Floodplain Rehabilitation and the Future of
Conservation & Development
Adaptive management of success in Waza-Logone, Cameroon
Paul Scholte
Wageningen University and Research Centre
Wageningen Université et Centre de Recherche
Paul Scholte
Résumé
Depuis 1979, l'hydrologie de la plaine inondée de Waza-Logone, située dans la zone semi-aride du Cameroun, a été
affectée par la construction en amont d'un barrage et d'un remblai, provoquant la réduction de l'intensité des
inondations sur une aire de 1500 km2, y compris le Parc National de Waza. Comme conséquence, les graminées annuelles
ont envahi les prairies permanentes et productives, réduisant la capacité de charge du site pour le bétail, la pêche et
la faune. En 1994, et suite à des consultations locales, un cours d'eau bloqué par le remblai le long du fleuve de Logone
a été rouvert, remettant en état une surface de 600 km2 des plaines inondées qui se sont desséchées. La surveillance a
montré une réinstallation progressive des prairies permanentes, pratiquement accompli en 2003, quoique avec une
composition floristique quelque peu différente des années 70. Les effectifs de la plupart des oiseaux d’eau ont doublé,
particulièrement lorsque leurs colonies ont bénéficié de la protection des communautés locales, tandis que des colonies
des oiseaux piscivores ont été fréquemment détruites. Les antilopes des plaines inondées ont montré une première
augmentation qui n'a pas persisté, en raison de la concurrence croissante avec le bétail dont les densités ont
augmenté du simple au triple comme conséquence de l'arrivée de nouveaux groupes pastoraux d'une part et du séjour
prolongé des troupeaux fréquentant traditionnellement le secteur. Ces développements indiquent un problème lié au
partage des avantages des inondations récurrentes entre la faune et les populations qui répondent rapidement aux
nouvelles opportunités, ce qui interpelle à une planification et gestion intensifiées pour prévoir ce succès inattendu.
Un dialogue a été institutionnalisé entre les autorités des aires protégées du secteur et les communautés locales.
La conservation efficace nécessite une participation plus importante et durable du personnel des aires protégées
du secteur dans le processus de planification et l’approche participative. Les programmes de formation continue ont
montré un grand potentiel pour améliorer les connaissances et qualifications du personnel actuel. En outre, des
changements institutionnels sont nécessaires pour assurer l'augmentation nécessaire des effectifs du personnel du parc
et leur rotation, non seulement au sein et entre les parcs nationaux mais également avec des instituts de formation. Le
succès de la réhabilitation des plaines inondées dépend finalement de la surveillance écologique et socio-économique
à long terme, avec des individus et des établissements capables de traduire les perspicacités prochaines en pratique.
Floodplain Rehabilitation and the
Future of Conservation & Development
Abstract
Since 1979, the hydrology of the Waza-Logone floodplain in semi-arid Cameroon has been affected by the construction
upstream of a dam and embankment, reducing the flooding intensity in an area of 1500 km2, including Waza National
Park. As a consequence, annual grasses invaded productive perennial grasslands, reducing the carrying capacity of the
area for livestock, fisheries and wildlife. In 1994, after local consultations, a watercourse blocked by the embankment
along the Logone river was re-opened, reinstating the natural flooding regime in an area of 600 km2. Monitoring showed
a gradual recovery of perennial grasslands, virtually completed in 2003, albeit with a species composition somewhat
different from the 1970s. Numbers of most waterbirds increased two-fold, especially when their colonies benefited from
local communities’ protection, whereas colonies of piscivorous birds were frequently destroyed. Floodplain antelopes
showed an initial increase that has not continued, probably because of increasing competition with livestock which
densities increased three-fold due to the arrival of new pastoral groups and prolonged stay of herds traditionally
frequenting the area. These developments indicate a problem in the sharing of the reflooding benefits between wildlife
and people that rapidly respond to new opportunities, calling for intensified management planning to anticipate this
unexpected success. A dialogue was institutionalised between protected area authorities and local communities.
Effective conservation called for a stronger and lasting involvement of protected area personnel in planning and
community conservation. Developed training courses showed good potential to upgrade necessary knowledge and
skills of present personnel. In addition, institutional changes are needed to assure the necessary increase in numbers of
park personnel and their rotation, not only in and amongst national parks but with training institutes as well. The success
of floodplain rehabilitation ultimately depends on long-term ecological and socio-economic monitoring, with individuals
and institutions capable to translate upcoming insights into practice.
67
Tropical Resource Management Papers
Documents sur la Gestion des Ressources Tropicales
67
FLOODPLAIN REHABILITATION AND THE FUTURE OF CONSERVATION
2005
&
DEVELOPM ENT
Tropical Resource M anagement Papers, No. 67 (2005);
ISBN 90-6754-953-3
Also published as t hesis (2005), Leiden Universit y
ISBN-10: 90-9019780-X
ISBN-13: 978-90-9019780-7
Floodplain Rehabilitation and the Future
of Conservation & Development
Adaptive management of succes in
Waza-Logone, Cameroon
Paul Scholt e
Preface
I first learnt of the Waza-Logone floodplain rehabilitation plans in 1992 and was
immediately thrilled by its perspectives. I expected that reflooding could trigger a
cascade of developments and provide a new élan in a rather paralysed environment.
At that time I was working in nearby Chad in a Sahelian environment where improved management without substantial inputs was out of scope under the reigning economic conditions. In Waza-Logone, an area still endowed with spectacular
wildlife, we could start this unique experiment, improving the livelihoods of fishermen and pastoralists, rehabilitating wildlife habitat and, ‘on top’, triggering an
improved management system. This study follows the quest of preparing, implementing, observing, discussing and analysing floodplain rehabilitation and, as
proved necessary, developing the human capacity to assure that both wildlife and
humans would benefit from the rehabilitation interventions.
This study reflects the development of my activities in the Lake Chad Basin in the
1990s. My assignment from 1990 till 1993 at the Programme Ecologie Pastorale
at the Laboratoire de Farcha (Chad) was an excellent preparation for the research
I carried out in neighbouring Cameroon at the Waza-Logone Projet (Maroua, 19931997) and at the Ecole pour la Formation des Spécialistes de la Faune (Garoua,
1998-2000). Temporarily back in the Netherlands, various organisations requested my presence in the region, through short-term assignments, in 2000-2003,
allowing me to keep track of some of the latest developments.
The following chapters also reflect my personal development. I started in the Lake
Chad basin as an ecologist with an interest for social and development activities,
developing into an ‘environmental scientist’ in the CML-Leiden tradition. Yet, I did
not neglect my field biological roots and was stimulated by the request to contribute
to the ‘Important Bird Areas of Africa’ and ‘Mammals of Africa’. I discovered another challenge when teaching an MSc course on rangeland management at the University of Ndjamena. I was captured by the exchange of experiences in professional
education. In 1998 I was asked to initiate Garoua’s community conservation curriculum. Supervising students during their field research has been amongst my
most rewarding experiences.
Presenting our Waza-Logone reflooding studies at congresses and in scientific journals, I realised how special our experiences actually were. The size and the impact of
reflooding were at a different scale than floodplain rehabilitation studies in Europe
or the USA. Reflooding was an excellent tool to test concepts of ecological change
that have seldom been assessed at such a scale in practice. The landscape scale also
widened the scope of this study from the ecological impact of floodplain rehabilitation to encompass also management issues such as planning and capacity building.
The process of w riting the papers and this thesis
In the mid-1990s, I started reporting preliminary results of our work, discussing the
set-up of our monitoring and presenting a series of base-line studies and inventories,
which laid a foundation for the later reported studies. This thesis has slowly grown
through the elaboration of individual papers that, although based on the WazaLogone and Garoua work, became small projects with their own objectives and conclusions. This allowed me to take some distance from the project and concentrate on
linkages with upcoming scientific discussions. In this process, a number of papers
were ultimately discarded from this thesis, especially the older ones that were largely descriptive and the ones that did not sufficiently contribute to the connecting (scientific) thread of this study, see annexed list of background publications.
The first half of this study ‘Impact of reflooding’ is predominantly based on ecological science. Chapters vary in scope, set-up and pretensions. Presented work on
vegetation (Chapters 3,4) and pastoralist responses (Chapter 7) was developed and
carried out as part of systematic research. Chapters on waterbirds- and antelope
dynamics (5,6) on the other hand, compile a large number of surveys, carried out
at various times for various purposes. The variety of survey methods and personnel as well as their large temporal and spatial scale, prevent a strong analysis on
causal relations. These latter chapters aim, based on best available science, to draw
broad conclusions on developments in bird and antelope populations, the area’s
‘conservation assets’. Moreover, they are of value for the total picture to be drawn
in Chapters 8 and 12.
The second half of this study, ‘Enhancing Conservation – Development Integration’, is only occasionally linked with theory. This second half is needed however
to lay the foundation for my aim to address the issue of adaptive management that
is required to understand the role of floodplain rehabilitation and how it may
enhance conservation-development integration. This study obliged the use of different research styles: ecological science, qualitative participatory methods and occasional steps into social and education science. I do not pretend to master these
latter, and limited myself to a straightforward methodology based on case-based
reasoning, greatly assisted by exchanges with subject specialists and reviewers.
Hydrology and fisheries’ ecology also called for attention in this floodplain environment. At the Waza-Logone project, hydrologists and fishery researchers with whom
I closely co-operated, have reported their findings in a series of field reports. A scientific synthesis would require more time and a different expertise than I possess.
The title I initially had chosen for this study was ‘Rehabilitating the floods, for people or wildlife?’ as this issue had remained at the forefront of my field research.
This question was already answered in the late 1990s when finalising a first version of Chapter 8 (‘time-bomb’). At that time, my work was already aimed at the
development of ways to redirect the conservation-development balance. The present title reflects this ongoing quest.
Remarks on terminology
The studies presented are the result of the work of a large number of people with whom I
worked in the field, as authors or as reviewers. Throughout this book I use ‘we’ to stress
common work and ideas of my colleagues and I, and ‘I’ when I refer to my own specific
opinion.
Throughout this book, indicated years of ecological monitoring refer to the rainy and flooding season and not necessarily the calendar year when the consequences of the reflooding
were monitored. Vegetation composition monitored in May 1994 for example, was referred
to as 1993 vegetation because 1993 was the year of the relevant growing (= rainy + flooding) season. To allow comparison of flooding differences, also bird and antelope counts
have been referred to the relevant rainy and flooding season. The January 1995 waterbird
counts and late April 1995 wildlife counts were thus indicated as 1994 counts. In descriptions not specifically related to ecological monitoring (Chapters 2, 8-11), years are indicated as calendar years.
Nomenclature of woody plants follows Geerling (1982), grasses van der Zon (1992), and
other plant species the second edition of the Flora of West Tropical Africa (Hepper, F.N.
1954-1972). Names of birds follow the checklist of the area (Scholte et al. 1999), names of
mammals follow Kingdon (1997).
Despite the multitude of more or less similar expressions for Community Conservation, I
have used the term (Integrated) Conservation-Development (Project) as most relevant term
for the activities conducted by the Waza-Logone Project.
Table of Contents
PART I – Introduction
1
General Introduction and Study Outline
13
2
The Ecological History of Waza-Logone: Constructing a reference
image for floodplain rehabilitation
65
PART II – The Impact of Reflooding in Waza-Logone
3
4
5
6
7
Impact on Vegetation
Floodplain Rehabilitation in North Cameroon: Impact on vegetation
dynamics
89
Maximum Flood Depth Determines Above-ground Biomass in African
Seasonally Shallowly Flooded Grasslands
107
Impact on Wildlife
Waterbird Recovery in Waza-Logone (Cameroon), resulting from
increased rainfall, floodplain rehabilitation and colony protection
129
Antelope Populations in Waza National Park (Cameroon) from 1960
till 2001: Impact of changes in rainfall, hydrology and human pressure
145
Impact on Pastoralists
Pastoralist Responses to Floodplain Rehabilitation in North Cameroon
161
PART III – Enhancing Conservation – Development Integration
by Management Planning and Training
8
Risks: the overshoot of success
Immigration: A potential time bomb under the integration of
conservation and development
185
9
10
11
Grip on the Whole: management planning
At the Interface of Legislation and Wildlife Management: A decade of
experience in consensual protected area management planning
in Cameroon
203
Foundation: development of human capacities
Curriculum Development at the African Regional Wildlife Colleges,
with special reference to the Ecole de Faune (Cameroon)
233
Wildlife Managers’ Perceptions of Community Conservation Training
in West and Central Africa
251
PART IV – Synthesis
12
Floodplain Rehabilitation and the Future of Conservation & Development:
Synthesis and concluding remarks
269
References
299
Summary
321
Samenvatting
329
About the Author
337
List of Background Publications
338
Acknowledgments
340
The Photographic Outline is found between Chapter 1 and Chapter 2
33
PART I
Int roduct ion
1
General Int roduct ion and St udy Out line
2
The Ecological Hist ory of Waza-Logone: Const ruct ing a ref erence
image f or f loodplain rehabilit at ion
1
General Introduction and Study Outline
15
1.1
Introduction to African floodplains and their rehabilitation
‘From a European’s point of view Nuerland has no f avourable qualit ies, unless it s
severit y be count ed as such, f or it s endless marshes and w ide savannah plains
have an aust ere, monot onous charm. It is t hroughout hard on man and beast ,
being f or most of t he year eit her parched or a sw amp. But Nuer t hink t hat t hey
live in t he f inest count ry on eart h and, it must be admit t ed, f or herdsmen t heir
count ry has many admirable f eat ures’
Evans-Prit chard (1940) int roducing t he ecology of t he Sudd (Sout h Sudan), t he
largest st ret ch of seasonally f looded grassland in Af rica.
African floodplains
The fertility of floodplains is legendary and some of humankind’s main civilisations, such as in Egypt and Mesopotamia have developed around them. Floodplains
are fed by the regular spilling, once or several times a year, of river water, loaded
with sediments over the levees into the surrounding plains. The great fluctuations
in water level create a seasonal cycle of flood and drought, allowing a high primary
production, abundant wildlife and often a high human population density.
Nearly all African rivers have fringing floodplains, but the vast areas of seasonally flooded grasslands, which are so important to the African economies and wildlife,
are associated with rivers that have seasonal rainfall catchment areas (Denny
1993). The largest are the Inner Niger Delta in Mali, Lake Chad and floodplains
associated with its tributary rivers (Cameroon, Chad, Nigeria), the Sudd in Sudan,
the Zambezi valley floodplains with the Kafue Flats and Barotse plains in Zambia
and the Okavango Delta (Botswana) (Fig. 1.1). These floodplain rivers have diverse
and abundant fish faunas, which support some of the richest inland fisheries
(Wellcomme 1979). After the floods recede, herbivores, wild as well as domestic,
find abundant and high quality grazing in the floodplain grasslands and reach
densities seldom encountered in other habitats (de Bie 1991). Other expressions
of these floodplains’ richness are the large number of waterbirds, resident as well
as migratory. All major sub-Saharan African seasonally flooded grasslands (Fig.
1.1) feature prominently in the list of Important Bird Areas (Fishpool & Evans
2001). This biological abundance notwithstanding, ‘African wetlands may not
qualify other than second-order hotspots especially as concerns endemism’ (Myers
1997), indicating the frequent bias of the ‘biodiversity’ concept towards species
diversity (Denny 1994; Scholes & Biggs 2005). Species-richness assessments of
African habitats therefore tend to neglect the role of seasonally flooded grasslands
(e.g. Fjeldsa et al. 2004) and relatively few of them have an effective protection status (WCMP 2004).
16
Part I – Introduction
No and name
Country
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Senegal
Senegal
Mali
Nigeria
Nigeria
Ghana
Nigeria
Cameroon
Cameroon
Chad
Congo
Zambia
Zambia
Botswana
Malawi
Congo
Congo
Sudan
Kenya
Senegal Delta
Senegal Valley
Inner Niger Delta
Niger fringing plains
Niger Delta
Volta River
Benoué River
Benoué River
Logone Floodplain
Chari and Logone
Congo River
Barotse plain
Kafue Flats
Okavango
Shire River
Kifakula Depression
Kamulondo
Sudd
Tana Delta
Area at peak flood (km2)
8 000
5 000
30 000
5 000
36 000
8 500
3 100
1 000
5 000
63 000
?
10 750
4 300
17 000
1 000
1 500
12 000
92 000
1000
Adapted from Drijver and Marchand 1986 and Welcomme 1979.
Fig. 1.1 – African Seasonally Flooded Grasslands
Floodplain type
Coastal delta
Fringing floodplain
Internal delta
Fringing floodplain
Coastal delta
Fringing floodplain
Fringing floodplain
Fringing floodplain
Fringing floodplain
Fringing floodplain
Fringing floodplain
Fringing floodplain
Fringing floodplain
Internal delta
Fringing floodplain
Fringing floodplain
Fringing floodplain
Fringing floodplain
Coastal delta
1 – General int roduct ion and st udy out line
17
Since about 5000 BC, when the earliest systematic colonisation of the Nile, Mesopotaminan and Indus rivers occurred, there has been concentrated human efforts
to tame the floods for agriculture and other purposes. Not all these developments
have been sustainable. Overexploitation of renewable resources led to the collapse
of some ancient societies, such as in Mesopotamia with its irrigated agriculture
(Janssen & Scheffer 2004). Land uses that did not interfere with the hydrological
regime such as grazing and associated burning and extensive agriculture, have
also had a profound influence on the natural vegetation cover (Drijver & Marchand
1986).
The hydrology of the wetlands of sub-Saharan Africa remained, until about 50 years
ago, relatively unchanged explaining their general natural appearance (Denny
1993). More recently, however, demand for water and electricity has resulted in an
increasing number of water management projects (Drijver & Marchand 1986).
Off-shoots off these developments were varied and include expansion of inland
fisheries and irrigation schemes. Large man-made lakes not only flood valleys but
also regulate the water flow in catchment and drainage areas, thus depriving the
seasonally flooded grasslands of their main inputs of water and sediments (Denny
1993). The loss of seasonally flooded grasslands often meant the loss of the resource base on which local communities and wildlife depend, with devastating
impacts (McCully 2001).
Few of the large African seasonally flooded grasslands are effectively protected, although recently efforts have been undertaken to designate large parts of the Inner
Niger Delta, Lake Chad and others as Ramsar sites (WCPA 2004). But even if
these floodplains have an internationally recognised protected area status, such as
Waza National Park and the Kafue Flats in Zambia, this has not necessarily stopped
the upstream construction of dams. Lack of impact assessments was blamed for
many of these interventions (Drijver & Marchand 1986). Even though such assessments are increasingly undertaken, as the one of the Jonglei canal in Sudan (Howell
et al. 1988), large scale hydrological interventions have continued on the African
continent (McCully 2001).
Floodplain rehabilitation
A number of studies has shown the economic irrationality of large-scale hydrological interventions in African floodplains, showing that their costs, including environmental ones, largely exceed benefits (Drijver & Marchand 1986; Marchand
1987). These studies and rising popular protests (McCully 2001) have led to an
emerging tendency to discuss the rehabilitation of these floodplains to their original functions.
18
Part I – Int roduct ion
Unfortunately the knowledge about African wetlands remains poor (Denny 2004)
and the few studies on the impact of processes, such as water releases from reservoirs, do not surpass a ‘black box’ stage (Hollis et al. 1993). One is therefore inclined
to look at experiences with floodplain rehabilitation experiments in Europe and
the USA that have lately taken flight (INTECOL 2004). These nature restoration
programmes are often carried out on a sub-landscape level only, in areas measured in hectares and not in square kilometres. With such applied scales it is not
surprising that most restoration programmes are based on a sectoral nature conservation approach, studied by predominantly natural sciences. Larger-scale rehabilitation efforts almost always associate people, and require interactions of natural and human sciences. The subsequent increasing complexity makes it particularly difficult to transpose experiences from one place to another.
These observations were earlier voiced by Schouten (2001) and Nienhuis (1998)
in their comments on rehabilitation experiments in the Netherlands and Europe
respectively:
‘In many cases, nat ure rehabilit at ion cannot be isolat ed f rom t he rehabilit at ion
and management of t he landscape...nor can t he landscape be considered in isolat ion f rom various socio-economic f unct ions. The int egrat ion of t hese layers of
int erest f orms one of t he biggest and most urgent challenges of spat ial planning
in t he Net herlands’ (Schout en 2001).
and
‘...experiences in Europe w it h t he rehabilit at ion of large rivers are rare, relat ive
t o smaller st reams, due t o t he large numbers of int erw oven societ al demands on
t he river, t he connect ed economic cost s and t he complexit y of t he physical and
biological syst ems involved. Proposals and concept s f or large river rest orat ion are
much more abundant t han demonst rat ions in t he f ield. The empirical large-scale
t est ing of t he models, connect ed t o t he polit ical w illingness of cat chment -scale
rest orat ion, are f orming t he real bot t lenecks’ (Nienhuis et al. 1998).
The Everglades (USA) hosts the best-known large-scale (sub) tropical wetland
2
rehabilitation program, aiming at restoring parts of the 5000 km degraded area
(Kiker et al. 2001). Compared to African seasonally flooded grasslands, the Everglades’ oligotrophy, the dominance of Cyperaceae and the limited role of wild and
domestic large herbivores is striking. Most striking differences are related with
the area’s socio-economic context as can be expected from its location in the USA.
This is illustrated by the Everglades restoration programme’s budget of approximately 10 billion USD. In contrast, the budget of the Waza-Logone floodplain reha2
bilitation totals 20 million SFr, i.e. 1% of the recovery cost per km of the Everglades (Kiker et al. 2001; IUCN 1999). Despite the multitude of technical studies,
1 – General int roduct ion and st udy out line
19
many questions remain on the impact of restoration on the Everglades ecosystem.
This has led to a still rather basic restoration vision ‘to mimic as closely as possible the appearance and behaviour of the system as if drainage and development
had not occurred’ (Davis & Ogden 1994). The lack of a common vision is further
illustrated by the repeated delays in the implementation of the Everglades restoration program (Grunwald 2002).
One can conclude that for the large-scale rehabilitation of tropical floodplains in
developing countries, not only demonstrations hardly exist, but also the concepts
and proposals (‘models’) have yet to be formulated. The present study analyses a
floodplain rehabilitation test at a landscape level in a tropical floodplain ecosystem
in North Cameroon and aims to contribute to the development of these concepts
and proposals for future rehabilitation trials in other African floodplains.
1.2
Introducton to ‘Conservation & Development’ and capacity
building needs
The community conservation panacea
The image with which conservation organisations in Africa present themselves
has changed over the last 20 years from wildlife protection to people-oriented conservation (e.g. Adams & Hulme 2001). The protection focus of the colonial and
early independence period was reflected in autobiographies of the last European
directors in charge of protected areas in Sub-Saharan Africa with titles such as
1
‘Mourir pour les éléphants’ (Verschuren 1970), ‘Les guardiens de la vie sauvage’
2
3
(Dupuy & Dupont 1984 and ‘An impossible dream’ (Parker & Bleazard 2001).
In the early 1990s, publications such as ‘The Myth of Wild Africa’ (Adams &
McShane 1992) and ‘Out of Eden’ (IIED 1994) reflected the focus of an emerging
conservation brand that refuted the concept of a ‘wild Africa that had to be protected against Africans’. The new ‘leitmotif’ was ‘conservation with, instead of
against people’ and ‘win-win scenarios’ (Inamdar et al. 1999). This new conservation paradigm, so-called Community Conservation, was considered a means of reconciling Conservation and Development by ensuring that the interests and knowledge of local communities were taken into account (Adams & Hulme 2001). Local
communities were to be provided with trade-offs of the protected area through a
share in tourism and hunting revenues or the (legal) use of specific natural resources. Alternatively, compensations should be provided for the opportunity (non-
1 Former director of wildlife of DRC (former Zaire).
2 Former director of wildlife of Senegal.
3 Game wardens in Kenya at the end of the colonial period who dedicate their book to ‘a Pleistocene
Africa which we so enjoyed and sought to preserve, but which is gone’.
