section. As will
be seen later, the Ndoinet Forest has been occupied by irregular settlements
that have been progressively removed during the 2000s. The cleared areas are
now dominated by pioneer species, such as Tabernaemontana stapfiana, Syzygium
guineense, and Neoboutonia macrocalyx, while the less impacted areas of forest are
home to indigenous species, such as Olea capensis, Prunus africana, Albizia gum-
mifera, and Podocarpus latifolius (BirdLife International, 2018; KFS, 2018). The
native forest section provides a habitat for a rich fauna, including mountain
bongo (Tragelaphus euryceros), buffalos (Syncerus caffer), elephants (Loxodonta africa-
na), yellow-backed duikers (Cephalophus silvicultor), dik-diks (Madoqua Kirkii), the
colobus guereza (Colobus abyssinicus), and honey badgers (Mellivora capensis).
The Southwest Mau is bordered on its northwestern edge by tea plantations,
which cover more than 20,000 ha in Bomet and Kericho counties. These areas
are mostly owned by two transnational companies (Unilever and Finlays), thus
reflecting the enduring relevance of the colonial legacy. East of the forest sec-
tor are located settlement schemes that are divided from the forest by a 32 km-
long boundary, and by a 10 km-long tea buffer zone planted in the last few years
to physically demarcate a separation between the protected area and the villages.
The settlement schemes are defined by an orthogonal geometry of small farms
(0.5 ha–2 ha) centered on the cultivation of a small number of products (maize,
potatoes, beans, peas, cabbage, and occasionally tea) and livestock rearing (cat-
tle, sheep, goats, and poultry). On the farms are also common fruit trees and
planted exotic trees (pines, cypresses, eucalyptus) used to meet the energy needs
of the inhabitants, which are based almost exclusively on wood, coal, and ker-
osene. In the southern section of the Ndoinet Forest between Kipkongor and
Kiptagich is a 2000-ha area occupied by settlements, despite being located with-
in the boundaries of the protected area; this is a remnant of previous irregular
settlements (see Fig. 2.2).
Finally, close to the Ndoinet Forest section in an area called Chematich, in
2016, the Italian company Cooperativa Muratori Cementisti (CMC) started the
construction of the Itare dam with the purpose of diverting water from the
Lake Victoria basin to the Rift Valley watershed through a 114-km tunnel to
supply the city of Nakuru with safe water (800,000 beneficiaries)6. Work on
the dam is currently suspended due to the economic difficulties of the Italian
company and to a legal action concerning the construction of two dams by
CMC in Kenya.

6 The dam is a medium-sized (280 ha) artificial reservoir the design of which was presented by
the Rift Valley Water Services Board and is one of the cornerstones of Kenya Vision 2030,
the national long-term development plan (Rift Valley Water Service Board, 2015).
 Research area 25

2.3 A brief history of Mau Forest settlement schemes

Settlement schemes in Kenya started in the last years of colonial dominion
(1955–1963) and were then frequently adopted by the subsequent Kenyan gov-
ernment to address poverty and boost economic development by allocating
land to specific groups of people (Shaffer, 1967; Wayumba, 2019). In the case
of the Southwest Mau and East Mau sectors, officially the government purpose
was to allocate land to the Ogiek who lived informally within the Mau Forest.
It was a measure suggested by KIFCON (1991–1994) to respond to the deg-
radation of the Mau Forest and to find a permanent solution (Bateson, n.d.;
KIFCON, n.d.). In fact, since the first decades of the 20th century the Mau
Forest has been inhabited by communities that the colonial officials identified
as Wadorobo Ogiek7 that were known to be hunters, gatherers, and semi-no-
madic beekeepers (Kratz, 1980; Sang, 2001).
Archival documents reporting government officers’ correspondence illus-
trate that since the 1970s the Ogiek were not the only group living in the for-
est. Mau Forest was a place of life for thousands of other people who lived
scattered about the area and who found in that environment the possibility of
surviving on small-scale agriculture and animal husbandry (RVPA, FOR 13/5
Vol. 6). In the East Mau, small human centers grew in relation to the cultivation
of tree plantations through a system known at the time as Taungya, predecessor
of today’s PELIS (see sub-chapter 5.2; see the cartography of East Mau, 1967).
The protected area was also subject to cycles of migration from neighboring
districts of the Rift Valley Province that occurred when rumors of possible per-
manent land allocation spread through the country, as in the 1980s in Southwest
Mau (RVPA, FOR 13/5 Vol. 6 and 7) or at times of ethnic and political unrest,
as occurred during the 1992 elections (Daily Nation, 1993; Moorehead, 1993).
In the Mau Forest Complex, the government intended to allocate two big
portions of the forest reserves to the creation of the settlement schemes of
Saino, Ndoinet, Tinet, and Kiptagich Extension in the Southwest Mau (25,000
ha)8 and of Baraget, Mariashoni, Nessuet, Sigotik, Teret, Sururu, and Likia
in the East Mau (35,000 ha)9. The initiative was identified as the “Wadorobo
Resettlement Programme” as an indication of its official purpose to settle the
Ogiek groups and the people affected by political violence (Mau Forest Task
Force, 2009) and was guided by the District Administration of Nakuru. Despite

7 “Ndorobo” or “Dorobo” was a derogatory nickname used by the Maasai people in referring
to the people currently known as Ogiek. This name was adopted by the colonial government
to generically refer to the hunter-gatherer groups in the colony. Today, in the Mau region
members of this group call themselves “Ogiek”.
8 These settlement schemes were anticipated by the establishment within the boundary of the
Southwest Mau of the Korau Settlement schemes (2,733 ha) with a forest excision in 1985.
9 See the archival documents located at the RVPA, FP 8/7 C3/NE/SU/GE; FOR 13/5/1 Vol.
2; FOR 13/5/1 Vol. 3; FP 8/27 LA/RVP/122.
26 Forest and Communities

the title, the program concealed a very specific political intent—the alteration
of the ethnic composition of the district aimed at strengthening political sup-
port for President Moi (Klopp, 2012). During the years, thousands of Kalenjin
people (the same ethnic group as President Moi) looking for land came from
the neighboring Rift Valley districts (Kericho, Bomet, Baringo, and Eldoret) and
settled on Mau land. While this left the Ogiek with a sense of injustice suffered
in seeing their ancestral land allocated in large amounts to outsiders, the forest
land redistribution to Kalenjin families would have worked to strengthen the
political power of the KANU government in the district.

Fig. 2.2 - Mau forest settlement schemes (edited by the authors, GoK and UNEP, 2008;
Department of Survey, 2018; Department of Resource Surveys and Remote Sensing, 2019)

The creation of Mau settlement schemes in the 1990s coincided with a po-
litical transition in the country (1992–2002) from an authoritarian one-par-
ty system centered on President Moi (1978–2002) to a plurality of political
parties (Hornsby, 2012). It was a decade of crisis triggered by the imposition
of so-called “good governance” reforms by western donors in sub-Saharan
African countries (Beekers and van Gool, 2012). The repercussions of this re-
gime change would have been felt in the climate of inter-ethnic tension (the
political violence of 1992 and 1997) and in the management of natural re-
sources. Land was allocated to build consensus in the Kalenjin electoral base to
create new constituencies and to benefit the ruling elite (as in the case of the
Kiptagich Extension settlement scheme) in order to forge personal alliances
between politicians, businessmen, administrators, and the military (Commission
 Research area 27

of Inquiry into the Illegal/Irregular Allocation of Public Land, 2005; Boone,

2012; Albertazzi et al., 2018; Cavanagh, 2018).
During the 1990s, the land allocations continued uncontrolled for some years
and went far beyond the designated areas, resulting in a severe impact on Mau
Forest cover. Furthermore, the allocations led to the eviction of those who
were settled on Mau Forest land and who lived in the most diverse situations—
from those who possessed land allocation documents to those who were simple
squatters awaiting some sort of amnesty. Since 2013, there has been a caveat
regarding the settlement areas of Mau that prevents the erection of perma-
nent structures and prohibits commercial transactions on the land. As a conse-
quence, those who live in settlement schemes and have not received property
titles in the past find themselves in possession of a frozen asset that places them
in a fragile situation. Due to this intricate past made up of the occupation of
forest land, forced evictions, and settlement validations documented and then
challenged, Mau’s settlement schemes are an extremely sensitive topic from a
political and an ethnic point of view.
3 Deforestation in the Mau complex: a
remote sensing analysis

3.1 General overview

The three forests have undergone different deforestation processes, very lim-
ited in the Koibatek Forest and more intense in the Kiptunga Forest and the
Ndoinet Forest. The decisive change occurred in the 1990s when 60,000 ha of
forest land (14% of the Mau’s protected area) were converted to settlements
(see sub-chapter 2.3) and then legally sanctioned through the 2001 degazette-
ment (GoK and UNEP, 2008). To understand the complex deforestation pro-
cess that characterized the central sector of the Mau complex, we conducted a
remote sensing analysis using satellite imagery of the Southwest sector.
Land cover changes can be caused by long-term natural climate change, eco-
logical and geomorphological processes, human-induced vegetation and land-
scape transformation, and inter-annual climate variations. Each of these phe-
nomena manifest at different scales, have different intensities, are characterized
by a certain degree of reversibility, and produce a specific effect on environ-
ment and society (Lambin and Ehrlich, 1997).
One ways these changes can be investigated is through remote sensing anal-
ysis, the collection of information about an object on the Earth’s surface with-
out direct physical contact with it, for example using satellite images1. Satellite
imagery provides three levels of information for the study of forests. The first
is land cover dynamics, thanks to the different spectral responses of forest and
non-forest areas, and the possibility of detecting fires and burn marks on the
ground. The second concerns the type of forest, which is mostly defined by
the tree canopy, thus distinguishing, for example, tree plantations from indig-
enous forests. The third is to gather information on biophysical and biochem-
ical properties that give indications regarding the quality of forest resources.
Some of the most used indices are the Simple Ratio, the Normalized Difference
Vegetation Index, and the Leaf Area Index (Boyd and Danson, 2005). Thus, it
is possible to estimate the amount of biomass, the degree of biodiversity, and
the impact of drought on forest areas (Foody, 2003).

1 A crucial role in this sense was played by the launch in 1972 of the Landsat program (by
the National Aeronautics and Space Administration and the US government), while the
European Space Agency and the European Union undertook the Copernicus–Sentinel pro-
gram in 2015.
30 Forest and Communities

At the end of the 1980s, Myers (1988) stressed the great opportunity of us-
ing remote sensing in the study of deforestation, highlighting the possibility of
having systematic and objective data. While this is true, many methodological
issues may affect the analyses. Lambin (1999) suggested a meticulous approach
to the study of forest degradation at a regional level, with the aim of avoiding
errors in monitoring due to ecological reversibility and the strong impact of
climatic variability that plays an important role in Africa south of the Sahara.
Since the 1990s, it has been possible to measure forest changes using satellite
imagery. There are many studies investigating forest area changes (Sánchez-
Azofelfa et al., 2001; Zhang et al., 2005; Torahi and Rai, 2011; Folega et al., 2014;
Pellikka and Alshaikh, 2016). Some of these studies do not limit themselves to
surveying land conversion but relate this to different variables, such as biophys-
ical variables (e.g., slope, altitude, soil type, vegetation, rainfall, and hydrology),
demographic variables (e.g., population growth), and economic variables (e.g.,
distance to markets, land price, and agricultural subsidies) (Chowdhury, 2006;
Reis, 2008).

3.2 Methodology
The analysis of the satellite images was carried out in two steps, a visual in-
terpretation of the satellite images in false colors and a land cover supervised
classification2. Six Landsat satellite images (30x30 m resolution) were used,
covering the significant years of 1986, 1995, 2000, 2003, 2008, and 2014, and
a Sentinel image (10x10 m resolution) was used for 2019. The images for each
year were taken in the dry season between January and the beginning of April,
which is preferable due to the marked contrasts between the forest vegetation
and the land being worked for sowing (in the case of maize, the main crop in
the area). It is also the time period in which most of the fieldwork was conduct-
ed and therefore suitable for the use of global positioning system (GPS) points
taken on the ground.
After the image acquisition, pre-processing was carried out, which consist-
ed of rectification, atmospheric corrections, and topographic normalization3.
Finally, a multispectral image was created for each year of classification con-
sisting of 13 layers—layer/bands 1–6 (band 1 = blue, band 2 = green, band 3
= red, band 4 = near infrared, band 5 = short-wave infrared, band 6 = thermal

2 In this chapter, we will mainly show the quantitative results of the supervised classification
(Fig. 3.8 and Table 3.1) to understand the cover changes of the three different classes used in
the analysis.
3 These processes were conducted by Dr. Hari Adhikari (University of Helsinki), who is also
credited with the design of the supervised classification procedures adopted in this analysis.
 Deforestation in the Mau complex: a remote sensing analysis 31

infrared), layer 7 = NDVI, layer 8 = TCT 1, layer 9 = TCT 2, layer 10 = TCT 3,

layer 11 = RSR, layer 12 = SAVI, and layer 13 = NPVND4.
False-color images were then analyzed to guide the visual interpretation of
deforestation processes. The false-color images result from the variable combi-
nation of three bands, the use of which is frequent in forest studies to detect
the differences between areas with dense vegetation and grasslands or agricul-
tural fields. In particular, we are showing multispectral maps in false colors (4
red [R]), 3 (green [G]), 2 (blue [B])5. The multispectral map displayed later is a
classic false-color representation in which the vegetation is bright red because
the green vegetation in good condition brightly reflects band 4 (NIR)6.
Regarding land cover classification (see an example in Fig. 3.1), the start-
ing hypothesis considered deforestation as the result of an expansion of hu-
man settlements, attested by the presence of crops within the protected area.
Therefore, it was decided to use three land cover classes—dense forest, transi-
tion forest, and cultivated/bare soil. The class “dense forest” represents closed
canopy cover, an indication of the presence of primary forest. The class “tran-
sition forest” includes bamboo forest, shrubs, and grasslands. The class “cul-
tivated land/bare soil” indicates the presence of crops or soil not covered by
vegetation, such as land awaiting sowing and erosion areas7.
For the Landsat analysis, considering the impossibility of collecting a large
number of GPS points directly in the field in different areas of the forest sector,
it was decided to simulate the presence of the operator using high-resolution
images from Google Earth, with 2014 as the reference year. This decision was
motivated by the fact that in the last 5 years the Southwest Mau sector has en-
joyed a stable situation, and there has been no acute loss of area. Furthermore,
the classification of past years’ land cover for the study area is difficult due to
the lack of maps, GPS surveys, and historical Google Earth images with good

4 NDVI stands for Normalized Difference Vegetation Index and is calculated as the ratio of
the difference to the sum of the reflected radiation in the near infrared and red. TCT stands
for Tasseled-Cap Transformation and is a conversion of some values present in the bands
and useful in studies of vegetation changes; TCT 1 refers to soil brightness, TCT 2 to the
degree of greenness of vegetation, and TCT 3 to the moisture present in the soil. RSR stands
for Reduced Simple Ratio, SAVI for Soil-Adjusted Vegetation Index, and NPVND for Non-
Photosynthetic Vegetation Normalized Difference.
5 In the overall analysis we also used multispectral images in false colors for short infrared
(SIR), near infrared (NIR), and red (R) bands and multispectral images in false colors for
NIR, SIR, and blue (B) bands.
6 The false-color image 4, 3, 2 is given by the display of the near infrared (NIR), red (R), and
green (G) bands.
7 The classification followed two different processes. We used the software ERDAS Imagine
2015 for Landsat images and the Semi-Automatic Classification Plug-in available in Qgis for
the Sentinel image. In both cases, were also used Qgis Version 3.8 and Office Excel to work
on data.
32 Forest and Communities

resolution, which makes it impossible to derive points to use in the classification

of pre-2014 years.
Next, 1000 points were manually selected on Google Earth (2014 reference
image) and assigned to one of the three classes. The 1000 points of the year
2014 were then associated with the values of the vegetation bands and indices
present in the normalized raster image for 2014. Finally, 70% of the points
were used as training points and 30% as test points. Thus, six Excel files were
obtained for each year of classification, two for each class (i.e. “training points”
and “test points”); altogether, 6,000 points were screened.

Fig. 3.1 - Example of land cover classification, 2019

(source: Stefania Albertazzi, 2020)

As the points belonging to the same area have the same spectral signature, it
is possible to notice the presence of one or more trends for each class and to
take advantage of this property. Therefore, one possible method to conduct a
multi-temporal supervised classification is to use the spectral signature trends
of each class of the 2014 image as a reference. It is then possible to eliminate
those points that in previous years show a trend differing from that of the class
 Deforestation in the Mau complex: a remote sensing analysis 33

to which the Excel file belongs, which means that in the year under consid-
eration that point did not belong to the class to which it belongs in the 2014
reference year. In this way, it is possible to detect changes in the land cover of
the point. In the end, the final points of each class—that is, after eliminating
all points with a trend differing from that of the class in the Excel file—are
then loaded into the software and will serve as training and test points for the
classification.

3.3 The case of the Southwest Mau Forest sector

The research covered the whole Southwest sector, of which the Ndoinet
Forest section is part. The total area before the 2001 excision of the eastern
portion (25,000 ha) was about 85,000 ha. As our hypothesis ascribes forest loss
to the conversion to crops, our analysis focused on highlighting the presence
of agricultural land or land not covered by vegetation within the forest reserve.
The visual interpretation of satellite images can provide useful insights regard-
ing the unfolding of the process.

1995
The false-color image (Fig. 3.2) depicts the primary forest in dark red, the
bamboo forest and shrubs in less bright red/purple, and the land without veg-
etation in light blue. These may indicate eroded areas, the presence of which is
frequent along the forest slopes, or areas where forest vegetation is absent due
to the presence of human settlements or cultivation (which in the month of
the image survey are being worked in preparation for sowing and are therefore
without cover). The image informs us of the state of the forest in the mid-
1990s, a period for which few sources are available. The forest sector is not
a single continuum of primary forest, and the vegetation varies between the
western, central, and eastern sections, where there is evidence of human pres-
ence. In particular, in the central–eastern edge of the sector, an area of ​​2,733 ha
known as the Korau Settlement Scheme (see subchapter 2.3) had been devoted
to the settlement of about 700 people in 1985. This was the first official popu-
lation settlement within the Southwest Mau sector and can be considered a sort
of anticipation of the following settlement schemes that would be created by
the government in the 1990s.
34 Forest and Communities

Fig. 3.2 - The Southwest Mau in 1995 (source: Stefania Albertazzi, 2020)

2000
In the false-color image (Fig. 3.3), as in the previous one, the dense forest is
represented in dark red, the secondary forest (with shrubs and bushes) in less
bright red/purple, and the land without vegetation cover, such as bare soils or
agricultural land, in light blue. This is a fundamental image to understand the
deforestation process of the Southwest sector. The situation looks very differ-
ent from 1995. The eastern section suffered a drastic change, and the expan-
sion of agriculture seems to be starting in the south, although it appears more
scattered than in the northern areas. In 2001, the Minister of the Environment
announced the official degazettement of the eastern section, the legal modifi-
cation of the boundary, and the allocation of degazetted areas to settlements,
for a total of approximately 25,000 ha. The Southwest Mau was reduced to
approximately 60,000 ha.
 Deforestation in the Mau complex: a remote sensing analysis 35

Fig. 3.3 - The Southwest Mau in 2000 (source: Stefania Albertazzi, 2020)

Fig. 3.4 - The Southwest Mau in 2003 (Stefania Albertazzi, 2020)

36 Forest and Communities

2003
The false-color satellite image (Fig. 3.4) illustrates the land cover changes
that occurred with the progressive settlement of people in the eastern section
of Southwest Mau. To the east of the 2001 boundary (represented by the white
line), the forest areas (dark red) are almost non-existent; a few exceptions are
located on the fringes along the edge of what was the previous boundary.
The alarming data concerns the strong pressure that remains on the forest
despite the recent degazettement of the eastern section. The widespread pres-
ence of light blue and blue/dark green spots indicates extensive areas without
forest cover west of the 2001 boundary. The presence of agriculture or bare
surfaces within the protected area is recorded at a distance of 9 km from the
boundary in the northern area and 14 km from the boundary in the southern
area.

Fig. 3.5 - The Southwest Mau in 2008 (source: Stefania Albertazzi, 2020)

20088
The false-color image (Fig. 3.5) captures the worst state of the Southwest
Mau over the past three decades. The loss of forest area (dark red) west of the
border (white line) reaches its maximum. The eastern part of the protected
area, the core of what is now the Ndoinet section, is completely devoid of tree

8 The white stripes on the image are due to a problem with the sensor of the Landsat 7 satellite.
 Deforestation in the Mau complex: a remote sensing analysis 37

cover. The area lying between 5 km and 10 km from the border is an extensive
area without tree cover (in light blue) interspersed with a few forest spots. This
forest loss is the result of spontaneous encroachments of the population that
lived on protected land in the wake of previous amnesties. The emergency sit-
uation focused the attention of the UNEP and of the Kenyan government on
creating the premise for the future launch of the Water Towers Initiative, the
program for the rehabilitation and protection of mountain forests that serve as
the country’s water reservoirs. East of the 2001 boundary (white line), in the
settlement area, we record a total absence of tree cover.