20
Part I – Int roduct ion
exploitation) costs of protected area resources by the initiation of small-scale development projects. For some, community conservation was also considered a pragmatic way of dealing with increasingly poor and often corrupt governments (Adams
& Hulme 2001). Community Conservation (CC) embraces a range of initiatives
such as Integrated Conservation-Development Projects (ICDPs), Community-Based
Conservation (CBC), Community-Based Natural Resource Management (CBNRM)
(Adams & Hulme 2001). These expressions are often used as synonyms, despite
their different (regional) origins.
The marriage of Conservation with Development has become an attractive target
for international funding that, at the end of the 1990s, totalled annually hundreds
of millions of dollars (El Ashri 2001). The development of the Community Conservation concept in Africa was based on experiences of which the CAMPFIRE ini4
tiative on wildlife utilization in Southern Africa is one of the best known (Child
1995). The community-based mountain gorilla tourism in Rwanda and Uganda
with its high revenues per tourist funnelled into local communities was another
experience that stimulated the early development of Community Conservation in
Africa (Weber & Vedder 2001).
Community conservation, the debate
In the late 1990s, however, critical reports on Conservation-Development projects
began to emerge (Berret & Arcese 1995; Kramer et al. 1997; Terborgh et al. 2002).
Some biologists went further and questioned the underlying principles of Community Conservation and pleaded for a return to the ‘fortress Conservation’ approach (Oates 1999; Spinage 1998). This so-called ‘pro-park lobby’ is in an unlikely alliance with critics who detect in community conservation a shallow façade to
hide old-style preservation (Adams & Hulme 2001). Although few protected area
managers seem to follow this criticism, it has led to an emerging consensus that
Community Conservation is not the panacea that it initially had appeared. Community Conservation has its limits when wildlife resources or socio-economic circumstances do not allow their sustainable use in sufficient quantities compared
to the opportunity costs of the protected areas (Adams & Hulme 2001). The potential of tourism and safari hunting has, most notably, shown to be more limited
than previously assumed (Wilkie & Carpenter 1999ab). The practical organisation
of community conservation has also proven to require long-term investments and
stable and decentralised institutional environments (Adams & Hulme 2001).
Sometimes, development near protected areas may do more harm than good. Experiences from West Africa, the Central African Republic and the one presented
in this study have shown that development in the vicinity of protected areas may
appear to be a Trojan Horse jeopardising conservation (Noss 1997; Oates 1998;
4 Former director of wildlife of Zimbabwe.
1 – General int roduct ion and st udy out line
21
Scholte 1998). Outside Africa, the Galapagos islands are an example of the risks
of developing ecotourism in a prime conservation area thus triggering massive
immigration and over-exploitation (Grenier 2000). Some, like Oates (1999), have
therefore pleaded for a total ban on Community Conservation.
Developing tools
One may argue that not enough has been done to develop the tools for Conservation-Development approaches that should also take into account the risks of
development. At present Conservation-Development approaches through Integrated Conservation Development Projects remind us of the expression ‘If your
only tool is a hammer everything looks like a nail’. Or, as Adams & Hulme (2001)
concluded pragmatically ‘The real issue is not whether conservation should be
done with people, but how’.
Developing the capacity of park managers
The changes mentioned above in conservation image notwithstanding, one has
the impression that daily African wildlife management practice has, since the days
of the last European wardens, little changed in its pre-occupation with poaching
control as described by Parker & Bleazard (2001). The title ‘Wildlife Wars’ of the
autobiography of the former director of the Kenya Wildlife Service (Leakey & Morell
2001), covering the period in the 1990s, is meaningful. I attribute this ‘business
as usual’ at least partly to the limited efforts to involve protected area personnel in
the design and implementation of conservation-development (see also Western
2003). ‘Approaches to natural resource management... tend to assume that the
manager is outside the system being managed. However, where the objectives
include long-term sustainability, ... the managers [are] integral components of the
system’ (Walker et al. 2002). With the increasing international attention for Community Conservation, the role of protected area managers has been further eroded into a caretaker with an agenda (i.e. management plan) filled in at the nation’s
capital. This situation contrasts the early days of independence when African Park
5
wardens were amongst the best-educated personnel in their country. Early graduates of the regional African wildlife colleges in Mweka and Garoua became later
in their careers minister or occupied other high-ranking posts in their countries,
a situation increasingly difficult to imagine. Their high-ranking positions allowed
them to influence developments beyond the borders of the protected area.
In this study I develop a pragmatic approach on Conservation-Development. The
study further aims at understanding the role of management planning and the
potential of raising the capacity of protected area managers, to offer a perspective
of a ‘future for Conservation & Development’.
5 Parker and Bleazard (2001) show that although colonial park wardens in Africa were materially not
well off, they enjoyed high status. After independence this status has dropped gradually.
22
Part I – Int roduct ion
1.3
Reflooding in Waza-Logone6
This study discusses the quest for adaptive management that was started in 1993
with reflooding parts of the Waza-Logone floodplain in North Cameroon. The reflooding aimed to rehabilitate vegetation, birds, wildlife, pastoral grazing and fisheries that had come under pressure as result of the construction of an upstream
dam and embankment in 1979. Initially, our attention was aimed at the monitoring of the impact of the reflooding and the design and discussion of large scale
reflooding options. During this process, our attention shifted to the mitigation of
existing and rising conflicts in the floodplain, in particular the one between Waza
National Park and the surrounding communities. Learning by doing not only has
given insight on the impact and consequences of reflooding but also highlighted
the need to know more about the condition of the floodplain prior to the construction of the dam in 1979.
Following the widespread famine during the 1970s in the Sahel, the Government
of Cameroon, assisted by several donor countries, initiated a number of large-scale
irrigated rice schemes. One of them was a gravitation irrigated rice scheme, for
which a dam and an embankment along the Logone river were constructed to
form the Lake Maga reservoir (Fig. 1.2). In combination with lower than average
7
rainfall, the depth, duration and extent of downstream inundations were reduced,
which led to serious ecological degradation of the Logone floodplain, both in and
outside Waza National Park (see Fig. 1.2). Annual grasses invaded the productive
perennial grasslands, limiting regrowth in the dry season and reducing carrying
capacity for wildlife and livestock (Scholte et al. 1996a). Fishing resources also
dropped dramatically, provoking an emigration of a significant part of the human
population from the floodplain (Drijver et al. 1995). Leiden University, in collaboration with the Garoua Wildlife College, initiated several impact studies that showed
not only the devastating impact of the rice scheme but also its failure in food production. In the 1980s, protests of local populations were still considered taboo;
when desperate floodplain inhabitants tried to reopen a watercourse they were
deterred by the army. Despite all investments, rice cultivation did not become successful and in the early 1990s half of the irrigated rice scheme area was left unused. A climate was thus slowly opened in which measures could be discussed to
mitigate and even counteract the impact of the dam and the embankment.
This allowed the start of the Waza-Logone project in 1992, a collaboration between
the Cameroonian Ministry of Environment and Forestry and the World Conservation Union (IUCN), the Institute of Environmental Sciences (Leiden University),
the Netherlands Development Organisation (SNV) and WWF-The Netherlands.
6 The Waza-Logone area will be introduced in Chapter 2 and the photographic outline.
7 As measured since the 1930s (Beauvilain 1995).
1 – General int roduct ion and st udy out line
Fig. 1.2 – The Waza-Logone Area
23
24
Part I – Int roduct ion
The overall goal of the project was to assist the government of Cameroon in pursuing the integrated management of natural resources in the Waza Logone region,
so as to provide a sustainable livelihood for the local people and to maintain the
diversity of the biological resources of the floodplain.
8
Amongst the specific objectives were:
• To investigate, design and implement an operational plan for the rehabilitation
of the hydrological regime of the Logone floodplain
• To formulate and test an operational action programme for the management
and development of the area around Waza National Park (IUCN 1999; Loth &
de Iongh 2004).
Studies of the Waza-Logone project in 1993 identified a number of reflooding options, which were expected to have a major impact on the floodplain, including its
human population. When in early 1994, the hydrologists Emmanuel Naah and
Hans Wesseling, topographer Paul Kouamou and I visited the floodplain after a
year of reviewing studies on hydrology, ecology and land use, we realised how little we knew on the possible impact of reflooding. The extremely flat area, with a
slope of no more than a few centimetres per km, made that even basic questions
such as those on the direction of flooding had to remain unanswered, let alone the
impact of reflooding on vegetation or land use. The plan described in the project
2
document, to prepare and implement a 1000 km reflooding in one step, looked
increasingly grotesque.
Although not planned and budgeted for, subsequent discussions held at the WazaLogone project lead to a strong backing of a pilot release as a learning by doing
experiment. An ideal candidate, without the risk of flooding habitations and cropland, was the opening of the Petit Goroma, a watercourse that was cut off by the
embankment along the Logone river (Fig. 1.2). After discussions with communities and authorities the pilot floodplain reflooding was initiated by breaching the
2
embankment, reflooding from September 1994 onwards 180 km of desiccated
2
floodplain and raising water levels in an additional area of approximately 600 km .
The impact of the reflooding, already in the first year, on especially water levels, fish
production and grazing exceeded our expectations. But should we attribute this to
8 Although formulated as specific objectives, these are in fact activities or outputs of the project. Also
mentioned:
• To contribute to ongoing government efforts towards regional development planning by means of
field evaluations of the environmental impacts of existing development projects.
• To develop a methodology for the design and implementation of conservation and development
activities in other situations in the Sudano-Sahelian region.
• To strenghten institutional capacity to manage the natural resources of the Waza-Logone region, in
particular through training of national staff participating in the project.
1 – General int roduct ion and st udy out line
25
the opening of the embankment or were there other factors, such as the favourable
rainfall, that interfered? The large impact of the pilot reflooding stressed the need
to strengthen the management of Waza National Park and its direct surroundings
to cope with the rapidly changing situation. Especially the attraction of pastoralists
and fishermen to exploit the newly reflooded areas drew our attention. At the
same time a hitherto relatively strong governmental presence, including a relatively well equipped and staffed Waza National Park, was disappearing. Solutions
had to be developed to mitigate the negative impacts of these changes. Through
the formulation of a management plan, such measures were formulated and programmed with local communities, sedentary as well as nomadic, protected area
authorities, traditional and administrative authorities and the Ministry of Environment and Forestry. Experiences with the implementation of this management
plan highlighted, however, the low capacity of park personnel to deal with new
forms of collaboration with local communities. This motivated the training of not
only park guards of the national parks in north Cameroon, but also at the Garoua
Wildlife College, thus increasing the impact beyond the Waza-Logone area.
The Waza-Logone pilot reflooding was a unique large-scale experiment in tropical
ecosystem and land use recovery. The induced changes in floristic composition and
vegetation productivity challenged vegetation dynamics theory and generally assumed flooding – productivity relationships. Monitored changes in waterbird and
antelope numbers questioned a simple recovery mechanism and highlighted interactions of wildlife with competing land uses, in particular pastoralism that
showed a steady increase in grazing intensity. With an impact zone of several hundred square kilometers, the reflooding covered site-specific and landscape issues,
linking local impacts, e.g. on vegetation composition, with region-wide changes,
such as on transhumant pastoralism patterns.
1.4
Research questions and study outline
The present study is organised in four parts. The first introduces the area and available information sources. The second part discusses the impact of the reflooding
induced in 1994 on the Waza-Logone floodplain. The third part questions the
risks of this floodplain rehabilitation that were mitigated by management planning and training. The fourth part syntheses these experiences.
In more detail, and mentioning the respective chapters where the information can
be found, the research questions and the outline of this study is as follows.
26
Part I – Int roduct ion
PART I – Introduction
How has our reference image of Waza-Logone developed in time?
The present knowledge of the ecology of Waza-Logone has developed in a period
of widely fluctuating climatic conditions and a rapidly changing land use. I start
this study with a presentation of the main sources of information on Waza-Logone
in a time sequence over the last 150 years (Chapter 2). The presented historical diagram shows the link between available information and the reference image for
floodplain rehabilitation. Presented information also stresses the ecological significance of Waza-Logone, even in its degraded post-dam period, justifying the rehabilitation efforts. The papers cited in the list of background publications provide
additional information on these issues.
PART II – The Impact of Reflooding in Waza-Logone
The overall research question is: Does reflooding lead to the restoration of the WazaLogone floodplain to its pre-dam structure and conservation and development functions?
In other words, How resilient is this floodplain environment after the dam construction? Has it switched into another state without possibility to return to its pre-dam
state?
Scientists active in the Waza-Logone area in the 1980s and early 1990s attributed
much of the observed degradation of the Waza-Logone area to the desiccation following the Maga dam construction, a process that has taken over 15 years and was
still under way at the start of the 1994 reflooding. Moreover, it occurred in a period of drought stress and with a rapidly changing land use. We did not take for
granted that reflooding would automatically reverse this process. In the meantime,
population pressure in and around the floodplain had increased considerably and
it was not excluded that rehabilitated resources could rapidly become overexploited again. In this study the question of ecosystem restoration is answered along three
lines, (1). vegetation composition and productivity, (2) numbers of waterbirds and
antelopes and (3) pastoral recovery. The question of part II is answered under
these latter three subheadings and their respective sub-questions, in five chapters:
Does reflooding lead to a 100% perennial grass cover with the same floristic composition
and production that existed prior to the Maga dam construction? What are the mechanisms with which these changes take place?
1 – General int roduct ion and st udy out line
27
Vegetation composition
In Chapter 3 we synthesise the main findings of unpublished studies in the 1960s
and 1970s, prior to the Maga dam construction and after the dam in the 1980s on
the changes in vegetation composition, that were our de facto reference image. My
colleagues and I systematised this monitoring to study the vegetation changes
triggered by the 1994 pilot reflooding. We expected that reflooding would induce
the restoration of perennial vegetation in five years to its pre-dam situation, characterised by the perennial grasses Echinochloa pyramidalis, Vetiveria nigritana,
Hyparrhenia rufa and Oryza longistaminata (Wesseling et al. 1994). This optimism
2
was based on observations in a 40 km area in the central, relatively well flooded
floodplain where the water level was raised by the construction of a small dam (Drijver & Kouahou 1995). Studies on hydroseries, the sequence of plant species along
a flooding gradient, explained the impact of this water raising (van der Zon 1992).
This use of hydroseries is, however, not uncontested (e.g. John et al. 1993; Leendertse et al. 1997). The area downstream of the Maga dam had become entirely
desiccated and a slower or only partial vegetation recovery could alternatively be
expected.
Above questions are linked to the discussion on ‘ (Gleasonian) Succession versus
State and Transition’ alternatively called ‘gradual changes’ versus ‘catastrophic
shifts in ecosystems’ (Scheffers 2001). This discussion has taken flight in rangeland management in the 1980s (Westoby 1980; Westoby et al. 1989, see also
Rietkerk et al. 1996), followed by aquatic and other ecosystems (e.g. Scheffer et al.
1993). Floodplains have been considered environments where cyclic (Gleasonian)
succession takes place; the ‘lineair’ Clementsian succession model (evolving into
a single state) is obviously of little use (Van der Valk 1992; Middleton 1999). The
alternative State and Transition model predicts a different scenario, in which sudden switches into different vegetation communities will take place only once a
threshold has been passed. This unpredictably threshold may be difficult to pass
hampering necessary interventions to reach the desired state. Following this model,
reinstalling the pre-dam flooding regime would not automatically lead to the
restoration to the pre-dam vegetation.
Vegetation productivity
Many floodplain plant species are said to have higher rates of production in less
flooded conditions (Middleton 1999), but this reasoning may be biased towards
temperate zones. We expected that perennial grasses in Waza-Logone would show
an increased production with reflooding, but no reference information existed on
the impact of the Maga dam on vegetation production. We further hypothesised that
in case of increased production there would be a time lag between the moment of
water raising and the full expression of vegetation production increase, because of
the predominance of perennial rhizomatous grasses of which a large part of the
biomass is in their rhizomes. In Chapter 4 I therefore linked maximum flood depth
28
Part I – Int roduct ion
with above-ground biomass at the end of the wet season allowing an assessment
of the impact of the 1994 pilot reflooding.
Does reflooding lead to the recovery of birds and wildlife to their pre-dam numbers?
Waterbirds
Waterbirds constitute one of the principal conservation assets of the Waza-Logone
area (Fotso et al. 2001), but have hitherto received little attention other than occasional, qualitative observations. The only quantitative information on the occurrence of waterbirds prior to the Maga dam was on Black-crowned Crane, one of
the area’s flagship species that numbered at least 10 000 individuals in the 1970s,
almost as much as the present entire subspecies population. The resident Blackcrowned Crane depends on moist grassland habitat during the crucial nesting and
dry season periods. Its population drop to about 2000-2500 individuals in the
early 1990s was therefore attributed to the Maga dam construction (Scholte et al.
1999; Sinibaldi et al. 2004). A comparison of the dominant habitats showed the
importance of perennial grasslands and 20-40 cm deep water for waterbirds
(Scholte et al. 2000a). Against this background, Chapter 5 investigates whether
reflooding, that increased the extent of shallow water and allowed the recovery of
perennial grasslands, has resulted in increased numbers of waterbirds. The relatively favourable rainfall that characterised the period since the reflooding, may
also have influenced the observed waterbird numbers. This also holds for protection, especially of vulnerable colonies.
Antelopes
Waza National Park is one of the last remaining areas in dryland West-Central
Africa with abundant wildlife. Antelopes are amongst the wildlife most prominently present in the area, but the floodplain antelope Kob and Korrigum (‘Topi’)
in particular have undergone a dramatic drop in population numbers after the Maga
dam construction. Chapter 6 assesses the development of these antelope populations from the early 1960s, allowing appreciating their possible (partial) recovery
to the spectacular population size in the early 1970s. As discussed for waterbirds,
the positive influence of the relatively favourable rainfall since the mid-1980s on
the one hand and the increasing human pressure on the national park on the other
hand was expected to influence possible reflooding impacts.
Does reflooding lead to full recovery of pastoral use of the floodplain?
Pastoralists
Transhumant pastoralists are a wary population group in Waza-Logone, which has
been largely neglected by developers and authorities. However, grazing of their estimated 200 000 cattle constitutes one of the main land uses in the Waza-Logone
area. Our activities with pastoralists were based on confidence building through
1 – General int roduct ion and st udy out line
29
the organisation of discussion sessions with authorities in the pursuit of solutions
to the reigning insecurity that troubled their daily life. In Chapter 7, we present the
impact of the 1994 pilot reflooding on the pastoral migration pattern and the
resulting increasing cattle grazing intensities. Three scenarios were formulated to
capture potential reactions to reflooding: an overshoot scenario with a higher grazing intensity than resources available, a territorial scenario with a lower grazing
intensity and an ideal distribution scenario with a grazing intensity in line with
the availability of resources. An important justification of the pilot reflooding was
that the rehabilitated floodplain would become an alternative for grazing inside
Waza NP. In the late 1980s and early 1990s incursions of cattle, searching for dry
season forage and water, constituted a major source of conflict with the Waza
National Park authorities.
PART III – Enhancing Conservation-Development Integration by
M anagement Planning and Training
The overall research question is: If a rehabilitation is ecologically successful, how may
then the ecosystem’s functions that underlie the balance between Conservation and
Development be assured?
Risks: The overshoot of success
Chapter 8 shows that pastoralists and fishermen, attracted by the newly available
resources, compete with waterbirds and floodplain antelopes, thus threatening the
balance of the pursued conservation-development integration. I subsequently analyse if comparable risks exist in other Integrated Conservation-Development Projects
(ICDPs) and how these are dealt with. The immigration risks motivated the development of a policy, based on a categorisation of local communities, of which some
first promising results are presented. The experiences from Waza-Logone are of
great interest for other Conservation-Development Projects whose impacts are
often difficult to distinguish from other development impacts.
Grip on the w hole: M anagement planning
Participatory studies showed that exploitations from inside Waza National Park,
despite their illegality, have continued under the presence of, till recently, fair numbers of park guards. These observations refute the ‘victim image’ that sometimes
is held of local communities living around protected areas (IIED 1994). Moreover,
the continuing exploitation of park resources, reinforced by the results of the
reflooding monitoring, question if the discussed ecosystem rehabilitation also leads
to the recovery of wildlife and improved land use, the reference image sketched in
Chapter 2. A new management policy on the status of the park villages was one of
the outcomes of the Waza NP management plan, the first of its kind in Cameroon
and indeed Central Africa. The plan further offered a consensual interpretation of
the 1994 environmental law to control the continuing exploitation of resources
30
Part I – Int roduct ion
from inside Waza National Park. The plan also allowed the creation of a management committee in which, for the first time, park authorities and local communities discussed problems related with the park management. I discuss the formulation process of this management plan and review its outcome based on an independent evaluation in 2002. These lessons are put in a perspective of some of the
key issues of ongoing management planning in African protected areas (Chapter 9).
Foundation: Development of human capacities
The policies analysed were facilitated and driven by third parties, i.e. international NGOs and expatriates mainly. Protected area managers played a rather limited
role only, due to lack of training and experience, undermining their involvement
in the development of conservation-development integration. I address necessary
training of protected area personnel on two levels. In Chapter 10 I address training requirements on an institutional level and discuss curriculum reforms at the
Garoua regional wildlife college for West-Central Africa, through a comparison
with the other regional African wildlife colleges in Eastern and Southern Africa. In
Chapter 11, I evaluate courses addressing community conservation and park planning. These pilot courses were developed at the Garoua Wildlife College during
the Waza management plan formulation and implementation. Presented experiences are also relevant for other protected areas in West and Central Africa.
PART IV – Floodplain Rehabilitation and the Future of ConservationDevelopment: Synthesis and Concluding Remarks
In the concluding Chapter 12, I review the outcome of the reflooding. I compare
the reflooding responses that, because of their different response times, caused a
domination of human supported resources. I continue analysing the contributions of management planning and capacity building to correct these undesired
side effects of the otherwise successful reflooding. Based on these experiences I
comment empirical concepts of ecosystem change most notably resilience and
hysteresis. Especially the time dimension, that played such an overruling role in
outcome of the floodplain rehabilitation, receives special attention. The normative
ecosystem approach, is subsequently used to present the key findings of this study
in a wider context.
Monitoring observations and insights allowed the reformulation of the floodplain
rehabilitation expectations and hypotheses postulated above and discussed in the
respective chapters. The consequences for the overall floodplain rehabilitation targets as programmed in Waza-Logone (IUCN 1999) and possibly in other (African)
floodplains, are discussed in section 12.6. I conclude with reviewing the overarching adaptive management approach that is introduced below.
1 – General int roduct ion and st udy out line
1.5
31
The quest for adaptive management
Adapt ive management , can be loosely def ined as t he learning by doing, relies
on an accumulat ion of credible evidences t o support a decision t hat demands
act ion (Walt ers & Holling 1990).
In addition to the discipline specific discussions introduced in the preceding paragraphs, I would like to present this book as test of an integrated approach of tropical floodplain rehabilitation.
The 1980s and 1990s witnessed a blossoming of concepts and theories which tried
to capture the ecological as well as social aspects of the management of ecosystems. The Ecosystem Approach (Mitchell 2002), Ecosystem Based Management
(Pirot et al. 2000), Resilience management (Walker et al. 2002) are some of the
more recent expressions, inspired by Adaptive Management, introduced by Holling (1978). Adaptive management was justified by the difficulties to assess the
control of all aspects of ecosystem management motivating ‘a shift to adaptive
management, which relies on the flexible, diverse and redundant regulation, monitoring for responsiveness and experimental probing’ (Holling 1995). Key desirable
attributes highlighted are the capacity for learning and to be flexible and adaptive,
hence the link that is often made with ‘learning organizations’ (Mitchell 2002;
Salafsky et al. 2001).