Fig. 3.6 - The Southwest Mau in 2014 (source: Stefania Albertazzi, 2020)

2014
The false-color image (Fig. 3.6) represents the end of the emergency situ-
ation. West of the boundary (white line), we can see the regrowth of forest
vegetation (in purple), interspersed with light blue patches that represent the
traces of previous encroachments that have ended. Nevertheless, an uncertain
situation remains in the southern end of the forest sector, where the boundary
of the protected area is unclear and the presence of agriculture is recorded.
East of the border, in the settlement area, the situation seems to be improv-
ing as well, compared to previous years. We note the widespread presence of
38 Forest and Communities

forest patches (in purple), a sign of the reforestation that is taking place along
the waterways and of the spread of private woodlots on farms.

Fig. 3.7 - The Southwest Mau in 2019 (source: Stefania Albertazzi, 2020)

2019
The false-color image (Fig. 3.7) is a recent photograph of the situation of
the Southwest Mau. The forest reserve continues in the positive trend that was
visible in the 2014 image. The dense forest (dark red) and the secondary forest
(light red/purple) are expanding, while several eroded areas can be seen within
the reserve (in light blue). The southern edge of the forest sector is still affected
by the presence of agriculture, a sign of an unclear boundary and of unresolved
land issues.
East of the border, in the settlement area, the positive reforestation and the
spread of private woodlots that had been observed 5 years earlier seem to have
diminished. This is worthy of attention, as reforestation in the settlement areas
is critical for the protection of the streams that cross the forest reserve and that
have their sources in the upstream settlements. Furthermore, the presence of
private woodlots on farms has significant implications for the pressure exerted
by the population on the protected area regarding the collection of firewood
because it reduces the community dependence on forest resources and defuses
a potential threat to the forest.
 Deforestation in the Mau complex: a remote sensing analysis 39

Discussion
Using the maps resulting from the land cover supervised classification, it is
possible to estimate the area of each class to produce a diachronic picture of
changes in forest cover in the Southwest sector and compare it with the other
sources used in this research9. It is worth remembering that since 2001 the
boundary of the forest reserve has been changed. The years 2003, 2008, and
2014 keep representing the original extension, even if the eastern section was
degazetted in 2001 and converted to settlements.
The graph and the table below show the area (ha) of each class and the
corresponding percentage compared to the total extension of the forest sector
(84,685 ha).

Fig. 3.8 - Variations of land cover in the Southwest Mau, 1986–2014

(source: Stefania Albertazzi, 2020)

The graph informs of two key moments in the history of the forest, name-
ly 1995 and 2008 (Fig. 3.8). In 1995, a symmetrical trend of loss of forest
(“dense” and “transition”) and growth of the “cultivated/bare soil” class start-
ed, culminating in 2008 when a reversal trend began.
Since 1995, the “crops/bare soil” class has experienced a steady increase
from about 14,500 ha to about 37,000 ha in 2008. This figure is consistent with
the events previously mentioned. In the period 1996–2001, the creation of gov-
ernment settlement schemes within the boundary of the protected area took

9 The land cover classification accuracy assessment, resulting by the cover class correspond-
ence of training points and test points, is 91,74% for 2014; 89,18% for 2008; 91,71% for
2003; 97,18% for 2000; 92,67% for 1995; 95,27% for 1986.
40 Forest and Communities

place, while in 2008 awareness of the deforestation process emerged, creating

the basis for the launch of the rehabilitation and conservation phase.
The “transition forest” class seems to be the class most affected by the ex-
pansion of cultivation. This data informs us that the deforestation process has
mainly affected this type of forest cover, which experienced the greatest impact
in 2008, with its surface area at about 15,000 ha compared to about 32,000 ha in
1995. The “dense forest” class seems to be the least affected by the changes in
cover, with an area of about 29,000 ha in 2008, the year of greatest loss, com-
pared to about 38,500 ha in 1986.
Considering the percentage changes in class composition, it is confirmed
that the “dense forest” class was the least affected by the deforestation process,
suffering a maximum of a 9.9% decrease (1986–2008) and a 6.1% loss in the
overall period 1986–2014. The class “transition forest” experienced a maximum
change of -19.4% (1995–2008) and a total loss of 2.2% compared to 1986 and
2014 values. The class “crops/bare soil” experienced a maximum increase of
28.5% (1995–2008) and an 8.2% gain in the whole period.

Table 3.1 - Land use change by class area, 1986–2014

(source: Stefania Albertazzi, 2020)

Year Dense forest Transition forest Cropland/bare soil

1986 45,40% 36,70% 17,90%

1995 44,80% 38,10% 17,20%
2000 41,90% 28,20% 29,90%
2003 37,30% 24,70% 38,00%
2008 35,50% 18,70% 45,70%
2014 39,30% 34,50% 26,10%

The table underlines the relevance of the years 1995 and 2008. In the former,
about 45% of the entire Southwest Mau is occupied by “dense forest”, while in
2008 the same percentage represents the presence of “crops and bare soil”. The
year 2014 seems to indicate the establishment of a reversal in the deforestation
process, following the implementation of protection and restoration measures
for the Mau complex.
In conclusion, we can highlight three major points concerning the deforest-
ation process in the southwest sector of the Mau Forest. First, the proximate
cause of the deforestation that hit the forest sector is to be found in the con-
version of forestland into settlements. From 1995–2008, the agricultural front
proceeded from the eastern edge of the forest reserve towards the west at the
expense of forest cover.
 Deforestation in the Mau complex: a remote sensing analysis 41

Second, the loss of forest area started from 1995 and became an emergency
in 2008. From that year, the situation improved and was marked by a process
of regeneration that concerned the forest reserve and the ​​settlements (for a
certain time).
Finally, the Ndoinet Forest section was strongly affected by the human oc-
cupation following 2001. Compared to the entire Southwest Mau sector, this
portion of forest was largely converted into agricultural land and experienced a
massive human presence until 2008. There is still evidence of this in the forest
vegetation, in particular in the abundant presence of grasslands and shrubs, a
direct result of human action on nature.
4 Wildlife

The Mau Forest, one of the few remaining forests in east Africa, has a re-
markable heritage of animal biodiversity. One of the objectives of our research
was the participatory mapping of local homeotherm wildlife. The communities
living around the forest were directly involved in the process of recording the
species present in the area, mapping their ranges at a local level and overlap-
ping them to identify hotspots of species richness and conservation value. The
methodology was adapted by the World Wildlife Fund (WWF)’s global work of
ecoregional assessment (Dinerstein et al., 2000) and subsequent applications at
a more local scale (Bogliani et al., 2009; Pompilio et al., 2018), with a reduction
in terms of the scale of chosen taxa (from classes or macro-groups to single
species) and of the study area (from the continental level to the local level).
In the three forests, the participatory mapping work was always divided into
two phases—data collection through focus groups and transect forest walks,
while a camera-trap sampling session was added in Kiptunga. Before the map-
ping exercise, the informants (on average, 15–20 per site, both genders rep-
resented, with a majority of men) were asked to locate their house on a map
to test their ability to read it. The three rapid biodiversity assessments were
carried out by asking the community to list and locate the species thought to
be present. The assessments focused only on homeotherm fauna (mammals
and birds), as the main aim of the work was to use a few taxa as indicators of
ecological integrity rather than a complete zoological checklist.
For game wildlife and other mammals, informants were able to draw a po-
lygonal range at a local level. For every bird species, we asked informants to
indicate one to three locations where they spotted the species, finally getting
a monospecific cloud of points comparable to a polygon. An initial checklist
of bird species was selected by intersecting two different criteria. The first was
the list of the birds detectable in the Important Bird Area1 of the Mau Forest
Complex, and the second was a forest quality indicator based on Bennun et al.
(1996), who clearly divided a list of forest birds from Kenya and Uganda into
three categories of sensitivity to the forest habitat quality. The sample shown
to the population was a set of birds of high-quality forest, plus a few species
of lower habitat quality used as a control. Informants were not made aware of
Bennun’s ecological distinction during the data collection in order to avoid any
influence in the process of localization of the species.
For mammals, the original checklist was simply based on species potentially
present in the area, based on relevant literature on the actual or recent dis-
tribution of the species (Rodgers et al., 1982; Kingdon, 1997; IUCN, 2016).

1 The Mau Forest Complex is the Important Bird Area KE051, as listed by Birdlife International.
44 Forest and Communities

Informants were first asked to confirm or deny the presence of the species
from the original list in the forests. In order to ensure an individual’s ability to
recognize the species, all of them were identified in the local language. In the
different sectors of the indigenous forest and in the area dominated by tree
plantations, the community was then asked to give an index of “detectability”
of each species or of the ecosystem service enjoyed. The latter refers to the
benefits that people get from ecosystems and relates to natural products (fire-
wood, fruits, water, etc.) or cultural value (aesthetical, spiritual, etc.) or regulat-
ing activities (for water and air regulation, pollination, etc.) (MA, 2005; see chap-
ter 8). The index value ranges from 0 if absent, to 1 if rare, and to 2 if common.
As the index is a potentially subjective indicator, no value was accepted before
it was agreed by the whole group of informants (n = 15–20, depending on the
forest), thus representing the result of a common discussion.
The average probability across all forest areas was used to rank species ac-
cording to sensitivity, and the average value across the species was used to esti-
mate the degree of the ecological integrity of the forest sectors. The commu-
nity was asked for some additional information about their perceptions regarding
population trends, utility (e.g., food, commerce), the level of human–wildlife
conflict, and the hunting of single species. Hunting is illegal within the pro-
tected area but is still practiced by local communities, although it is difficult to
assess with what intensity. Once the data collection phase was completed, some
visits were made to the forests with local informants to detect direct or indirect
animal signs, such as feces, fur, or footprints. In addition, some camera traps
were placed in Kiptunga.

4.1 Koibatek Forest

The Koibatek section was divided into four main zones—the southern area,
where the temporary agriculture of the PELIS program (see below, chapter 5)
prevails; the north-eastern area dominated by shrubs due to recent fires (to-
gether defined as impacted areas); the central and forested area; and the western
area where bamboo prevails (later defined as not impacted areas). In the field,
the Kiptuget Forest area immediately north of the Koibatek section was also
considered an influential ecological factor. In fact, the Koibatek Forest is the
southern part of an ecosystem that has its core in the Kiptuget Forest, and
wildlife populations from the two forests are likely to be connected as a whole.
Analyzing the deep forest means, ecologically speaking, to take into consid-
eration the potential return of locally extinct wildlife under an adequate level
of protection, and it is therefore of strategic value in terms of restoration.
However, the Koibatek Forest acts as a buffer area for this deeper forest, which
probably represents the true ecological value of the whole area and which
should be protected through the adequate management of the buffer area.
 Wildlife 45

Fig. 4.1 - Koibatek Forest map resulting from a participatory workshop (Trivellini,
2018). The blue line and the star indicate the footpath taken with community inform-
ants during a forest walk; the red line indicates the Kiptuget Forest extension; and the
black lines indicate the boundaries of each section.

Mammals
The results of the focus group on the mammals are summarized in Table
4.1. In particular, we can highlight three main aspects. First, from the original
checklist of 32, the community confirmed only 15 species as present (plus an
undefined group of rodents). Out of 16 elements, 12 have a kind of utility for
the community (the informants indicated a direct use for 11). All species with a
decreasing population trend (or undefined due to rarity, such as the forest hog
(Hylochoerus meinertzhageni), were game species or carnivores that create conflict
(serval, Leptailurus serval). Of the 16 species, 11 (69%) show a certain form of
conflict with human activities, and 6 out of 8 (75%) of the increasing species
conflict with agriculture, as their range overlaps with farming areas.
The elephant (Loxodonta africana) and the forest hog, whose presence de-
pends on the thick forest habitat, are defined as uncommon and can only be
found in indigenous forest areas. The most detected species was the bushpig
(Potamochoerus larvatus), which can live in agricultural areas and poses serious
problems for crops (Seydack, 2013). The Harvey’s red duiker (Cephalophus har-
veyi), a game species typically linked to a forest habitat, was more common
in less impacted areas, while the olive baboon (Papio anubis), the blue monkey
(Cercopithecus mitis), and the black and white monkey (Colobus guereza) were de-
scribed as very common, as in the rest of the Mau Forest. In addition, the
bongo (Tragelaphus eurycerus, IUCN Near Threatened) and the buffalo (Syncerus
 Table 4.1 - Mammals species in Koibatek Forest according to the community informants (Trivellini, 2018)
Forest and Communities
46
 Wildlife 47

caffer), recorded in other forest sectors as a species very sensitive to human pres-
ence2, were not mentioned in Koibatek.
No significant difference was found in the number of reported species be-
tween the four sectors (Steel–Dwass test: pairwise comparisons, p always >
0.05). When clustering the sectors in impacted (agriculture and burnt areas)
and unimpacted (bamboo and forest areas) habitats, no difference was found
between the groups (Wilcoxon test, W=43, p> 0,05).
Although the indigenous forest was the area with the highest absolute num-
ber of species, other areas showed a certain species richness (particularly the
burnt area due to the presence of regrowing secondary vegetation). The lowest
number of species was detected in the southern anthropized area where PELIS
is practiced. The northern area of Kiptuget Forest, previously indicated as a
potential home to bushbuck and bongo, was walked in search of wildlife signs
and showed traces of several species (Fig. 4.3).

Fig. 4.2 - Mammals ranked by average detectability index in Koibatek Forest,

used as a proxy for the sensitivity of the species (impacted versus unimpacted
habitats) (Trivellini, 2018)

2 Kiptunga Forest communities have indicated the buffalo as a species that abandons an-
thropized areas in favor of thicker forest, probably because of its importance as a game
species.
48 Forest and Communities

Fig. 4.3 - Map of wildlife traces recorded during the forest walks in Koibatek Forest
and Kiptuget Forest, with forests analyzed using a remote sensing analysis of the land
cover (source: Trivellini, 2018)

Birds
Following Bennun et al. (1996), birds were divided into two groups, a more
specialist one (FF), indicator of a higher-quality forest, and a more generalist
one (f), indicator of a more anthropized forest. While the first group is biologi-
cally expected to be found in the indigenous forest, the second was expected by
researchers to be found in the margins or in the most impacted areas.
The mapping work reported the community’s idea that birds are rare in the
bamboo area and reaffirmed the ecological role of the dense forest (Fig. 4.4).
In the more anthropized areas, in the south and northeast, informants reported
a mixed ecological situation, with some “high-quality forest” and “low-quality
forest” indicator species overlapping. According to the informants, high-quality
forest species birds (FF) appeared to be present in degraded (burnt) areas also,
while low-quality forest species (f) were almost absent in the indigenous forest.
 Wildlife 49

Fig. 4.4 - GIS map of bird species as indicated by community members (Trivellini, 2018).
The green dots represent the specialized bird group and the red dots represent the gener-
alist group.

Overall, the picture emerging both from the analysis of the distribution of
mammals and birds describes a territory that is much more agricultural than
in other sectors in the Mau Forest Complex (e.g., the Ndoinet Forest section).
The human–wildlife conflict revealed by questionnaires mainly concerned
agriculture rather than cattle breeding, as almost no large predators were left3.
Koibatek Forest appeared to be much more anthropized than other study areas.
People, in particular, live in the southern area where the forest ecosystem is
highly fragmented by tree plantations and crop agriculture, without any relevant
contact with wildlife. Most of the species reported to be increasing were small-
er size species (medium and small mammals), species of little hunting interest
that carry a low level of conflict and therefore are more anthropophilic. The
wildlife communities testify to the lack of a real integrity of the landscape, the
existence of a residual forest, and a clear anthropic influence in the area of
tree plantations and agricultural activities. Species listed by the community were
fewer than elsewhere in the research area, and animal biodiversity, especially in
the southern agricultural area, appeared to be already compromised.

3 When asked about the presence of the leopard (Panthera pardus), the community answered “If
we knew that it was present, we would get organized to go and kill it”.
50 Forest and Communities

4.2 Kiptunga Forest

In the case of Kiptunga Forest, we proceeded in the same way as for Koibatek
Forest, asking the local community to list and describe the existing species of
homeotherm wildlife. In order to ensure individuals’ ability to recognize the
species, all of them (18 birds and 21 mammals) were first identified in the local
languages, Ogiek and Maasai4. For participatory mapping, Kiptunga Forest was
divided into six areas—four indigenous forests, Chebuin, Olengape, Kiboyet,
and Kaamweu, and two highly anthropized zones (tree plantations and open
areas)5. The average community index of detectability was used to estimate the
likely habitat quality of the different sectors (Table 4.2)

Fig. 4.5 - Map of the Kiptunga Forest edited through aerial pictures (Google maps),
fine scale mapping (Original GIS data of the WRI 2007) (Trivellini, 2018): bamboo (light
green), indigenous forests (dark green), open areas (yellow), and tree plantations (brown).

Mammals
The results of the participatory survey were coherent for birds, mammals,
and provisioning ecosystem services (see chapter 8 below). For the wildlife taxa,

4 To the south, the Kiptunga Forest borders Narok County, mostly populated by Maasai com-
munities.
5 Tree plantations occupy nearly 20% of the Kiptunga Forest section. Agriculture within
the protected area is attributable to a few PELIS initiatives and to the fact that within the
Kiptunga boundary are located the only Ogiek settlements allowed in Mau Forest.
Table 4.2 - Mammals’ detectability index in Kiptunga Forest (Trivellini, 2018)
Wildlife
51
52 Forest and Communities

the four areas of indigenous forests showed a similar average value of the de-
tectability index, with a crash in anthropized areas (Fig. 4.6). For mammals,
indexes were significantly lower than the one recorded in the forest sectors both
for the exotic plantations area and the open/grazed areas (Kruskal–Wallis test
for multiple comparison, p always < 0,05).
Table 4.2 and Fig. 4.6 show without possible misunderstanding that, accord-
ing to the community, outside the indigenous forest biodiversity and ecosystem
services (see chapter 8) collapse. This means that the population of mammals
in Kiptunga could be fragmented and, in the long run, genetically threatened,
with possible population reduction and localized cases of extinction (Dixo et
al., 2009; Wan et al., 2018; Lino et al., 2019). In a study about forest fragmen-
tation effects on the genetic diversity of different species at different latitudes
and longitudes, Lino et al. (2019) found that species with larger body mass
are the most negatively affected by fragmentation, that terrestrial and arboreal
mammals are more negatively affected than flying species, that herbivores suffer
consistent negative effect of fragmentation, and that forest-dependent species
are the most susceptible to the negative effects of fragmentation.

Fig. 4.6 - Average participatory detectability indexes for mammals, birds, and ecosys-
tem services in Kiptunga Forest (see also Albertazzi et al., 2018).

This scenario, applied to the participatory analysis performed in Kiptunga

Forest, explains well how the fragmentation of the forest sector could have
a relevant impact on the viability of the animal populations of the area.
Consistent with this framework, larger game species were reported by the com-
munity to have abandoned most of the sector (Table 4.2), remaining only in
the largest section of the forest (e.g., forest buffalo, Syncerus caffer in Chebuin).
 Wildlife 53

Many species had an average lower index of detectability in impacted areas

(Fig. 4.6), and almost all the species were perceived by the community to have
a decreasing population trend.
As for Koibatek, participatory mapping data were tested with fieldwork, in-
cluding transect walks and images collected during a fast camera trapping cam-
paign. Four study areas were sampled—the three indigenous forest areas of
Olengape, Chebuin, and Kiboyet, plus small forest patches located in the area
with tree plantations (Fig. 4.7). We undertook extensive walks with community
members, allowing the collection of qualitative data on animal presence through
indirect signs (feces, fur, footprints, and predation signs). Two camera traps were
placed in each forest sector at a minimum distance of 150 m from each other and
were used for 72 hours (in total eight traps and three days of continuous sampling
in December 2013). In total, there were 576 hours of sampling.
The hyena (Crocuta crocuta) was detected only in the area of tree plantations,
while the bushbuck (Tragelaphus scriptus) was detected in all three indigenous
forest areas. The hyena is widely considered to be anthropophilic, often fol-
lowing livestock and creating human–wildlife conflicts. It is defined as “an
opportunistic carnivore wherever there is animal waste, resulting from feeding
other carnivores or humans” (Kingdon, 1997). The bushbuck (as the name
explains) is an indicator of a forest habitat. Kingdon (1997) reports the species
is “dependent on dense vegetation”, while IUCN (2022) describes the species
as occurring “wherever there is cover to conceal it”.
Moreover, livestock (both sheep and cattle) were spotted in the most an-
thropized areas—the forest remaining around the tree plantations and
Olengape, the most degraded of the three forest areas, and the closest to the
most deforested part of Kiptunga.
Considering together data from camera traps and other indirect signs of
this rapid sampling session aimed at placing camera traps (a few days of walk-
ing), we found just 3 wildlife species in the plantations area and 9 in the forest
areas. In the plantations area, we found traces of hyena (Crocuta crocuta), wart-
hog (Phacochoerus africanus), and blue monkey (Cercopithecus mitis), quite adaptive
species living in primary and secondary forests (IUCN red list, 2022). In the
indigenous forest area, we found signs of blue monkey, bushbuck (forest sec-
tors of Olengape, Kiboyet, and Chebuin), leopard (Panthera pardus, in Kiboyet
and Chebuin), black and white colobus (Colobus guereza, in Olengape), bushpig
(Potamochoerus larvatus, in Olengape), aardvark (Orycteropus afer, in Kiboyet), and
buffalo (Syncerus caffer, only in the largest sector of Chebuin). Although the
leopard has a large range of habitats in Africa, it remains a large carnivore de-
pendent on a forest environment that provides sufficient cover and higher prey
densities (Balme et al., 2007). The buffalo (Syncerus caffer), traditionally found all
over the Kiptunga Forest, was reported by an elder to nowadays remain only in
Chebuin, the largest patch of forest that allows the existence of species now
disappeared from smaller forest areas.
54 Forest and Communities

Fig. 4.7 - Camera traps positioned in Kiptunga Forest in a fragmented forest patch
(rough grain) located close to tree plantations (small grain) (Trivellini, 2018).

Table 4.3 - Camera trap shots and indirect signs of wildlife in Kiptunga Forest
(Trivellini, 2018)
Table 4.4 - Forest birds used as an indicator of biodiversity in Kiptunga Forest (Trivellini, 2018)
Wildlife
55
56 Forest and Communities

Birds
Regarding the animal distribution, the community from Kiptunga drew
polygonal ranges for 20 mammal species (Artiodactyla: n=6; Carnivora: n=5;
Hyracoidea: n=2; Primates=3; Rodentia: n=3; Tubulidentata: n=1). Informants
(unaware of any ecological distinction) then added the point data regarding 18
bird forest species, later clustered according to the biological indicator used in
Koibatek Forest as high-quality (FF) or low-quality (f) forest indicator species
(Bennun et al., 1996).
Biodiversity hotspots were identified in a GIS environment as the result of
overlays of species ranges, ranging from 2–17 overlapping layers/species/per
pixel. The participatory exercise revealed the occurrence of high levels of spe-
cies overlay in the largest areas of indigenous forest. The final map (Fig. 4.8)
illustrates the results, including data from mammals and bird species. While
most of the points indicating forest specialist birds (FF) fall in mammals’ high
overlay areas, points indicating forest generalist birds (f) are typically located on
the edge of the forest.

Fig. 4.8 - Overlay of mammals’ ranges and birds’ spotting sites as identified by the
local community in Kiptunga Forest (Trivellini, 2018). Range (light to dark blue), class
1: 1–3 species; class 2: 4–6 species; class 3: 7–9 species; class 4: 10–12 species; class
5: 13–16 species. For birds, black dots represent individuals of species as high-quality
forest indicators (FF), and white dots represent individuals of species classified as
low-quality forest indicators (f).

As in the case of Koibatek, in Kiptunga the ecological information emerg-

ing from participatory data was very consistent between the two investigated
wildlife groups (mammals and birds), and the difference between the tree plan-
tations and the indigenous forest areas was shown to be the most important
 Wildlife 57

distinction within the protected area. The issue of the ecological connectivity
of forests emerges as central; preserving the remaining corridors and restoring
the functionality of other key points will be crucial to the long-term conserva-
tion of Mau’s ecological system (see chapter 9).

4.3 Ndoinet Forest

The Ndoinet Forest section differs from the other two by the absence of
silviculture and PELIS schemes (see sub-chapter 5.2) and the presence of a
degraded “transition forest”, which is the outcome of a complex history of hu-
man occupation and forced evictions (see subchapter 2.3). In this case, the most
relevant human activity for the forest is cattle and sheep grazing, evidencing an
increasing gradient of forest integrity from east to west.

Fig. 4.9 - Ndoinet Forest map modified by the community (Trivellini, 2018). The lines
indicate grazing ridges (white) and the indigenous Ndoinet Forest (blue) and its ad-
ministrative section boundary (red).
58 Forest and Communities

To map the presence of animal species within the forest, the community
first chose to identify land use patterns connected to grazing activities (Fig. 4.9).
Therefore, the study area was divided into two macro-areas—in the western part,
an inner zone where indigenous vegetation prevails and in the eastern section,
an area closer to the settlements that corresponds to the one that was occupied
in the 1990s and 2000s, where grasslands are used for grazing. This second area
is functionally divided into 12 sections, following the grazing ridges identified by
the community members. As above for Kiptuget Forest and Koibatek Forest, the
two forest sections west of the Ndoinet Forest (Itare and Mara Mara forests) but
ecologically connected to it are considered a potential source area for biodiversity
and are therefore considered in our ecological evaluation.

Mammals
The results of the focus group meeting with the community concerning the
presence of mammals in the Ndoinet Forest are presented in Table 4.5 and can
be summarized as follows. According to the community’s perceptions, of 24
present species, 12 have a decreasing population trend and seven (30%) have
an increasing trend6. Of the 12 declining species, eight (67%) suffer from both
overharvesting and habitat loss, three species (25%) are declining only due to
overharvesting, and only one species (the carnivore Leptailurus serval) is declin-
ing due to habitat loss alone.
Twelve species were shown to have a potential direct use by the community
or an economic utility (meat, skin, ivory), which explains the existence of some
hunting activity. Of this sample of “useful” species, 92% (11 of 12 species)
have a declining population trend, and in this group only the dik-dik (Madoqua
kirkii) was reported to have an increasing trend, consistent with the fact that
the species is generally known to be anthropophilic (Brotherton, 2013). At the
same time, 6 of the “non-useful” species (55%) have an increasing population
trend. Of the 11 species apparently in conflict with human activities, 6 that are
considered “useful” are declining, while 5 species without a specific utility are
not perceived as declining.
The depicted situation emphasizes the importance of the variable “utility”
in determining population decline, more than what happens, for example, for
the variable “human–wildlife conflict”. Therefore, hunting activity, particularly
small-scale poaching for subsistence food, seems to play a significant role in
determining negative population trends.
As for the other forests, the community was asked to state the perceived
probability that the investigated species were absent, rare, or common (see de-
tailed methodology at the beginning of the chapter) in the different sectors of

6 Four species characterized by poor information are not considered in the percentage calculation.
Table 4.5 - Mammals present in Ndoinet Forest (Trivellini, 2018)
Wildlife
59
60 Forest and Communities

the forest station and then to draw a polygonal species range for mammals or
spotting points for birds.
This participatory data analysis showed a clear difference between the 12 grazing
areas and the inner ungrazed portions of indigenous forest (Ndoinet, plus the deep
forest of Itare and Mara Mara). In general, the focus group data on detectability
indexes (Fig. 4.10) proved to be very consistent with the species range maps drawn
by the community, indicating that the grazing areas degraded by the encroachments
of the 1990s and 2000s are marked by the lack of many wildlife species.

Fig. 4.10 – Mammals’ detectability index in Ndoinet Forest (Trivellini, 2018). Data are
sorted according to the detectability in grazed areas as an indication of ecological ad-
aptability to anthropized areas.

With reference to Fig. 4.10, it should be noted that almost all animals have
Least Concern (LC) as an IUCN category; the only Near Threatened (NT)
species are the bongo (Tragelaphus eurycerus) and the yellow-backed duiker
(Cephalophus silvicultor), which have a minimal probability of occurrence. For
most species, the community detectability index seemed to be usable as a key
to read the species’ sensitivity. However, data on the presence of the elephant
(Loxodonta africana, IUCN CR) does not indicate it is widespread. Due to the
total lack of site fidelity and the species’ well-documented ability to move across
the largest ranges (Ngene et al., 2017, among many), they may also visit areas
of secondary quality, but their presence occurs very rarely as part of a complex
migratory pattern involving areas larger than the entire southwest forest sector.
The lack of occurrence of some ungulates in the eastern section is rep-
resentative of their status as typical game species, as they are likely targeted
 Wildlife 61

by poachers. The high probability of the occurrence of the dik-dik (Madoqua

kirki) is consistent with the fact that this has been described as the only “use-
ful” species with an increasing population, due to its high level of adaptability
(Brotherton, 2013). The high average index of detectability of monkeys is relat-
ed to their effective adaptability to a wide number of habitats, testified by their
involvement in agricultural human–wildlife conflicts and by the existing litera-
ture (Cercopithecus mitis: Butynski, 1990; Colobus guereza: Fashing and Oates, 2013;
Cercopithecus Ascanius: Cords and Sarmiento, 2013; Papio anubis: Zinner, 2013).
When considering the division between the grazing sectors (east) and the in-
digenous forest (deep forest in the Ndoinet western section, plus the Itare and
Mara sectors), it is evident that some species are absent in the eastern sectors.
The bongo, the yellow-backed duiker (both IUCN NT), the suni, the bushpig,
the Harvey’s duiker, and the buffalo are ungulates of high hunting interest,
while the leopard, the serval, and the hyena are predators of the former, which
determines the presence of the latter. In conclusion, while the western part of
Ndoinet Forest is a hotspot for different mammals and a refuge for the most
hunted ones, due to the absence of anthropogenic disturbances, the eastern
part is de facto the grazing area of the Southwest sector where many species are
no longer present due to greater anthropic pressure.

Birds
Table 4.6 presents the common, scientific, and local names of avian fauna
and the results of the data collection. The avian fauna are distinguished accord-
ing to the already identified two classes of birds, indicators of a higher forest
quality (FF) or a lower forest quality (f).

Fig. 4.11 - Yellow-backed duiker (red) and bongo (blue) ranges in Ndoinet Forest
(Trivellini, 2018). The selection by these species between the anthropized and not an-
thropized areas is clearly visible.
62 Forest and Communities

Table 4.6 - Birds present in Ndoinet Forest (Trivellini, 2018)

From the table, we can see that out of 10 FF species (good forest indicators),
6 are perceived as decreasing by the informants and 3 are perceived as increas-
ing. At the same time, all 4 “f ” species (low-quality forest indicators) (100%) are
perceived as increasing and very common.
The participatory mapping of birds confirmed the clear distinction between
the eastern and the western parts of the forest7. Forest specialist bird (FF) data
points fall typically in the western area, while data points on low-quality forest
birds (f) fall in the eastern edge of the forest (Fig. 4.12).
It is important to emphasize that while deforestation can potentially result
in a habitat loss for the first group, it can result in a habitat gain for the second
group. It should be noted that while specialist bird species (FF) are still present
in the most impacted areas, low-quality forest species (f) were never mapped in
the western dense forest. This is consistent with the fact that the ecotonal area
where grazing is practiced is produced by the loss of a dense forest and not by
the filling of a previously open area.
Birds suffer fragmentation much less than mammals (Lino et al., 2019) and
can have a different spatial perception compared to large mammals. Forest

7 While we also considered Itare and Mara as biodiversity sources, the community was working
only on Ndoinet Forest and therefore mapping species only up to the western boundaries of
Ndoinet Forest (outermost slanting line).
 Wildlife 63

edges still host large patches of indigenous forest (see chapter 3). While the
size of these areas is probably sufficient for the presence of some high-quality
forest (FF) bird species, it is too small to safely accommodate large wild mam-
mal species that need to hide in the deep forest and that are more sensitive to
human presence due to hunting or other anthropogenic disturbances.

Fig. 4.12 - Map of mammal and bird species in Ndoinet Forest (Trivellini, 2018). The
polygons indicate the areas where there are bushbuck (light blue), buffalo (red), bongo
(blue). Point data indicate the presence of specialist birds (green dots) and generalist
birds (white dots).
 Part 2
The natures of Mau Forest

The public narrative of the Mau Forest tends to produce an image of ho-
mogeneity (the complex is usually cited as a single entity) focused on its natural
component, a “primary” forest threatened by deforestation (GoK and UNEP,
2008). On the contrary, the reality is that the Mau complex is characterized by
great heterogeneity, and its features, far from being “primigenial”, are the out-
come of historical interaction with human communities.
To bring order to this heterogeneity, we have identified two categories of var-
iables that define different forms of interaction with the forest. First, the for-
est has relatively low anthropized areas (where indigenous vegetation prevails)
alongside completely artificial ecosystems (e.g., monospecific plantations for
commercial forestry). Along with this first typological distinction, we wanted to
include a second class of variables that refers specifically to the actors involved.
Therefore, we have parts of the forest in which the transformation (or conser-
vation) initiative comes predominantly from local communities and other areas
in which the initiative is primarily exogenous (national and international compa-
nies and organizations). This dual articulation produces a matrix of four forest
types that define different forms of co-evolution between society and nature,
identifiable as productive systems connected to a particular nature in the Mau
Forest. We have used the term “socio-ecology” to highlight the inseparable link
between society and nature that gives rise to relatively stable relationships over
time between human and non-human elements (Albertazzi and Bini, 2021).
The first type of socio-ecology is the socio-ecology of forestry, defined
by the cultivation and trade of timber grown in monospecific plantations of
pine, cypress, and eucalyptus trees. These are completely artificial ecosystems
(although they are considered forests according to the Food and Agriculture
Organization [FAO] definition; FAO, 2020) in which land transformation is
driven by exogenous actors, particularly national companies such as Timsales.
Tree plantations are developed within the boundaries of the protected area on
state-owned lands administered by the KFS. In the research area, silviculture is
particularly developed in the sections of Koibatek (Mt. Londiani sector) and
Kiptunga (East Mau sector). Today, it is a socio-ecology in crisis, due to the
imposition in 2018 of a government ban on logging in state-owned forests.
The second type of socio-ecology, family farming in settlement schemes, re-
fers to an agro-pastoral production system in which agriculture and livestock ac-
tivities are practiced almost exclusively for subsistence purposes. These systems
exert considerable pressure on the forest, and the outcome is, like in the previous
case, a sort of man-made forest, although here the main actors are the local com-
munities. This production system can be found in the villages adjacent to the
66 Forest and Communities

forest whose presence was formalized and legitimized in the Moi government’s
settlement plan (1993–2001). Family farming generally focuses on maize produc-
tion and animal rearing on farms of about 2 ha in size. An important element is
livestock grazing within the protected area, particularly in open forest spaces, such
as in the Ndoinet section where grasslands are used as common use space.

Fig. II.1 - Distribution of the four socio-ecologies in the research area

(Albertazzi and Bini, 2021)

The third type of socio-ecology we identified is that of conservation, which is

characterized by little anthropization and a strong exogenous presence. Particularly
prominent here are the multinational tea companies Finlays Kenya and Unilever
Kenya, which practice large-scale, export-oriented agriculture in Kericho County.
In Kenya, tea (Camellia sinensis) production areas are located adjacent to upland
forests because these environments provide ideal conditions for plant growth
(humidity and climate regulation, rainfall; GoK and UNEP, 2008). This produc-
tion system is particularly dependent on the ecosystem services provided by the
forest. For this reason, multinational tea corporations are funding a conservation
project that is highly restrictive of forest use (ISLA-IDH, 2018) and that aims to
produce an intact and “wild” forest where no signs of human presence are found.
The last type of socio-ecology we analyzed is a sort of renewed indigenous
agro-forestry system that is based on honey production and sustainable tour-
ism activities. This can be particularly found in the eastern forest sector of Mau
(Kiptunga section), where there are some Ogiek villages, the only ones allowed
within the boundary of the entire Mau Forest. This socio-ecology is not charac-
terized by a purely functional relationship with the forest, as in the case of family
farming; it is an integration of the natural and cultural components that simulta-
neously ensures the existence of Ogiek society and the protection of the forest.
5 Silviculture

The FAO definition of forest does not make a distinction between natural
and planted forest (FAO, 2020), but tree plantations are a special form of rela-
tionship between society and nature and therefore deserve a separate analysis.
In the case of the Mau Forest, the bond with silviculture is profound, both in
quantitative terms and in relation to the historical role played by this activity
in the area. Monospecific plantations are widespread in the forest, especially
in the central and northern part. In the Koibatek and Kiptunga sections, for
example, silviculture occupies about one-third (3,000 ha out of 9,000) and one-
fifth (2,000 ha out of 10,300 ha) of the forest, respectively. In these areas, tree
plantations occupy state-owned lands and are administered by the KFS or given
in concession to private companies, such as Timsales. Historically, the classifi-
cation of the area as “forest reserve” is directly connected to the need by the
colonial administration to preserve wood as an economic good and to develop
a profitable activity of silviculture. It is therefore necessary to briefly retrace the
colonial strategy to understand the evolution of this part of the forest.

5.1 History
The story of the Mau Forest is indissociable from the presence of the
Mombasa–Kisumu railway, “the origin and spine of what we now call Kenya”
(Wainaina, 2011, p. 40). Not only was Mau’s timber extensively used in con-
structing the railroad and to meet its energy needs, but the very economic devel-
opment of silviculture activity is strictly connected to this crucial infrastructure
that still marks the landscape of this area (Ofcansky, 1984).
In the first decades of the 20th century, plantations of exotic trees started to
replace the natural vegetation cover1. The planting of fast-growing species was
due to the desire to hinder the pressure exerted on the forest by agriculture, to
respond to the demand for raw materials, and to encourage the development
of a market for such products. In this context, forestry grew rapidly in the first
phase, with the development of many small sawmills in the Mau area. During
the Great Depression, the sector experienced a moment of crisis with the fail-
ure of some companies but later evolved into a more stable organization that
was consolidated by the independent state. In this sector, therefore, there is a
remarkable continuity between the colonial period and the post-colonial period.
During the European occupation, all the material and symbolic elements that
still characterize this form of land organization were fixed.

1 District Forest Office of Elburgon (n.d.), available at the Kiptunga Forest station.
68 Forest and Communities

From a material point of view, the main transformation brought about by

colonial forestry concerns the territorial order imposed on the forest. In fact, it
was in this phase that the “forest reserves” were demarcated, and their bounda-
ries have remained substantially unchanged up to now, limiting the usage rights
of the indigenous communities. From an environmental point of view, the col-
onizers replaced the complex nature of the primary forest with monospecif-
ic plantations and started a form of “scientific” management of the territory
made up of measurements and mapping built on an orthogonal plot that is
almost the symbol of colonization itself.
The colonial origin is also evident in the organization of the actors involved
in forestry. First of all, in this phase emerges the close relationship between
the state, owner of the forest lands, and private companies. Timsales, the main
company operating in the sector in Kenya, was established in 1932 with the
merger of seven sawmills created a few years earlier and obtained the manage-
ment of large portions of the forest2. With decolonization, control of the com-
pany was taken over by local management, but the strong link with the political
system remains (particularly the Kenyatta and Moi families; Ndii, 2018). Several
agreements repeatedly entrusted the private company with the management of
large portions of the forest reserve. Timsales progressively became a crucial
economic actor in the region, at the same time shaping the landscape of the
forest through its plantations, and outside the protected area, contributing to
the development of the city of Elburgon, where Timsales is headquartered and
which became a sort of company town that now has a population of approxi-
mately 30,000 (KNBS, 2019).
The second arrangement that originated in the colonial era and that still ex-
ists, albeit in slightly different forms, is the intersection between forestry activi-
ties and family agriculture. This occurred in the shamba system, in which farmers
can cultivate the fields in which the trees intended for forestry have been plant-
ed for a few years. This technique, quite widespread within the British Empire,
was first adopted in Kenya in 1910 (Fanstone, 2020) and has been revived today
under PELIS (see below).
The last area of attention concerns the symbolic dimension. Although the
terms have changed somewhat and the word “sustainable”, for example, has
been used in its current meaning only since the 1980s, there is substantial con-
tinuity in the ways in which forestry is described—essentially in terms of a ra-
tional organization of the land that allows its use in the long term and opposes
the “irrational” uses of local populations and immigrant farmers. In coloni-
al Kenya, in fact, alongside the common conflict between the colonizers and
the local population, there was a tension between the foresters who aimed at
the valorization of the forests through scientific forestry and the settlers who

2 https://timsales.webflow.io/about
 Silviculture 69

wanted to rapidly exploit the forest’s timber or simply convert these areas into
agricultural land.
The conflict between the timber industry and the Forest Department is well
represented by the case of the Lembus section in the central part of the Mau
Forest. This 26,000-ha forest has crucial environmental value for the Rift Valley
region because it hosts the source of the Perkerra river, one of the most im-
portant waterways in the region. The area was acquired by the English busi-
nessman Ewart Grogan in 1904 with the purpose of developing the timber
industry through his Equator Sawmills company. The tension with the Forest
Department lasted for decades until the company ceased its operations in 1935
due to the consequences of the international economic crisis, and the area re-
turned to the control of the Forest Department (Anderson, 2002).