Walters & Holling (1990) classified adaptive management into three categories,
ranging from trial and error, to passive and subsequently active adaptive management. For some floodplain resources, i.e. waterbirds, lack of any quantitative reference information allowed us initially only to pursue a trial and error approach
as outlined in the previous paragraph. For other resources, most notably perennial vegetation and floodplain antelopes, we started with a single hypothesis ‘requiring a natural habitat’, i.e. with as target ‘back to the pre-dam situation’. Over the
years this passive adaptive management evolved into a more active adaptive management and alternative hypotheses of floodplain rehabilitation were formulated,
amongst others related with favourable rainfall conditions and unfavourable conditions related to human pressure.
This thesis follows a thread of adaptive management: From [I] the conduct of inventories and overview of studies (Chapter 2 and annexes), allowing the formulation
of an initial reference and target image and rehabilitation hypotheses (Chapter 1) to [II]
probing, pilot reflooding & monitoring, impact studies on vegetation, waterbirds,
floodplain antelopes and pastoralists (Chapters 3-7), to [III] problem analysis on the
risks of conservation-development integration (Chapter 8). Subsequently solutions are planned through management planning and reviewing (Chapter 9) and
capacity building is undertaken for future Conservation & Development (Chapters
32
Part I – Int roduct ion
10, 11). [V] In the synthesis the contributions of the undertaken approach and the
initially postulated hypotheses are evaluated and reformulated.
Floodplain Rehabilitation and the Future
of ‘Conservation & Development’
A PHOTOGRAPHIC OUTLINE
This story of Waza-Logone is long and has loops and many details. The following
photographs and their captions illustrate the essential aspects of the story. They
touch upon each of the chapters in the same sequence as the text.
PART I – Introduction
The ecological background of the floodplain
The Adamaoua mountains,
situated 500 km south of WazaLogone, are the main catchment
area of the Logone River.
The Mandara mountains, 100 km
south-west of Waza-Logone, no
longer drain directly into the
Waza-Logone floodplain.
The dam and embankments of
Lake Maga, built in 1979 for
irrigated rice cultivation, block
flooding of the Logone river and
from streams of the Mandara
Mountains into the southern
Logone floodplain.
34
Part I – Introduction
Within weeks of the floods
receding, the area changes from
a carpet of green lush grass into
scorched earth.
The annual cycle of flood and drought
characterises African floodplains. This
thesis examines the ecological effects of
this annual cycle, and what happens
when is no longer occurs and when it is
reinstated again. The reinstatement of
the annual flood and drought cycle is
what is meant by ‘reflooding’.
The ecological background of the floodplain
35
Acacia seyal shrubland borders the floodplain. It is not only an important source of food for wildlife, but
also provides Arabic gum and firewood to local communities
The rocks near Waza town tower above Sclerocarya birrea woodland that dominates the sandier, slightly
elevated parts of the western Waza-Logone area. Here wildlife concentrates during the rainy and flooding seasons.
36
Part I – Introduction
Three species of vultures with Giraffes near a waterhole in the western part of Waza NP. The importance
of Waza NP was recognised in its designation as an UNESCO-Man and Biosphere Reserve.
Concentrations of waterbirds and Kob in
the eastern floodplain part of Waza NP.
The national park and floodplain were
selected as Important Bird Area and presented as the first Ramsar Site of
Cameroon.
Lion escaping the heat. Spectacular wildlife such as
Elephant and Lion attracts more tourists to Waza NP
than to than any other protected area in Central
Africa.
The ecological background of the floodplain
37
The floodplain people and the resources they use
Ramparts of Zina, a Kotoko town in the centre of the floodplain. For centuries, the Kotoko used to be
the rulers of the floodplain, which has a long history of conflicts over floodplain resources.
The Kotoko used to be the exclusive owners of
fish canals. This unique fishing technique drains
depressions towards the main watercourses at the
end of the flooding season. At the end of each
canal the fisherman places his nets and traps
almost the entire fish population that lived the
previous season on the floodplain.
Fish canals from the air, taking receding waters
towards the Logomatya, a branch of the Logone River that can be seen in the background.
38
Part I – Introduction
Dieguéré, a Musgum village on the edge of Waza NP. The
Musgum are generalists who cultivate, fish and keep livestock.
Most of them settled in the floodplain in the early 1900s.
Fishing on the plains near Zwang at the
onset of the floods.
Sorting fish. In WazaLogone, fisheries are second only to grazing in its
economic importance.
Andirni, an agropastoral village created in 1917 because of the proximity of several pools. Many FulBe
and Arab agropastoralists settled in villages like Andirni south of Waza NP after the droughts of the
1970s.
Inside the house of Fannie, wife of Moussa Barka, a park guide living in Andirni. On the left the decorated bed, showing the family’s transhumant roots. For years, Moussa and Fannie were our wonderful
hosts.
40
Part I – Introduction
Young men herd the cattle of the agropastoralists.
FulBe and Arab transhumants move with their families. This harsh life is increasingly being abandoned
for life in villages and towns with health, marketing and schooling facilities.
The floodplain people and the resources they use
41
The ecological impact of the Maga dam
Deserted quarter of Mahé, in the northeast corner of Waza NP. After the Maga
dam construction approximately one third
of the sedentary and mobile population of
the floodplain left.
Dead Giraffe in Talabal waterhole.
Dead Kob antelope. Following the
Maga dam construction the Kob population crashed from 20 000 to 5 000.
Its population further declined to
2000 by the end of the 1985 drought.
Rinderpest hit the Kob hard during
those years as well.
42
Part I – Introduction
Since the completion of
the Maga dam, waterholes
are filled by tankers
towards the end of the
long dry seasons to help
keep wildlife alive.
The only Vetiveria nigritana
left in the western floodplain,
where prior to the Maga dam
this tussock grass used to be
dominant. It is not the lack of
water but especially the drop
in perennial grass production
that has caused the ecological
catastrophe following the
dam construction.
The ecological impact of the Maga dam
43
The lead-up to the reflooding and the first water coming through
The Maga rice scheme with large
parts of its area left fallow. In
the early 1990s, it had become
increasingly clear that the Maga
dam was an ecological as well as
economic failure. The continuing
import of Asian rice, available in
the whole of northern Cameroon, testifies the failure of the
Maga irrigated rice cultivation
scheme. With the start of the
Waza-Logone Project in 1992,
the issue of reflooding was no
longer a taboo.
Consultation of local people showed unanimous support for the reflooding. Local communities had
never been consulted before and were hostile to the Maga dam.
44
Part I – Introduction
Breaching of the embankment in 1994.
View from the Logone at the reopened ‘Petit Goroma’ in flooding, with on the left the remaining
embankment. From September 1994 onwards, 20 000 litres water per second flowed additionally onto
the floodplain for three months of the year.
The lead-up to the reflooding and the first water coming through
45
Car stuck near Andirni
The Waza-Logone project has put much effort
in monitoring the impact
of the (pilot) reflooding
to prepare large-scale
reflooding. Transport
conditions were often
difficult, it took one day
to travel by car and boat
to Zina, the centre of the
floodplain. From there it
was another two days by
canoe or foot to the furthest monitoring sites.
Tracks become navigable for canoes during
the flooding season.
Following the monitoring transect from Zina
to Waza NP on foot.
46
Part I – Introduction
PART II – The Impact of Reflooding in Waza-Logone
Impact of the reflooding on vegetation dynamics
Installation of a monitoring
grid, in the heart of the
reflooded area where flooding and vegetation dynamics
were intensively monitored.
In 1994, prior to the reflooding, the grid was largely covered by Sorghum arundinaceum, a stout annual grass
that invaded the floodplain
during the absence of flooding in the mid-1980s.
Sorghum arundinaceum survived several years of flooding
but it ultimately disappeared from the reflooded area.
With its hard-to-digest reed-like stems containing few
nutrients, it is of limited grazing value.
Aerial view of the eastern part of Waza NP. In 1996
Sorghum arundinaceum still covered large stretches of
the floodplain, especially around waterholes, such as at
Tchikam. (the woody vegetation is dark green, the shorter
mostly biannual grasses, are clear green, whereas the
stout Sorghum can be recognised by its beige-greenish
colour).
The transect running from Zina into the heart of Waza National Park was installed in 1984 by students
of Leiden University. Covering always flooded, reflooded since 1994, and desiccated floodplain since the
Maga dam, the transect was an important base for (qualitative) vegetation monitoring.
The typical perennial floodplain vegetation,
with the rhizomatous perennial grasses wild
rice (Oryza longistaminata) and Echinochloa
pyramidalis. The cover of perennial grasses
increased between 1993 and 1999 from 41
to 75% of the area affected by the reflooding. In 2003, almost the entire reflooded
area was covered with perennial grasses
again.
In the foreground (bi-) annual vegetation with its large open spaces, which
allowed a recolonisation by the perennial floodplain grasses through lateral rhizomatous growth (from the background).
Vetiveria nigritana, a tussock grass, has not been able
to recover in the reflooded area despite its good condition elsewhere in the floodplain. We attribute this
set-back to the changed direction of flooding, with a
lower quantity of sediments it carries.
48
Part II – The impact of reflooding in Waza-Logone
Impact of the reflooding on vegetation production
This study clarified the link between maximum
flood depth and vegetation production. With a
20cm rise in maximum waterlevel through reflooding, above-ground biomass increased with approximately one third.
Measuring dry season regrowth following fire.
Regrowth, important because of its high quality
and availability in the dry season, is expected to
increase only after several years when sufficient
rhizome biomass has been accumulated.
Kob antelopes gathering in area with fresh regrowth after burning.
Impact of the reflooding on vegetation production
49
Impact of the reflooding on waterbirds
Foraging crane family on perennial floodplain grassland. Estimated numbers of Black-crowned Crane
dropped from over 10.000 in the early 1970s to 2000-2500 during the 1990s. The latter number still
represents one sixth of the world population of the western subspecies of the threatened Black-crowned
Crane. Black-crowned Cranes are present in Waza-Logone during the entire year and strongly depend on
flooding for both breeding and foraging.
Black-crowned Crane nest built on an inundated Vetiveria nigritana tussock, the grass species that
showed a dramatic decline after the Maga dam construction and has not yet shown any recovery.
50
Part II – The impact of reflooding in Waza-Logone
Waterbird counting team of the Garoua Wildlife College and Waza-Logone project. Total numbers of
observed waterbirds increased between 1992 and 2000 from 60 000 to 105 000. A combination of factors
was found to be responsible for this increase, including improved rainfall (especially ducks and geese),
floodplain rehabilitation (omnivorous storks and herons) and protection measures (some locally breeding storks and herons).
White-faced Whistling Duck. The increase in ducks and
geese in Waza-Logone corresponds to their recovery
over most of West Africa following the droughts in the
1980s.
Impact of the reflooding on waterbirds
The increase of Black-headed Heron (an omnivorous heron) was attributed to a combination of floodplain rehabilitation and colony
protection.
51
Black-headed Heron colony in Andirni. Because of protection by the village and floodplain rehabilitation,
this colony has increased from 750 in 1993 to the exceptional size of 2500 nests from 1999 onwards.
Large fish-eating birds such as Yellow-billed Storks did not benefit from the reflooding because of repeated destruction of their
colonies. Young birds are increasingly taken for consumption,
not hindered by fishermen who believe that fish-eating birds
reduce their own catch.
Destroyed Marabou Stork colony at Halé in 1997. Afterwards
Marabou Storks moved to Kalya in the centre of Waza NP where
the colony increased to more than 500 nests but was destroyed
again in 2002.
52
Part II – The impact of reflooding in Waza-Logone
Impact of the reflooding on antelopes
Park guard counting animals at Gobe waterhole in Waza NP. At the end of the dry season, only a handful of water points remain, allowing fairly accurate counts of animals that drink daily, such as most
antelopes.
Kob antelopes. The initial increase in number of Kob after the reflooding did not continue beyond 1997,
possibly because of competition from increasing numbers of livestock.
Red-fronted Gazelle, a Sahelian species, the only species that has benefited from the drying out of the
floodplain.
Impact of the reflooding on antelopes
53
The responses of pastoralists to the reflooding
Discussions with agropastoralists from Fadaré. Responses of mobile pastoralists to the floodplain rehabilitation programme were assessed through interviews with leaders of over 100 pastoral camps, which
were held at the end of each grazing season from 1993 to 1999.
Park warden questioning
a herder inside Waza NP.
One of the goals of the
reflooding was to stop
the incursions of pastoralists into the park to
reduce the competition
between wildlife and
livestock for dry season
grazing.
54
Part II – The impact of reflooding in Waza-Logone
Signing of the contract between the
Fadaré pastoralists, Waza NP authorities and Waza-Logone project
(December 1997). This formalised
the verbal commitment of Yerima, a
respected pastoral leader, that he
and his followers would not longer
send their herds into Waza NP.
Grazing intensity tripled from 1993 to 1999 due to a sharp increase in number of pastoralists migrating
onto the floodplain each year. Because of this increase in livestock numbers, the condition of individual
cattle has hardly improved. This leaves few incentives for pastoralists to refrain from grazing the good
pastures inside Waza NP.
The responses of pastoralists to the reflooding
55
PART III – Enhancing Conservation – Development Integration by
Management Planning and Training
The risks of a Conservation & Development project that is too successful
Lougouma, village on the border of
Waza NP. If the park was really completely closed for the exploitation of
natural resources such as wood, thatch
grasses and pastures, the 15 villages
surrounding it would not be able to
stay where they are.
Newly built quarter at
Lougouma village hosting
people arriving after the
improved flooding.
The Sultan of Logone Birni in front of his palace. Traditional authorities have lost much of their power
and no longer control settlements in ‘their’ area. It is in the interest of village chiefs to allow an increase
in population, as it gives them more power and financial means. As also the government has little control over these settlements, there is a need for developing alternative local management structures.
56
PART III – Enhancing conservation – development integration
Fishermen returning from fishing in Waza NP. After the reflooding, the number of fishermen increased
by a third, whereas the number of seasonal fishermen increased even more.
Kalamaloué National Park, where the disappearance of wildlife due to human encroachment is a bleak
perspective for nearby Waza NP.
The risks of a Conservation & Development project that is too successful
57
The difference that management planning can make
Practising a new fire policy in Waza NP. Traditionally, management plans focus on technical and ecological issues, which were also included in the Waza Management Plan.
Removing mud out of the
waterhole at Kalya, in the
centre of Waza NP. As long as
the flooding remains inferior
to the pre-dam period, the
maintenance of waterholes
will be an essential management activity in Waza NP.
Elephant poached inside Waza NP.
Also with increased local community
involvement in conservation, antipoaching remains an important activity because of the continuing presence
of well-armed commercial poachers.
58
PART III – Enhancing conservation – development integration
Triggered by Participatory Rural Appraisals in each of the villages around Waza National Park, a dialogue was started between local communities and national park authorities. This was an integral part of
the Waza management planning, building consensus on sensitive issues such as demarcation of the
national park and the exploitation of resources such as straw and Arabic gum in the park.
Meeting of the Waza NP committee, with representatives of park villages and transhumant pastoralists.
One of the most significant results of the Waza management plan was this platform where local communities are consulted on park management issues.
The difference that management planning can make
59
Building the new village of
Halé, outside the national park.
Whereas decades of oppression
towards Baram, the village in
Waza NP, only caused frustration, the new more balanced
policy of the ‘carrot and stick’
resulted in the voluntary movement of more than half of its
population to Halé.
Water pump that was installed at
Halé with the support of the WazaLogone project.
A new future.
View of Halé village
60
PART III – Enhancing conservation – development integration
The importance of developing capacities of protected area managers
Mweka wildlife college. The three African regional wildlife colleges together have trained more than
4000 protected area managers many of who are presently in charge of protected areas. Protected area
managers generally have little input in protected area planning and community conservation. Their lack
of capacity in these disciplines was not unique to Waza-Logone but was also identified as a major constraint for Integrated Conservation & Development Programmes elsewhere in Africa.
Examination at the Garoua Wildlife College. This francophone college has trained protected area personnel from over 20 West and Central African countries since 1970. Its curriculum was outdated however.
A new curriculum was therefore developed and implemented in the late 1990s. Our study analyses the
factors that influenced this curriculum reform in the college’s endeavour to respond to new developments in African wildlife management.
The importance of developing capacities of protected area managers
61
Trainees expressed their interest in field trips and exercises as the most effective type of training,
addressing skills and complementary to the predominantly knowledge-based classroom training.
Field training, counting wildlife in a protected
area. Planning and financial problems are the main
reasons why field trips rarely cover the quarter of
the training time they are programmed to occupy.
Botanical field training.
Computer exercises at
the Southern African
Wildife College. After
some initial hesitation,
computer science has
become an integral part
of the wildlife curricula
at the three African
wildlife colleges.
62
PART III – Enhancing conservation – development integration
Anti-poaching training for Waza National Park Guards. This was their first training for over a decade in
these essential skills.
Garoua students practising Participatory Rural Appraisal, a field exercise in the newly developed Community Conservation course. This study reviews the trainees’ evaluations of this course to appreciate its
relevance and support its further development. Their reactions suggest that protected area personnel
are not ‘attitude limited’ as often suggested. Their constraints to develop a more people-oriented work
style lie largely in the areas of knowledge and skills. These findings motivate increased efforts to implement training for protected area personnel in community conservation.
The importance of developing capacities of protected area managers
63
Exhausted after monitoring the transect from Zina to Tchikam and continuing to Waza
town (with Moussa Barka, October 1997). This three-day trip by foot during the flooding
season was my last with the Waza-Logone project.
2
The Ecological History of Waza-Logone:
Constructing a reference image for
floodplain rehabilitation
‘...the country, however, became exceedingly interesting and pleasant when we
reached one of the numerous water-courses of these African Netherlands...., an
open and clear river about seventy yards broad, which being fringed on each
bank with a border of slender deléb palms or kamelutu, in the clear, magnificent
morning sky afforded a most pitoresque view’
Barth (1857), the first European visitor to Waza-Logone on 13 January 1852
67
2.1
Historical periods and main ecological characteristics
Rainfall and flooding are the dominant factors, which trigger changes in vegetation composition and production (Chapter 3, 4), waterbird and large antelope
population sizes (Chapter 5, 6) and pastoral migration and livestock densities
(Chapter 7). They should be known in order to interpret the changes induced in
1994, subject of this thesis, in a wider, spatial and chronological/temporal context.
Below, I will show that our reference image of the ‘pre-Maga dam’ period when
the ‘floodplain was still intact’ was biased by the period from which information
was consulted. Changes in hydrology, land use, vegetation and especially wildlife
have often occurred already years earlier, but in several cases, the system’s resilience has led to their recovery. The intensity and speed of changes in the post
Maga dam period were, however, unprecedented in recent times, triggered by lower
than average rainfall and man-induced drought.
I distinguished eight historical periods, based on criteria such as rainfall, hydrology and the impact of human interventions and, above all, the quantity and quality
of available sources of information. Information sources prior to the 1930s encompassed indirect information from excavations, travel descriptions and oral history,
whereas scientific descriptions, contacts with scientists and local communities
constitute our more recent sources of information.
Rainfall data was collected at nearby N’djamena (Chad) from 1905 till 1914 and
from 1931 onwards (Figs. 2.1, 2.3). We therefore also have to rely on less detailed
information on climatic periods in the region, derived from Lake Chad water levels during the last millennium (Fig. 2.2), and other geological, palynological and
historical sources (Maley 1981; Nicolson 1986).
1 Prior to ±1820
Information sources
Sources of information on Waza-Logone prior to 1820 are essentially archaeological
studies and travel descriptions, hand-written manuscripts in Italian or Arabic,
which were translated and reinterpreted before being accessible (Hopkins & Levtzion 1981; Rauchenberger 1999). One often has the impression that these secondary sources are illustrative of the interpreter’s own perception and interest, making
it difficult to draw conclusions on the state of the environment.
Rainfall and hydrology
Wet and drought spells have occurred in Waza-Logone throughout the last millennium, their reported frequency is often a function of the quality of information
(Fig. 2.2). The 17th century was particularly wet and floods probably inundated the
entire floodplain (Zeltner 1997). From approximately 1700 onwards, information
Interventions
HIGH
LOW
2000
1980
1970
1960
1950
-500
0
500
Vegetation & wildlife
Communities/pastoralists
Waza NP management
change of park warden (Saleh)
Move of Baram (park village)
Waza Committee created
Management plan approved
3 -1
reflooding 2 (7-10 m s )
3 -1
2
reflooding 1 (20 m s ) enlarged with 180 km
start of vegetation monitoring
Change of park warden (Habibou)
Start WL-project
none
none
none
frequent crop raiding by Elephants
low Kob numbers
Civil war in Chad (1982-1987)
start of dam impact studies
extinction Waterbuck
emigration out of floodplain
Designation as Biosphere reserve (1979 or 1982?)
New park warden (Badjoda)
Park warden killed
first reflooding study (Wit & Grijs)
2
Maga dam
reduced with 1500 km
extinction Cheetah
end of FAO project
construction of upstream embankments (1971-1977)
limiting drainage into the Logone plain
road construction none
0 = long-term average
FAO assistance project
Ecole de Faune created
Waza NP created
Elephants destroying crops
crop protection unit created
park village Zeila burned
Independence of Cameroon
Flizot park warden (or earlier?)
‘first’ Elephants appearing
Part I – Introduction
1990
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1987
1986
1985
1984
1983
1982
1981
1980
1979
1978
1977
1975
1974
1973
1972
1970
1969
1968
1965
1964
1963
1962
1960
1959
1958
1957
1956
1955
1954
1953
1952
1951
1950
Flood
68
Rainfall
HIGH
LOW
1950
1950
Dragesco 1952
1947
‘first Elephants since memory in Waza’
1941
1938
1940
1936
1935
1932
Zina-Waza hunting reserve created
1930
no elephants observed
1929
1926
2 – Reference images
1944
Dieguere village created (Musgum)
1923
1920
1920
creation of Andirni and Badadaye villages (Arab and Fulbe)
1917
1914
1911
1913
major drought
1910
no elephants observed
rhino observed
creation of Mbili-2 by Musgum from Pouss
1908
1985
-500
0
500
1900
Rabah defeated / colonisation
rise in ivory commerce
conquest of Rabah, massacre in Mahé
rinderpest?
(park village)
0= average
?
1890
?
1880
?
1870
buffalo present
elephant dead, Niwadji
Zina brisk market for ivory
Anané large village
?
1860
?
1850
elephant prolific
69
Figure 2.1 – Historical chart, events in Waza-Logone from 1850 onwards, with periods of major upheaval indicated (cf. collapsing panarchies,
Chapter 12.4)
70
Part I – Introduction
t1
285 m
t2
t3
283 m
281 m
279 m
z1
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
1950
Year
t1, t2, t3: spilling into Bahr el Ghazal
z1: complete drying up of lake Chad
Note: siltation caused the rise of the average level of the lake bottom (280 → 281) during the last millenium.
Figure 2.2 – Levels of Lake Chad (adapted from Maley 1981, 1993).
is sufficiently detailed to suggest the occurrence of drought-wet cycles with a length
of approximately three-quarter of a century (Fig. 2.2).