5.2 The nature of silviculture

Generally, in the Mau Forest silviculture occupies areas with moderate slopes
(less than 30%), the most suitable for the mechanical operations that character-
ize this activity. In the Koibatek section, which covers the southern versant of
Mount Londiani, it means that the plantations are concentrated in the lowest
part of the forest. In the case of Kiptunga, which occupies the ridge between
different watersheds, silviculture alternates with natural forests.
The nature of this socio-ecology consists of allochthonous species of cy-
press (Cupressus lusitanica), pine (Pinus patula, Pinus radiata), and some varieties of
eucalyptus (e.g., Eucalyptus Saligna). These are fast-growing species whose timber
can be used 10–15 years after planting for the pole and energy production va-
rieties (mostly eucalyptus) or after 25–30 years for varieties used for processing
by the timber industry (cypress and pine) (KFS, 2014).
The organization and the planting time define the characteristics of these
forests and shape the landscape. First, trees are placed linearly and close to-
gether, averaging 500–1500 per acre of land, an arrangement that serves to
maximize yield and available space. Second, during its growth the plantation is
subjected to cyclical and planned operations of pruning and thinning. The first
term means the removal of branches, which is necessary to allow the plant to
develop without impediments by enlarging the trunk in height and width. The
second term identifies the progressive removal of trees to allow the remaining
ones to increase their biomass. This means that of the initial amount of plants
present, one-sixth will remain3. Tree plantations are organized and managed in
a rather strict way through 10-year plans and detailed mapping. In some cases,
they are controlled and monitored to prevent the theft of timber4.

3 Personal communication, KFS forest manager (2022).

4 Personal communication, KFS ranger (2022).
70 Forest and Communities

The presence of tree plantations creates a soil poor in fertile substances, with
an almost non-existent undergrowth and equally poor biodiversity. Research
has shown that the only animal species found on tree plantations are anthropo-
philic (cattle, sheep, hyenas). For this reason, planted forests are called “silent
forests” due to the absence of noise that differentiates them from natural for-
ests (Maathai, 2011; Trivellini and Lindon, 2014).
The peculiar ecosystem that characterizes these monospecific forests is com-
plemented by the nature produced by PELIS (Fig. 5.1). In the framework of
this program, the farmers cultivate the plots, generally about half an acre in
size, cleared by forestry activities, during a period of 4 years until the cano-
py of the new trees hinders the proper development of the crops. Koibatek
and Kiptunga PELIS areas mostly host monocultures of maize and potatoes,
and agriculture is conducted in a conventional way with the systematic use of
chemical inputs. Among the chemicals used are fertilizers (e.g., Cropsta multi-
purpose, a typical NPK foliar fertilizer), insecticides (e.g., Atom and Oshotion,
both manufactured in India and based on deltamethrin and malathion, respec-
tively), fungicides (e.g., Zetanil, manufactured in Italy and based on Mancozeb
and Cymoxanil), and herbicides (e.g., Moto Plus, manufactured in China and
based on Metribuzin).

Fig. 5.1 - Agriculture in a PELIS area, Koibatek Forest (source: Valerio Bini, 2022)
 Silviculture 71

On the whole, therefore, the PELIS program produces a heavily anthropized

environment that contributes to the fragmentation of the forest ecosystem.
Silviculture segments the forest in a direct way because plantations are ac-
cess-controlled areas but also in an indirect way because they are low-complex-
ity environments in which the ecosystem services (MA, 2005) are significantly
less than in the indigenous forests where people practice beekeeping, collect
medicinal herbs, and perform rituals. Ecologically, plantations are also spaces
in which the complexity and relationships of living organisms are kept to a
minimum. They are controlled and managed by the company to make resources
economically efficient, thus promoting rapid growth and maximum biomass
volume.
It has been highlighted that plantations are the outcome of a reorganization
of the living world in which organisms are dissociated from their original ecol-
ogies and are reproduced in identical forms to maximize replication speed and
efficiency (Tsing, 2015; 2018). On the Mau Forest plantations, two processes
typical of this spatial form are found. First, identical single-species organisms
(eucalyptus, pines, cypress) are assembled together in a bounded space. Second,
these organisms are alien and isolated from the surrounding ecologies, thus
producing a fragmentation in the ecology of the indigenous forest.

5.3 The narrative of sustainable forestry

As pointed out earlier, the narrative of silviculture as the rational way to
manage a forest has played a crucial role in the development of this activity
in the Mau Forest since colonial times. Bryant (1996) has shown how colonial
forestry services used the narrative of “progress” to introduce forms of forest
management that tried to preserve the economic value of the forest in a “scien-
tific” way that today we would call “sustainable”.
Over the decades, many different narratives have accompanied the evolution
of forestry on a global scale. Singer (2015) identifies five successive cycles of
ideologies that guided tropical forest management—ecological conservation,
participatory management, good governance, sustainable management, and the
forest–climate nexus (particularly through the REDD+ program5). The con-
temporary narrative of silviculture within the Mau Forest makes particular ref-
erence to the last two principles and thus invokes the idea of the economic
exploitation of forest resources that can last over time and emphasizes the role
that planted forests play in absorbing carbon dioxide.

5 REDD+ stands for Reducing Emissions from Deforestation and Forest Degradation.
Through REDD+ initiatives, financial incentives are provided to developing countries to
protect, manage, and responsibly use their forest resources with the aim of combating cli-
mate change.
72 Forest and Communities

On the Timsales website, the environmental dimension is addressed in a

specific section in which the company’s role in the fight against deforestation
is emphasized:

The vigorous re-afforestation program supported by Timsales Limited will ensure

that the forest has a continued existence for many years to come. The depleted
forest areas are now replenished on a sustained basis. The firm’s re-afforestation
rate is much higher than the rate of deforestation, aiding in achieving the firm’s
objectives in conservation6.

This narrative is based on a specific idea of nature strongly connected to the

scientific paradigm of ecosystem services within which ecosystems are analyzed
in a segmented way, identifying the single functions performed with respect to
human societies, with particular reference to the carbon cycle and the regulation
of the hydrological regime.
In the climatic context, the forest is essentially narrated as a “carbon sink”,
and in this sense forestry responds to the need to increase the forest area of the
planet, regardless of the qualitative characteristics of the forest itself. Regarding
the regulation of the water cycle, this is a central issue for a state with serious
problems of access to water resources, such as Kenya. The country framed the
defense of forests essentially as a tool for the protection of the strategic water-
sheds (the forests considered as “water towers”). In this direction, the Kenyan
government has created the KWTA, the government agency entrusted with the
management of the country’s upland forests.
In these perspectives based on the reduction of forests to mere providers of
specific ecosystem services, there is an equalization between planted and natural
forests and therefore a sort of “fungibility” between the two environments. At
the international level, this equivalence is rooted in the FAO approach that con-
siders monospecific tree plantations as part of reforestation programs (FAO,
2020).
On the national scale, we find the same idea of equivalence between indig-
enous forests and plantations. In the mapping of East Mau produced by the
KWTA (2019, p. 28), for example, there are no differences in the representa-
tions of the various forest covers (forest, open forest, grassland), although they
perform very different functions from a social and environmental perspective.
The same is true in the KFS’s multi-year strategic plan (KFS, 2017, p. 15) where
the first goal listed is the rehabilitation of the country’s five main forests, includ-
ing Mau, and among the various strategies is listed planting with exotic species.
The “discourse” of forestry is thus centered on the combination of sus-
tainability and ecosystem services that drives many national and international
policies and frames the forest essentially in terms of its economic value, thus

6 https://timsales.webflow.io/about
 Silviculture 73

equating naturally growing forests and artificial ecosystems, such as monospe-

cific plantations.

5.4 Silviculture: what future for monofunctional forests?

Silviculture in the Mau Forest is more than just a way of managing forests;
its history is intertwined with the evolution of this area to such an extent that in
some portions of the Eastern sector one has the feeling of being in a “compa-
ny forest”. Timsales company controls large portions of the forest, its factory
shapes the landscape of the town of Elburgon, and for decades it has provided
work for the majority of wage laborers in the area. As is often the case in such
situations, its role has extended far beyond the economic sphere, investing in
society as a whole through the construction of “Timsales schools” and even the
creation of the local soccer team (Timsales F.C.).

Fig. 5.2 - Planted forest managed by Timsales in Kiptunga (source: Stefania

Albertazzi, 2022)

The artificial forest that is the outcome of this now centuries-old activity is
a monofunctional space, simplified in its material and discursive characteristics
and extremely fragile because its structure is monostable and thus exposed to
the fluctuations of the markets on which it depends. This aspect is crucial to
understand the present situation of this part of the forest. In 2018, the gov-
ernment imposed a logging ban in public and community forests triggering a
74 Forest and Communities

structural crisis that still persists. The ban, officially motivated by a desire to
protect tree cover, was partially lifted in 2020 but still directly affects the activity
of the forest sections that are more linked to silviculture, such as Koibatek and
Kiptunga. The action also appears to be linked to a need for government inter-
vention in the commercial timber sector to address mismanagement brought
to light by a KFS investigation (Ministry of Environment and Forestry, 2018).
The consequence of this stop to logging activities produced effects in at least
four ways. First, many forest sections depend financially on the benefits com-
ing from forestry timber sales. In Koibatek Forest station, for instance, in the
period 2009–2012 the benefits deriving from timber (160 million KES for the
4 years) covered 98% of the total income from forest activities. Second, at the
moment the plantations lie in a state of semi-abandonment, as no maintenance
can be carried out. These activities used to involve the members of the CFAs7
that consequently have lost economic opportunities and thus bear the brunt
of this decision. Third, the ban jeopardized the PELIS program, progressively
reducing the areas available for farming. The KFS had set aside spare areas
within the plantations that allowed the program to function in recent years, but
the process now seems to have reached its limits8. Finally, the town of Elburgon
was directly affected by the stop of logging activities with thousands of layoffs
and the paralysis of the city’s economic activities.
The future of this part of the forest is thus strictly connected to the evolu-
tion of silviculture, which is at the moment still uncertain. What is at stake now
is not only the plantation areas in the forest but the whole socio-environmental
structure produced by the silvicultural system.

7 The Community Forest Associations (see sub-chapter 7.2) are village-based organizations
composed of local members who gather spontaneously to co-manage a state or a community
forest with the Kenya Forest Service.
8 Personal communication, KFS forest manager (2022).
6 Settlement schemes and the forest as
commons

6.1 The new geography of settlement schemes

In the 1990s, at the time of the establishment of government settlement
schemes within the boundary of Southwest and East Mau, the areas were cov-
ered by various natural forests, grasslands, shrubs, and plantation forests (in
East Mau). The Ogiek population and other forest-dweller communities were
living off shifting agriculture and animal rearing, accompanied in the case of
the Ogiek by wild fruit collection, hunting, and beekeeping.
The settlement schemes established across 60,000 ha of Mau Forest have led
to a fundamental shift in the geographies of Mau. In particular, in Southwest
Mau thousands of people migrated from Kalenjin districts of the Rift Valley
Province, while in the case of East Mau the people settled in a formerly low-pop-
ulated area surrounded by small towns and infrastructure (roads, railway).
The creation of settlement schemes brought forest lands under the control
and rule of the government, with the construction of formal villages, under
an administrative order, defined through territorial units and constituencies.
Specifically, in the case of Southwest Mau the creation of settlement schemes
was not connected to the provision of public service (dispensaries, schools,
roads); rather, they were initiated by community members and later supported
by or included within the state administration, even if in the most remote places
they are still lacking today1. We can say that there occurred a process of state
making (Scott, 1998) that put previously unmanaged and partially uncontrolled
forestlands under the jurisdiction of the state.
However, three movements of people occurred that had an impact on forest
cover and that had significant social relevance. The first relevant movement was
the eviction of forest dwellers prior to the beginning of the settlement program
and the gathering of the displaced people near the forest stations that at that
time served as waiting points for the registration of the land allocation to future
beneficiaries (Southwest Mau). This process led to the abandonment of the
informal villages dispersed within the forest boundary, leaving traces of settle-
ment and open forest areas surrounded by functional tree species (e.g., cypress;
sodom apple, Calotropis procera). The second relevant movement was the spon-
taneous but politically mediated migration of thousands of people from the
Rift Valley Province to Southwest Mau and East Mau that led to a process of

1 Interviews, 2019.
76 Forest and Communities

continued and uncontrolled land allocation. This dynamic created the premise
for the institutionalized settlement of individuals beyond the designated area of
settlement schemes for a certain time and subsequently drove vegetation cover
changes inside the natural protected area (Southwest Mau). The third relevant
movement was the definitive sedentarization of forest dwellers, particularly the
Ogiek, who came to be farmers and ended up living permanently on farm-
lands. They claim that the forced sedentarization caused sickness in people and
livestock, causing a sharp increase in deaths and a difficult adjustment to the
environment (different from the forest) for both humans and non-humans2.
From another point of view, the allocation of public forest lands has gen-
erated longstanding hostilities between communities and toward the state that
are still relevant today and make Mau Forest much politicized. Intra-community
hostility can be traced back to the migration of people to the Nakuru district
from neighboring areas that induced a feeling of land dispossession, particu-
larly for the East Mau Ogiek community, who accused other ethnic groups
(Kipsigis, Tugen, Maasai) of appropriating what belongs to them by right (an
informal right derived by an occupation of several centuries). The relationship
with the state is marked by animosity and suspiciousness for different reasons.
On one hand, some Ogiek individuals think the government has favored spe-
cific tribes with whom there are privileged and stronger patronage ties. On the
other hand, there is a general resentment towards certain figures (surveyors, ad-
ministration officers, politicians) who are considered corrupt and who are held
responsible for the allocation to second and third parties of the same parcel
of land3. Finally, there is a certain sense of abandonment due to the continued
presence of the caveat on land transactions, which forces thousands of people
into a situation of vulnerability and exposes them to possible eviction4.
A final element to focus attention on is that the creation of 60,000 ha of set-
tlement schemes with the establishment of permanent populations surround-
ing the most important water tower of the country put unprecedented environ-
mental pressure on the forest. In fact, 170,000 individuals or 37,000 households
(according to the last census; see KNBS, 2019) pose a major challenge to the

2 Interviews, 2022. There are different views on the sedentarization of the Ogiek in settlement
schemes. While some members blame the government for the severe impact it had on hu-
mans and animals, others remember that period with gratitude and as an important time that
laid the foundations for the hoped-for economic development of Ogiek communities.
3 A serious land management issue in Mau’s settlement schemes concerns the fact that often a
land allocation document reported two or three different individuals as beneficiaries. In other
cases, some found that the land they lived on was owned by others and sold by them to third
parties without their knowledge.
4 On the matter of eviction organized and threatened by the Kenya Forest Service, a person in
Kiptunga (2022) stated that “They cannot evict us, because we were here before them”. The
sentence remarks the claim of the Ogiek communities of a centuries-long land occupation
of Mau.
 Settlement schemes and the forest as commons 77

relationship between society and the forest. Of the various topics involved in
this matter (agriculture, infrastructure, social services, etc.), the most pertinent
for this research are the issues of energy supply and grazing that allow us to
see the dynamics that characterized the ecology of settlement schemes and that
produced a specific nature in the Mau Forest.

Fig. 6.1 - Tea buffer zone (left) and farms close to the Ndoinet boundary (source:
Drone Adventures, 2019)

6.2 Settlement schemes’ metabolism

The rural areas of Kenya are strictly dependent on biomass for their energy
needs. Data shows that 90% of people utilize firewood or charcoal for cooking
and heating (in the rainy season), while at the national scale the average falls to
70% (UNEP, 2009). The latest 2019 census (KNBS, 2019b) for the Kuresoi
South and Kuresoi North areas confirms the figures and attests the use of fire-
wood as cooking fuel by 86% of the population in the two sub-counties, along
with the use of charcoal (11%).
The settlement scheme areas are no exception, and households rely largely
on firewood from the forest. For lighting, use of the electricity grid or photo-
voltaic micro-installations is growing, but for cooking and heating the burning
of biomass remains almost the exclusive source of energy supply (KFS, 2015;
KFS, 2018). Interviews show that all the members of local communities close
78 Forest and Communities

to the Ndoinet Forest section (Southwest Mau) (n=1055) use firewood as their
primary energy source. Of them, it was found that 52% (n=55/105) were ex-
clusively sourcing firewood within the boundaries of the protected area (the
major sector of Southwest Mau and the small forest of 32 ha surrounding
the Ndoinet Forest station6) with varying frequency, depending on the cooking
technologies used.
The common cooking system is a high-energy one. It consists of three
stones positioned close together, on top of which are placed the pots used
for cooking. Evaluation research conducted for a project funded by the Italian
Agency for Development Cooperation (Njoroge, 2020) shows that households
(with an average of seven members) using the traditional three-stone method
consume about 60 kg of firewood per week, corresponding to four trips into
the forest with loads of about 15 kg. Multiplied by the number of families living
in the protected area, this figure corresponds to approximately 115,000 tons
of firewood consumed per year. It is a critical figure, also taking into account
the impact on the lives of women, who are mainly responsible for transporting
firewood. The activity of searching for and accumulating firewood generally
lasts one hour, takes place in the open areas within 1 km of the forest reserve
boundary (47% respondents, n=7/15), and is strenuous.
An alternative biomass source of energy is charcoal (for 25%–30% of
households; see Njoroge, 2020 and County Government of Nakuru, 2018), the
consumption of which is estimated at 15 kg/week per household at a cost of
approximately 300 KES (2.3 €). Charcoal is purchased or self-produced locally,
either legally from plantation wood or illegally within the forest from more val-
uable wood (e.g., cedar, podocarpus).
The general rule for firewood collection imposed by the KFS and advocated
by the CFA itself is the prohibition of cutting live trees and damaging the for-
est. Furthermore, each individual is obliged to pay a small fee to get collection
permission from the KFS (100 KES or 0.8 €/month). Therefore, open forest
areas play a crucial role in the local communities’ social metabolism, provid-
ing an essential contribution to household energy needs. In this sense, if we

5 Using semi-structured interviews, we investigated the relationship between the Ndoinet

Forest and the local communities in the sub-locations of Chemare, Chematich, Tinet,
Kapnanda, and Ararwet (Nakuru County) during the period 2018–2020. We met 105 peo-
ple with the aim of researching a multiplicity of aspects—the socio-territorial context; the
process of settlement on the plot; and the uses of the forest by the communities (grazing,
wood collection, beekeeping, collection of medicinal products, rituals). The interviewees
were mostly members of the CFA of Ndoinet, and in carrying out the work we were assisted
by a local collaborator and accompanied by one or two local members who organized the
interviews in the villages according to the criteria we requested.
6 The Southwest Mau consists of a main body of 60,000 ha (in which the Ndoinet section is
located) and a small forest of 32 ha located around the Ndoinet Forest station and 5 km away
from the main forest sector.
 Settlement schemes and the forest as commons 79

visualize the energy relationship between forest and settlement schemes, we

would see a unidirectional flow of matter from the former to the latter, reflect-
ing a rigid spatial division between forest and cultivated fields, which in turn
echoes the division between nature and society.
However, to facilitate the energy supply and alleviate human pressure on
the forest, NGOs are supporting the establishment of private woodlots in the
village, distributing fast-growing species (pine, eucalyptus, and cypress) usually
planted along the boundary of the farm. The initiative meets a legislative obli-
gation of (RoK, 2009) that imposes a minimum of 10 percent forest cover on
private plots, in order to support agroforestry and energy self-sustenance. In
7–10 years, a family will be able to harvest the first tree branch, thereby lessen-
ing the supply from the protected forest.

6.3 Grazing and grassland

A similar dynamic is detectable for access to pastures within the bounda-
ry of Mau Forest. In this region, the agro–pastoral system predominates, and
livestock (cows and sheep) enjoy considerable importance. The farms located
adjacent to the natural protected area in the settlement schemes’ zone have a
limited extension (2 ha) almost entirely devoted to agriculture (maize, sorghum,
vegetables). In these units, it is difficult to reserve a space for grazing or pro-
ducing fodder, and stable farming can be found only in pilot projects funded
by international cooperation7. For this reason, having grazing areas available is
essential for the farmers/herders of the Mau region.
Grazing within the forest is allowed during the day upon a fee payment
(100 KES/cow/month), but overnight stays are not permitted. Of the live-
stock present, sheep are raised for selling or local slaughter (a sheep can be
valued at 4,000 KES–6,000 KES or 30 €–45 €), often on special occasions
(e.g., Christmas, weddings). Cattle are a key source of livelihood for families
and an element of significant cultural importance to the Kalenjin people. Cows
are used for milk production and for the safekeeping of wealth and are to be
sold only if necessary (e.g., payment of school fees)8. The cows owned by the
people interviewed are mostly indigenous or hybrid breeds (crosses of local and
non-local species) and therefore produce small quantities of milk, about 1 L–3
L per day9. Despite the limited amount, only 40.5% (n=17/42) of the respond-
ents produce milk exclusively for family needs, while 59.5% (n=25/42) manage

7 We visited six farms involved in the stable grazing project of SNV NGO, part of the ISLA-
IDH partnership.
8 A cow can be worth 15,000 KES–30,000 KES (115 €–235 €), and often a family sells one per
year for the reasons mentioned above (74% respondents in our investigation, n=14/19).
9 This is in contrast to cows of the Jersey or Friesian variety that produce 7 L–15 L/day, de-
pending on the quality and quantity of their feed.
80 Forest and Communities

to sell half or all the production. Both these values highlight how fundamental
milk is for subsistence and for a small market economy in areas marked by high
self-consumption.
The product is either sold locally (e.g., to neighbors or commercial centers)
or to processing companies (e.g., Coorta, Brookside Dairy Ltd, the latter owned
by the Kenyatta family) that collect the milk themselves or to whom the milk is
delivered by motorbike taxi drivers (boda boda). Prices vary, and there is a general
decrease from 30 KES/L to 22 (even 15) KES/L.
More than half of our respondents use the Ndoinet Forest (Southwest Mau)
for grazing and access to pastures (54%, n=57/105). As previously mentioned,
the first few kilometers from the forest reserve limit (and particularly the areas
close to the boundary) into the interior are characterized by the presence of
grasslands, the most suitable spaces for grazing. It is worth underlining that it
is the history of human occupation and abandonment of the forest reserve,
expressed in forward and backward movements of the population, that has cre-
ated these open areas. Through the actions of cutting down trees and digging
soil to establish homes and small farms, which led to the formation of informal
small centers, and the grazing of animals and the trampling of the ground,
these grasslands have been formed and maintained in the Ndoinet Forest sec-
tion (Southwest Mau).