Archaeological sources
Dwelling mounds, presumably man-made, are widely dispersed on the Waza-Logone floodplain. Some of them are at present still inhabited, such as the district
centre Zina and the villages bordering Waza NP in the East and North. It is assumed that all dwelling mounds used to be inhabited during wet spells, as they
were the last spots of non-inundated land in a sea of inundated land. Inside present-day Waza National Park pottery remains can be found on virtually all dwelling
mounds that witness the existence of (temporary) fishing camps. Several settlement
mounds have been the subject of archaeological surveys. Daima in nearby Nigeria
gives insight into the subsistence economy of its inhabitants during the last two
and a half thousand years (Connah 1981). The large majority of bones found were
from fish, suggesting a predominance of fish protein in daily consumption, as is
presently also the case. Remaining bones found were from birds and domestic cattle, a few from small stock, presumably goats, and a small number of wild animals,
including Bohor Reedbuck, Kob, Warthog, Gazelle and Elephant. It is likely that
sorghum was already grown from the earliest occupation, but remained for some
time below archaeological threshold visibility (Connah 1981).
Travel accounts
Amongst the first Arabian visitors to West Africa, Ibn Abi Zayd al-Qayrawani
(922-996) and Al-Maliki (around 1057), pronounced themselves against the trade
in elephants tusks: ‘It came from trading in ivory and I was loath to be involved in
something which had received adverse comments from men of learning’ (Hopkins
2 – Reference images
71
& Levtzon 1981). Yet 800 years later, increasing demand for ivory in Europe would
ravage the elephant herds in Waza-Logone. Leo Africanus passed the northern
part of the Waza-Logone area in the dry season of 1513-1514 (Rauchenberger 1999).
From the Lake Chad area he mentioned slaves as main export product but only
once noticed the export of ivory, in contrast with Dangala (Sudan) he visited and
described as home to renowned elephant hunters.
Human populations
The Sao, ancestors of present Kotoko (Lebeuf 1969), were predominantly sedentary and inhabited the various dwelling mounds in the floodplain such as present
day Mbili, Baram and Mahe in the north and eastern part of Waza National Park.
Fish was their predominant source of protein, probably supplemented through
hunting. After the wet 17th century (Fig. 2.2), the first nomadic Fulbe and Arab
Choa entered the area, diversifying the human exploitation by their pastoral way
of life (see also Chapter 7). It seems likely that with their arrival local breeds of Bos
taurus cattle, possibly related to the Kuri cattle still surviving in the Lake Chad
(CIRAD-EMVT 1997), were replaced with the now dominant Bos zebu breeds well
adapted to transhumance.
2 ±1820 – ±1900
Information sources
The first European travelers visiting the Waza-Logone area in the early 19th century found an area that had just undergone major changes with the arrival of Fulbe
and Arab pastoralists bringing, amongst other things, the islamic culture into the
area (Chapter 7). Denham & Clapperton (1826) travelled in the Lake Chad area
and visited the sultan of Logone Birni in February 1824, who, though with far less
authority, still reigns over the floodplain (Lebeuf 1969; van Est 1999). Unfortunately, Denham’s account provides hardly any environmental information of the
floodplain. Thanks to later travellers, especially Heinrich Barth (1857), detailed
first-hand descriptions of the floodplain environment became available. In March
1872, Nachtigal visited like Denham and Barth, Logone Birni and collected detailed,
yet mostly secondary information on the floodplain on the eve of the turmoil
caused by rinderpest and the conquest by Rabah that would change the area dramatically.
Rainfall and hydrology
The 19th century had both droughts and wet spells, ending the 19th century as it
started with a relatively wet period (Fig. 2.2).
Vegetation
Descriptions of Barth (1857) suggest that the levees along watercourses were much
more densely wooded than nowadays. The same plant species seem to have been
72
Part I – Introduction
dominant (Celtis africana, Ficus spp, Tamarindus indica), although some Sudanian
savanna tree species, rather scarce in the area nowadays, were mentioned as well
(Vitellaria paradoxa). The vegetation of the lower parts of the floodplain was largely comparable to the present one, although explorers paid little attention to its floristic composition.
Birds and large mammals
Birds did not attract the attention of most European explorers and their observations
are therefore limited to either conspicuous species as Ostrich or to abundant and
frequently eaten species such as Guineafowl (Barth 1857).
Nachtigal, who traveled several years in the Lake Chad Basin, remarked on his
passing through Sudan in 1874: ‘When one recalled the large number of antelopes
which the traveller encounters on all sides in the regions of Bornu, even in the
neighborhood of inhabited places, the difference was astonishing’ (Nachtigal 1987).
Even today, densities of larger mammals seem to be higher than in most other parts
of the Sahel (Scholte & Hashim 2005). Of these large mammals, Giraffe attracted
the attention of Barth (1857) who mapped these observations, showing the early
importance of present Waza NP). Barth (1857) frequently mentioned Hippopotamus from the floodplain, where it is now reduced to Lake Maga and its direct
environs.
Elephant is amongst the most conspicuous wildlife species of Waza-Logone, easily
traceable by the deep prints they leave in moist soil, remaining well visible till the
next rainy season. Elephant prints are still the fear of any driver and one of the
horses with which Barth’s company travelled fell and his rider broke his arm.
Without doubt the interest of the explorers for Elephants can also be explained by
the increasing demand for ivory during the 19th century. UK ivory imports from
Tripoli, through which most ivory from Central-West Africa was transported by
camel, shot up in the early 1850s to some 70 tons a year remaining fairly constant
and only reducing at the end of the 1890s (Johnson 1978). Barth (1857) described
several encounters with Elephant herds in Waza-Logone and mapped some observations and prints. Barth and his company frequently ate Elephant flesh, said to be
the most commonly available. During the passage of Nachtigal, Elephant hunting
had become so important that he described Zina as a ‘brisk market for ivory’ (Nachtigal 1987). The slaughter of Elephants was also reflected by the naming of the village of Niwadji that, created in the late 19th century south of Waza NP, signifies
the ‘place of a dead Elephant’.
Human populations: turmoil
Whereas the first part of the 19th century passed under relatively stable conditions,
the second part of the 19th century was marked by continuing slaving raids, rinderpest, the devastating passage of the army of Rabah and ended with the colonisation
2 – Reference images
73
of the entire sub region (Zeltner 1997, 1988). Apart from the resulting human tragedy, these developments have had a profound impact on the ecology of the area,
still visible today.
From 1890s till deep into the 20th century, northern Cameroon was hit by rinderpest that wiped out entire herds of livestock, not only upsetting the lives of pastoralists, but the entire agro-ecology of the area (Beauvilain 1989). Losses of wildlife
due to rinderpest have only been reported in the second part of the 20th century,
but it is assumed that it was responsible for the start of the decline of African Buffalo. ln Sudan, a link was laid between rinderpest and increased Elephants hunting to reconstitute the depleted cattle herds (Cunnison 1960).
The spreading of rinderpest was triggered by the movements of people prior to
and with the arrival of Rabah’s army that swept from East to West through the
Lake Chad Basin and reached Waza-Logone in 1893. Rabah’s ‘Sudanese’ or ‘Arab’
soldiers targeted especially Kotoko villages of which 33 out of the 43 in the Logone
sultanate were completely destroyed. In Mahé, a small village on the northeastern
corner of Waza NP, 60 adult men were killed, nearby Garlé lost 20 adults and its
remaining population was driven in slavery (Beauvilain 1989). Anané was mentioned as an important village by Nachtigal (1987), whereas there remains at present only a waterhole surrounded with some Balanites aegyptiaca trees in the Acacia
seyal zone of Waza National Park. The army of Rabah lived mostly off looting (Zeltner 1988), and combined with the increasing availability of firearms may have
contributed to the decline of several large mammals, most easily detectable by the
Elephant population dynamics (see below).
3 ±1900 – 1950
Information sources
Scattered reports by European travelers appeared, some of them with considerable
zoological details (Herzog zum Mecklenburg 1912; Jeannin 1936; Zwilling 1940),
yet none with the genuine observation quality of the earlier travelers Barth and
Nachtigal. Our PRA-studies and other oral history surveys (Chapter 9; van Est
1999) clarified much of the local history.
Rainfall
The first part of the 20th century has known both droughts, around 1913 and 1934,
and a relatively wet period from ± 1936 till 1945 (Figs. 2.1, 2.3).
Birds and large mammals
Bates (1927) was the first ornithologist to publish his observations of his passage
through the Waza-Logone area on his trip to Lake Chad. None of his observations
has been included in the checklist of the area due to the lack of accuracy on obser-
74
Part I – Introduction
vation locations (Scholte et al. 1999). Gide (1927-1928) and Allegret (1993) who
travelled in 1927 downstream the Logone river, described in passing a conspicuous
avifauna with abundant Black-crowned Cranes and ducks. No species were mentioned which do not exist today anymore and their description gives an ‘impression’ rather similar to the one that is presently available.
Jeanin (1936), a French North-Cameroon-based veterinarian, wrote the first account
on the mammals of Waza reserve, and commented on their decline since 1900.
For some larger animals his descriptions are sufficiently detailed to allow a semiqualitative comparison with later estimates. Giraffe for example, had declined from
1900 onwards when groups of 30-40 Giraffes could still be observed (like at present), to a maximum groups size of 15 in 1935. The entire reserve was estimated to
number 100 Giraffes only, compared with 1200-1500 in the 1970s and 1990s
(Esser & van Lavieren 1979; Tchamba & Elkan 1994). Early 1900, African Buffalo
Syncerus caffer used to be common in the floodplain: an inhabitant of Zina was said
to have hunted over 100 specimens during flooding when animals concentrated
on the non-inundated mounds. In 1935, African Buffalo was already rare in the
area (Jeannin 1936). Jeannin (1936) observed Korrigum (‘Topi’) herds of 100-150,
occasionally up to 250, like the present situation (see Chapter 6). In 1935, however,
Korrigum’s distribution spanned the entire area between the Mandara Mountains
and the Logone river, whereas nowadays Korrigum is largely restricted to Waza
NP and the bordering floodplain. Kob was observed in ‘groups of 5-50 individuals’
and ‘near the villages of Mahé and Baram hundreds can be observed in a morning’. Its population was under heavy hunting pressure, however. Once the area was
gazetted as hunting reserve, numbers of Kob increased dramatically to the tens of
thousands that could be observed the 1950s-1970s (see Chapter 6).
Herzog zum Mecklenburg (1912) did not observe Elephants on his journey through
present northern Cameroon in 1910-1911 and remarked that Elephants’ preferential habitat was ‘with sufficiently large trees in steppe land with drinking places’.
He added that former Elephant hunters around Lake Chad had changed their way
of living because of the lack of Elephants. A decade later, Zwilling (1940) did not
observe Elephants in north Cameroon either. Twenty years later Jeannin (1936)
described the presence of a few Elephants in Northern Cameroon, but not in the
Waza-Zina reserve. Flizot (1948), the first Waza National Park warden, described
that in 1947 one single Elephant roomed in the Mandara Mountain area, with Elephants increasing in the Benoue valley further south due to supposed immigration from Nigeria (Flizot 1948). The number of Elephants in Waza National Park
increased from nil in the 1940s to 50-80 in 1954 (Robin 1954) to 600 in 1968 (Flizot 1968) and over 1000 in the 1990s, mostly as the result of immigration from
Chad (Tchamba 1996).
75
2500
wet
wet
d
average
d
a
dry
w
Rainfall with:
2000
increasing line > average
horizontal = average
1500
decline < average
no flooding
1000
500
1998
1994
1990
1986
1982
1978
1974
1970
1966
1962
1958
1954
1950
1946
1942
1938
1934
0
1930
cumulative deviation of mean annual rainfall (mm)
2 – Reference images
-500
Figure 2.3 – Cumulative rainfall 1930-2001
Human populations: settling
The relative security that arrived with the colonial rule triggered from the early
1900s onwards the migration of Musgum groups into the Logone floodplain. Rich
fishing grounds and fertile soils for agriculture were the principle attractions.
Temporary fishing camps developed into settlements on the dwelling mounds
deserted after the turnoil of the previous decades. In several cases the Musgum settled in Kotoko villages, but created slightly separate quarters. The arriving Musgum conformed themselves to the authority of the Kotoko and their leaders till
deep into the 20th century (van Est 1999).
Fulbe and Arab transhumant pastoralists settled increasingly at the sandier southern edge of the present Waza National Park. Most of these villages, such as present-day Badadaye, Andirni and AmKodje were already used as transitory camps
because of the presence of water in their immediate surroundings.
Management
The ‘Waza-Zina Hunting Reserve’ was created in 1935 (CCTA 1953). Discussions
in villages, with the sultan of Logone Birni and with Badjoda (former park warden)
stressed the tensions in the negotiations of colonial authorities with the sultan of
Logone Birni on the creation of the reserve. This resulted in a status-quo, lasting
till present, of the presence of villages right on the border on the north and east
side of the (present-day) national park (Chapter 9). In contrast, the lamido of
Pette, reigning over villages on the southern part of Waza NP, agreed on the des-
76
Part I – Introduction
ignation of the hunting reserve and assisted in the retreat of villages from the borders, creating the only de-facto buffer zone (see Chapter 9).
4 The 1950 and 1960s
Information sources
After the Second World War, there was an increasing attention of French agricultural and veterinary officers for the Waza-Logone area, studying vegetation (Vaillant
1956; Letouzey 1968), wildlife (Robin 1954; Flizot 1962) and birds (Dragesco 1960,
1961).
Rainfall and hydrology
The 1950s and 1960s were characterised by abundant rainfall and floods (Figure
2.3). The hydrological situation of Waza NP described by Flizot (1962) illustrates
the recent changes in environmental conditions (the post 1979 situation): Water
has completely retreated at the end of February or March (December-January),
afterwards the dry herbs are being burned and the subsequent regrowth forms an
excellent pasture for the herbivorous animals. Approximately twenty water holes
used to contain water till the first rains in May (March). The Yaérés give room to
the entire fauna from mid March till June (from December onwards), with notable
exception from Elephant and Giraffe, who remain dependent on the forage sources
in the Acacia seyal zone (no changes).
Vegetation
Letouzey (1968) in his general account on the vegetation of Cameroon, described
a dominance of Echinochloa pyramidalis, with Vetiveria nigritana as second and
Oryza barthii (= O. longistaminata) and Hyparrhenia rufa as co-dominant. He further indicated an inundation depth of 1 to 2 metres (<1 metre) and gave a list of a
number of species, which can still be found in association with before mentioned
species, namely Sorghum arundinaceum, Dinebra retroflexa, Pennisetum ramosum
and Ischaemum afrum.
Birds and Large Mammals
Dragesco (1960, 1961) reported his, mostly qualitative, ornithological observations,
some of species (eg. African Skimmer Rynchops flavirostris) which have not been
observed in recent times. Behind Waza camp, Dragesco observed Black-crowned
Cranes Balearica pavonina pavonina, whose numbers he estimated in December
1958 at over 5000. On the same spot I observed them regularly during different
periods in 1995-2000, yet never counted more than 1000.
In 1954, the first estimates on large mammals in Waza NP were published, probably based on ground total counts by its warden Flizot (Robin 1954, see also Chapter 6). The 1962 total ground count by Flizot (1962) showed an increased number
77
2 – Reference images
of large mammals compared to the earlier reports by Jeannin (1936) probably
because of the above-average rainfall and impact of protection measures.
Human populations
When we discussed the former floodplain conditions with people, they often referred, with nostalgia, to this period. In 1960 Cameroon became independent and
in subsequent years the state was further consolidated. It was not only peace and
prosperity however. Zeila, Garle and other Kotoko villages that had rallied against
the ruling party, were burned down and moved to neighbouring villages and towns.
Zeila, inside Waza NP, has remained a deserted dwelling mound with its occasional lion pride. Badadaye, a Fulbe village, moved further from the Waza National
Park boundary.
Management
Tourist numbers to Waza NP rose steadily up to 4000 persons per year, and made
Waza amongst the best-known West-Central African protected areas. The number
of park guards rose during the 1950s and 1960s steadily till fifteen, reflecting the
increasing presence of the state in the Waza-Logone area (Fig. 2.4). Apart from this
governmental personnel, Flizot appointed young men as so-called park guides in
Waza and Andirni village near the main entrances of Waza NP. Their main task
was to guide tourists inside the park for which they received a modest remuneration. In return for their exclusive position, park guides assisted park guards in
incidental tasks such as anti-poaching campaigns and were held to inform the
park authorities on any infraction in their area.
number of park personnel
50
bar: park guards (armed)
40
+ village
guards
(unarmed)
line: park guides (unarmed)
30
20
10
Figure 2.4 – Number of Waza National Park guards
2002
2000
1998
1996
1994
1990
1988
1985
1982
1980
1978
1974
1970
1965
1962
1953
0
78
Part I – Introduction
5 1970 – 1979
Information sources
In the 1960s-1970s, wildlife conservation had captured the personal interest of
the first president of Cameroon who guided several of his international visitors to
Waza National Park where he possessed a lodge. The president’s interest in wildlife
conservation was also expressed by the choice of the northern capital of Garoua for
the regional francophone Wildlife College that was created in 1970s (Chapter 10).
Staff and students of the Garoua Wildlife College would play a crucial role in the
initiation of the first substantial ecological studies of Waza National Park. In the
1970s, a Garoua based FAO project supported the development of national parks,
in particular Waza and Benoue NP for which several base-line studies would be
carried out (Vanpraet 1977).
Rainfall and hydrology
The 1970s were characterised by the great Sahelian droughts of 1973-75. These
droughts coincided with the construction of the Waza-Kousseri road, that cut off
some areas of flooding of the Logone river as well as the ephemeral streams that
drain the northern part of the Mandara mountains into the Waza area. At the same
time, parts of the Logone upstream of the later Lake Maga were embanked, on
both the Cameroonian and Chadian sides, further reducing flooding.
Vegetation
Staff of the Garoua Wildlife College conducted several studies on the vegetation of
Waza NP and produced Floras on woody species and grasses (Geerling 1982; van
der Zon 1989). Throughout this thesis, I refer to the discussions with Garoua ecologists whose experiences have been used to reconstruct a qualitative image of the
former vegetation (photographs and personal communications of Geerling, van
Lavieren and Wit). Vegetation maps of respectively the entire area and Waza NP
were produced based on 1:50.000 aerial photographs from the mid-1960s (Gaston
& Dulieu 1976; Wit 1975). In the description of Gaston & Dulieu (1976), Hyparrhenia rufa was indicated as dominant in the western parts of the floodplain with
Vetiveria nigritana as second only. Amongst the communities of the wetter parts
of the floodplain, the lack of Echinochloa pyramidalis was striking. Their description differed from Wit (1975), who distinguished five plant communities for the
floodplain of Waza NP:
1 Vetiveria nigritana community which was the most common type of grassland
(observations confirmed by our discussions with pastoralists);
2 Sesbania pachycarpa community found on the slightly sandier parts of the
floodplain (area of Zeila and Baram, NE Waza NP);
3 Hyparrhenia rufa that was restricted to ridges along drains near Tchede and
Lougouma, contrary to Gaston & Dulieu (1976). Since their field work was car-
2 – Reference images
79
ried out in subsequent years, this difference is probably due to different locations the authors have visited;
4 Echinochloa pyramidalis vegetation community was found to be locally dominant, especially on the Acacia seyal side of the plain, subject to less inundation,
with large numbers of dead Vetiveria nigritana tussocks;
5 Forb communities: Just east of the Acacia seyal zone, on the relatively higher
parts of the floodplain, a vegetation community without grasses was found.
Vetiveria nigritana used to be common as well, given the fact that numerous
dead tussocks were found.
Birds and large mammals
Waza-Logone was frequently visited in the late 1960s and early 1970s by mostly
Nigerian-based ornithologists, coinciding with the construction of a tourist lodge
(Fry 1970; Broadbent 1971; Greling 1972ab; Mundy 1972; Holmes 1972, 1974;
Pettet 1976; Kavanagh 1977). Their visits were limited to the more accessible western part and lasted a few days only. Only Greling (1972a) carried out a longer study
of bird densities in the Acacia seyal shrubland and of the birds of the wooded
savanna of the southern part of Waza NP. Thiollay (1978) was the first ornithologist who reported the ornithological significance of Waza, based on his systematic
raptor observations. Louette (1981) compiled most of these early observations from
Waza-Logone in the first checklist of birds of Cameroon.
The staff of Garoua Wildlife College (see below) initiated wildife studies in Waza
NP including the first published systematic large mammals census of Waza NP
(Esser & van Lavieren 1979, see Chapter 6), as well as an ecological study on Giraffe
(Ngog Nje 1983, 1984).
Human populations
The first Sahelian drought after a long period of good rainfall struck hard not only
in Waza-Logone, but especially further North in Northern Nigeria and Niger. Increasing land pressure caused a steadily influx of pastoralists from these areas into
Waza-Logone.
Management
The designation of the Waza-Zinah Faunal reserve into Waza National Park in 1968,
a few years later followed by the construction of a tourist camp near Waza village,
triggered a rapid development of the park’s management infrastructure. Additional park guards were recruited and their number rose steadily till the mid-1980s
(Fig 2.4). In the early 1970s, the number of tourists rose to 5000-6000, remaining constant till the early 1990s (Fig. 2.5). The number of park guides increased
from 4 to over 20 (Fig. 2.4).
Part I – Introduction
8000
number of visitors
6000
4000
2000
no
information
available
0
19
66
19
68
19
70
19
72
19
74
19
76
19
78
19
80
19
82
19
84
19
86
19
88
19
90
19
92
19
94
19
96
19
98
20
00
20
02
80
Year
Figure 2.5 – Number of visitors to Waza National Park
6 Post Maga dam period, 1979 – 1993
Information sources
Garoua Wildlife College continued to play an important role in the initiation of
ecological studies in Waza-Logone, joined in the mid-1980s by Leiden University
through its field station with the University of Dschang, CEDC.
Rainfall and hydrology
The Maga dam construction in 1979 proceeded a relatively long period of belowaverage rainfall, that some called the worst drought in the area’s history. From
1987 till 1989 rainfall improved and subsequently deteriorated somewhat (Fig. 2.3).
1
Vegetation
In the early 1980s, staff of the Garoua Wildlife College reported that Vetiveria nigritana was dying massively in the western part of the floodplain, especially inside
Waza NP. This motivated the start of vegetation studies in 1984-1985 by Cameroonian and Dutch students, under the supervision of Carel Drijver (Leiden University) and Ton van der Zon (Ecole de Faune-DGIS). They installed a transect from
the centre of Waza National Park to Zina in the heart of the floodplain and included a grid near Tchikam in the centre of the no longer flooded floodplain.
1 For more specific information on vegetation, see Chapter 3.
2 – Reference images
81
Birds and large mammals
Contrasting the interest expressed during the 1970s, the only ornithological studies undertaken in the 1980s were aerial waterfowl counts on the newly created
Lake Maga and other wetlands of Far North Cameroon (Roux & Jarry 1984, 1986,
1987) as well as a terrestrial count of Ruff Philomachus pugnax on the rice fields
of Maga by OAG Münster (1991). In the early 1990s, Njifortii (1997) conducted a
Ph.D. study on the potential of harvesting Guineafowl that has not had an implementation so far.
Crop damage by the expanding Elephant populations attracted more and more the
attention of authorities and became a political issue in the 1990s, making the
provincial head of wildlife say: ‘when the Elephants are out it is the minister of
environment who calls me, the moment Elephants are back into the park, I have
to call him for any discussion.’ The CEDC initiated a Ph.D. study on the migration of Elephants during the rainy season out of Waza NP towards the south, and
during the dry season towards Kalamaloué NP (Tchamba 1996). Forage scarcity
inside Waza NP was assumed to be the main reason for this migration, possibly
enhanced by the lack of flooding, although it remains speculating on the exact reason (Thouless et al. 1995).
Human populations
To counter some of the negative impacts of the Maga dam construction the Cameroonian government mobilised a series of development interventions in the area.
Village waterpumps and large pastoral drinking ponds were installed during this
period. Nevertheless, the Maga dam construction, followed by the severe 1980s
drought triggered a considerable out migration of the floodplain, of sedentary as
well as of mobile population groups. The newly created Lake Maga and Lake Lagdo
further south attracted fishermen from Waza-Logone. Some pastoralists moved
into neighbouring Chad, from where even larger numbers of people left because
of the long lasting civil war.
Management
Following the rapidly degrading situation in Waza NP due to the construction of
the Maga dam, the Government of Cameroon launched the ‘Sauvetage Waza’ campaign, digging a large number of waterholes. Although not addressing the real
cause of the problem, the reduced flooding, nor its major consequence, the reduction in forage resources, these water holes have played an indispensable role in the
management of Waza’s wildlife during the 1980s-1990s droughts. Outside Waza
NP, comparable water holes were dug for livestock.
In the mid-1980s, the Cameroonian government recruited additional park guards,
of who most were on duty when we provided their first training course in 1996.
Despite the reduced flooding and the virtual collapse of several of its wildlife pop-
82
Part I – Introduction
ulations, Waza NP continued to attract large numbers of tourists (Fig. 2.5). In 1979
or 1982 Waza NP was designated as Biosphere reserve (Batisse 1992), although
this has remained a designation on paper only with no consequences for its management (Chapter 9).
In 1985, a small dam was constructed by the livestock department to raise the
flooding level in an area downstream of Zilim, north of Zina (Fig. 1.2), which was
2
still relatively well flooded, of approximately 400 km . Studies were undertaken to
assess the state of its vegetation, livestock keeping and fisheries after the intervention. Although no reference existed from prior to the intervention, positive
effects were observed, as well as increased conflicts between fishermen and pastoralists. This experience would inspire the formulation of the Waza-Logone project (Drijver & Kouahou 1995).
7 Pilot reflooding and consensual park management planning, 1993 – 2003
The subject of this thesis; the reader is referred to Chapter 1 for an introduction to the
research questions and discussions of this study. Below I emphasise the links with the
topics discussed in this overview.
Information sources, the rationale of pilot reflooding
In 1994, an initial study into the potential of reflooding concluded that only by pilot
testing the impact of reflooding could be assessed (Wesseling et al. 1994). This
pilot testing should include hydrological monitoring (the extremely flat area made
predicting rather difficult), ecological-monitoring and the monitoring of land use
recovery, integrated into a cost – benefit analysis of the various reflooding options.
The pilot reflooding was initiated in 1994 by breaching an embankment that had
blocked a former watercourse, the ‘Petit Goroma’ (Fig. 1.2). Through this newly
3
opened watercourse an additional flow of approximately 20 m /s streamed into the
floodplain. In 1997 another watercourse, the Areitekele, located just south of the
3
Petit Goroma was opened, triggering a further 8-10 m /s inflow into the floodplain.
Rainfall and hydrology
In 1994, the first year of reflooding, rainfall was favourable, followed by three years
of below average rainfall. From 1998 onwards rainfall has been above-average (Fig.
2.3). The 1994 flooding was abundant, more than could have been expected from
the rainfall, strongly setting everybody’s mind on the impact of the breaching of
the embankment. The reflooding was strongly and unanimously appreciated by
people from the Waza-Logone area, even though its strength led to the destruction
of crops and houses.
2 – Reference images
83
Vegetation
From 1993 onwards, vegetation monitoring was re-activated by the Waza-Logone
project in the Waza-Zina and Tchikam transect which were earlier used in 19841985. Starting prior to the pilot reflooding, floristic composition was also monitored in over 100 plots in a grid that was situated in the centre of the pilot release
impact zone (Fig. 1.2). Vegetation production was assessed as a function of flooding depth (Chapter 4).
Birds and large mammals
Despite the previous, mostly isolated ornithological publications, Waza-Logone
was poorly known for its rich birdlife. Based on the 1970s studies and our own
fieldwork, we elaborated a checklist with 379 bird species, more than most other
areas in West-Central Africa, attributed to the area’s varied habitat and study
intensity (Scholte et al. 1999). Systematic waterbird counts were initiated in 1993
(Wetten & Spierenburg 1998) and followed up by the Waza-Logone team in subsequent years (Chapter 5). The resulting, quantitative information motivated the
inclusion of Kalamaloué and Waza National Park, the Logone floodplain and Lake
Maga each as Important Bird Area (Fotso et al. 2001). The Cameroonian Government also presented Waza National Park as the country’s first Ramsar site.
Human populations
The economic crisis that started in the late 1980s led to the devaluation of the local
currency in 1994 with an impact on the functioning of central and local governments. Although Cameroon has remained a rather stable country in a troubling
region, lack of security became a major problem in the Waza-Logone area (Chapter 7). Interventions of the Waza-Logone project and its offshoots encompassed
besides the reflooding, small scale development interventions (Chapter 9).
Management
The recruitment stop into the Cameroonian civil service from the early 1990s onwards, said to be dictated by international financing bodies, had serious consequences for the management of Waza and other national parks in Cameroon.
Most of their park guards were recruited in the late 1970s and mid 1980s, and either
approached the age of retirement in the late 1990s or became eligible for mutation to their region of origin. Although Waza NP had a reasonable number of park
guards in 1990, in the following years their numbers would drop dramatically and
at the end of the 1990s, Waza NP was back to its early 1960s protection level (Fig.
2.4).
Park guides, appointed to guide tourists and researchers, are not armed nor formally trained for anti-poaching and park outreach activities. Yet their commitment
and continuous presence in villages on various corners of the park has made them
over the years indispensable for the protection of the park, first complementary,
84
Part I – Introduction
but due to shrinking number of park guards, more and more as the only representatives of the park authorities. This experience motivated the initiative of the
appointment of so-called village guides who play a more formal role in patrolling
the park. It should be stressed however, that true anti-poaching control can only
be carried by armed and properly trained park guards.
8 Envisaged large scale reflooding
In Chapter 12, the consequences of the present study for envisaged large scale reflooding
are discussed. Below I address generalities of the proposed reflooding options.
Rationale
Large scale reflooding aiming to rehabilitate the quasi-totality of the Logone floodplain, was initially programmed to take place in 1998, postponed till 2000 and has
at the moment of writing not started. The Waza-Logone project distributed a project document, summarising the pilot reflooding results, including some of those
presented in this thesis, to attract the necessary 12 million euro (IUCN 1999).
Despite the economic efficiency, as suggested by the various versions of the CostBenefit analysis (IUCN 1999; Emerton 2004), no follow-up has been given by any
of the contacted donors.
Information sources
The impact study of the reflooding was mainly based on systematic monitoring.
Unfortunately some of this monitoring infrastructure is no longer in place due to
lack of maintenance, allowing in future, without considerable efforts, only qualitative observations.
Rainfall and hydrology
Three reflooding options were described, amongst which the maximum one totals
2
640 km that includes not only flooding from the East (direction Logone river) but
also through the Mayo Vrick (spillway Lake Maga) with an impact also in the western floodplain (IUCN 1999).
Management
2
Originally the 180 km reflooding, subject of this thesis, was aimed as a pilot to
develop insights on the impact of large scale reflooding on the hydrology, ecology
as well socio-economics of the floodplain. Results of this study will show that the
integration of these subjects into management recommendation to mitigate potential conflicts has proven to be an even more important challenge.
2 – Reference images
2.3
85
Reference and target image of the Waza-Logone reflooding
Throughout this book I discuss floodplain rehabilitation against the background
of a reference image of the area that we constructed in 1993 at the start of the
Waza-Logone project (Wesseling et al. 1994). This, in our initial perception, ideal
situation served not only as source of inspiration (Lenders et al. 1998), but also
seemed, in retrospect with some naiveté, to be an achievable goal. This reference
image was formed by impressions from above-mentioned information sources
from the 1960s and 1970s. Simplified, our target image for a completely restored
floodplain was characterised by a dominant cover of perennial grassland vegetation with Echinochloa pyramidalis and Vetiveria nigritana on the plains (Wit 1975),
abundant wildlife like in the 1960s and early 1970s e.g. with 10 000 Black-crowned Cranes (Dragesco 1960, 1961, Chapter 5), 25 000 Kobs and 20 000 Korrigum
(Fizot 1962; Chapter 6), both in and outside the national park, and prosperous
pastoralists and fishermen with functioning (conflict resolution) management
institutions (Chapter 7; van Est 1999).
The ecosystem as described in the 1960 and 1970s was, however, the result of a
long period of above-average rainfall (Fig. 2.1). When considering the time-span of
the last millennium, the 1960s reference image has been just part of the returning wet-drought spell cycles (Fig. 2.2). The severe rainfall deficiencies that the area
experienced in the 1970s and especially in the 1980s, after the Maga dam construction, was by no means exceptional, neither in recent history (1913 drought)
nor during earlier periods (Fig. 2.2).
Spells of wet and drought periods are characteristic for the Waza-Logone ecosystem, to which its plant and animal species are thought to be well adapted. This
ecosystem resilience became apparent with the recovery of antelopes and Giraffe
between 1935 and 1960 after lifting much of the hunting pressure (Jeannin 1936)
and assisted by above-average rainfall. The resilience of the Elephant populations
was even more spectacular. Hunted to local extinction in the late 19th century and
absent from the Waza-Logone area till the 1940s, Elephant has recovered to such
an extent that it is now locally considered almost a problem animal (Tchamba
2
1996). Yet the extinction of animal species during the early 20th century, most
notably Western Black Rhino (Diceros bicornis longipes) and Painted Dog (Lycaon pictus) followed by African Buffalo, Cheetah (Acinonyx jubatus), Leopard (Panthera pardus) and Red-flanked Duiker (Cephalophus rufilatus) in the 1970 and 1980s (Chapter 6) may be more difficult to be reversed. The spectacular return of Waterbuck,
observed in 1998 and 2003, shows the difficulty to make such predictions though.
2 One should realise that the concentration of Elephants in Waza-Logone, is the result of pressure in
almost the entire Lake Chad Basin and a rather artificial situation.
86
Part I – Introduction
Changes in wildife and livestock densities have had an impact on vegetation as
well. The late 19th century extinction of the Waza-Logone Elephant population,
concurring with the drop in numbers in cattle and several Antelope species due to
the ravaging rinderpest, may have created a window of opportunity for the regeneration of trees, in analogy to the situation in East Africa (Prins & van der Jeugd
1993). The Sclerocarya birrea woodlands of southern Waza and the Balanites aegyptiaca and Tamarindus indica woodlands on the dwelling mounds in the floodplain
can possibly be traced to this period. Although still speculative, this would explain
the lack of regeneration of these woodlands that has preoccupied ecologists working in the area since the 1960s (Flizot 1968; Vanpraet 1977; MINEF/IUCN 1997).
The disappearance of Hippopotamus (Hippopotamus amphibius) from the Logone
floodplain, still present till the end of the 19th century, may have had an important impact on the floodplain environment. In the somewhat comparable Okavango
delta, hippos create a network of small channels between their day-time retreat
and non-inundated grazing grounds causing diversion of water and sediment to
the backswamp areas (McCarthy et al. 1998).
The floodplain rehabilitation hypotheses, postulated in Chapter 1, were based on
above presented reference and target images. During the 1993-2000 study period
these images were gradually developed as discussed in more or less detail in each
of the following chapters. In the syntheses, Chapter 12, I reformulate these floodplain rehabilitation hypotheses based on the overall study results, leading to new
floodplain rehabilitation objectives.
PART II
The Im pact of Ref looding in
Waza-Logone
3
4
5
6
7
Impact on Veget at ion
Floodplain Rehabilit at ion in Nort h Cameroon: Impact on veget at ion
dynamics
M aximum Flood Dept h Det ermines Above-ground Biomass in Af rican
Seasonally Shallow ly Flooded Grasslands
Impact on Wildlif e
Wat erbird Recovery in Waza-Logone (Cameroon), result ing f rom
increased rainf all, f loodplain rehabilit at ion and colony prot ect ion
Ant elope Populat ions in Waza Nat ional Park (Cameroon) f rom 1960
t ill 2001: Impact of changes in rainf all, hydrology and human pressure
Impact on Past oralist s
Past oralist Responses t o Floodplain Rehabilit at ion in Nort h Cameroon
3
Floodplain Rehabilitation in North
Cameroon: Impact on vegetation
dynamics
Paul Scholt e, Philippe Kirda, Saleh Adam and Bobo Kadiri
Updat ed version, including 1998, 1999 and 2002 dat a, f rom Applied Veget at ion
Science 3: 33-42, 2000.
Abstract
Since t he const ruct ion in 1979 of a dam in t he Logone f loodplain in t he SaheloSudanian zone of Cameroon, annual inundat ions have decreased, reducing
perennial veget at ion as import ant grazing f or nomadic herds and w ildlif e during
t he dry season. Present ly, possibilit ies exist t o release excess w at er f or f loodplain
2
rehabilit at ion. In 1994, a pilot release w as execut ed ref looding 180 km t o verif y
predict ed advant ages. Veget at ion has been st udied f rom 1984 onw ards along a
t ransect covering f looded, recent ly ref looded as w ell as desiccat ed part s of t he
f loodplain. Since 1993, t he f lorist ic composit ion has been monit ored in a grid in
t he cent re of t he impact zone. Cover of perennial grasses, most not ably Echinochloa pyramidalis and Oryza longist aminat a increased again f rom 41 t o 75% in
t he ref looded zone. Vet iveria nigrit ana, a t ussock grass t hat used t o be dominant , disappeared slow ly af t er t he dam const ruct ion and has not show n a comeback. The cover of annual species, most not ably Sorghum arundinaceum, a dominant annual grass only since t he mid-1980s, decreased in t he ref looded zone
f rom 58% t o 23% . If t he observed conversion rat e of annual int o perennial
grassland is ext rapolat ed, a recovery t ow ards a ‘100% ’ perennial st at e w as most
likely reached af t er t he 2003 f looding season, as indicat ed by qualit at ive observat ions in 2002. Apart f rom f avourable climat ic condit ions, recovery t o t he predam sit uat ion might be dependent on t he rest orat ion of soil f ert ilit y, limit ing an
approach f ocusing on f looding dept h only.
91
3.1
Introduction
Seasonally flooded grasslands are amongst the most productive ecosystems known
in sub-Saharan Africa (Denny 1985). They are essential habitats for water birds
and can harbour large concentrations of antelopes. Many people have adapted to
life on the floodplains and are exploiting them with increasing intensity, competing with wildlife. Extreme cases of competition occur when the water supply is cut
off for irrigation or hydroelectric purposes, causing abrupt environmental changes
due to desiccation and lack of silt input (Breen et al. 1988; Middleton 1999). After
widespread food shortages in the early 1970s, such water control projects were undertaken in almost all major river systems in sub-Saharan Africa (Drijver & Marchand 1986). Following a change in general attitudes in which the value of downstream, small-scale human activities as well as of wildlife is recognised, more diversified approaches of river basin development are currently being developed (Adams
1996). In several river systems possibilities exist for the controlled release of water
to deliberately flood downstream areas, potentially rehabilitating the degraded environment and restoring its functions. Such controlled flood releases have been tried
on the Phogolo river in South Africa (Bruwer et al. 1996), in the Senegal valley (Hollis 1996) and in the Hadejia-Nguru wetlands in Nigeria (Polet & Thompson 1996).
They are reported as being ‘successful’, but remain ecological ‘black boxes’ with
available information restricted to water flows (in) and local production increases
(out).
In 1979, the Waza-Logone area has followed the fate of many African floodplains,
with the construction of a dam, forming Lake Maga, and an embankment along the
Logone river, as part of a parastatal rice scheme (Fig. 3.1). In combination with
lower than average rainfall for the last two decades (Beauvilain 1995; Fig. 2.3), this
dam has led to a reduction in depth and extent of flooding over an area of 1500
2
km . Previously productive perennial grasslands were replaced by stands dominated by annual-grass dominated stands, limiting regrowth in the dry season and
reducing the carrying capacity for wildlife and livestock (Scholte et al. 1996a,b). A
pilot reflooding was started in 1994 by the opening of a watercourse, formerly
closed by the embankment along the Logone River, reflooding an area of ca. 180
2
km . Based on initial positive impressions, options were formulated for a further
2
1000 km of floodplain rehabilitation.
No ecological information following reflooding has been reported from elsewhere
in (West) Africa (John et al. 1993). Expectations of the Waza-Logone rehabilitation
were, however, optimistic based on the impressive dynamics of floodplains (Breen
et al. 1988). In the reflooding appraisal study (Wesseling et al. 1994), we assumed
that by reflooding, disturbed floodplain vegetation in the impact zone would fully
rehabilitate into its previous ‘perennial’ state within five years. One of the objec-
92
Part II – Impact of ref looding
Fig. 3.1 – The Waza-Logone area
93
3 – Veget at ion dynamics
tives of the pilot reflooding experiment was to test these assumptions, summarised
in the following questions:
1 How did the floodplain vegetation change during the desiccation process following construction of the Maga dam?
2 What is the rehabilitation capacity of the disturbed vegetation?
The answers to these questions will be pivotal in assessing the viability of the
planned large-scale reflooding in Waza-Logone and for similar releases elsewhere.
Moreover it may indicate possible bottlenecks in floodplain rehabilitation and inspire actions to solve them.
3.2
Description of study area
Hydrology
2
The Waza-Logone area covers ca. 8000 km in the Far North province of Cameroon (Fig.3.1) and has a mean annual rainfall of 650 mm (Beauvilain 1985). The
rainy season is from June to September, which is followed, in parts of the area, by
flooding from the Logone River and its branches until December. The floodplain
is dominated by heavy cracking clay soils (vertisols), with loamy soils on the levees along watercourses (Brabant & Gavaud 1985). The area is extremely flat, intersected by only a few developed watercourses. Most of the flooding occurs over the
land surface, also called creeping flow (Howell et al. 1988).
No systematic information is available on water levels on the floodplain prior to
the reflooding. Information on former flooding extent was obtained by Landsat
MSS satellite image interpretation and discussions with local people employed for
hydrological monitoring. The year 1994 was considered as a standard for rainfall
and river discharges for reasons of comparison as only slight differences were
detected in the period 1994-1997 (internal reports Waza-Logone project). In 1997,
a second watercourse was opened, increasing flooding beyond the 1994 floodline.
Although the reflooding aimed to (partly) restore the pre-dam hydrological regime,
major differences remained. Flood water could now only arrive from the Logone
River in the east, whereas in the past it also arrived from the west, through water
courses draining the nearby Mandara Mountains. These waters are now trapped
by Lake Maga. This change of water direction has the following consequences:
• A shorter flooding season, due to the relatively late arrival of flood waters from
the Logone River (mid-August onwards), which also takes a long time before
arriving in the downstream, western part of the floodplain. This is particularly
apparent in the Tchikam plots and the western part of the transect, where flood
water does not arrive until late October, well after the area has dried up following the rainy season.
94
Part II – Impact of ref looding
•
Flooding water now has to travel over 40 km before arriving in the western area.
It passes through dense vegetation that reduces its speed and probably leads to
a loss of sediment load, resulting in less fertile soil conditions downstream
(Breen et al. 1988).
Vegetation history
Barth (1857) and Nachtigal (1889) gave the impression that in the middle of the
19th century the levees along water courses were far more densely wooded than
nowadays, most probably with the same dominant species (Celtis africana, Ficus
spp, Tamarindus indica), although some Sudanian tree species, scarce nowadays,
were mentioned as well (Vitellaria paradoxa). Floodplain vegetation composition
seems to be comparable to the present.
Flizot’s (1962) description indicated a drying of the floodplain 1-2 months later
than presently observed. This is due to the relatively abundant rainfall and floods
of the 1950 and 1960s (Beauvilain 1995). Letouzey (1968) mentioned the dominance of Echinochloa pyramidalis, with Vetiveria nigritana as second and Oryza
barthii (= O. longistaminata) and Hyparrhenia rufa as co-dominants. He indicated
an inundation depth of 1 to 2 metres (at present <1 metre) and mentioned Sorghum arundinaceum, Dinebra retroflexa, Pennisetum ramosum and Ischaemum afrum,
which can still be found in the desiccated floodplain.
The period of 1970-1979 was characterised by the great Sahelian droughts of 19731975, coinciding with limited changes in the hydrological infrastructure due to the
construction of the Waza-Kousseri road (Fig. 3.1). Vegetation maps of the entire
area and Waza National Park, based on aerial photographs from the mid 1960s
(Gaston & Dulieu 1976 and Wit 1975, respectively), showed that presently inundated floodplain communities used to extend 10-15 km further west. According to
Gaston & Dulieu (1976), Hyparrhenia rufa was dominant in the western parts of
the floodplain with Vetiveria nigritana as second. The lack of Echinochloa pyramidalis in the floodplain is striking. This description differed from Wit (1975) who
found Vetiveria nigritana the most common grass species in the western floodplain
and Hyparrhenia rufa restricted to ridges along drains. Wit (1975) found
Echinochloa pyramidalis locally dominant, especially on the western side of the
plain which is subject to less inundation, with the presence of large numbers of
dead Vetiveria nigritana tussocks. Sorghum arundinaceum was present in Acacia
seyal thickets above the flood line in the western part of Waza NP.
The presently inundated floodplain vegetation is typical of African seasonally
flooded grasslands with its species-poor perennial grasslands (Denny 1993). Oryza longistaminata and Echinochloa pyramidalis constitute single (or two) speciesstands on the floodplain, but Vetiveria nigritana is abundant on the levees of the
drainage lines and on the higher parts of the intact plain. Hardly any perennial
95
3 – Veget at ion dynamics
grasses can be found on the clay soils above the present flood line, where annual
grasses (in particular Sorghum arundinaceum) and herbs occur as well as, locally,
thickets of Acacia seyal and Piliostigma reticulatum.
3.3
M ethods
Vegetation monitoring
Maga dam period, 1984-1985: A 35 km long transect, with six permanent plots (P)
each 20x50m, was established. It started at the western edge of the floodplain,
which is no longer inundated, with exclusively annual vegetation where Vetiveria
nigritana used to be dominant (Wit 1975). The transect ended in the East in the
well-inundated floodplain, dominated by perennial grasses, little changed since
Gaston & Dulieu (1976) and Wit (1975) (Table 3.1; Fig 3.1). In the desiccated floodplain, 23 permanent 10 x 10m plots were installed in a 1 x 0.5km area around
Tchikam (Table 3.1; Fig. 3.1). Data was collected along the transect in 1984 and
1985 and from the Tchikam plots in 1985, forming a reference for the present study.
Table 3.1 – Observat ion met hods
Observat ion
Locat ion
plat f orm
(Fig. 3.1)
Flooding
1
No of
Lengt h
Plot size
Observat ion
permanent
(km)
(m)
years
period
2
Observat ion
plot s
Transect
Zina → Tchikam F (D)
16 (Z)
8
6x6
1993-97, 2002
f lood
R (D)
18 (Z)
9
6x6
1993-97, 2002
f lood
F R (D)
4 (P)
13
50 x 20
1984-85 ,
(east )
3
f lood
1993-97, 2002
Transect
Tchikam → w est
D
2 (P)
5
50 x 20
(w est )
Tchikam
Tchikam
D
23
0.96 x 0.54
10 x 10
plot s
3
1984-85 ,
mid-dry
1993-97, 2002
season
4
1985 , 1993-96,
mid-dry
1999
season
1993-97-99
early dry
22/9/94
Grid
FRD
f loodline
20 x 16
6x6
above
under
season
28
108
lat e dry season
1993
1
Flooding t ypes F: alw ays w ell f looded (also af t er 1979); R: ref looded f rom 1994 onw ards; D: desiccat ed since 1979,
2
excluding lat e f looding ; ( ) marginally present ; Throughout t his paper w e ref er t o t he f looding years w hich w ere st ud3
ied, and not t o t he observat ion period; Based on Oijen & Kemdo (unpubl. M Sc t hesis 1986) and t en Cat e (unpubl. M Sc
4
t hesis 1988); Based on Brummelen & Specken (unpubl.M Sc t hesis 1986).