Fig. 6.2 - Livestock fence in “Ndoinet transition” Forest (source: Valerio Bini, 2018)
 Settlement schemes and the forest as commons 81

In addition, we can say that the farmers living adjacent to the Mau Forest
consider its pastures a sort of farm extension into the natural protected area.
First, following informal rules and the geo-morphological shape of the high-
lands, the farmers of a village usually cross the natural protected area boundary
and access the pastures of the same ridge, which is renamed according to the
village located in the adjacent settlement schemes. The division into ridges is a
system of territorial organization probably derived from the Ogiek culture in
which the konoito is a strip of land of a few kilometers delimited by streams,
over which the lineage had certain rights to use resources, particularly for bee-
keeping (Blackburn, 1970; 1982).
Second, the forest open areas in which the livestock graze are still identi-
fied by the names of the informal villages established in previous decades (see,
e.g., the areas of Kapkiwaron, Kimesto, Kipkosor, Cheptalukiat, Kiptegelde,
Kipsengwet, and Tilolwet cited in the archival documents of the 1990s as evic-
tion centers that are still present in the cartography of Ndoinet forest10). In this
sense, the toponymy of pastures is a valuable indicator for tracing the environ-
mental history of the forest. All these elements together emphasize the agency
and the role of farmers/herders and their livestock in the production of the
forest’s nature in general.
The presence of pastures is so important that their availability is probably
regulated through fire, an illegal but frequent activity. The extension of the
phenomenon is limited and seems not to represent a danger for the forest. A
remote sensing analysis for the period 2010–2019 in the Ndoinet Forest section
shows that this practice has been intensifying in the last years, with 135 ha of
burnt land in 2017 and 550 ha in 2019. The location of the signs of fire varies
spatially. They are mostly located near the cutline, between 1 km and 3.5 km,
and between 7 km and 9 km, distances that could respectively indicate the grass-
land used for daily grazing and for the illegal overnight stay of livestock (Fig.
6.2)11. It is worth noting that over the years, some of the burnt areas overlap,
particularly in the central-southern section of the Ndoinet Forest. This, togeth-
er with the fact that the areas are generally located close to each other, seems
further confirmation that they are not random fires. Instead, fire is probably
used to renew the grass or burn the ferns (Pteris aquilina) that suffocate and pre-
vent the generation of plants and that can be poisonous to people and animals
(KFS, 2019; personal communications, 2019).
Another aspect involved in this relationship with the forest concerns en-
croachment beyond the protected area’s boundary. Although the physical sepa-
ration between forest and settlements is today more recognizable and respected
than in the past, in some locations in the East Mau (Mariashoni and Sigoin

10 See the cartography available at the Ndoinet Forest station.

11 The figures are consistent with what emerged in some interviews in 2019 and 2020.
82 Forest and Communities

villages), farms press on the protected area and erode land to turn it into culti-
vated fields. This process is facilitated by the fact that the vegetation immediate-
ly close to the boundary of the protected area is often characterized by a grass-
land cover or a mix of shrubs, isolated trees, and re-growing species, precisely
because these areas are more subject to anthropogenic pressure and harvesting.
The response of the KFS is often eviction and a rapid planting of fast-growing
exotic species (pines, cypresses) to re-appropriate the area and prevent future
encroachment. The reasons behind the encroachment into the forest reserve’s
boundary are varied. In some cases, these actions are part of an intricate past
of previous occupations with respect to which there is a perceived continuing
claim. In other cases, they are the work of new immigrants to the area, whose
illegal behavior is condemned by other community members12.
It is important to focus on the fact that this relationship between forest and
settlement schemes is specifically centered on open forest areas where people
can find firewood and grass within 2 km–3 km from the boundary. The local
people approach the dense forest in Ndoinet (Southwest Mau) with tall trees
and closed canopy with caution. Slightly less than half of the respondents in our
investigation (47%, n=26/55) visited dense forest areas in the last year, mainly
to collect medicinal herbs13. During the interviews, it emerged that the dense
forest is a frightening place because of the presence of wildlife, which is reason
enough not to visit it. Several respondents reported that they had not walked
into a dense forest for many years, thus showing how the relationship with the
forest has changed with the implementation of the settlement schemes.

12 Personal communications in 2020.

13 The fact that the recurring motivation for visiting the dense forest is to collect medical prod-
ucts may suggest that there is a high biodiversity within this canopy that is not found else-
where (chapter 4). However, several respondents state that they are planting some herbs on
their farm, especially those that grow further away from the forest boundary, to have them
available at all times.
7 Forest conservation

Forest conservation defines a particular relationship between nature and so-

ciety that produces a specific geographical configuration. We identified three
modalities of conservation with recognizable characteristics that produce dif-
ferent socio-ecological relations—fortress conservation, community conserva-
tion, and conservation with direct political purposes.

7.1 Fortress conservation and plantation agriculture

The first kind of nature conservation emerging in the Mau Forest focuses on
the link between plantation agriculture (tea) and dense forest. It is strictly con-
nected to the logic, interventions, and narratives developed in colonial times.
The tea plant (Camellia sinensis) was introduced from India and established in
Kenya at the beginning of the 1900s, while its trade began two decades later
(KHRC, 2008). In the 1920s, some British companies established their planta-
tions in the city of Kericho, close to the Southwest Mau Forest, where the pres-
ence of Finlays’ and Unilever’s plantations (over 20,000 ha of land) still mark
the landscape. Over the last years, Kenya has always been the world’s leading
exporter of this product (Intergovernmental Group on Tea, 2022), but recently
companies have started to raise the alarm about a production decrease linked
to climate change in the area, referring particularly to precipitation volatility
(University of Cambridge, 2012; ISLA-IDH and Finlays Kenya Ltd., 2019).
The main cause of these environmental changes is found in the deforestation
that has characterized the area, and subsequently the solution of conservation
emerged as a priority. From this perspective, population and its activities (graz-
ing, firewood collection, charcoal production) are considered the major threats
to the forest, and interventions have been put in place to reduce, at minimum,
the use of forest by local communities1. While in the literature a common pat-
tern of deforestation relates to conversion from forest to plantations (e.g., in
South America; see Hosonuma et al., 2012), in this case the large-scale tea es-
tates are driving a specific form of forest conservation2.

The historical background

“Fortress conservation” has a long history in the Mau region. Its roots can
be found in the forest management strategies of colonial times, when the Forest

1 Communication of the Forest Manager of Ndoinet Forest (Southwest Mau) (2019).

2 Tea plantations are not the only kind of cultivation present here; small-scale tea growers have
spread, but financial support and planning for conservation can only be traced to multina-
tional companies.
84 Forest and Communities

Department governed protected forests for timber production and watershed

protection, strictly restricting use by local populations (Mwangi, 1998). This
approach to nature conservation is characterized by a top–down perspective
based on surveillance and enforcement that denies or severely restricts local
communities’ access to the forest. The “fortress” reference underlines the de-
fensive and exclusive logic aimed at the protection of a hypothetical “wilder-
ness” that can only be preserved by separating the forest from humans (Adams,
2004).
Decades later, in post-colonial Kenya, KIFCON—a development coop-
eration project (£4.86 million) funded by the UK’s Overseas Development
Administration within a World Bank program known as Forestry IV—proposed
a similar strategy. The project was implemented in the period 1990–1994 and
was guided by the assumption that there was a fast decline of Kenya’s forests
because of the high demand for land and resources, exacerbated by population
growth (KIFCON, 1993). The project aimed to conserve the indigenous forests
of the country and to introduce joint forest management with the forest-de-
pendent population (KIFCON, 1991). The project was conceived on two lev-
els—the national level and the level of pilot areas. With the first one, KIFCON
aimed to conduct extensive data collection, forest mapping, planning, and pol-
icy development. With the second, the project covered three areas of direct
intervention, including the Southwest sector in the Mau complex. In particular,
the project saw the population as the most significant factor putting pressure on
the Southwest Mau. This view was motivated by the fact that at the beginning
of the 1990s Mau was inhabited by dwellers consisting of Ogiek communities
and squatters who had been living in the forest for various reasons, including
rumors about a possible government forest excision for a conversion to set-
tlement3 (Bateson, 1993; RVPA, EC 1/6/14) and the return of the landless
people after the 1992 political clashes in the Molo area (Daily Nation, 1993;
Moorehead, 1993).
The intervention planned in the Southwest Mau was the resettlement of all
the eligible squatters in the eastern part of the forest sector (24,000 ha); at the
same time, the remaining part of the Southwest Mau (approximately 60,000
ha) would have been conserved with the cooperation of the local communities
(Bateson, n.d.; KIFCON, n.d.). The decision to target the eastern section was
due to the fact that this was considered by KIFCON as a “degraded” area with
a low-value vegetation cover described as consisting mainly of grasslands and
shrubs (Jackson and McCarter, 1994). In reality, the area was characterized by
a fragmented and intermixed vegetation (Davies, 1993; Blackett, 1994), a detail

3 In 1985, 2,733 ha of the Southwest Mau sector were excised to allow the settlement of forest
squatters in the Korau Settlement Scheme. The event created a significant precedent about
the re-proposition of a similar opportunity.
 Settlement schemes and the forest as commons 85

that makes KIFCON’s decision questionable and stresses the preference for a
specific kind of “nature” (a high and dense tree canopy) that still persists.
In this framework, KIFCON was responsible for the census of the for-
est dwellers (1991–1993) in collaboration with the Kenyan Provincial
Administration and under the supervision of the Office of the President and
of the British High Commission (KIFCON, 1991; KIFCON, 1992). From the
exercise, it emerged that in the forest were living 3,793 households or 18,044 in-
dividuals, of which 15% had been categorized as “Ogiek” and the remainder as
“Non-Ogiek” whom KIFCON believed to belong to the Kipsigis ethnic group
(Bateson, n.d.). In the end, the resettlement in the Southwest Mau never took
place under the KIFCON project but in the following years (1996–2001) was
led by the District Administration of Nakuru (see sub-chapter 2.3).

The ISLA-Kenya Project

Recently, a nature conservation program that recalls the “fortress” approach
has begun operating in the Southwest Mau led by an international public–private
partnership called Initiative for Sustainable Landscape-Initiative for Sustainable
Trade (ISLA-IDH). The initiative is promoted by IDH, a foundation created
by the Dutch Government in 2008 to foster public–private initiatives linked to
sustainable development. In particular, ISLA focuses on promoting sustainable
relations between society and the environment, mainly in tropical forests close
to agricultural areas where big companies, particularly Dutch companies, have
relevant interests. In the case of the Southwest Mau, the initiative is linked to
the presence of tea plantations owned by the Anglo–Dutch company Unilever.
This initiative reflects two significant political changes. One is in internation-
al development cooperation, where the private sector is progressively more in-
volved in what has been called a post-aid era (Mawdsley et al., 2013). The second
is a shift in nature conservation where, since the 1990s, the private sector has
entered a field that for decades was dominated by the state (Brockington et al.,
2008).
The ISLA-Kenya program was established in 2016 to protect the Southwest
sector of Mau with an initial budget of €3.7 million funded half by IDH and
half by a large network led by Unilever Kenya and Finlays Kenya4. The ap-
proach is aimed at improving farm productivity around the forest, diversifying

4 For more information, see also https://www.idhsustainabletrade.com/initiative/isla-kenya/.

The partnership is composed of the government organizations of the Kenya Forest Service,
the Kenya Wildlife Service, the Kenya Water Tower Agency, the Water Resource Management
Authority, and the Nyayo Tea Zone Development Corporation, with companies such as the
Kenya Tea Development Agency, KENGEN, Safaricom, and the Timber Manufacturers
Association. A few international cooperation agencies are present (GIZ, UK AID,) together
with conservation organizations (SNV, Rhino Ark), a research institution (CIFOR), and local
community associations (CFAs, WRUAs, Ogiek Council of Elders).
86 Forest and Communities

income-generating activities for the neighboring communities, and improving

forest conservation (ISLA-IDH, 2018).
ISLA-Kenya’s objective is to restore and conserve the Southwest sector that
is considered to be alarmingly degraded due to several factors, particularly the
pressure coming from the settlement area, although over the past decade this
forest sector has not suffered from severe pressure and its dense forest cover
is increasing, according to remote sensing analysis (see chapter 3) and the local
community members interviewed.
ISLA and the partners work to reduce the animal presence in the forest in
two ways. One is to regulate the current number of livestock (cattle and sheep)
by implementing a grazing plan. The other is to promote stable livestock farm-
ing as a replacement for grazing within the protected area. According to ISLA’s
figures, the Ndoinet Forest hosts approximately 17,000 (KFS, 2019) or 22,500
animals (Korir, 2016) everyday, numbers considered unsustainable by the pro-
gram. The first report quoted indicates in 6,104 tropical livestock units5 (TLUs,
corresponding to about 8,716 cows) the amount appropriate for the carrying
capacity of the forest (KFS, 2019). The organization consequently proposes a
progressive reduction in the number of cattle and the replacement of indige-
nous or hybrid cattle species with more productive dairy breeds to be raised in
the families’ stables.
However, these figures probably slightly overestimate the number of animals
normally present in the area. With the aim of conducting a partial census of
livestock using the Mau’s pastures and in partnership with the non-profit organ-
ization Drone Adventure, in 2019 we conducted a drone remote sensing survey
of 8,000 ha in the Ndoinet Forest section (total area 19,000 ha)6. The analysis
first focused on the grasslands area located close to the forest boundary, and
the first results (for a sample of 2,000 ha) allow us to highlight some impor-
tant elements. First, just under 1,900 livestock heads have been counted, with
slightly more sheep than cows. This is important, as the conservation initiatives
currently in place focus solely on the grazing of cows. Second, there is no signif-
icant evidence of the permanent overnight stay of cows in the forest, another
issue claimed by the conservation organizations. We recorded three traces of
posts used as night shelters for the herders or of paddocks, the presence of
which is prohibited by law. These kinds of structures are certainly present, as
stated by some interviewees (2019), but probably occur to a much lesser extent
than believed by NGOs and the KFS.

5 Tropical Livestock Unit is a unit of measurement used to compare the nutritional needs of
different species of animals and is used to reference the amount of food fed to a 300 kg–400
kg adult cow.
6 The analyses of the drone images were carried out by Lucrezia Boiani, Andrea Colombo,
Marco Cortesi, Giulia Gussoni, Armelle Mfoupou, Vittoria Olgiati, and Gaia Soldano.
 Settlement schemes and the forest as commons 87

The remote sensing analysis indicates the presence of less than one (0.95)
animal per ha, a figure that differs significantly from the ISLA-IDH estimates
(KFS, 2019) that counted 17,263 animals (cows, sheep, goats, and donkeys)
grazing over 6,460 ha of grasslands and glades7, which is an average of 2.67
livestock heads per ha. Something that usually escapes quantitative analyses
based on carrying capacity is that the use of the forest for grazing is also influ-
enced by the seasons (dry or rainy), the specificity of the cows (age, milked or
kept only for meat), and the condition of the farm (the possibility of a surplus
for feeding the cattle). In addition, one might think about organizing the grass-
land into grazing zones, programming more precisely the use of pasture and
imposing livestock surveillance to better manage the activity.
Another intervention planned by ISLA-IDH concerns the creation of a
physical boundary to the forest reserve, which does not exist today, with the
aim of regulating and limiting the access of people and animals. The construc-
tion of an electric fence along the eastern boundary of the protected area (41
km8, with 14 entrance gates) has been envisaged for this purpose. This artifact
is officially justified by the need to resolve conflicts between humans and wild-
life (e.g., elephants), while at the same time providing a physical separation that
facilitates control by the Forest Service, particularly regarding illegal activities as
timber extraction, charcoal production, poaching (Butynski and de Jong, 2016),
which are present to a very limited degree9.
The two main measures explained above are justified by an alleged link be-
tween forest degradation and human and livestock uses, the causality of which
is weakly demonstrated by the research conducted. Therefore, the narration of
the “grazing threat” seems to be used to justify fortress conservation measures
that have been decided on a political rather than a scientific basis. The reduction
of livestock units in the forest and the erection of the fence show an interest in
creating a wild, intact forest cover that has never existed, at least in the last 100
years (Bally, 1946).
Drawing from the examples of KIFCON and ISLA-IDH, we identify the
existence of a functional relationship between plantation agriculture and a spe-
cific nature (dense forest). This de-humanized forest is built materially through
conservation programs and symbolically through an ecosystem services narra-
tive that focuses on the hydrological and climatic functions of tropical forests.

7 The research considered the Ndoinet Forest section plus the area beyond the Southwest Mau
boundary, for a total of 28,000 ha, but the animals mostly graze in open areas.
8 In the feasibility study, the fence does not exactly follow the boundary of the forest but
delimits the northern part of the West Mau sector and the southern part of the Southwest
Mau from the settlements. For this reason, the length indicated is longer than the length of
the forest boundary (32 km).
9 Personal communication with the local forest manager and forest rangers of the Ndoinet
Forest (2018, 2019).
88 Forest and Communities

Fig. 7.1 - Electric fence in the Maasai Mau Forest sector (Narok County)
(source: Stefania Albertazzi, 2022)

In this socio-ecology, grasslands, shrublands, and intermix vegetation cover

appear to be a minor part of the nature, disregarded as an expression of a hu-
man presence or as evidence of livestock use. This “environmental narrative”
becomes functional in relation to a neo-Malthusian approach (Ross, 2017) that
blames the local communities and marginalizes them from the forest.

7.2 Community conservation

The second strategy is related to the community conservation approach de-
veloped around the 1990s as an alternative to “fortress conservation” and its
heavy impact on local communities, in terms of dispossessions, human rights,
and induced poverty (West et al., 2006). Community conservation is character-
ized by a variety of interventions that focus on the link between conservation
and resource management through the active inclusion of the local community
and with the aim to generate positive local impacts (Adams, Hulme, 2001).
The local communities around Mau Forest enter the forest to access several
resources that are fundamental for their livelihood, particularly pastures and
firewood. Other important forest activities include placing beehives on specific
indigenous trees (see below, chapter 8), collecting medicinal herbs, collecting
mineral salts for livestock, accessing cultural sites (caves, waterfalls), and hunt-
ing wildlife (which is an illegal activity).
 Settlement schemes and the forest as commons 89

Therefore, local communities consider the forest a sort of a resource pool

containing communal products that each individual can access according to
some formal and informal local rules (for this institutional approach to the
study of natural resources, see Gibson et al., 2000; Haller, 2019). The legis-
lation has regulated this human–nature relationship (RoK, 2005; RoK, 2016;
KEFRI, 2016), creating the CFA and approving a number of official docu-
ments that codify the legitimate use of the forest’s resources. In particular, for-
est community members can create a CFA with the aim of participating in the
conservation and management of a state forest with the KFS by developing
a Participatory Forest Management Plan (PFMP), which, together with a manage-
ment agreement, guides future conservation and management measures. It is the
signing of the agreement between the CFA and KFS that makes the PFMP
operational, setting out the direction and the objectives of forest management,
conservation, and use.
The problem is that the PFMP is an instrument the drafting of which is al-
most prohibitively expensive for local communities, being a technical document
prepared by private consultants10. The plan presupposes conducting research
on the biological and socio-cultural value of the forest, an in-depth investiga-
tion of the status of the forest and the way the local community uses it, and
a final draft. Therefore, generally, the redaction of these plans is only possible
with a contribution by external donors.
The approach outlined by the legislation and its implementation in the Mau
Forest defines specific forest uses rather than a real power delegation reflecting
a proper decentralized approach11 (German et al., 2010). This relationship fo-
cused on resources translates into the CFA’s structure and organization. At the
lower level are user groups made up of people who practice the same activity or
have a common interest. However, the forest user groups cover a wide range of
activities, in some cases only marginally related to the forest. For example, in
the village of Ndoinet we find the “Ndoinet farmer self-help group” that brings
together farmers to discuss cultivation techniques, product trade, and tools on a
monthly basis. There is also a “Table baking group” of 10–20 people who give
each other money loans (e.g., to buy livestock or pay school fees). In several
cases, participation in the CFA is motivated by a need to use the forest rather

10 Interview with D. G. from National Alliance of Community Forest Associations (NACOFA)

(2020) and with G. N. from Kenya Forests Working Group (KWFG) (2020). The overall
cost of research, writing, and officially proposing of the PFMP is probably around 15,000 €
(personal communication, 2019).
11 The KFS is the legal manager of the public forests and grants management or conservation
permits. It can terminate a management agreement or prohibit the enjoyment of a user right if
it considers it necessary for conservation or if the CFA violates the terms and conditions of
the agreement.
90 Forest and Communities

than by a marked interest in its management and conservation, highlighting a

sort of functional approach to the forest.
At the same time, this extractivist relationship between forest and communi-
ties is fostered by the KFS itself. The KFS, for example, opposes the construc-
tion of permanent houses in Kiptunga Forest, protesting against the inhabit-
ants when they see the replacement of thatched roofs with sheet metal roofs,
which are perceived as more permanent12. Community members themselves are
involved in several forest activities. For example, they patrol the forest to report
fires, tree cutting, or charcoal production and are involved in the care of the
tree nurseries, in reforestation activities, and in the management of grazing13.
Community members are used as mere human resources to overcome the lack
of personnel and to ensure an adequate payment of fees to the KFS.