96
Part II – Impact of ref looding
Pilot reflooding, 1993-1999(-2002): From 1993 onwards, prior to the pilot reflooding, vegetation monitoring was re-activated in the same Tchikam plots as in 1985
(Table 3.1). The Tchikam plots were only marginally affected by the pilot release
because of the late arrival of flooding water. Their position above the present flood
line makes them complementary to the ‘above flood line’ part of the grid.
In 1993, the number of permanent plots in the transect was increased from six to
40, the new 6 x 6 m plots (Z) were located at each 0.5 km (Table 3.1). Floristic composition was monitored in a newly installed 20 x 16 km grid, situated in the centre of the pilot release impact zone, upstream from the transect and Tchikam plots
(Fig. 3.1). The grid had 136 (6m x 6m) plots. Vegetation in the grid was initially
composed of 40% perennial grasses (mainly in the eastern part, bordered by the
water course which brought in the release water) and 60% annual species in the
centre and west, influenced by the Maga dam. Here the dominance of the 3-4 m
tall Sorghum arundinaceum was striking. Initially aimed at hydrological monitoring, these 136 permanent plots were located at a distance of 0.5 or 1 km (Table 3.1).
Data collection and analysis
Based on the 1984-1985 methodology (Tchikam plots and transect), the following
parameters were determined in the three observation areas (see Table 3.1 for additional details):
Tchikam plots
• Number of individuals of Echinochloa pyramidalis, Ischaemum afrum, Vetiveria
nigritana, and Pennisetum ramosum. From 1993 onwards, Panicum anabaptistum and Sorghum arundinaceum were also counted.
• From 1993 onwards, Oryza longistaminata cover was estimated
Transect (east and west)
• Dominance ranking of Echinochloa pyramidalis, Vetiveria nigritana and Oryza
longistaminata (the latter only since 1993)
• To detect possible trends of Vetiveria nigritana, its tussock condition was classified in 1984, 1985, 1996, 1999 and 2002 (adapted from Riney 1963, see Fig.
3.2).
Grid
• Estimation of total ground cover (except 1993)
• Estimation of cover for each species and life form (except 1993)
• Ranking of species according to dominance (1993, 1994).
Observations in the grid in 1993 were only possible at the end of the dry season,
when fires had passed, preventing reliable cover estimations. General dominance
assessments were still possible, however, and were compared with the 1994 early
97
3 – Veget at ion dynamics
% belonging to condition class
100
80
60
40
20
0
Year
Dominance
Plot no.
84 85 96
99 02 84
85 96 99
2 ? 0
02 84 85
0 0 1
96 99 02
1 12 5
P1
84 85 96
0 1 2
P3
99 02 84
0 0 0
85 96 99
1 1 3
P4
4 3 1
02 84 85
2 3 3
P5
96 99 02
3 2 2
P6
3 3 3
P7
1
2
3A
3B
4
4
1
degrading
tussock
condition
Figure 3.2 – Transect : Vet iveria nigrit ana vigour class development (1984-2002) (adapt ed f rom
Riney 1963)
1
base of t ussock f irm w hen kicked gent ly
2
cent ral part of base dead, of t en charred st umps remain; deat h is commonly f rom t he cent re out w ards
3A living ext erior is not int act anymore
3B f ew minor living part s remain on t he ext erior
4
only dead t ussock base st ill remaining
dry season records. Cover per species or life form is expressed as percentage of
total vegetation cover (relative cover) to limit bias related with the observation period. In the grid, plots were separated according to their position to the 22 September 1994 flood line, the maximum height of the floods in the first year of reflooding (Table 3.1). As explained above, from 1997 onwards the actual floodline
expanded into the area above the 1994 floodline, preventing quantitative use of
these data beyond 1997.
Monitoring results of all plots have been treated as paired samples, eliminating
plots where data for one or more years were missing. Given the general non-normal distribution of species and plant form cover, 1993-1997 results were analysed
with Friedman two-way ANOVA test (for multi, paired, non-parametric variables)
or, in a few cases, Wilcoxon signed ranks test (for two, paired, non-parametric variables) (SPSS 1993).
98
Part II – Impact of ref looding
3.4
Results
Changes in vegetation attributed to the impact of the M aga dam
In 1984 Vetiveria nigritana was the dominant species in four of the six studied
plots of the transect, and was second in dominance in the two extreme plots in the
west and east (Fig.3.2). Between 1984 and 1985, during a severe drought, deterioration of its condition in the transect was observed and it lost its dominant position to Panicum anabaptistum and Echinochloa pyramidalis. In the 1990s, only one
individual remained in the western part of the transect, but was found dead in 2002.
In the eastern part of the transect, which was still well flooded, Vetiveria nigritana
remained strongly present although it had lost its dominant position to Echinochloa pyramidalis and Oryza longistaminata (Fig. 3.2; Table 3.2). Its condition seemed,
however, to be improved in 1996 and 2002 compared to 1984 and 1985. In the Tchikam plots, Vetiveria nigritana also disappeared between 1985 and 1993 (Fig. 3.3).
Information on Echinochloa pyramidalis and Oryza longistaminata grasses indicated a reduction of their cover following the Maga dam construction. In the Tchikam plots, Echinochloa pyramidalis disappeared between 1985 and 1993 (Fig. 3.3).
On the transect it disappeared from two plots between 1985 and 1993 (P3, P4),
although further to the east its presence continued in 1993 in the lower laying
parts of the desiccated plain (Z24, P5, Z26-29) (Table 3.2). In the still well flooded part of the transect (P6, P7) both species remained dominant (Table 3.2).
number of individuals (per 100 m2)
200
1985
1993
1994
1995
1996
1999
150
100
50
0
Pennisetum
ramosum*
Sorghum
Panicum
arundinaceum anabaptistum*
Ischaemum
afrum*
Vetiveria
nigritana*
Echinochloa
pyramidalis*
Oryza
longistaminata
(relative cover)
Figure 3.3 – Tchikam plot s: Variat ion in number of individuals per species (1985-1996, 1999)
* : signif icant diff erence 1993-1996 (p < 0.05), see Table 3.6
99
3 – Veget at ion dynamics
Table 3.2 – Abundance of Echinochloa pyramidalis and Oryza longist aminat a along t he
t ransect
Part no.
Flood
1
Z2
Z3
Z4
Z5
P7
Z6
Z7
Z8
E
F
F
F
F
F
F
F
Z9 Z10 Z11 Z12 P6 Z13 Z14 Z15 Z16 Z17 Z18 Z19 Z20
F
F
F
F
F
F
F
R
D
E
E
E
D
0
Echinochloa pyramidalis
2
1984
1
1985
1
4
1
1993
0
2
3
1
1
1
1
1
1
2
2
2
1
2
4
0
0
0
0
0
1994
0
2
2
2
2
2
2
1
1
0
2
?
1
2
?
0
0
0
0
0
0
1995
0
2
2
2
2
2
2
1
1
0
2
4
2
2
1
0
0
0
0
0
0
1996
0
1
1
2
1
2
2
2
1
0
1
2
2
2
1
2
0
0
0
0
0
1997
0
2
2
2
2
2
2
2
1
0
2
3
2
2
2
4
0
0
0
0
0
2002
0
1
2
2
1
2
1
2
2
0
2
3
2
2
1
0
0
0
3
0
0
Oryza longist aminat a
1993
0
1
2
2
?
2
2
2
?
1
1
1
1
1
1
0
0
0
0
0
0
1994
0
1
1
1
1
1
1
2
2
1
1
1
2
1
0
0
0
0
0
0
0
1995
0
1
1
1
1
1
1
2
2
1
1
1
1
1
3
0
0
0
0
0
0
1996
0
2
2
1
2
1
1
1
2
1
2
1
1
1
2
0
0
0
0
0
0
1997
0
1
1
1
1
1
1
1
2
1
1
1
1
1
1
0
0
0
0
0
0
2002
0
2
1
1
2
1
2
1
1
1
1
1
1
1
2
2
0
0
0
0
0
Z21 Z22 Z23 Z24 P5 Z25 Z26 Z27 Z28 Z29 Z30 P4 Z31 Z32 Z33 Z34 Z35 P3
P1
Part no.
Flood
1
R
R
R
R
R
R
R
R
R
R
D
R
R
R
R
R
R
D
D
0
Echinochloa pyramidalis
1984
2
4
3
1985
2
3
2
0
0
0
1993
0
0
0
2
2
0
3
1
2
4
0
0
0
0
5
0
0
1994
0
0
0
3
0
0
0
?
?
5
0
0
0
0
0
0
0
0
0
1995
0
0
0
2
4
0
0
2
3
2
0
0
0
0
0
0
0
0
0
1996
0
0
0
2
5
0
0
2
4
2
0
0
0
0
0
0
0
0
0
1997
0
0
0
2
5
0
0
2
3
2
0
0
0
0
0
0
0
0
0
2002
1
1
1
2
1
3
2
2
2
2
0
0
0
3
0
0
0
0
0
Oryza longist aminat a
1993
0
0
0
0
3
0
0
0
0
1
0
0
0
0
3
0
0
0
0
1994
0
0
0
0
5
3
5
0
0
3
0
0
0
0
0
0
0
0
0
1995
0
0
0
0
0
0
0
1
0
1
0
0
0
0
5
4
0
0
0
1996
0
0
0
0
0
0
5
1
2
1
0
5
5
0
2
3
0
0
0
1997
0
0
0
0
0
0
5
1
1
1
0
5
0
4
3
2
0
0
0
2002
2
0
0
1
4
0
1
1
1
1
0
3
1
1
1
1
1
0
0
1
F:
2
Flooded, R: Ref looded, D: Desiccat ed, E: Exondat ed (sit uat ion 1994 onw ards).
Dominance: 0 not present , 1 dominant , 2 second, 3 t hird, et c. (see Table 3.3 f or P-values)
100
Part II – Impact of ref looding
Table 3.3 – P-values of shif t in abundance along t ransect , see Table 3.2
(Wilcoxon M at ched-pairs Signed ranks t est , Anon. 1993).
Echinochloa pyramidalis
1993
1994
1995
1994
0.92
1995
0.66
0.79
1996
0.55
0.50
0.80
1997
0.87
0.78
0.16
2002
1996
Oryza longist aminat a
1997
1993
1994
1995
1996
1997
0.30
0.035*
<0.0001* <0.0001* <0.0001* <0.001*
0.001*
0.44
1.00
0.39
0.15
0.02*
0.87*
0.31
0.14
0.41
0.013*
<0.001*
<0.001*
<0.001*
<0.001*
* : signif icant at 0.05 level
Sorghum arundinaceum, was again found in the Acacia seyal thickets above the
flood line in the western part of Waza NP (Wit 1975). In 1984 and 1985 it was not
mentioned in the plots on the transect although it dominated in several of them
from 1993 onwards (including two of the 1984 plots). It was also not mentioned in
the Tchikam grid either, but it remains unclear whether or not it did occur there.
Changes in vegetation after reflooding
Tchikam plots
Results showed the high inter-annual fluctuations in the number of plant species
as well as in the cover of Oryza longistaminata (Fig. 3.3). The appearance of Panicum anabaptistum in 1995, also observed in the grid above the flood line, is remarkable. As observed in the grid above the flood line, the number of Sorghum
arundinaceum individuals did not change significantly between 1993 and 1997,
contrasting a remarkable drop between 1996 and 1999 (Fig. 3.3).
Transect
On the transect, Echinochloa pyramidalis lost its dominant position to Oryza longistaminata in the parts with the highest water levels from 1994 onwards (Table 3.2).
In the period 1994-1997, Echinochloa pyramidalis colonised only one plot, situated
at the edge of the flooded plain. In the same period Oryza longistaminata, appeared
in eight plots, but did not persist in two of them, and disappeared from one plot.
The annual fluctuation in dominance of both rhizomatous grasses, apparently unrelated to inundation dept, is striking. This is also observed for Echinochloa pyramidalis between 1996 and 1997 and Oryza longistaminata between 1995 and 1996
(Table 3.2). Changes in the transect accelerated between 1997 and 2002 when Echinochloa pyramidalis and Oryza longistaminata appeared in seven respectively six of
the 16 reflooded plots, each of which became the dominant species in three plots.
101
3 – Veget at ion dynamics
In addition, Oryza longistaminata became the dominant species in four reflooded
plots where it was already present. At the end of 2002, 13 out of 16 reflooded plots
were dominated once again by rhizomatous grasses, in the remaining three, these
grasses had appeared and were expected to dominate soon.
Grid
From 1993 onwards, a total of 67 annual grasses, eight perennial grasses, five perennial herbs and eleven woody species were found. Presented data on woody, annual, perennial and total cover are based on all species. We also present the results
in detail of all perennial grasses (with exception of three rare ones), Acacia seyal,
which was reported to be regenerating after the Maga dam construction, and the
two dominant annual grasses (Table 3.4). This selection is based on the previous
pilot research (Table 3.1), Gaston & Dulieu (1976) and Wit (1975) and forms the
base of the floodplain grazing exploitation (Scholte et al. 1996b). The contribution
of this selection to the total cover in the flooded grid from 1994 to 1999 was 67,
77, 72,82, 72 and 76 % respectively.
Table 3.4 – Select ed species cont ribut ing 67-82 % of t ot al cover in t he f looded grid f rom 1994
t o 1999 (see Text )
Species
Plant t ype
Acacia seyal
w oody perennial
Penniset um ramosum
annual grass
Sorghum arundinaceum
t all annual grass
6
Panicum anabapt ist um
(short living) t ussock grass
5
Ischaemum af rum
(short living) t ussock grass
4
Vet iveria nigrit ana
t ussock grass
3?
Echinochloa pyramidalis
rhizomat ous grass
2
Oryza longist aminat a
rhizomat ous grass
1
1
2
Hydrosere posit ion
2
8
7
1
In t he st udy area bi-annual f orms w ere observed in very low densit ies
Series, w it h increasing inundat ion f rom 8 t o 1, grasses adapt ed f rom van der Zon (1992)
In 1994, after the first reflooding season, Echinochloa pyramidalis, Ischaemum afrum,
and Vetiveria nigritana were found to be less dominant in the reflooded zone than
in 1993 (Table 3.5). In the reflooded part of the grid, relative cover of perennial
grasses increased from 41% in 1994 to 62% in 1997 and 75% in 1999, a 7% annual average (Fig. 3.4). The change in perennial grass cover can largely be attributed
to the relative cover increase of Oryza longistaminata and Echinochloa pyramidalis
(Fig. 3.5). The limited number of observations in the grid above the 1994 flood line
did not allow a clear distinction in the perennial species’ increase, which seems to
be initially merely Panicum anabaptistum and Ischaemum afrum, Fig.3.5 (but see
under ‘Tchikam plots’). In the reflooded part of the grid, annual grasses and herbs
Part II – Impact of ref looding
80
cover %
102
60
40
20
Total cover (real)
Annuals*
Perennials*
Woody
0
1994 1995
1996 1997
1998
1999
Figure 3.4 – Development in lif e f orm cover (1994-1999) in grid plot s
* : signif icant diff erence 1993-1997 (p < 0.05), see Table 3.6
Table 3.5 – Changes in dominance f rom 1993 t o 1994 in t he grid area
(Wilcoxon signed ranks t est , SPSS 1993)
Dominance in 1994
compared w it h 1993
P-values
Acacia seyal
>
0.80
Penniset um ramosum
>
0.51
Sorghum arundinaceum
<
0.18
Panicum anabapt ist um
<
0.34
Ischaemum af rum
<
< 0.001*
Vet iveria nigrit ana
<
0.004*
Echinochloa pyramidalis
<
<0.001*
Oryza longist aminat a
>
0.16
* signif icant at 0.05 level
decreased in relative cover from 58% to 23% from 1994 to 1999 (Fig. 3.4). Despite
high inter annual fluctuations, Sorghum arundinaceum’s relative cover decreased
in the flooded grid from 26 to 16% between 1994 and 1997 but above the 1994
flood line it fluctuated around 35% (Fig. 3.5). From 1998 onwards, coinciding with
103
3 – Veget at ion dynamics
left: below '94 floodline
50
right: above '94 floodline,
cover %
from 1997 onwards including
several flooded plots
1994
1995
1996
1997
1998
1999
40
30
20
10
0
Acacia seyal
Pennisetum
ramosum
* Sorghum
arundinaceum
Panicum
anabaptistum
Ischaemum
afrum
Vetiveria
nigritana
*Echinochloa
pyramidalis
*Oryza
longistaminata
Figure 3.5 – Changes in percent age cover of individual species (1994-1999) in grid plot s
* : signif icant diff erence 1993-1997 (p < 0.05), see Table 3.6
Table 3.6 – P-values Figs. 3.3-3.5 at Tchikam and Grid plot s (1993-1997)
(Friedman t w o-w ay anova t est , SPSS 1993)
Test ed variables
Fig. 3.3
Fig. 3.4
Fig. 3.5
under ’94 f loodline above ’94 f loodline
Acacia seyal
0.94
0.82
Penniset um ramosum
<0.0001*
0.99
0.94
Sorghum arundinaceum
<0.0001*
<0.0001*
0.56
Panicum anabapt ist um
<0.0001*
0.90
0.72
Ischaemum af rum
<0.0001*
0.89
0.88
Vet iveria nigrit ana
<0.0001*
0.60
1.0
Echinochloa pyramidalis
<0.0001*
0.040*
1.0
0.0053*
0.81
Oryza longist aminat a
Tot al cover
0.038*
0.24
Annual cover
<0.0001*
Perennial cover
<0.0001*
Woody cover
* : signif icant at 0.05 level
0.8
104
Part II – Impact of ref looding
the 1997 extension of the floodline, the cover of Sorghum arundinaceum above the
1994 floodline dropped dramatically as well. The increase in cover of Oryza longistaminata above the 1994 floodline was due to its dominance in four plots, situated below the 1997 floodline.
3.5
Discussion
Pre-M aga dam vegetation
Available information from the pre-dam situation is highly fragmented and sometimes contradictory (Gaston & Dulieu 1976). However, none of the old landscape
descriptions indicated the presence of plant species or plant communities that are
no longer encountered in the floodplain. A reduction in some of the floodplain
species (e.g. Vetiveria nigritana), has already started well before the Maga dam construction.
Impact of the dam
The limited number of observations did not allow a comprehensive assessment of
the disturbance process following the 1979 Maga dam construction. Data on Vetiveria nigritana showed, however, that changes took place over a period of six years
and were strongly influenced by climatic and hydrological fluctuations, most notably
the 1984-1985 drought. Floodplain rehabilitation is, obviously, subject to the same
variable environmental fluctuations and fits in the ‘succession and fluctuation’
model of Van der Valk (1985). In 1988, a year with rainfall well above the decade
average (Beauvilain 1995), local people considered the flooding conditions as comparable to ‘pre-dam averages’. This suggests that even before the reflooding flooding levels were still comparable to pre-dam levels. The Maga dam apparently caused
‘only’ a change in relative cover of existing plant communities, resulting in a different vegetation pattern. The slow changes following the disturbance in 1979
might be indicative of the future rehabilitation process.
The rehabilitation
The hydrosere observed by van der Zon (1992), and from other African seasonally flooded grasslands (Denny 1993; Howell et al. 1988) matched with our field observations (Table 3.4). The validity of such a hydrosere as a way of possible succession in West Africa was, however, strongly contested by John et al. (1993) and is
also contradicted by the presented rehabilitation succession. The lack of Vetiveria
nigritana in the rehabilitation succession is especially illustrative of the different
species’ regeneration capacity, irrespective of their position in the hydrosere. The
decrease in dominance of Echinochloa pyramidalis on the transect in the first reflooding year and its replacement by Oryza longistaminata showed, however, that
the hydrosere sequence may be of some help in predicting changes if dealing with
species with comparable regeneration strategies. Oryza longistaminata and Echi-
3 – Veget at ion dynamics
105
nochloa pyramidalis, both rhizomatous grasses, re-established relatively quickly by
lateral vegetative propagation (Van der Valk 1992). This ‘peripheric propagation’
was probably influenced by the proximity of depressions with remaining perennial vegetation, the ‘border’ effect (Barrat-Segretain & Amoros 1996). The fluctuating dominance of both rhizomatous grasses was surprising, given their perennial
nature and the period of relatively comparable rainfall and flooding conditions.
This fluctuation can only be explained by contraction and spreading speeds of at
least the plot length (6 m/year), which is less than the estimated lateral growth
-1
speed of 50 m year of Phragmites australis (van der Valk 1992). Some of the more
sudden appearances of Oryza longistaminata and Echinochloa pyramidalis can,
however, only be explained by propagule dispersal (Noble & Slatyer 1980; Cellot et
al. 1998), almost certainly during the flooding.
The tussock grasses Ischaemum afrum and Panicum anabaptistum, both depend on
generative propagation and appeared as quickly as the dominant rhizomatous
grasses. Vetiveria nigritana, also dependent on generative propagation, did not
regenerate at all. Given the dominance of Vetiveria nigritana in the study area prior
to the Maga dam construction and its importance as fodder for livestock and
wildlife, its lack of regeneration is a major set-back of the otherwise positive pilot
reflooding. The 1996, 1999 and 2002 monitoring showed the general good condition of Vetiveria nigritana tussocks, producing seeds in abundance in an area
where vegetation cover is very low, potentially permitting gap phase replacement
(Van der Valk 1992). Further west in the transect, where Vetiveria nigritana used
to be dominant and is still present in depressions, the lack of regeneration might
be attributed to the earlier described late arrival of flood water and/or the low sediment level of flood water. It remains, however, difficult to establish the characteristics of these safe sites (Van der Valk 1992). In the eastern part of the transect
where, between 1985 and 1993 and despite the continuing flooding, Vetiveria nigritana lost its dominant position to Oryza longistaminata and Echinochloa pyramidalis, it is assumed that the strong competitive nature of these species prevented
regeneration of Vetiveria nigritana.
We attribute the 1998 and 1999 changes in species composition in the area situated above the 1994 floodline (Figs. 3.3, 3.5), to the impact of the additional flooding from 1997 onwards. Above average rainfall in 1997-2000 (Fig. 2.3) may further have reduced the differences between flooded and non-flooded areas.
Speed of recovery
The decrease in dominance of Echinochloa pyramidalis, Ischaemum afrum and Vetiveria nigritana in the first flooding season contrasted with their later overall increase. We attribute this to the bias of the late 1993 observation period when after
bush fires, these species were relatively well visible. We speculated on the speed
of a 100% perennial grassland conversion and assumed that the increase would
106
Part II – Impact of ref looding
remain on the same level as in the period 1994-1999. Perennial grass cover in the
grid was initially 40% and with an annual 7% conversion, it would take nine years,
i.e. till after the flooding season of 2003, before the area would again be entirely
dominated by perennial grassland. This prediction is in line with presented qualitative observations on the transect (Table 3.2). These reported results are satisfactory compared to temperate wetland rehabilitation projects such as mire rehabilitation in Central Europe where return to ‘original stage’ is said to be ‘impossible within reasonable time spans’ (Pfadenhauer & Klotzli 1996). Three years after
reflooding wetlands in the prairie pothole region in the USA, its vegetation composition still differed significantly from natural wetlands (Galatowitsch & Van der
Valk 1996). Thoraug (1980) mentioned recovery rates of two to over 30 years in
restored conditions and at least double that time under natural conditions in restoration efforts in various ecosystems in the USA. The relative speed of wetland
rehabilitation in the Waza-Logone area may be explained by the dominance of
grasses (Thoraug 1980) and the productive tropical floodplain environment. A
further factor may be the short, 15 year time span between disturbance and rehabilitation (or 25 years when the climate-induced disturbances are included). This
resulted in relatively few changes in the hydrological infrastructure and the continued presence of all major floodplain species.