7.3 Nature conservation with direct political purposes

A third form of conservation is marked by the manipulation of access to
nature to pursue specific political or economic objectives. This approach is de-
fined by land dispossession, in some cases violent, and by a tight political–
economic entanglement. The narrative spread remarks the need to protect the
natural forest and the water catchment from the pressure of peasants, to be
realized through the eviction of people or the creation of buffer zones between
the local communities and the forest.

Political evictions
In June–July 2020, part of the village of Mariashoni, adjacent to the East
Mau Forest, was subject to an eviction implemented by the KFS, the Kenya
Wildlife Service, and the police. One thousand individuals were removed from
their homes, reclaimed by the three government agencies as part of the natural
forest to be recovered and reconnected with the rest of the Mau complex. The
operation targeted one of the settlement schemes created in the 1990s and was
characterized by violence, which was aggravated by the timing of the operation
that coincided with the Covid-19 pandemic and the rainy season. In the end, the
eviction had a brutal impact on the local population; houses were cleared and
burnt, people lost livestock and furniture, and some schools remained closed
for some time. The exercise lasted a few weeks, during which the agencies evac-
uated an area of 2,000 ha–4,500 ha14, and then it stopped because of a court
case initiated by a ward administrator in Nakuru County. Currently, the local

12 Personal communication in Kiptunga (2022).

13 CFA members can generate a small but significant income from a well-structured tree nurs-
ery and a small allowance from participating to reforestation activities.
14 The first figure given is the result of a personal estimate following data collected in the field
in 2022; the second figure refers to KNA (2020).
 Settlement schemes and the forest as commons 91

inhabitants can still enter their farmland to cultivate, but they are not allowed
to rebuild houses, even though some of them are erecting temporary accom-
modation (built with mud and straw, with a few essential pieces of furniture).
This operation is part of a growing governmental use of violent methods
against local communities. In fact, this eviction was anticipated by a similar but
huge event in the Maasai Mau sector. In November 2019, 15,000 ha of Maasai
Mau Forest (Narok County) were reclaimed, and about 10,000 families were
removed with the aim of recovering a part of forest that was inhabited for a
few decades with an uncertain legal status.
These interventions are part of a larger strategy via which, in the name of
conservation and water tower protection, the government pursues political ob-
jectives. In fact, the evictions are linked to the shifting of alliances in the central
government. The 2018 handshake between President Uhuru Kenyatta and pre-
vious rival Raila Odinga marked a distancing from Deputy President William
Ruto (The Standard, 2020), a member of the Kalenjin ethnic group whose main
electoral basis lies in this region.
In both the cases of the Eastern Mau and the Maasai Mau, the evictions were
accompanied by a new land survey and by the establishment of new beacons on
the ground with the aim of demarcating a new forest boundary. To that end, a
30-km electric fence has already been erected around Maasai Mau Forest.
The evictions not only impact the people removed from the villages but
create hostilities between the various ethnic communities living around Mau
(Ogiek, Kipsigis, and Maasai) that accuse each other of fostering evictions to
regain possession of what they think belongs exclusively to one community.
For this reason, in Mariashoni (East Mau area) and Kebenet (Maasai Mau area)
police and army checkpoints are located along the roads, guarding the territory
to prevent clashes15.

Buffer zones
A different kind of conservation is centered on the role of tea plantations
that, it is assumed, act as a buffer zone between smallholder farms and the
Mau Forest. This approach has a rooted history in Kenya where, with the sup-
port of the World Bank and the creation of a specific agency—the Nyayo Tea
Zone Development Corporation—since the end of the 1980s a 100-m wide tea
plantation was created along the boundaries of the forests to protect them and
foster tea production and local employment. The logic of the operation is to
clearly demarcate the natural protected area boundary that lacks a physical sep-
aration from nearby farmland. The origins of the idea can probably be traced
back to the implementation of the Integrated Conservation-Development

15 Observations and interviews during fieldwork (2022).

92 Forest and Communities

Projects16 approach at the international level, which used buffer zones as areas
of several kilometers where nature conservation and local community devel-
opment coexist in the proximity of natural protected hotspots where more
stringent protection could be implemented (Neumann, 1997).
In the case of Southwest Mau, this kind of intervention is visible around
Kuresoi village (in Kiptororo) where the buffer zone has been initiated but then
suddenly stopped after 10 km because the plantation spread far beyond the 100
m designated extension. The example of Mau has similarities with many other
precedent cases where the country’s élite (civil servants or politicians connected
to the KANU party) used buffer zones as a subterfuge to seize and then convert
protected areas into plantations and private property (Klopp, 2012)17.

Fig. 7.2 - Tea buffer zone along the Ndoinet Forest boundary
(source: Stefania Albertazzi, 2019)

We can trace a similar approach in the narrative of tea plantations in Kiptagich

(Southwest Mau area). In the words of a manager in a leading position at the
Kiptagich Tea Estate Limited, “The plantations play a dividing role between
the forest and the farms and act to protect it”18. Actually, the plantations were

16 Integrated Conservation-Development Projects emerged in the 1990s as one way to over-

come the “fences and fines” approach in conservation by promoting a program that simulta-
neously stimulates nature conservation and the rural development of communities.
17 It is interesting to know that the parastatal Nyayo Tea Zone Development Corporation was
directly managed by the Office of the President and that its highest office was for a long
time held by a Moi loyalist. This was Isaiah Cheluget (Hornsby, 2012), formerly Provincial
Commissioner of Nyanza Province (probably in the period from 1969 to mid-1980).
18 Personal communication (2022).
 Settlement schemes and the forest as commons 93

created on natural protected forest areas of Transmara (another sector of Mau)

in the second half of the 1980s and in Southwest Mau later on for the produc-
tion of the tea brand Asis, which is connected to the family of former President
Moi (Commission of Inquiry into the Irregular/Illegal Allocation of Public
Land, 2005). In this second case, we observe an appropriation of public forest
land for the personal benefit of President Moi, a process that later was used to
justify a narrative of nature conservation to protect against the pressure of the
peasants through a buffer zone. For both the evictions in Mariashoni and for
the Kiptagich tea plantations, the common element is the instrumentalization
of forest conservation for political and economic ends, taking advantage of
an ambiguous legal situation (settlement schemes of Mariashoni) or of an old
privileged (Moi family plantations), while at the same time blaming the peasants
for the degradation of Mau.
8 Indigenous agro-forest systems

The last form of interaction between society and nature that we present
refers to the indigenous agro-forest systems centered on the Ogiek people, the
historical inhabitants of the Mau Forest. Before colonial times, the Ogiek com-
munities used to interact, in terms of both conflicts and cooperation, mainly
with the Maasai and Kalenjin people1. Politically speaking, like other hunter–
gatherer groups in Africa, the Ogiek people were strongly marginalized by the
government during and after the colonial period (Kimaiyo Towett, 2004).
The 1904 and 1911 Maasai agreements, which granted land to the British
colonizers while forcing the Maasai to move off traditional lands, intensified
competition for land in the Rift Valley region. The Ogiek started to be evicted
from the forest (1911, 1926, 1932), and their land was declared Crown Land
(1930s) or was allocated to white settlers or other tribes (in Nakuru, Naivasha,
and Narok). Finally, their identity as a tribe was not recognized by the colonial
administration, and there were repeated attempts to assimilate them into the
largest ethnic groups, such as the Maasai or Kalenjin (Cavanagh, 2017). First
under colonial rule and later under the independent government, they were
marginalized and discriminated against because of their low number and irrel-
evant political power (Sang, 2001). After three decades of peace, a new phase
in the socio-environmental conflict began in 1977. In that year, the national
authorities moved against the Ogiek in Tinet (Southwest Mau Forest), arresting
members of the community, destroying their houses, and accusing them of
being illegal squatters. Since that time, the process of sedentarization described
earlier (see sub-chapter 2.3) has intensified.
The Ogiek (about 52,000 in the country) mainly inhabit the Mt. Elgon area
and the Mau Forest Complex (KNBS, 2019b), particularly the central sectors;
in fact, Kiptunga is the only forest section in the entire Mau complex that has
a number of villages (Kiptunga, Satellite, Tertit, Songi) within it (KFS, 2015).
This is an exception, as permanent human presence in forest reserves is pro-
hibited by legislation (GoK, 2016). In the last two decades of the 20th century,
agriculture and animal husbandry began to play an increasingly prominent role
in Ogiek subsistence, partly as a result of stable population settlement in the
government-identified areas (35,000 ha in East Mau) within the Mau Forest
(1994–2001).

1 The Ogiek ethnic group is usually considered close to the Kipsigis, a sub-group of the
Kalenjin tribe, with whom they share a similar language.
96 Forest and Communities

8.1 An indigenous nature

What we call now the “primary” Mau Forest is actually the result of co-evolu-
tion between nature and the indigenous Ogiek people who helped shape it into
its present form through activities such as shifting cultivation, hunting, herb
and fruit gathering, and beekeeping (see sub-chapter 9.2), probably the core
activity of the Ogiek territorial system, both in symbolic and material terms.
Therefore, the plant and animal biodiversity found in the indigenous forest
is deeply connected with the Ogiek socio-territorial organization. In terms of
vegetation, the species of greatest importance are those favored by bees (par-
ticularly the Dombeya (Silibwet/Dombeya goetzeni), those used as medicines or
as building material for hives (Mororta/Allophylus Abyssinica, Saptet/Podocarpus
latifolius, Aonet/Polyscias Kikuyuensis), and those that are important symbolically,
such as the Podocarpus (Saptet), an extremely long-lived tree that marks the
landscape with its large size. Gathering herbs, leaves, bark, and roots for phyto-
therapeutic use is a typical activity of all peoples living in contact with areas of
forests and thus also for the Ogiek. Trees such as the staddo (Rhamnus staddo),
the magic guarri (Euclea divinorum), the African cherry (Prunus africana), and the
East African olive (Olea capensis), for example, have multiple medicinal uses for
the local community. Wild fruit gathering is also common and is an important
supplement to a diet that is otherwise rather poor, based mostly on maize, po-
tatoes, cabbage, and milk.
Animal biodiversity remains significant in Kiptunga indigenous forest areas.
These areas are still of great importance in the country, despite being affected
by deforestation and intense anthropization in recent decades. Significant spe-
cies mentioned by locals during participatory research include striped tragela-
phus (Tragelaphus scriptus), the blue monkey (Cercopithecus mitis), the yellow-backed
duiker (Cephalophus silvicultor), leopards (Panther pardus), and buffalos (Syncerus
caffer). In addition, African forest elephants (Lexodonta cyclotis) are occasionally
reported.
Hunting, practiced by men with bows, arrows, and dogs, is a fundamental
part of the traditional socio-territorial organization. Hunting is prohibited in
the protected area, so the information gathered on the subject is only partial.
However, participatory mapping indicated as hunting areas the same areas that
had been identified for honey and medicinal herb collection, namely, indig-
enous forest areas (see below, sub-chapter 8.3). Some informants reported a
weekly hunting frequency and others a monthly frequency. The reported fre-
quency may be different from the actual frequency, but it is probably true that
the number of hunters is decreasing because game is no longer considered a
primary food source.
 Indigenous agro-forest systems 97

Table 8.1 - Plants gathered as food and medicines in Kiptunga Forest (Trivellini, 2018)

Usually, hunter–gatherer populations place great symbolic value on the

forest, and the Mau complex is no exception. Indeed, this is where initiation
ceremonies take place and where places of the highest cultural value to the
community are located. The situation in the three protected areas analyzed is
very different. In Koibatek Forest, which is more anthropized and has a lower
percentage of Ogiek population in the neighboring villages, no cultural ecosys-
tem services are mentioned. On the contrary, regarding Kiptunga Forest and
Ndoinet Forest, the indigenous forest is still an important place from a symbol-
ic point of view. Caves, waterfalls, and trees of particular sizes, particularly Ficus
thonningii/natalensis and African cedar (Juniperus procera), are mentioned among
the most relevant aspects.
Therefore, the main characteristic of these parts of the Mau Forest from a
material and symbolic point of view is that it is simultaneously a naturally grow-
ing forest with considerable environmental value, and an inhabited forest, as
opposed to more rigid conservation practices. The strong interpenetration be-
tween nature and society emerges particularly clearly in the case of beekeeping.
98 Forest and Communities

Fig. 8.1 - Cave used for rituals in Kiptunga Forest (source: University of Milan, 2020)

8.2 Beekeeping
Honey plays a key role in the culture of forest-adjacent people. In addition
to its use as a staple food, the Ogiek people also use it as a medicine and in
community rituals. Traditional beehives are set up on indigenous trees (e.g.,
Dombeya goetzenii) at a height of 5 m–10 m, and twice a year honey is harvested
by climbing the trunks. This is a complex process. The beekeeper produces a
small fire using local moss, and then this moss is carried up the tree in a leather
bag so the harvester can drive the bees away with smoke (Fig. 8.2).
 Indigenous agro-forest systems 99

Traditional honey harvesting is deeply related to the quality of the vegeta-

tion. First, traditional hives with a typical cylindrical shape are constructed by
carving the trunk of some indigenous species, particularly Saptet (Podocarpus
latifolius), Silibwet (Dombeya torrida), Aounet (Polyscias Kikuyuensis), and Tenduet
(Prunus africana) (Albertazzi, 2020; Zocchi et al., 2020). Second, the interviews
showed that the areas most used for beekeeping are those characterized by na-
tive vegetation and that are less fragmented by fields and exotic tree plantations.
However, more recently beekeeping activity has expanded to more anthropized
areas where people use modern hives, especially close to waterways where rich-
er and more diverse vegetation remains.
Lately, a number of international cooperation projects have supported
traditional and modern beekeeping activities by enhancing local groups. In
the research area, the most structured cooperative is undoubtedly Macodev
(Mariashoni Community Development), which operates in the village of
Mariashoni (Nakuru County) and was formed in 2013 in the framework of an
international project implemented by a local NGO (Necofa) and three Italian
organizations (Mani Tese, Slow Food, and WWF). Macodev is a kind of um-
brella organization officially consisting of 97 members belonging to eight self-
help groups. The cooperative promotes local development in the village of
Mariashoni, and its main activities are in three areas—the production and sale
of honey, the promotion of ecotourism in the forest, and the management of
an Ogiek radio station created in 2020. Since its establishment, Macodev has
been involved in different projects, and this continuity has allowed the coop-
erative to participate in beekeeping training and business management and has
enabled the organization itself to receive equipment and resources to promote
its activities.
During the period 2013–2020, Macodev appeared to be continuously active,
although very low production was recorded for 2 years (2018 and 2020). The
quantities of honey produced were the highest in the first years of the cooper-
ative’s existence, which coincides with the project support and the conferment
of the title “Slow Food Presidium”2. Over the 2014–2017 period, the amount
of raw honey processed by Macodev was between 1,400 kg/year and 1,600 kg/
year, allowing the production of 1,100 kg–1,250 kg of refined honey, generating
an annual profit between 215,000 KES and 250,000 KES (1,700 €–2,000 €).

2 A Slow Food Presidium is an award given by the Slow Food Organization to traditional
products that are distinguished by the richness of the heritage of knowledge and culture
from which they are produced, by the distinctiveness of the area of production, and by the
mode of production. The Presidium project was born in 2000 as part of the organization’s
strategies to safeguard traditional culinary products placed at risk of extinction, to preserve
their existence and practices of production, and to enhance the territory in which they are
located. Currently, there are 600 Presidia in 70 countries (https://www.fondazioneslowfood.
com/it/cosa-facciamo/i-presidi/).
100 Forest and Communities

In 2019, 1,300 kg of raw honey was estimated to be received, with a profit of

195,000 KES (about 1,500 €); while 2020 was a difficult year in the Mau Forest
area, with the delivery of only 500 kg and a refined honey production of 375 kg,
for a profit of 75,000 KES (just under 600 €). Honey market networks are ex-
tensive and benefit from Macodev’s many partnerships, the favorable location
of the store in the village next to the offices of the KFS and the local Chief,
and contacts generated by national and international organizations. Honey con-
sumers include government officials and staff of partner organizations, tour-
ists who come to Kiptunga Forest, and attendees of national and international
events in which Macodev participates.
In the Southwest Mau sector the Ndoinet Honey Producers’ Cooperative
(NOHPA) has been active since 2018. The cooperative evolved in close rela-
tionship with the CFA, the association that brings together villagers living near
the forest as part of another international cooperation project funded by AICS.
As in the case of Macodev, the operation of the cooperative is quite simple. It
buys honey from members (200 KES/kg) and refines it to sell it in the same
production area at a higher price (600 KES/kg). The 179 members own 900
hives and pay a fee of 100 KES (0.80 €) to join the cooperative. Production
increased over the three project years but remained at a very low level, stopping
in March 2020 at the end of the project. Honey harvesting increased from 50 kg
in 2017 to 150 kg in 2019, while the value of refined honey sold has increased
from 8,000 KES to 24,000 KES per year (60 €–180 €). The causes of the dif-
ficulties are diverse but generally relate to the cooperative’s dependence on the
project that fostered its genesis.
The area is not a particularly relevant market center, but it had attracted some
interest partly because it is close to the construction area of the Itare Dam, a
major infrastructure project that started in 2016. The initiative mobilized about
1,000 workers, some from the area and some migrants from elsewhere, who
were expected to constitute, at least for a time, a potential market for the coop-
erative. However, construction work on the dam stopped in 2018 and has not
yet resumed, leading to a significant drop in the number of people in the area
and less willingness to buy on the part of the resident population left without
work. The end of the project also meant that the staff of the organizations
involved, which represented a significant portion of the buyers, were in the area
less frequently. To these critical issues must also be added climatic difficulties;
prolonged rains between the end of 2019 and the beginning of 2020 hindered
the normal harvesting times (generally in the months of February/March).
Although the members installed beehives with specific rain cover, these diffi-
culties weakened the cooperative’s foundation that had been laid by the project.
Interviews with cooperative managers revealed an intention not to abandon the
project and to revive it with the 2022 harvest, but the structural conditions of
this cooperative remain extremely fragile.
 Indigenous agro-forest systems 101

Finally, in 2020 the Koibatek CFA began to develop a beekeeping cooper-

ative with the support of the project Agrichange funded by AICS and imple-
mented by Necofa and the Italian NGO Mani Tese. Today, the cooperative has
150 hives, some traditional and some modern, and produces 150 kg–200 kg of
refined honey per year. The project ended in 2022, and the major challenge for
the future is to assure economic sustainability.

Fig. 8.2 - Honey harvesting in Kiptunga Forest (source: Stefania Albertazzi, 2022)

8.3 Participatory mapping of ecosystem services

In Kiptunga, we conducted rapid participatory research to assess the dis-
tribution of indigenous activities within the boundary of the protected area
(Albertazzi et al., 2018) and thus to identify the most important spaces for the
local community. Concerning hunting activity, informants indicated that the
102 Forest and Communities

hunting zones cover an area of about 4600 ha. Furthermore, villagers reported
that some species were not locally available anymore, after moving to larger are-
as of the forest (namely the buffalo, Syncerus caffer, moved to Chebuin). The are-
as of the rainforest indicated for beekeeping have a surface area of about 4360
ha. Regarding this activity, some comments underlined a certain mortality rate
of bees due, according to the community, to the relatively recent use of fertiliz-
ers in local agriculture. Mapped plant gathering areas for medical, cooking, or
other non-timber uses have a surface area of 2590 ha. The community reported
a list of 25 forest products (plants). For gathering activities, the informants
reported the generic need for the community to take much longer walks in the
forest to find the useful plants compared to the past.
Of the 5600 ha identified as a source for any of the three ecosystem services,
1860 ha (33%) were mapped as important for the use of all three services, 2030
ha (36%) for the use of at least two of them, and 1710 ha (30%) for only one of
the three services provided by the forest to the community (see also Albertazzi
et al., 2018).

Fig. 8.3 - Hotspots of ecosystem services in Kiptunga (Albertazzi et al., 2018). The
polygons indicate one mapped ecosystem service (light grey); two overlapped ecosys-
tem services (dark grey); and three overlapped ecosystem services (black).
 Indigenous agro-forest systems 103

8.4 Ecotourism
The last activity that directly refers to the indigenous forest is the develop-
ment of ecotourism. Tourism in the Mau Forest has great potential but suffers
from competition from areas of great wildlife value, such as the savannah parks
in the southern part of the country. The elements of interest are the naturalistic
scenery and the cultural heritage of the Ogiek communities. For this reason, the
small ecotourism initiatives that have been developed have sought to enhance
the distinctiveness of certain ecosystems and thus the indigenous forest areas.
In the research area, only the Kiptunga section is actively involved in ecot-
ourism, due to the presence of a strong Ogiek community and to the relative
accessibility of the area. We can highlight two major initiatives—the actions by
Macodev centered in Mariashoni and the projects developed by the CFA in the
area around Satellite, one of the Ogiek villages located within the boundaries
of the protected area.
In the Mariashoni area, the development of responsible forms of tourism
was fostered by the international cooperation project mentioned earlier (see
sub-chapter 8.2, 2014–2016). Thanks to funding from an Italian philanthropic
foundation (Fondazione Cariplo), it was possible to set up a number of trails
and vistas within the forest, and a local travel agency called Terra Madre was
started, which is still the main carrier of tourists in the area.
As part of the same project, an accommodation facility was built in
Mariashoni, which also serves as an Ogiek cultural center. Ten community
members were trained as operators and to engage in various other activities to
promote the local economy and protect the forest. Numbers are limited to a
few hundred visitors per year, but still relevant is the activation of local devel-
opment dynamics that are associated with the Macodev cooperative and that are
focused on forest conservation and the enhancement of Ogiek culture.
The second initiative in this sector is promoted by the Ogiek community
itself and aims at developing ecotourism in the forest by enhancing the Ogiek
cultural center located in the forest, close to the village of Satellite. Currently,
some members of the Ogiek community work as guides for occasional ecotour-
ism excursions in the indigenous forest organized mainly by the Terra Madre
travel agency. However, in these cases tourists rely on accommodations outside
the forest, such as the ecolodge in Mariashoni. The project, started in 2021,
aims at using the space of the cultural center as a base for ecotourism in the for-
est, with the development of ecocamping within the boundary of the protected
area. It is too soon to know if the project can be successfully implemented, es-
pecially in the tourism sector that has been jeopardized by the pandemic crisis.
However, the initiative shows that the Ogiek community is trying to reimagine
its cultural and environmental heritage, connecting it to international networks.
9 Conclusion

The Mau Forest represents a socio-environmental heritage of international

significance and is a fundamental environment for the communities living with-
in and around the protected area. Therefore, its protection and the development
of a sustainable relationship between human communities and the forest must
be a priority for all institutions on both the local and international scales. In the
framework of the United Nations’ Sustainable Development Goals (SDGs),
the protection of the Mau Forest is directly linked with the achievement of
Goal 13 (fight against climate change) and Goal 15 (protection of terrestrial
ecosystems), but it is also strategic for the pursuit of SDGs 1 (fight against
poverty), 2 (fight against hunger), 6 (access to water), and 7 (access to energy).
Finally, it should not be forgotten that the area has been the site of major polit-
ical conflicts, and therefore proper management of this strategic space also ties
in with the achievement of Goal 16 (peace).
Today, the forest is not undergoing major deforestation processes after a
period in which it saw a remarkable reduction in its area, culminating in 2001
with the conversion of 60,000 ha of forest to agricultural land. Along with this
“institutional” deforestation, however, there has also been a more informal de-
forestation involving uncontrolled private use of the protected area.
State policy has played a leading role in the evolution of the protected area,
first with the contested allocation of forest land for agricultural purposes and
then, with the change of government in 2002, with a renewed protection policy
that also led to forceful actions, such as the removal (2001–2008) of irregular
inhabitants from the forest. The result is that today the forest is indeed stable
but also highly vulnerable, threatened above all by its fragmentation that hinders
biodiversity conservation efforts. This fragmentation is the result of several
factors—an institutional structure of the forest divided in 22 sectors, of which
only 16 are contiguous; the settlement schemes that eroded the protected area
and produced new forms of anthropogenic pressure on the forest; and the
presence of significant percentages of tree plantations that, while playing an
important economic role, do not guarantee real continuity in ecosystemic terms.
Today, the main anthropogenic pressure, particularly in the Ndoinet area,
comes from the use of the forest as a source of firewood and as pasture for
animals, particularly cattle. From the energy point of view, the area under study
is still largely dependent on firewood from the forest. If, in fact, people are
beginning to use the electricity grid or photovoltaic micro-plants for lighting,
cooking, and heating, biomass burning remains the almost exclusive source
of energy supply (KFS, 2015; KFS, 2018). During the research conducted in
Ndoinet, all members of the local community surveyed relied on wood as their
106 Forest and Communities

primary energy source, meaning that the forest remains the primary source for
this wood. Our estimates indicate that, with a population of 37,000 households,
traditional cooking (i.e., three-stone stoves) consumes more than 100,000 tons
of wood a year, a significant amount both environmentally and socially, given
the fact that women are usually in charge of collecting wood in the forest.
As for cattle, the most problematic situation is in the Ndoinet Forest, an area
regularly used as pasture by the nearby populations. Here, estimates vary con-
siderably. The ISLA program that aims for a substantial reduction of grazing
in the forest has indicated there are about 17,000–22,000 animals present daily
in the Ndoinet section. We consider these numbers to be excessively high and
estimate the presence of cattle in the forest during the dry season (the one with
the highest presence) to be less than 10,000 heads.
In any case, anthropogenic pressure in the three sections considered does
not seem to be such as to jeopardize the existence of the vegetation, which
appears on the whole to be stable (Koibatek, Kiptunga) or regrowing (Ndoinet)
and properly co-managed by the KFS and local communities (CFAs). The sit-
uation appears more difficult when projected into the medium and long terms,
particularly in a context of intense demographic growth: the local population
increased by about 40% in the decade 2009–2019 – the first period after the
formal settlement of the communities – and now has a density of about 300
dwellings/km2, a high value for rural areas on the continent. Therefore, in the
coming decades the pressure of the population living around the forest and
also the demand for arable land at the national scale will be increasing. In the
absence of a rethinking of agricultural development strategies, this dynamic
represents a major risk factor for the forest.
Along with these political and demographic processes, the evolution of in-
frastructure in the region will also need to be monitored. While the construc-
tion of the Itare dam (now suspended) does not appear to pose a danger in
the southwestern sector of the forest, the impact of the Bosto Dam project1,
currently under study, appears decidedly more critical.
The Mau Forest thus presents itself as a socio-environmental hotspot, an
area of great natural and cultural significance and simultaneously an extremely
vulnerable space in the face of multiple socio-political risks. The protection
of this heritage is therefore a strategic factor for any sustainable development
policy on a national and an international scale.
However, to promote this strategic protection it is first necessary to remove
the forest from the narrative of a homogeneous and pristine space, restoring to
it its own image of complexity. Actually, the Mau Forest hosts many different
forms of forests that are the outcome of a co-evolution between nature and the

1 In 2017, an Environment and Social Impacts Assessment Study Report was published for the
construction of a 252-ha reservoir on the Kipsonoi River within the Southwest Mau protect-
ed area (National Water Conservation & Pipeline Corporation, 2017).
 Conclusion 107

actors living in and around the protected area. Indeed, only by recognizing this
complexity will it be possible to construct specific conservation and develop-
ment policies tailored to the specific characteristics of the forests and the actors
involved. The following proposals therefore seek to identify different strategies
in the different areas researched.

1. Guaranteeing the forest boundaries

Ensuring that no further loss of forest area occurs is essential in order not to
further parcel out an already fragmented protected area. The three forests are
equally exposed to the risk of formal excision (as happened in 1985 and 2001
with a state-promoted degazettement) or informal stable human occupation
of the protected area (as occurred in the years 2001–2008). In the case of the
Ndoinet Forest, for example, the eastern boundary in contact with the agricul-
tural area seems to hold and be respected. However, in the southern part close
to former President Moi’s plantations in Kiptagich, the boundary of the pro-
tected area is not clearly defined, and encroachments are present in an area of
about 2,000 ha. Fires are also more frequent in the southern area, a possible in-
dicator of stronger pressure from agriculture and grazing in the forest reserve.
Beyond the specific case, it is important to highlight the importance of a clear
boundary-safeguarding policy, not in terms of physical patrols or army defense,
but in terms of safeguarding the current forest reserve extension. This would
be essential to address a possible increasing pressure on resources that could
lead to the creation of informal settlements or state-legitimized settlements,
particularly in the previously occupied Ndoinet area.

2. Reconnecting or densifying the forest

To reverse the fragmentation process, it is possible to foster the connec-
tion between the different forest sectors by carrying out targeted reforesta-
tion works, even in limited areas. Within the research area, two sections seem
particularly prone to a fragmentation problem—the forest between Mount
Londiani and the western sector and some western portions of the East Mau
section (Fig. 9.1). In these areas, there would be a need to negotiate with local
actors the replacement of small portions currently in agricultural or forestry use
with areas of indigenous reforestation that would allow reconnection between
crucial sectors of the forest.
In the degraded sections, forest can be densified with reforestation process-
es, starting from internal areas (west to east, in the case of Ndoinet Forest).
The process can start filling the smallest gaps, thus maximizing the dense forest
areas with a minimum effort, then proceed to larger degraded areas, only after
having secured a large area of stable forest.
108 Forest and Communities

With the intention of supporting spontaneous regrowth of vegetation, a mi-

cro-fencing practice has been implemented in Ndoinet, in which a few hectares
of forest are fenced off by CFA members and closed to grazing for a limited
period of time (3–5 years) with the aim of encouraging the rehabilitation of
vegetation cover.

Fig. 9.1 - Example of forest fragmentation and identification of contact points for
ecological corridor drawing in the East Mau Forest (Trivellini and Lindon, 2014)

3. Local communities at the center

Local communities are an unavoidable actor in building a sustainable fu-
ture for the Mau Forest. In the past, colonial and post-colonial governments
implemented authoritarian forms of conservation that severely disadvantaged
communities living around and within the forest. Even some contemporary
conservation initiatives pursue top–down conservation strategies, such as im-
plementing physical barriers between the forest and the communities. In our
view, such arrangements are a mistake because they institutionalize a division
between human groups and the protected area that hinders the development of
shared land management and the integration of environmental protection and
human activities.
 Conclusion 109

With the establishment of CFAs, Kenya has embarked on a path of co-man-

agement of protected areas that has several interesting elements from a formal
point of view. At the practical level, however, such co-management often re-
mains to be built, the redaction of the participatory plans remains too expen-
sive, and CFAs are still very weak. Therefore, there is a need to work on insti-
tutional strengthening so that these communities can take the lead in a renewed
relationship between humans and nature that will lay the foundations for the
future of the Mau Forest. For this to be possible, it is crucial to support local
initiatives to diversify the production model in the key sectors of agriculture,
energy, forestry activities, and sustainable tourism.

4. A renewed agriculture
At present, the communities living around the forests analyzed base their
livelihoods essentially on a production model based on small-scale properties
(2 ha on average) that combine the cultivation of maize, potatoes, and cabbage
with extensive cattle and sheep farming. Usually, this consists of conventional
agriculture supported by chemical inputs that are dumped downstream, even
if some experimental organic farms are developing in the area. With current
population growth rates, this model risks becoming unsustainable, pushing to-
wards new forms of forest appropriation and degrading water quality in the Rift
Valley and Lake Victoria watersheds.
To ensure sustainable development for these communities, it is necessary to
accompany a transformation of production strategies oriented more toward
product quality rather than quantity. Diversification of agriculture, agroforestry,
organic farming, and diversification and improvement of livestock husbandry
are possible strategies to pursue. For this to be possible, however, it is necessary
for the products of these activities to find a connection with the rapidly devel-
oping regional and national markets.
The state-owned areas of the PELIS program currently devoted to con-
ventional and undiversified agriculture could function as a laboratory aiming
toward agriculture with less environmental impact and greater economic and
social value added.

5. New energy
Local communities living around the forest often lack access to the power
grid, mainly because they are too poor to afford a connection. This issue is par-
ticularly present in less accessible areas, such as those adjacent to the Ndoinet
section. While working on improving the living conditions of communities,
initiatives that aim to reduce firewood consumption through technologies with-
in the reach of households should be supported. Improved stoves, micro- and
mini-solar panels, and biogas are some concrete examples.
110 Forest and Communities

6. A new forest economy

For centuries, the Ogiek people have used the forest for economic (honey
and fruit harvesting) and socio-cultural (gathering of medicinal herbs) activi-
ties. A part of the strategy to diversify the economy could be to enhance these
activities. International cooperation projects have led to the Slow Food Presidia
recognition of some local food products from the Mau Forest area, such as
Ogiek honey, dried nettles, Molo sheep, and mushunu chicken. Regarding honey
in particular, it was possible to start a marketing activity managed by the local
community that now functions independently. More recently, however, the local
community has been trying to enhance the medicinal herbs that are harvested in
the forest. These activities alone do not allow families to be self-sustaining, but
when linked with the others mentioned above they can contribute to the devel-
opment of a sustainable relationship between local communities and the forest.
Such initiatives take on particular value in areas where the roots of local com-
munities are strongest, such as Kiptunga Forest and Ndoinet Forest, where the
Ogiek people have the largest and most active presence.

7. A new tourism
Sustainable tourism is not an easy and immediate answer to the needs of the
population. Today, many basic infrastructures (roads, lodging, trails) are lacking,
and the Mau Forest is in fact competing with world-renowned tourist hotspots
located only a few tens of kilometers away (e.g., Lake Nakuru, Maasai Mara
Park). However, with limited and targeted investment, it would be possible to
attract niche tourism to the area that is interested in forest excursions and in
encountering the local communities. Such a tourist presence is compatible with
the protection of the forest and could help enhance its conservation. Not all
areas lend themselves to this type of activity, as only areas of indigenous veg-
etation and strong socio-cultural cohesion (e.g., Kiptunga and Ndoinet) have
sufficient elements of attractiveness to compete in an already highly developed
market such as Kenya.
References
When not explicitly stated, all websites were last accessed in March 2023.

Achard F., DeFries R., Eva H., Hansen M. and Mayaux P. (2007). Pan-tropical
monitoring of deforestation. Environmental Research Letters, 2: 1-11.
Achard F., Stibig H.-J., Eva H.D., Lindquist E.J., Bouvet A., Arino O. and Mayaux
P. (2010). Estimating tropical deforestation from Earth observation data. Carbon
Management, 1, 2: 271-287.
Adams W.M. (2004). Against Extinction. The Story of Conservation. London/Sterling:
Earthscan.
Adams W.M. and Hulme D. (2001). If community conservation is the answer in
Africa, what is the question? Oryx, 35, 3: 193-200.
Albertazzi S. (2020). Usi comunitari e conservazione della natura nell’area protetta
di Ndoinet (foresta Mau, Kenya): elementi di conflitto. Geography Notebooks, 3, 2:
53-72.
Albertazzi S. and Bini V. (2021). La produzione della natura nella postcolonia: la
foresta Mau (Kenya). Rivista Geografica Italiana, Special Issue, 128, 1: 21-36.
Albertazzi S., Bini V., Lindon A. and Trivellini G. (2018). Relations of power driving
tropical deforestation: a case study from the Mau forest (Kenya). Belgeo - Revue
Belge de Géographie, 2: 1-18.
Anderson D. (2002). Eroding the Commons. The Politics of Ecology in Baringo, Kenya
1890s-1963. Suffolk: James Currey.
Bally P.R.O. (1946). Coryndon Museum Expedition to the Mau Forest. Journal of the
Africa Natural History Society, 19, 3-4: 81- 91.
Balme G.A., Hunter L. and Slotow R. (2007). Feeding habitat selection by hunting
leopards Panthera pardus in a woodland savanna: prey catchability versus
abundance. Animal Behaviour, 74, 3: 589-598
Bateson J.D. (n.d.). The registration of the forest dwellers of the South Western Mau Forest
Reserve 1991-1994. Nairobi: KIFCON Project. (Not published).
Bateson J.D. (1993). Korau Settlement Scheme – Appraisal Report. Nairobi: KIFCON
Project. (Not published).
Beekers D. and van Gool B. (2012). From patronage to neopatrimonialism. Postcolonial
governance in Sub-Sahara Africa and beyond. ASC Working Paper 101. Leiden: African
Studies Centre.
Bennun L., Dranzoa C. and Pomeroy D. (1996). The forest birds of Kenya and
Uganda. Journal of East African Natural History, 85, 1:23-48.
BirdLife International (2018). Important Bird areas factsheet: Mau Forest Complex.
http://www.birdlife.org on 17/01/2018.
112 Forest and Communities

Blackett H.L. (1994). Forest Inventory Report N. 1. South Western Mau and Trans-
Mara. Nairobi/Chatama: KIFCON Project, Natural Resources Institute. (Not
published).
Blackburn R. (1970). A Preliminary Report of Research on the Ogiek Tribe of Kenya
Nairobi: Institute for Development Studies.
Blackburn R. (1982). In the land of milk and honey: Okiek adaptations to their
forests and neighbors. In Leacock E. and Lee R. (edited by), Politics and History in
Band Societies. Cambridge: Cambridge University Press: 283-305.
Bogliani G., Agapito Ludovici A., Arduino S., Brambilla M., Casale F., Crovetto
G.M., Falco R., Siccardi P. and Trivellini G. (2007). Aree prioritarie per la biodiversità
nella Pianura Padana lombarda. Milano: Regione Lombardia e Fondazione
Lombardia per l’Ambiente.
Boone C. (2012). Land conflict and distributive politics in Kenya. African Studies
Review, 55, 1: 75-103.
Boyd D.S. and Danson F.M (2005). Satellite remote sensing of forest resources:
three decades of research development. Progress in Physical Geography, 29, 1: 1-26.
Brockington D., Duffy R. and Igoe J. (2008). Nature Unbound. Conservation, Capitalism
and the Future of Protected Area. London/Sterling: Earthscan.
Brotherton P.N.M. (2013). Madoqua kirkii Kirk’s Dik-dik. In Kingdon J. and
Hoffmann M. (edited by), The Mammals of Africa. London: Bloomsbury
Publishing.
Bryant R.L. (2001). Political Ecology: A critical Agenda for Change? In: Castree
N. and Braun B. (edited by), Social Nature. Theory, Practice, and Politics. Oxford:
Blackwell: 151-169.
Bryant R.L. (1996). Romancing colonial forestry: the discourse of ‘forestry as
progress’ in British Burma. The Geographical Journal, 162, 2: 169-178.
Butynski T.M. (1990). Comparative ecology of blue monkeys (Cercopithecus mitis)
in high- and low-density subpopulations. Ecological Monographs, 60: 1-26.
Butynski T.M. and de Jong Y.A. (2016). Game-proof Barrier Feasibility Study, Report prepared
for ISLA/IDH by Rhino Ark Charitable Trust. https://www.idhsustainabletrade.
com/uploaded/2016/11/Butynski-De-Jong-SWMauReport20Oct16-mk-1.pdf
Castree N. (2001). Socializing Nature. Theory, Practice, and Politics. In: Castree
N. and Braun B. (edited by), Social Nature. Theory, Practice, and Politics. Oxford:
Blackwell, 1-21.
Cavanagh J.C. (2017). Anthropos into humanitas: civilizing violence, scientific
forestry, and the ‘Dorobo question’ in eastern Africa. Environment and Planning D,
Society and Space, 35, 4: 694-713.
Cavanagh J.C. (2018). Land, Natural Resources and the State in Kenya’s Second
Republic. In Adeniran A. and Ikuteyijo L. (edited by), Africa Now! Cham: Palgrave
Macmillan: 119-147.
 References 113