3.6
Conclusions
The changes in the period 1993 till 1999 showed partial recovery towards a perennial dominated vegetation due to the 1994 pilot release. The ‘100%’ perennial state,
probably reached in 2003, differs, however, from the pre-Maga dam vegetation,
which was the result of the relatively humid 1940-1970 period. Most remarkable
is the lack of Vetiveria nigritana recovery in the reflooded downstream areas. Its
comeback may depend on brief periods of entirely different environmental conditions, possibly not occurring in the space of a decade. Another unknown factor in
the rehabilitation is the possible lack of sediments brought in with floodwater in
the downstream areas.
4
M aximum Flood Depth Determines
Above-ground Biomass in African
Seasonally Shallow ly Flooded Grasslands
Paul Scholt e
Abstract
Af rican seasonally f looded grasslands are an import ant source of f orage f or
w ildlif e and livest ock. Flood dept h has been f requent ly used t o explain t he
dist ribut ion of plant species, but it s relat ion w it h veget at ion product ion has
remained ambiguous. I st udied t he relat ion bet w een f looding and above-ground
biomass at t he end t he f looding season, as w ell as t he dry season t o assess t he
impact of ref looding on t he shallow ly inundat ed Logone f loodplain, Cameroon.
Above-ground biomass of a combinat ion of all species and of t he individual
perennial grasses Oryza longist aminat a and Echinochloa pyramidalis show ed a
posit ive linear relat ionship w it h maximum f lood dept h up t o at least one met er.
These relat ions became st ronger during t he t w o years f ollow ing t he w at er raising, show ing t he lag in react ion t ime t o f loodplain rehabilit at ion. Above-ground
biomass dat a f rom ot her major f loodplains in t he t hree main Af rican geographic
regions show ed a similar relat ion w it h maximum f lood dept h up t o one met er.
Dry season regrow t h, import ant because of it s high nut rient qualit y in a period
of f orage scarcit y, w as not direct ly relat ed w it h maximum f lood dept h, possibly
because of it s dependency on t he period of burning and residual soil moist ure.
Alt hough t he mechanism of t he maximum f lood dept h - above-ground biomass
relat ion is not yet f ully underst ood, present ed dat a allow an assessment of t he
impact of (re-) f looding on veget at ion product ion.
109
4.1
Introduction
African seasonally flooded grasslands have drawn widespread attention as source of
forage for wildlife (Rees 1978a; Howell et al. 1988) and livestock (Evans-Pritchard
1940; Hiernaux & Diarra 1983; Breman & De Wit 1983). Flooding intensity characterises floodplain vegetation composition (Ellery et al. 1991) as has become apparent
by the construction of upstream dams (Marchand 1987; McCully 2001; Scholte et al.
2000b). The raising of waterlevels by flood releases, as practised in the Logone floodplain in North Cameroon, showed the reversibility of these changes on floodplain
vegetation composition (Scholte et al. 2000b). The impact of the flooding regime on
the productivity of seasonally flooded grasslands remained unknown, as only average productivity characteristics, not linked to flood depth, are reported. In relatively
well-studied temporate zones, where flooding usually occurs outside the growing
season, it is assumed that the lower the inundation the higher the production
becomes (Middleton 1999). Although this may also hold for forested tropical floodplains (Hughes 1990), scattered data of seasonally flooded grasslands in Africa suggest a higher production with increasing flood depth (Francois et al. 1989; Hiernaux
& Diarra 1983; Ellenbroek 1987). Only under rapid and extremely deep flooding,
more than several meters, tropical seasonally flooded grasslands have shown a relatively low production (Ellenbroek 1987; Junk et al. 1989; Morton & Obot 1984).
Flood pulsing, i.e. periods of flood followed by a pronounced dry season (Junk et al.
1989; Middleton 1999), allows fire and grazing, reducing above-ground biomass
and triggering regrowth of perennial grasses. This regrowth is marked for its nutritional quality and is the base of wildlife and livestock production (Howell et al. 1988;
Hiernaux & Diarra 1983). Regrowth production has been linked with above-ground
biomass (Breman & de Ridder 1991), but not with flood depth so far.
To assess the impact of increased flooding on vegetation production, I studied the
relationship between above-ground biomass and regrowth of grasses with maximum flood depth. Fieldwork was conducted in the Logone floodplain that is part
of the Lake Chad seasonally flooded grasslands (Fig. 4.1). I compared these findings with the main seasonally flooded grasslands in West, North-East and Southern Africa (Denny 1993). Above-ground biomass of depression and stream communities (Hiernaux & Diarra 1984a; Ellenbroek 1987; Francois et al. 1989), with
a maximum flood depth of over one meter will not be discussed because of their
limited occurrence in Logone. Because of the floodplain’s importance for grazing,
Logone assessments included key forage quality parameters allowing a link with
monthly nutritional quality assessments of comparable grassland communities in
the Sudd (Howell et al. 1988). Changes in local variations in soil properties such
as cracking depth were assessed as they were expected to lead to differences in soil
moisture and influence (regrowth) production. Tobias & Vanpraet (1981) and Ellenbroek (1987) who concluded that pedological differences in the Logone and Kafue
floodplain were insignificant compared to the impact of flooding on vegetation,
limited their observations to general soil characteristics.
110
Part II – Impact of ref looding
Figure 4.1 – Waza-Logone area and posit ion of t he Transect .
4 – Veget at ion product ion
4.2
111
M aterials and methods
Study area
The Logone floodplain has a Sahelo-Sudanian climate, with an average annual rainfall of 650 mm that falls from June to September (Beauvilain 1995). In the study
period 1994-1996, rainfall in Zina (Fig. 4.1) was 691, 656 and 761 mm respectively.
The Logone floodplain is situated in the alluvial plains of the former Mega Lake
Chad (Schneider 1994; Brabant & Gavaud 1985). The lacustrine and more recent
fluvial deposits of the Logone river attain a thickness of hundreds of meters and
largely consist of fine textured smectitic clays. Soils have been classified as undifferentiated vertisols (Brabant & Gavaud 1985). The eastern and central parts of the
area are subject to flooding, from September till December, by the Logone River
that has its origin in the volcanic Adamaoua highlands, 500 km further south. The
nearby predominantly gneissic Precambrian Mandara Mountains feed seasonal
rivers, draining, from July till September, into the south-western part of the study
area. The area is extremely flat, intersected by a few little developed watercourses.
Most of the flooding therefore occurs over the land surface, still saturated by the
rains, also called creeping flow (Howell et al. 1988).
The Logone floodplain vegetation with its species-poor perennial grasslands is typical of African seasonally flooded grasslands (Denny 1993; Ellenbroek 1987; Howell
et al. 1987). Oryza longistaminata and Echinochloa pyramidalis constitute one- or
two species-stands on the floodplain, whereas Echinochloa stagnina and Vossia cuspidata dominate depressions and stream borders. Hardly any perennial grass is
found on the clay soils above the present flood line, where annual grasses (especially Sorghum arundinaceum) and herbs occur, as well as scattered thickets of
Acacia seyal and Piliostigma reticulatum. Vetiveria nigritana used to be abundant on
the higher parts of the floodplain, but was replaced, due to desiccation (see below),
by Ischaemum afrum and Panicum anabaptistum.
The Logone floodplain severely degraded when it was cut off from its main water
supply through the construction of a dam upstream (Fig. 4.1). Built in 1979, this
dam has caused the disappearance downstream of the perennial floodplain vege2
tation in an area of 1500 km . This had a negative impact on livestock and fisheries, as well as on large mammals and birds (Drijver et al. 1995; Scholte et al.
1996a). To counter the ongoing degradation, an embankment was breached in
1994, opening a watercourse that had been closed off since 1979 (Fig. 4.1). This
2
brought back the annual floods in an area of 180 km and triggered recovery of
perennial floodplain vegetation that was expected to have fully covered the reflooded area in 2003 (Scholte et al. 2000b).
Part II – Impact of ref looding
Description of sample sites
In 1985, a 16 km transect was established starting at Tchikam in Waza National Park
(Fig. 4.1). Prior to the Maga dam its vegetation was dominated by Vetiveria nigritana (Wit 1975) subsequently replaced by Sorghum arundinaceum, Ischaemum afrum
and Panicum anabaptistum. This area has been reflooded since 1994 (Fig. 4.2),
and is gradually being occupied by Oryza longistaminata and Echinochloa pyramialis (Scholte et al. 2000b). Zwang village, situated on a rise on the border of Waza
National Park (Fig. 4.1), separates the reflooded western from the eastern part of
the transect that has always been flooded. The transect ended near Zina, in the wellinundated floodplain, dominated by Oryza longistaminata and Echinochloa pyramidalis (Annex 4.1; Fig. 4.1). In 1994, the transect’s 4 permanent (P-) observation
sites were increased with 28 (Z-) sites, equidistant at 0.5 km, followed by four (S-)
plots in 1995 (Annex 4.1), (Scholte et al. 2000b).
70
60
Max. flood depth (cm)
112
1994
1995
1996
flood duration 1994 (weeks)
50
40
30
20
10
0
sites: Z34 Z32 P4
r
r
flooding: r
West
Tchikam
Z29 Z26 S4
f
r
r
1 km
P5
r
Z22 Z20 S3
r
d
f
P6
f
Zwang
Z11 S2
f
f
Z8
f
P7
f
S1
f
Zina
East
Figure 4.2 – M aximum Flood Dept h in t he t ransect Zina-Zw ang-Tchikam (see also Fig. 4.1)
Sit es, see Annex 4.1 f or more inf ormat ion; Flood sit uat ion: f : f looded (bef ore and af t er 1994), r:
ref looded f rom 1994 onw ards, d: desiccat ed (bef ore and af t er 1994)
In 1994, 1995 and 1996, water depth was measured at each south-central site border, twice a month during the flooding period from mid-August till mid-November. In 1994, every second week flood duration (Le Houérou 1989; Van der Valk
1981) was assessed, but was discontinued in subsequent years because of the inaccessibility of the western part of the transect.
Table 4.1 – Diff erent iat ing soil charact erist ics
Gilgai 2
Root
OM %
pH
Dept h t o sandy
Surf ace
Cracking
CaCO3
Soil moist ure
Dominant plant
Layer
Top-sub
(KCl)
clay (loam)
cracks
dept h
nodules
mid/end dry
species3
(cm)
soil
layer (cm)
(cm)
In layer (cm)
season rat io
S1
limit ed
0-10
9.2-0.5
3.9-5.5
115
60
10-60
1.36
P7
limit ed
0-10
–4
–
122
60
60-120
1.67
Oryza
longist aminat a,
S2
limit ed
0-10
–
–
142
P6
pronounced
0-15
6.1–0.6
4.5–6.2
125
rare
72
no
1.57
100
no
2.54
Echinochloa
4 – Veget at ion product ion
aAlw ays f looded
Sit e1
pyramidalis,
Vet iveria
nigrit ana
S3
pronounced
no
6.4–0.3
5.0–6.8
115
65
no
5.21
P5
mod. pronounced
no
–
–
>180
65
no
3.34
Ischaemum
af rum,
ref looded
S4
mod. pronounced
no
–
–
150
common,
70
no
3.01
Panicum
anabapt ist um
w ell
developed
P4
no
no
6.4 – 0.5
4.7 – 6.3
110
40
no
2.01
Sorghum
arundinaceum,
Cichorus
olit orius5
1
see Figure 4.2
2
limit ed: 1:50 (w avelengt h: amplit udo); mod. pronounced: 1:25; pronounced: 1:15
3
see Scholt e et al. 2000b
not det ermined
5
a non-f orage f orb of w hich no nut rit ional qualit y w as assessed.
113
4
114
Part II – Impact of ref looding
In 1995, at the end of the dry season, eight soil pits were dug along the transect
(Annex 4.1), described following FAO guidelines (1977) with additional observations on gilgai (wavelength and amplitudo, Blokhuis 1993), Table 4.1. Soils were
classified as Eutric Vertisols (FAO 1988) as even CaCO3 concentrations in S1 and
P7 did not meet the 15% level for a diagnostic calcic horizon (Driessen & Dudal
1989). Of four selected pits, granular composition, pH-KCl, organic matter and
CaCO3 were analysed for each distinguished horizon (Table 4.1) at BLGG laboratories (The Netherlands). The upper soil layers had high organic matter contents
and low associated pH (KCl) values (Table 4.1), respectively higher and lower than
usually observed in vertisols (Blokhuis 1993; Bunting & Lea 1962), but comparable to seasonally flooded grassland vertisols in the Kafue Flats (Rees 1978b; Ellenbroek 1987). With their common and well developed cracks, sites P6 and S3 situated in the eastern, continuously flooded part of the transect, marked a transition
to soils characteristic of more limited flood depths (Table 4.1; Fig. 4.2). Soil moisture (‘residual’ moisture) was determined during the excavation at 20-40 cm, 6080 and around 120 cm. In 1996, soil moisture content was determined in February,
mid-dry season, in 16 unprotected plots at depths of 15, 20, 25, 30 and 35 cm (Annex 4.1). The 110-ml samples were oven-dried for 24 hours at 110°C and subsequently assessed soil moisture was expressed as percentage of dry weight.
Biomass
Above-ground biomass
In 1994, 32 sites were sampled at equal distances in the transect, reduced to eight
and 16 equidistant sites in 1995 and 1996 respectively (Fig. 4.2, Annex 4.1). Im2
mediately after the floods receded (see below), in each selected site 1 m randomly taken samples of above-ground biomass were hand clipped at ground level.
The difficulty to reach the western part of the transect in 1994 allowed only four
samples to be taken and another four additional replicates in each site to be visually estimated after cutting, which were corrected following double-sampling weightestimate (’t Mannetje 1987). In 1995 and 1996, ten samples in each selected site
were cut. In the laboratory, clipped biomass was separated according to species.
Samples were oven-dried at 65°C for 48 hours and expressed as dry weight percentage. The average of the above-ground biomass per site, expressed as DM (dry
-2
matter) gm , was used for comparison with maximum flood depth, below-ground
biomass, regrowth and soil moisture. Comparison of biomass between years was
based on the individual samples in each site.
In view of the low percentage of dead above-ground biomass (<5%), we assumed
that biomass sampled when sites dried up was also the maximum above-ground
biomass of the growing season. Data from the Kafue Flats (Annex 4.2) suggest
that this maximum might in some cases be reached somewhat earlier. The Logo-
4 – Veget at ion product ion
115
ne floodplain as well as the Inner Niger Delta, Sudd and Kafue Flats, has only one
growing season however, and with a relatively low regrowth production, standing
crop at the end of the main flooding season approaches total annual production.
Below-ground biomass
In 1996, three months after the recession of the floods, eight below-ground biomass samples of 0.5 litres were taken horizontally at depths of 0-15 cm and 15-50
cm respectively in the same 16 pits as used for the assessment of soil moisture
(Annex 4.1). Roots and rhizomes with a diameter larger than approximately 0.2
mm were separated by hand from the extremely firm soil, dry cleaned and ovendried for 48 hours at 65°C. No distinction was made between living and dead
below-ground biomass, as the latter was estimated to be less than 10%. The aver-1
-2
age per site was expressed as grams of root l soil, and converted into gm .
Regrowth after burning
In 1995 and 1996, eight and 16 sites respectively adjacent to the soil moisture and
below-ground biomass assessment sites (Annex 4.1) were burnt in late and mid
February respectively, as commonly applied in the area. In the 1995 and 1996
3
plots, 12 respectively eight, 1 m wire-netted exclosures had been placed to prevent
grazing on each of the sites. In April regrowth turned brown and was hand cut at
ground level, irrespective of species composition. For comparison we took ten random samples outside the exclosures at each site. Samples were treated as described
under above-ground biomass.
Forage Quality
In 1994 and 1995, 30 oven-dried samples were selected, targeting sites dominated by the six dominant grass species (Table 4.1). The entire plants were sampled
(cf. Mefit-Babtie 1983), allowing a comparison with the monthly measurements in
the Sudd (Howell et al. 1988) and an assessment of total N-yield (N% x Biomass).
N-content was analysed at the BLGG laboratory (The Netherlands) and converted
to digestible crude protein content, following DCP (‰) = 58.06 (%N) – 35.2 (Boudet 1983).
Data analysis
Normality has been assessed with Kolmogorov-Smirnov test (SPSS 1999), rejecting non-normality for water depth (p = 0.999, 0.59, 0.77 for 1994, 1995 and 1996
respectively), as well as for total standing crop (p = 0.10, 0.30, 0.20 respectively).
Above-ground Oryza longistaminata biomass, excluding 0-sites, non-normality
was rejected (p = 0.84, 0.34, 0.85), no longer when including all sites (p < 0.0001).
The situation with above-ground Echinochloa pyramidalis biomass was mixed: excluding 0-sites (p = 0.27, 0.001, 0.006, all sites (p < 0.0001). For regrowth in 1995
and 1996, non-normality was not rejected (p < 0.0001), nor for under-ground bio-
116
Part II – Impact of ref looding
mass (p < 0.0001). For soil moisture in 1995 at 20-40, 60-80 and 120 cm, p was
respectively 0.32, 0.003, 0.82, in 1996 non-normality was not rejected (p = 0.026).
Above-ground biomass observations from the same plots were compared amongst
years, depending on normality with t-tests or Mann-Whitney (SPSS 1999). The
calculations of correlation coefficients, Pearson’s or Spearmann’s and linear regressions followed SPSS (1999).
4.3
Results
M aximum flood depth and flood duration
In 1994, flooding lasted 5 to 15 weeks, prolonging the rainy season that started with
the first rains in June that ponded on the surface from July onwards. The rains
ended in early October, when floodwater had arrived in all sites. Some sites dried up
in the first week of November, most of the others in the last week of November; the
remaining three sites west in the transect during the second week of December.
Maximum flood depth followed the same pattern in the three observation years and
was in 1996 5.2 and 4.4 cm lower than in 1994 and 1995 respectively (1994*1995,
p = 0.68; 1994*1996, p = 0.007; 1995*1996, p < 0.001) (Fig. 4.2). Flood duration
was correlated with maximum flood depth (R = 0.55, p < 0.001), but with limited
amplitudes, partly due to the two-week interval of flood duration measurements
(Fig. 4.2).
Above-ground biomass
Total above-ground biomass (Fig. 4.3), as well as that of Oryza longistaminata (Fig.
4.4) and Echinochloa pyramidalis (Fig. 4.5), increased with increasing maximum
flood depth. In 1996, in the third year of the reflooding, this relation had become
more pronounced than in the first year 1994. In 1995 the number of observations
was too limited for such calculation. Total above-ground biomass in both 1996 and
1995 was higher than in 1994 (p < 0.0001) and in 1995 equal to 1996 (p = 0.82).
Looking at the individual species, the following can be observed. Sorghum arundinaceum a stout annual reed-like grass that had invaded the desiccated floodplain
in the mid-1980s, produced in 1994 substantial above-ground biomass till a maximum flood depth of 38 cm, in eight out of 32 reflooded sites. In 1996, its occurrence was reduced to two out of 16 reflooded sites with less than 15 cm maximum
flood depth. The limited number of observations due to sorghum’s absence in the
continuously flooded eastern part of the transect did not allow a functional relation to be drawn.
117
total above-ground biomass (g DM m -2)
4 – Veget at ion product ion
1600
1994
1995
1996
1200
y = 12.43 x + 263.7
R2 = 0.68
p <0.0001
1996
800
1994
400
y = 3.88x + 174.5
R2 = 0.25
p =0.004
0
0
20
40
60
80
max. flood depth (cm)
Figure 4.3 – Relat ion bet w een maximum f lood dept h and t ot al st anding crop (0-65 cm max.
f looddept h)
above-ground Oryza longistaminata
biomass (g DM m-2)
1000
1994
1995
1996
y = 11.45x - 48.4
800
R2 = 0.54 p =0.001
1996
600
1995
y = 9.85x - 100.6
R2 = 0.7014 p =0.009
400
1994
y = 5.45x - 64.9
200
R2 = 0.42 p <0.0001
0
0
10
20
30
40
50
60
70
80
max.flood depth (cm)
Figure 4.4 – Relat ion bet w een maximum f lood dept h and above-ground biomass of Oryza
longist aminat a (0-65 cm max. f looddept h)
118
above-ground Echinochloa pyramidalis
-2
biomass (g DM m )
Part II – Impact of ref looding
600
1994
500
1995
1996
y = 6.86x - 52.9
400
1996
R2 = 0.76 p <0.0001
300
200
1994
y = 3.25x - 47.4
R2 = 0.40 p <0.0001
100
0
0
20
40
60
80
max.flood depth (cm)
Figure 4.5 – Relat ion bet w een maximum f lood dept h and above-ground biomass of Echinochloa
pyramidalis (0-65 max. f looddept h)
For wild rice Oryza longistaminata, the dominant perennial grass, above-ground
biomass was a function of maximum flood depth in 1994, a relation that improved
in subsequent years (Fig. 4.4). In 1994, above-ground biomass of Oryza longistaminata was lower than in 1995 and 1996 (p < 0.0001) and in 1996 slightly higher than in 1995 (p = 0.042).
Echinochloa pyramidalis, co-dominant in inundated sites, showed an increasingly
pronounced relation with maximum flood depth (Fig. 4.5). In 1995, the limited
number of observations did not allow such calculation. In 1994, above-ground biomass was lower than in 1995 and 1996 (respectively p = 0.002 and p < 0.0001)
and in 1995 equal to 1996 (p = 0.50).
Below -ground biomass
-2
In the topsoil (0-15 cm), below-ground biomass varied from 8 g DM m (Z32, with
-2
annual vegetation) to 2750 g DM m , double the above-ground biomass (Z29, with
a dense rhizome layer of O.longistaminata). In the 15-50 cm layer, below-ground
-2
biomass was reduced to 3-165 g DM m . Below-ground biomass at both 0-15 and
15-50 cm was correlated with above-ground biomass and with the percentage of
perennials (Table 4.2). Maximum flood depth was correlated with especially belowground biomass at 15-50 cm (Table 4.2). The shoot/root ratio was negatively correlated with the percentage of perennials (R = -0.60, p = 0.014).
4 – Veget at ion product ion
119
Regrow th after burning
Regrowth was higher inside than outside the exclosures (p < 0.05) in one site (S1)
in 1995 and in 1996 in 5 sites (S1, Z8, Z11, P6, Z22). With one exception (Z22),
these sites were situated in the well-flooded eastern part of the transect, outside
Waza National Park (Fig. 4.2). The following presentation will be limited to the
exclosure data.
In 1995 regrowth was correlated with residual soil moisture still present at the end
of the dry season at a depth of 20-40 cm but not with residual soil moisture at 6080 cm, and at 120 cm (Table 4.2). The 1996 data indicated that regrowth was
hardly influenced by mid-dry season soil moisture. Both mid-dry season soil moisture and residual moisture was influenced by flood depth (Table 4.2). A positive
correlation existed between maximum flood depth, below-ground biomass and %
perennials, but these parameters had little influence on regrowth that was only
correlated with soil moisture at 20-40 cm (Table 4.2).
Forage quality and N-yield
Nitrogen content of aboveground biomass was between 0.38-0.74%, with minor
differences between species and not correlated with biomass (R = -0.051, n = 23,
p = 0.82), or with maximum flood depth (R = -0.27, n = 23, p = 0.21). Nitrogen
yield was not correlated with maximum flood depth either (R = 0.26, n = 9, p =
0.50). Regrowth had a N-content between 0.9 and 2.2 %; the limited number of
observations tend to have a negative correlation with above-ground biomass (R =
-0.55, n = 7, p = 0.21).