Chambers R. (1994). The origins and practice of participatory rural appraisal. World
Development, 22, 7: 953-969.
Chowdhury R.R. (2006). Driving forces of tropical deforestation: the role of remote
sensing and spatial models. Singapore Journal of Tropical Geography, 27: 82-101.
Commission of Inquiry into the illegal/irregular allocation of public land (2005).
Report of the Commission of Inquiry into the Illegal/Irregular Allocation of Public Land.
Nairobi.
Cords M. and Sarmiento E.E. (2013). Cercopithecus ascanius Red-tailed Monkey.
In Butynski T.M., Kingdon J. and Kalina J. (edited by), The Mammals of Africa.
Volume II: Primates. London: Bloomsbury Publishing.
County Government of Nakuru (2018). Nakuru County Integrated Development Plan
2018-2022. Nakuru.
Courtney Mustaphi C.J, Githumbi E.N., Mutua J., Muthoni R., Rucina S.M. and
Marchant R. (2014). Ongoing sedimentological and palaeoecological investigations at
Nyabuiyabui wetland, Kiptunga Forest Block, Eastern Mau Forest, Nakuru District, Keny.
A report to the Mau Forest Conservation Office, Kenya Forest Service, and the
National Museums of Kenya Palaeobotany and Palynology Section.
Daily Nation (1993). Why we cannot go back home. 31/03/1993. Daily Nation
Archive in Nairobi.
Daily Nation (2002). Squatters to be evicted today. 15/01/2002. Daily Nation
Archive di Nairobi.
Daily Nation (2002b). 600 families are evicted. 03/09/2002. Daily Nation Archive
di Nairobi.
Davies G. (1993). West Mau, South West Mau and Transmara Forest Reserves biodiversity
overview. Nairobi: KIFCON Project. (Not published).
Dinerstein E., Powell G., Olson D., Wikramanayake E., Abell R., Loucks C.,
Underwood E., Allnutt T., Wettengel W., Ricketts T., Strand H., O’Connor
S. and Burgess N. (2000). A workbook for conducting biological assessments and
developing biodiversity visions for ecoregion-based conservation. Washington, D.C., USA:
Conservation Science Program, World Wildlife Fund.
District Forest Office of Elburgon (n.d.). Management Plan for the Eastern Mau Forest
Reserve, 1967-1976. (Not published).
Dixo M., Metzger J.M., Morgante, J.S. and Zamudioc, K. R. (2009). Habitat
fragmentation reduces genetic diversity and connectivity among toad populations
in the Brazilian Atlantic Coastal Forest. Biological Conservation, 142: 1560-1569.
Fanstone, B. (2020). Shamba Forestry in Colonial Kenya: Colonial Dominance or
African Opportunity?. In Halterman I. and Tischler J. (edited by), Environmental
Change and African Societies. Leiden/Boston: Brill: 98-120.
FAO (Food and Agriculture Organization) (2020). Global Forest Resources Assessment.
Main report. Rome: FAO.
114 Forest and Communities

Fashing P.J. and Oates J.F. (2013). Colobus guereza (Black-and-white Colobus,
Abyssinian Colobus). In Butynski T.M., Kingdon J. and Kalina J. (edited by), The
Mammals of Africa. Vol. II. Primates. London: Bloomsbury Publishing.
Folega F., Zhang C.-Y., Zhao X.-h., Wala K., Batawila K., Huang H.-H. Dourma
M. and Akpagana K., (2014). Satellite monitoring of land-use and land-cover
changes in northern Togo protected areas. Journal of Forestry Research, 25, 2:
385−392.
Foody G.M. (2003). Remote sensing of tropical forest environments: towards the
monitoring of environmental resources for sustainable development. International
Journal of Remote Sensing, 24, 20: 4035-4046.
German L.A., Pierce Colfer C.J., Barrow E., Küchli C., Blaser J. and Wardojo
W. (2010). Forest Governance and Decentralization in Africa: Linking Local,
Regional and Global Dialogues. In German L. A., Karsenty A. and Tiani A.-
M. (edited by), Governing Africa’s Forests in a Globalized World. London/Sterling:
Earthscan: 1-25.
Gibbs H., Brown S., O Niles J. and Foley J.A. (2007). Monitoring and estimating
tropical forest carbon stocks: making REDD a reality. Environmental Research
Letters, 2: 1-14.
Gibson C.C., McKean M.A., and Ostrom E. (edited by) (2000). People and Forests.
Communities, Institutions, and Governance. Cambridge (MA), London: MIT Press.
GoK (Government of Kenya) and UNEP (United Nations Environmental
Program) (2008). Mau complex and Marmanet forests, environmental and economic
contributions, Briefings notes. Nairobi: UNEP.
Hadi Krasovskii A., Maus V., Yowargana P., Pietsch S. and Rautiainen M. (2018).
Monitoring deforestation in rainforests using satellite data: a pilot study from
Kalimantan, Indonesia. Forests, 9, 389: 1-26.
Haller T. (2019). Towards a new institutional political ecology: how to marry external
effects, institutional change and the role of power and ideology in commons
studies. In Haller T., Breu T., De Moor T., Rohr C. and Znoj H. (edited by), The
Commons in a Glocal World. Global Connections and Local Responses, London/New
York: Routledge: 90-120.
Hoffmann W.A., Schroeder W. and Jackson R.B. (2003). Regional feedbacks among
fire, climate, and tropical deforestation. Journal of Geophysical Research, 108: 1-11.
Hornsby C. (2012). Kenya. A History Since Independence. London/New York: I. B.
Tauris.
Hosonuma M., Herold M., De Sy V., De Fried R.S., Brockhaus M., Verchot L.,
Angelsen A. and Romijn E. (2012). An assessment of deforestation and forest
degradation drivers in developing countries. Environmental Research Letters, 7: 1-12.
Intergovernmental Group on Tea (2022). Current Global Market Situation and
Emerging Issues. FAO, Committee on Commodity Problems. https://www.fao.
org/3/ni282/ni282.pdf on 13/12/2022.
 References 115

ISLA-IDH (2018). Initiative for Sustainable Landscapes South West Mau. Building Our
Flourishing Future. Program Action Plan. ISLA-IDH. www.idhsustainabletrade.
com/uploaded/2018/08/ISLA-Kenya-Action-Plan.pdf.
ISLA-IDH, FINLAYS KENYA LTD (2019). The business case for a landscape approach
to sustainable tea production in Kenya and worldwide. Nairobi.
IUCN. (2016). The IUCN Red List of Threatened Species. https://www.iucnredlist.org.
Jackson J.H. and McCarter P.S. (1994). A profile of the Mau Forest Complex. Nairobi:
KIFCON Project. (Not published).
KEFRI (Kenya Forestry Research Institute) (2016) Summary of the Forest Conservation
and Management Act 2016. https://www.kefri.org/cadep/policydocuments/
SUMMARYOFTHEFOREST-CONSERVATIONANDMANAGEMENTA
CT,2016.pdf on 13/12/2022.
KFS (Kenya Forest Service) (2013). Koibatek Participatory Forest Management Plan,
2013-2017. Koibatek.
KFS (Kenya Forest Service) (2014). Kiptunga Forest Plantation Management Plan.
Nairobi.
KFS (Kenya Forest Service) (2015). Kiptunga Participatory Forest Management Plan
2015-2019. Kiptunga.
KFS (Kenya Forest Service) (2017). Strategic Plan 2018-2022. Draft. Nairobi.
KFS (Kenya Forest Service)(2018). Ndoinet Participatory Forest Management Plan. Final
Draft. Ndoinet.
KFS (Kenya Forest Service) (2019). Ndoinet Forest Livestock Management Plan. (Not
published).
KHRC (Kenya Human Rights Commission) (2008). A Comparative Study of the Tea
Sector in Kenya. A Case Study of Large Scale Tea Estates. Nairobi.
KIFCON (Kenya Indigenous Forest Conservation Programme) (1991). Progress
report 1. Nairobi. (Not published).
KIFCON (Kenya Indigenous Forest Conservation Programme) (1992). Quarterly
Report N. 2. Nairobi. (Not published).
KIFCON (Kenya Indigenous Forest Conservation Programme) (1993). Briefing
notes for the Hon. John Sambu, Minister of Environments and Natural Resources. Nairobi.
(Not published).
KIFCON (Kenya Indigenous Forest Conservation Programme) (n.d.), Phase 1
Report. Nairobi. (Not published).
Kimaiyo Towett J. (2004). Ogiek Land Cases and Historical Injustices 1902-2004.
Egerton: Ogiek Welfare Council.
Kingdon J. (1997). The Kingdon Field Guide to African Mammals. Princeton: Princeton
University Press.
Klopp J. M. (2012). Deforestation and democratization: patronage, politics and
forests in Kenya. Journal of Eastern African Studies, 6, 2: 351-370.
116 Forest and Communities

KNBS (Kenya National Bureau of Statistics) (2019). Kenya Population and Housing
Census Volume II. Distribution of populations by administrative units. Nairobi.
KNBS (Kenya National Bureau of Statistics) (2019b). Kenya Population and Housing
Census Volume IV, Distribution of populations by socio-economic characteristics. Nairobi.
Konoin Kabara Community Forest Association (2013). Itare Participatory Forest
Management Plan 2013-2016. (Not published).
Korir R. (2016). Beef value chain assessment for South-West Mau. Kenya. https://www.
idhsustainabletrade.com/uploaded/2017/07/Livestock-Assessment-Report-
South-West-Mau-Forest-Kenya-June-2017.pdf
Kratz C.A. (1980). Are the Okiek really Masai ? Or Kipsigis ? Or Kikuyu? Cahiers
d’Études Africaines, 20, 79: 355-368.
KWTA (Kenya Water Tower Agency) (2019). Kenya Water Status Report. East Mau
Revised. Nairobi.
Lambin E.F. (1999). Monitoring forest degradation in tropical regions by remote
sensing: some methodological issues. Global Ecology and Biogeography, 8, 191-198.
Lambin E.F. and Ehrlich D. (1997). Land-cover changes in Sub-Saharan Africa
(1982-1991): application of a change index based on remotely sensed surface
temperature and vegetation indices at a continental scale. Remote Sensing of
Environment, 61: 181-200.
Lino A., Fonseca C., Rojas D., Fischer E. and Ramos Pereira M.J. (2019). A meta-
analysis of the effects of habitat loss and fragmentation on genetic diversity in
mammals. Mammalian Biology, 94: 69-76.
Loftus A. (2012). Everyday Environmentalism. Creating an Urban Political Ecology.
Minneapolis: University of Minnesota Press.
Maathai W. (2011). Silent forests and famine in east Africa. The Guardian.
25/11/2011. https://www.theguardian.com/commentisfree/2011/nov/25/
silent-forests-famine-east-africa
MA (MILLENIUM ECOSYSTEM ASSESSMENT) (2005). Ecosystems and Human
Well-being: Synthesis. Washington, DC: Island Press.
Mau Forest Task Force (2009). Report of the Prime Minister’s Task Force on The
Conservation of the Mau Forest Complex. Nairobi.
Mawdsley E., Savage L. and Mi Kim S. (2013). A “post-aid world”? Paradigm shift
in foreign aid and development cooperation at the 2011 Busan High Level
Forum. The Geographical Journal, 180, 1: 27-38.
Ministry of Environment and Forestry (2018). Taskforce Report on Forest Resources
Management and Logging Activities in Kenya. Nairobi.
Mkawale S. and Gachui K. (2020). Why region near Mau is beset by perennial
land rows, flare-ups. The Standard. https://www.standardmedia.co.ke/
rift-valley/article/2001381723/why-region-near-mau-is-beset-by-perennial-land-
rows-flare-ups
 References 117

Harvey D. (2010). A Companion to Marx’s Capital. London: Verso.

Moore J. W. (2003). Capitalism as World-Ecology. Organization & Environment, 16,
4: 431-458.
Moorehead R. (1993). Annexes to the report on a consultation with the Mau forest dwellers.
Nairobi/London: KIFCON Project, International Institute for Environment
and Development. (Not published).
Morgan W.T.W. (1963). The ‘White Highlands’ of Kenya. The Geographical Journal,
129, 2: 140-155.
Morjaria E. (2012). Electoral competition and deforestation. Evidence from Kenya. Working
Paper Prepared for the World Bank.
Mwangi E. (1998). Colonialism, self-governance and forestry in Kenya: policy,
practice and outcomes. Research in Public Affairs. https://dlc.dlib.indiana.edu/
dlc/bitstream/handle/10535/5706/Colonialism%20self%20governance%20
and%20forestry%20in%20Kenya.pdf ?sequence=1
Myers N. (1988). Tropical deforestation and remote sensing. Forest Ecology and
Management, 23: 215-225.
National Water & Pipeline Corporation (2017). Proposed Bosto Dam Project, Bomet
County. Environmental and Social Impact Assessment Study Report. https://eawildlife.
org/resources/reports/ESIA%201381_%20STUDY%20REPORT%20
FOR%20PROPOSED%20BOSTO%20DAM%20IN%20BOMET%20
COUNTY.pdf
Ndii D. (2018). Crony Capitalism and State Capture: The Kenyatta Family Story.
The Elephant 07/07/2018. www.theelephant.info/op-eds/2018/07/07/crony-
capitalism-and-state-capture-the-kenyatta-family-story/
Neumann R.P. (1997). Primitive Ideas: protected area buffer zones and the politics
of land in Africa. Development and Change, 28: 559-582.
Ngene S., Makonjio Okello M., Mukeka J., Muya S., Njumbi S. and Isiche J. (2017).
Home range sizes and space use of African elephants (Loxodonta africana) in
the Southern Kenya and Northern Tanzania borderland landscape. International
Journal of Biodiversity and Conservation, 9, 1, 9-26.
Njoroge J.K. (2020). Final Evaluation report Imarisha!, Nairobi. (Not published).
Ofcanski T.P. (1984). Kenya forestry under British colonial administration, 1895-
1963. Journal of Forest History, 28 (3): 136-143.
Pellikka P. and Alshaikh A.Y. (2016). Remote sensing of the decrease of juniper
woodlands in the mountains of Southwestern Saudi Arabia - reasons and
consequences. Arabian Journal of Geosciences, 9: 1-12.
Peluso N.L. and Vandergeest P. (2001). Genealogies of the political forest and
customary rights in Indonesia, Malaysia and Thailand. The Journal of Asian Studies,
60, 3: 761-812.
118 Forest and Communities

Plumptre A.J. (2000). Monitoring mammal populations with line transect techniques
in African forests. Journal of Applied Ecology, 37: 356-368.
Pompilio L., Bionda R., Mosini A., Bogliani G., Casale F., Celada C., Rossini E. and
Soldarini M. (2018). Un approccio multitaxa ed expert based per l’individuazione
delle aree prioritarie per la conservazione della biodiversità nel Verbano Cusio
Ossola. Natural History Sciences, 5, 2: 41-56
Reis S. (2008). Analyzing land use/land cover changes using remote sensing and
GIS in Rize, North-East Turkey. Sensors, 8: 6188-6202.
Rift Valley Water Service Board (2015). Itare Dam Water Supply Project. Environmental
and Social Impact Assessment (ESIA) Study Report. Nairobi https://www.nema.
go.ke/images/Docs/EIA-1250-1259/EIA1257ItareESIAStudyReport2016.pdf
Rodgers W.A., Owen C.F. and Homewood K.M. (1982). Biogeography of East
African forest mammals. Journal of Biogeography, 9, 4: 1-54.
RoK (Republic of Kenya) (2005) Forest Act. Chapter 385. Nairobi.
RoK (Republic of Kenya) (2016) The Forest Conservation and Management Act.
Nairobi.
RoK (Republic of Kenya) (2009) The Agriculture Act (Cap 318). The Agriculture
(Far Forestry) Rules, Nairobi.
Ross E.B. (2017). Sub-Saharan Africa, Kenya and the Malthusian Paradigm in
Contemporary Development Thinking. In Pimbert M.P. (edited by), Food
Sovereignty, Agroecology and Biocultural Diversity. London: Routledge: 169-201
Sánchez-Azofeifa G. A., Harriss R.C. and Skole D.L. (2001). Deforestation in Costa
Rica: A Quantitative Analysis Using Remote Sensing Imagery. Biotropica, 33, 3:
378-364.
Sang J.K. (2001). Kenya. The Ogiek in Mau Forest. In Nelson J., Hosscak L. (edited
by), From Principle to Practice: Indigenous Peoples and Protected Areas in Africa. Moreton-
in-Marsh (UK): Forest Peoples Programme: 119-194.
Scott J. (1998). Seeing Like a State: How Certain Schemes to Improve the Human Condition
Have Failed. New Haven, CT: Yale University Press.
Seydack A.H.W (2013). Bushpig Potamochoerus larvatus. In Kingdon J. and Hoffmann
M. (edited by), Mammals of Africa. London: Bloomsbury Publishing.
Shaffer N.M. (1967). Land resettlement in Kenya. Yearbook of the Association of Pacific
Coast Geographers, 29: 121-139.
Singer B. (Ed.) (2015). L’homme et les forêts tropicales, une relation durable? Versailles:
Quae.
Smith N. (1984). Uneven Development: Nature, Capital, and the Production of Space.
Oxford: Blackwell.
The Standard (2020). A handshake that shook the whole country. The Standard.
https://www.standardmedia.co.ke/article/2001363600/a-handshake-that-
shook-the-whole-country
 References 119

Torahi A. and Rai S.C. (2011). Land cover classification and forest change analysis,
using satellite imagery - a case study in Dehdez Area of Zagros Mountain in
Iran. Journal of Geographic Information Systems, 3: 1-11.
Trivellini G. (2018). Evaluation of natural resource of conservation and tourism interest in the
south Northern MAU (Koibatek) Forest. Milano: Cooperativa Sociale Eliante.
Trivellini G. and Lindon A. (2014). Evaluation of natural resource of conservation and
tourism interest in the Northern MAU (Kiptunga) Forest. Milano: Cooperativa Sociale
Eliante.
Tsing A. (2018). Résurgence holocénique contre plantation anthropocénique.
Multitudes, 3, 72: 77-85.
Tsing A. (2015). The Mushroom at the End of the World. Princeton: Princeton University
Press.
Tucker C.J., Townshend J.R.G. (2000). Strategies for monitoring tropical
deforestation using satellite data. International Journal of Remote Sensing, 21 (6-7):
1461-1471.
UNEP (United Nations Environmental Programme) (2009). Kenya: Atlas of Our
Changing Environment. Nairobi: UNEP.
University of Cambridge (2012). Natural Capital Business Case Study: The Kericho Tea
Plantation. Cambridge: Programme for Sustainability leadership.
Wainaina B. (2011). One Day I Will Write About this Place. London: Granta.
Wan HY, Cushman SA and Ganey JL. (2018). Habitat fragmentation reduces genetic
diversity and connectivity of the Mexican spotted owl: a simulation study using
empirical resistance models. Genes, 10; 9, 8: 403.
Wayumba G. (2019). The White Highlands and the establishment of the African
settlement schemes in Kenya. International Journal of Innovative Studies in Sciences and
Engineering Technology, 5, 6: 45-56.
West P., Igoe J. and Brockington D. (2006). Parks and peoples: the social impact of
protected areas. Annual Review of Anthropology, 35:251-77.
World Resources Institute, Department of Resource Surveys and Remote Sensing,
Ministry of Environment and Natural Resources (Kenya), Central Bureau
of Statistics, Ministry of Planning and National Development (Kenya) and
International Livestock Research Institute (2007). Nature’s Benefits in Kenya, An
Atlas of Ecosystems and Human Well-Being. Washington, DC/ Nairobi: World
Resources Institute.
Zhang Q., Devers D., Desch A., Justice C. O. and Townshend J. (2005). Mapping
tropical deforestation in central Africa. Environmental Monitoring and Assessment,
101: 69-83.
Zinner, D. (2013). Olive baboon Papio anubis. In Mittermeier R.A., Rylands A.B. and
Wilson D.E. (edited by), Handbook of the Mammals of the World Volume 3: Primates.
Barcelona: Lynx.
120 Forest and Communities

Zocchi D.M., Volpato G., Mutiso D.C., Chalo P. and Fontefrancesco M.F. (2020).
Expanding the reach. Traditional ecological knowledge and technological
intensification in beekeeping among the Ogiek of Mau Forest, Kenya. Journal of
Ethnobiology and Ethnomedicine, 16, 57: 1-52.
Archival documents

EC 1/6/14 Forest General, Volume 9, historical period 1986–1987, Rift Valley

Province Archive, Nakuru.
FP 8/7 C3/NE/SU/GE Ndorobo settlement 1997–2004, historical period 1996–
2003, Rift Valley Province Archive, Nakuru.
FOR 13/5/1 Volume 1, Registration and Settlement of Dorobo, historical period
1991–1996, Rift Valley Province Archive, Nakuru.
FOR 13/5/1 Volume 2, Registration and Settlement of Dorobo, historical period
1993–1997, Rift Valley Province Archive, Nakuru.
FOR 13/5/1 Volume 3, Registration and Settlement of Dorobo, historical period,
1996–1999, Rift Valley Province Archive, Nakuru.
FP 8/27 LA/RVP/122 Land adjudication Ndoinet Chepkoburot Settlement
Scheme, historical period 1998–2001, Rift Valley Province Archive, Nakuru.
Forest and Communities
Deforestation, Conservation, and Socio-ecological
Relations in the Mau Forest, Kenya

Stefania Albertazzi, Valerio Bini, Guido Trivellini

The book analyzes the case of Mau Forest (Kenya), exploring the deforestation process that
has occurred and the controversial and changing relationships between a protected forest and
the communities living within and around its borders.
The volume contributes to the international debate on political ecology from a predominantly
geographical perspective, enriched by contributions more closely related to the natural scienc-
es. The study is based on a multi-year research (2017-22) that combines qualitative and quan-
titative methodologies: research in archives and government offices, field studies in the forest
area, semi-structured interviews, participatory mapping with local community members, and
satellite and drone remote sensing.

Cover Image: Electric fence in a contested area of the Mau Forest’s southern sector, by Stefania Albertazzi, 2022

ISBN 979-12-80325-78-5 (print)

ISBN 979-12-80325-80-8 (PDF)
ISBN 979-12-80325 71-6 (EPUB)
DOI 10.54103/milanoup.89