4.4
Discussion
Comparison w ith other African seasonally-flooded grasslands
Howell et al. (1988) considered the plains around Lake Chad too dry to be compared with the Sudd (Sudan), neglecting the Logone floodplain further south. In
the Nyanya study area in the Sudd, long-term rainfall was 100-150 mm higher than
in Logone (1981: 739 mm; 1982: 777 mm) (Mefit-Babtie 1983). Its species composition was comparable with Logone with Oryza longistaminata and Echinochloa
pyramidalis as dominant species and Hyparrhenia rufa and Vetiveria nigritana
occurring on the edges of the floodplain. In the Kafue flats (Zambia) that have a
-1
somewhat higher rainfall of 880 mm year and comparable soils, Echinochloa stagnina and Vossia cuspidata dominated the deeper flooded parts (> 1 m), as in Logone
(Ellenbroek 1987). The area with shallow floods (< 1 m) had however a different
vegetation composition from Logone with only occasionally Oryza longistaminata
as dominant species (Ellenbroek 1987). In the Inner Niger Delta (Mali), with a low
-1
rainfall of 380 mm year , Oryza longistaminata dominated the parts with a maximum flood depth of about 50 cm, with Echinochloa stagnina and Vossia cuspidata
Table 4.2 – Correlat ions bet w een maximum f looddept h, regrow t h, and int ermediat e paramet ers (1995, n = 7; 1996 n = 16)
1995
1996
1995
1996
Aboveground
Biomass
% Perennials
0-15 cm
1995
1996
1995
1996
M ax. f lood
dept h
R
Sig
1
1.00
1.00
Regrow t h
R
Sig
0.60
.16
.20
.47
1.0
1.0
Above-ground
Biomass
R
Sig
.49
.22
.83
<.0001
.41
.36
.158
.56
1.00
1.00
% Perennials
R
Sig
.70
.054
.69
.003
.49
.27
.44
.086
.46
.25
.63
.009
1.00
1.00
Below ground
Biomass
Below -ground Biomass
15-50cm
1995
1996
0-15
cm
R
Sig
.49
.052
–.011
.967
–
.56
.024
–
.63
.009
–
1.00
15-50
cm
R
Sig
.69
0.03
0.072
.791
–
.59
.016
–
.75
.001
–
.86
<.0001
3
2
3
2
3
2
3
2
2
1995
1996
–
1.00
3
Soil
moist ure
20-40 cm
1995
1996
Soil
moist ure
60-80 cm
1995
1996
Soil
M oist ure
20-40 cm
R
Sig
.54
.17
.53
.035
.88
0.009
.33
.22
.49
.22
.42
.10
.24
.57
.53
.036
–
.66
.006
–
.59
.016
1.00
1.00
Soil
M oist ure
60-80 cm
R
Sig
.16
.71
–
.22
.63
–
–
.31
.45
–
–.34
.42
–
–
–
–
–
.52
.19
–
1.00
–
Soil M oist ure
120 cm
R
Sig
–.68
.061
–
–.64
.12
–
–
–.024
.96
–
–.47
.25
–
–
–
–
–
–.47
.24
–
–.11
.81
–
1
2
3
Pearson Correlat ion Coeff icient (SPSS 1999); Soil moist ure at t he end of t he dry season (M ay); Soil moist ure half -w ay t he dry season (February)
–: not det ermined; in bold: signif icant (p < 0.05)
Soil
moist ure
120cm
1995
1996
1.00
–
Part II – Impact of ref looding
Flooding season:
Regrow t h
dept h
120
M ax f lood
121
4 – Veget at ion product ion
occurring at a maximum flood depth of 1-2.8 m, occasionally up to 4-5 m (Hiernaux & Diarra 1983, 1984a). Biomass measurements in these three areas generally followed the methodology described for Logone, for differences see Annex 4.2.
Information on maximum flood depth was not very accurate and limited the use
of data (Annex 4.2).
In this sample of African seasonally flooded grasslands, above-ground biomass
measured at the retreat of the floods was function of maximum flood depth as well
(Fig. 4.6). Together with the Logone 1996 data, a consistent relation for African
seasonally flooded grasslands, with a maximum flood depth of less than 1 m was
-2
found: Above-ground Biomass (g DM m ) = 14.8 Max. Flood depth (cm) + 242.6
2
(R = 0.69, p < 0.0001).
M aximum flood depth – flood duration
above-ground biomass (g DM m-2)
In addition to flood depth, flood duration has often been considered a main determinant of vegetation composition in wetlands (Ellenbroek 1987; Thompson 1985).
In the Inner Niger Delta (Hiernaux & Diarra 1984a), maximum flood depth was
more strongly correlated with flood duration than in Logone (R = 0.94, p < 0.0001
versus R = 0.545, p < 0.001). I attribute this difference on one hand to the large
number of deeply inundated sites in the Inner Niger Delta and on the other hand
to the large number of fish canals in Logone that accelerate draining (Drijver et al.
1995). The difficulties to measure flood duration and its low amplitude (see above),
3000
Sudd
y=30.1X- 8.94 R2 =0.97
Kafue
y=22.8X 504.9 R2 = 0.66
Inner Delta
y=19.2X + 386.5 R2 = 0.75
2000
combined: Y=15X+273.5
(R2 = 0.68, p < 0.0001)
1000
Logone:Y=12.4X+263.7
0
0
20
40
60
80
100
120
140
max.flood depth (cm)
Figure 4.6 – Relat ion bet w een maximum f lood dept h and t ot al above-ground biomass in t he Inner
Niger Delt a (M ali), Kaf ue f lat s (Zambia) and Sudd (Sudan) (0-120 cm max. f looddept h) (see also
Annex 4.2)
122
Part II – Impact of ref looding
made us concentrate on the maximum flood depth-above ground biomass relation,
discussed below.
Floodplain rehabilitation
In 1995 and 1996, above-ground biomass of all species together as well as of the
individual species Oryza longistaminata and Echinochloa pyramidalis were higher
than in 1994, the first reflooding year. The correlation coefficient between maximum flood depth and biomass also increased after 1994, showing the two-year
time lag in reaction of vegetation production to water level increase. We reported
earlier the initial vegetation stress, due to the sudden rise of the water level, leading to the lack of flowering of Oryza longistaminata and the limited increase in
cover of this species and Echinochloa pyramidalis compared to the following years
(Scholte et al. 2000b). In 1994, Sorghum arundinaceum was still widespread and
attained a high production in the re-flooded sites, but gradually disappeared in following years (Scholte et al. 2000b). This contrasted with the sudden die-off of
Sorghum sp. in standing water in Sudan, reported by Bunting & Lea (1962). Contrary to these changes in species composition and production was the lack of any
noticeable impact on the morphological characteristics of the (top) soils within
this three-year study period (Table 4.1). The impact of soil properties on (regrowth)
production remained therefore unclear.
M aximum flood depth – Above-ground biomass
Flood depth is a well-known environmental determinant of species distribution in
wetlands (Ellery et al. 1991). My results show that maximum flood depth was also
the main environmental factor explaining above-ground biomass production in
the shallowly inundated Logone floodplain.
The difference in intercept between 1994 and 1996 of Oryza longistaminata and
Echinochloa pyramidalis (Figs. 4.4, 4.5) can be explained partly by the 5-cm difference in maximum flood depth (Fig. 4.2). The Rain-Use Efficiency, based on the 0-1 -1
intercept of maximum flood depth (Fig. 4.3), was with 2.5 and 3.5 kg DM ha yr
-1
-1 -1
-1
mm in 1994 and 1996 respectively, well in line with the 1.5-6 kg DM ha yr mm
range noted for the Sudano-Sahelian transition subzone (Le Houerou 1989).
At a maximum flood depth of 60 cm, virtually all biomass consisted of Oryza longistaminata and Echinochloa pyramidalis (Fig. 4.3 versus Figs. 4.4, 4.5). Under deeper flooding, Echinochloa stagnina and Vossia cuspidata were dominant, but restricted to depressions that were not considered in this study.
Micro-topical heterogeneity, caused by gilgai, is expected to have influenced the
2
maximum flood depth – above-ground biomass relation that was assessed in 1 m
plots. An experimental set-up that takes this heterogeneity into account may lead
4 – Veget at ion product ion
123
to the establishment of a stronger maximum flood depth – above-ground biomass
crop relationship.
Regrow th
Mid-late dry season regrowth was very limited, but not neglected by grazing animals as shown by the difference in biomass in and outside the exclosures. The equation of Breman & de Ridder (1991) that relates above-ground biomass and regrowth
-2
would imply a regrowth of 42-147 g DM m with a maximum flood depth of 2060 cm. This is considerably more than the maximum regrowth measurements of
-2
-2
10 g DM m (1995) and 27 g DM m (1996), following mid-dry season burning
as commonly applied in the area. Regrowth in the Sudd area, under similar flood
depths, was also low (Annex 4.2). More regrowth is only produced with a maximum flooding depth exceeding 50-100 cm and early dry season burning (Table
4.4; Hiernaux and Diarra 1984b), contradicting the linear biomass-regrowth relation of Breman & de Ridder (1991). The role of above and under-ground biomass
and key soil properties in the production of regrowth remains to be clarified.
Biomass forage quality
Digestible crude protein (DCP) of regrowth (1.7-9.3%), of above-ground biomass
of Oryza longistaminata (0-0.75%) and of Echinochloa pyramidalis (0%) correspond
to the extremes of the biomass-nutritional quality relationship of Howell et al.
(1988) to which I refer for the development of forage quality throughout the year.
Herbivores select the more nutritious parts of the grasses and target nutritious
regrowth, maintaining a DCP content of 2.9% in their diet in the Inner Niger Delta
(Hiernaux & Diarra 1984b).
Wet season growth was N-limited, as virtually none of the standing crop samples
surpassed the physiological threshold of 0.5% N (Breman & de Ridder 1991;
Boudet 1984). Also the lack of correlation between N-content and above-ground
biomass points to this. In contrast, dry season regrowth has been largely water
limited, but irrigation experiments in the Inner Niger Delta showed that with
-2
higher regrowth (>110 g DM m ), production was greatly enhanced by NP fertilisation (Hiernaux & Diarra 1984a).
Functional explanation of the maximum flood depth – above-ground
biomass relation
Junk et al. (1989) emphasised the local character of a possible flood depth - production relation, arguing that differences in inundation rates regularly disturb
this relationship. The above-discussed time lag of vegetation production (2 year)
and species recovery (1-10 year, Scholte et al. 2000b) to the start of reflooding are
illustrative for such disturbances. Catling (2001), based on a large number of field
measurements, concluded that most floating rices are adapted to an optimum
flooding regime at which their full grain yield potential is expressed.
124
Part II – Impact of ref looding
The growing season is a function of soil moisture that was influenced beyond a
depth of 120 cm, even with shallow flooding. An assessment of the development
of soil moisture throughout the year was, however, beyond the scope of this field
study. The N-growth limitation notwithstanding, results of this study did not support a link between N-yield and maximum flood depth or with the vegetation’s
position in the transect, a measure for the distance water travelled and potentially
an indication of sediment load (Breen et al. 1988; Scholte et al. 2000b).
Flooding drives adapted plants to grow up and keep pace with the rising water level
(Ellenbroek 1987; Breen et al. 1988; Blom & Voesenek 1996; Cronk & Fennessey
2001). This shoot elongation may be practised by the deeply flooded Vossia cuspidata – Echinochloa stagnina communities with relatively weak stems that, supported by the buoyancy of the water, may reach 250 cm or more with a height of
the aerial parts of about 80-100 cm at all seasons (Ellenbroek 1987). The maximum flood depth-above-ground biomass relation and the N-content that reached
physiological threshold values suggest a comparable shoot elongation in the more
shallowly inundated studied communities. The discussed Echinochloa pyramidalis
has relatively strong fibrous shoots, however, that tower, already at the start of the
flooding season above the approaching floodwater, refuting this explanation. In
addition, Oryza longistaminata, the only important C3 grass on the floodplain, is
also known to photosynthesize under reduced light conditions and relatively low
ambient temperatures under water (Ellenbroek 1987; Cronk & Fennessey 2001).
4.5
Conclusion
It can be concluded that changing maximum flood depth, through water extraction or reflooding, not only has an impact on vegetation composition but also on
its production. Based on data from Central as well as West, North-East and Southern Africa, I showed that a rise in maximum flood depth with 1 cm, to a maximum
of 1 m, corresponds to an increase in above-ground biomass with approximately
1
140 kg DM ha . Regrowth after burning possibly increases as well, but is indirectly related with maximum flood depth, and only becomes substantial once maximum flood depth surpasses 50-100 cm. The factors, influenced by maximum
flood depth, that determine plant production such as the length of growing season, soil moisture and induction of shoot elongation, have yet to be clarified. An
experimental follow-up of this study should also take into account the relation of
maximum flood depth with sediment load and other parameters that might influence productivity.
125
4 – Veget at ion product ion
Annex 4.1
Overview of observation sites and measurements along the transect
(see Figure 4.1)
Flooding season
1994
St udy paramet er
1995
1996
Above-ground biomass,
Above-ground Biomass,
Above-ground Biomass,
Fodder qualit y
Regrow t h, Fodder qualit y,
Below -ground Biomass,
Soil charact erist ics and
Regrow t h
-moist ure
Soil moist ure
P7, P6, P5, P4
P7, P6, P5, P4
P7, P6, P5, P4
6x6m
–
SI, S2, S3, S4
SI, S2, S3, S4
6x6m
Z2, Z4, Z6, Z8, Z9, Z10, Z11,
–
Z8, Z11, Z20, Z22,
Plot size
50 x 20 m
Z12, Z13, Z14, Z15, Z17, Z18,
Z26, Z29, Z32, Z34
Z19, Z20, Z21, Z22, Z23, Z24,
Z25, Z26, Z28, Z29, Z30, Z31,
Z32, Z33, Z34
Tot al number
of sit es
32
8
16
M ax. Flood
M ax. t ot al Above
Sample size
dept h 4 (cm)
-ground Biomass (gm -2)5
Replicat es
Dominant species
replicat es
Echinochloa pyramidalis
17±88
0.5 m 2
Oryza longist aminat a
7±48
10 t imes
Oryza longist aminat a
9±58
Oryza longist aminat a
5±38
Echinochloa pyramidalis
7±48
Vossia cuspidat a
1679
(821±337)
Sudd 2
9
277±57
(231±45)
Sudd 3
Sudd 4
10
6
270 ±62
0.5 m 2
(234±63)
10 t imes
254 ±71
(210±70)
Sudd 5
13
283± 35
(102±49)
Kaf ue Flat s 12
120
2486
Echinochloa st agnina
Kaf ue Flat s 2
100
1198
1m
? t imes
Kaf ue Flat s 5
85
1547
? t imes
Vossia cuspidat a
2
Echinochloa st agnina
Oryza longist aminat a
–10
Echinochloa st agnina
–
Oryza longist aminat a
1m 2
Part II – Impact of ref looding
1100 ±220
Annex 4.2
Production characteristics of seasonally flooded grasslands in the Sudd,
Kafue Flats and Inner Niger Delta (see also Fig. 4.6)
36
Sample size
M ax gm -2
(idem dominant species)
Sudd 11
Regrow t h
af t er burning
126
Sit e no.
Sit e no.
M ax. Flood
M ax. t ot al Above
Sample size
dept h 4 (cm)
-ground Biomass (gm -2)5
Replicat es
Dominant species
Regrow t h
Sample size
af t er burning
replicat es
(idem dominant species)
70
17476
Kaf ue Flat s 7
60
350
Paspalidium obt usif olium
–
Panicum repens
Leersia denudat a
Kaf ue Flat s 8
40
4 – Veget at ion product ion
Kaf ue Flat s 6
M ax gm
-2
479
–
Panicum sp.
Set aria sphacelat a
–
Oryza longist aminat a
92.4±6.611
Inner Delt a 79-803
49
1220.5±124
Inner Delt a 80-81
9
471.3±121
1m 2
Oryza longist aminat a
9.2±0.411
Inner Delt a 81-82
24
541.3±37
24 t imes
Oryza longist aminat a
11.9±1.411
4 m2
Inner Delt a 82-83
0
279.1±18
Oryza longist aminat a
–
3 t imes
Inner Delt a
2
774±897
Vet iveria nigrit ana
10.9±3.212
Inner Delt a
49
1582.8±2557
Vet iveria nigrit ana
137.3±9.812
1
2
3
4
Int erpret ed f rom M ef it -Babt ie (1983), see also How ell et al. (1988); Ellenbroek (1987); Hiernaux and Diarra (1984a); Flood dept h has generally only
5
been assessed t w ice, result ing in an underest imat ion of maximum f lood dept h compared t o our st udy; M easured at t he end of t he main f looding sea6
7
8
son (as in Logone); 1747 (t op) → 1491 at moment of drying up; M easured t hroughout t he dry season; M easured mont hly, maximum value present ed,
9
10
11
12
normally 2-3 mont hs af t er burning; ± 2 mont hs af t er drying up, not burned; not det ermined; measured in June , 6 mont hs af t er burning; measured in June, 4 mont hs af t er burning
127
5
Waterbird Recovery in Waza-Logone
(Cameroon), resulting from increased
rainfall, floodplain rehabilitation and
colony protection
Paul Scholt e
Abstract
The impact of ref looding of a desiccat ed f loodplain on w at erbirds in Nort h
Cameroon w as monit ored by January t ot al count s f rom 1992-2000. Bird numbers
in t he dry season increased f rom 60 000 t o 105 000, w hereas t he number of
species surpassing int ernat ional 1% crit eria doubled f rom 6 t o 12. The increase
in Anat idae corresponds t o t heir recovery in West Af rica f ollow ing t he drought s
in t he 1980s. The increase in especially Ciconiif ormes in Waza-Logone w as not
paralleled by similar t rends in ot her main West Af rican f loodplains, suggest ing
t hat f loodplain rehabilit at ion has played an import ant role. The limit ed increase
of t he large piscivorous Ciconiif ormes Lept opt ilos crumenif erus, M yct eria ibis
and Pelecanus ruf escens is associat ed w it h repeat ed dest ruct ion of t heir breeding colonies, possibly leading t o t heir decline as in t he rest of West Af rica. In
cont rast , an Ardea melanocephala colony increased f rom 750 t o 2500 nest s
bet w een 1993-2003. The except ional colony size, a mult iple of t he next largest
know n colonies, implies t hat besides improved habit at due t o ref looding, prot ect ion also played a vit al role. It can be concluded t hat t he increase in w at erbirds
in Waza-Logone is due t o a combinat ion of f act ors: improved rainf all (especially
Anat idae); f loodplain rehabilit at ion (especially omnivorous Ciconiif ormes); and
prot ect ion measures (select ed Af rot ropical Ciconiif ormes).
131
5.1
Introduction
Studies on the population dynamics of waterbirds have centered on the impact of
climatic and hydrological variability. Their scale varied from continent-wide responses to erratic mass flooding in the arid zones of Australia (Roshier et al. 2001),
to reactions to local changes in water regime in temperate France (Paillisson et al.
2002). In line with these population dynamics are the responses of waterbirds to
man-made changes in the hydrology of wetlands, the subject of this chapter.
After the severe drought in Sahelian Africa in 1972-1974, its impact on numbers
of West European migratory waterbirds such as purple heron Ardea purpurea
became clear (Held 1981). In reaction to the 1980s Sahelian droughts, a range of
other studies appeared on the impact of the variation in West African rainfall on
Palearctic birds (Szep 1995; Foppen et al. 1999; Fasola et al. 2000; Schricke et al.
2001; Barbraud & Hafner 2001). The impact of these droughts has been more
obvious than have been the long-lasting effects of water control projects undertaken in many major African wetland systems. Triplet & Yesou (2000) attributed
negative trends in numbers of Anatidae in the Senegal Delta to reduced flooding
following the construction of embankments and dams in the Senegal River. The
impact of drainage works in the Everglades (USA) on breeding waterbirds is
amongst the best documented. Since the 1930s, the number of pairs decreased
from 50% in case of Great Egrets (Casmerodius albus) to 90% for Woodstork (Mycteria americana) (Bancroft 1989).
In response to these negative impacts, wetland rehabilitation efforts have received
increasing attention, especially in Europe and North America, but only on relatively small areas (Middleton 1999). Their impact on a large scale is still poorly
known, the potential importance of rehabilitation for waterbird numbers notwithstanding (Roshier et al. 2001). In Hadjia-Nguru, Nigeria, numbers of Anatidae
were correlated to the size of the area impacted by flood releases (Polet 2000).
Anatidae numbers also correlated with time, suggesting an alternative explanation
in the recovery of waterbird populations, as observed elsewhere in West Africa
after the mid-1980s droughts (Schricke et al. 2001). In addition, high water levels
are not necessarily an indication of suitable habitat for shorebirds and egrets
(Bancroft 1989; Butler et al. 2000), and may even have a negative impact (Crivelli
et al. 1995). An analysis of counts of waterbirds using temporarily flooded rainfed
lakes in Niger showed that bird densities were inversely correlated to the size of
the inundated area, with smaller lakes having higher nutrient loads (Mullié et al.
1999). In addition to these trophic related factors, nest-site availability may also be
a limiting factor for colonial waterbirds (Hafner 2000). These experiences show
the need to study the development of waterbird populations not only in relation to
flood levels, but also to other environmental and anthropological factors.
132
Part II – Impact of ref looding
Figure 5.1 – The Waza-Logone f loodplain and colony locat ions (see sect ion 5.3). Year of
colony dest ruct ion is indicat ed behind t he locat ion name.
133
5 – Wat erbirds
The Waza-Logone floodplain in North Cameroon (Fig. 5.1) was severely degraded
due to restricted water supply resulting from the construction of a dam and
embankment upstream. Built in 1979, these structures caused the disappearance
2
downstream of the perennial floodplain vegetation in an area covering 1500 km .
This habitat loss had a strong negative impact on livestock and fisheries, as well
as on antelopes (Scholte et al. 1996a). Lack of sound data prevented a quantitative
assessment of the effects of the embankment and dam on waterbirds except for
the western Black-crowned Crane Balearica pavonina pavonina. For this endangered species Waza-Logone is a stronghold with about 2500 individuals or 16% of the
estimated population (IUCN 2003; Scholte 1996; Beilfuss et al. 2003). Prior to the
construction of the dam, however, its population numbered over 10 000 individuals (Holmes 1972). Balearica pavonina is resident and the decreased population
size can be attributed to changes in the Waza-Logone area proper. Floodplain degradation following the construction of the dam is the most likely cause thereof.
To counter the ongoing degradation, an embankment was breached in 1994, rais2
2
ing flooding levels in an area of 600 km and rehabilitating 180 km of perennial
floodplain vegetation (Scholte et al. 2000b). Annual total waterbird counts, organized in the framework of the African Waterbird Census (Dodman & Diagana 2003),
formed the main tool to monitor the impact of the reflooding on waterbirds (Scholte et al. 1999a; Scholte et al. 2000a; Fotso et al. 2001).
In this chapter, I analyse the impact of long-term rainfall, floodplain rehabilitation
and protection measures to the observed increases in numbers of waterbirds. To
discern the impact of floodplain rehabilitation from long-term rainfall changes, a
comparison is made with counts from West Africa sites with a comparable rainfall history. An expanding colony of the Black-headed Heron Ardea melanocephala
allowed a complementary assessment of the impact of floodplain rehabilitation on
this terrestrial feeder preferring damp grassland (Brown et al. 1982; Kushlan 2000),
greatly extended with the reflooding (Scholte et al. 2000b). The developments of
Ardea melanocephala numbers are compared with those of other, but unprotected,
colonial breeders in the area. This allowed an appreciation of the importance of
colony protection for the recovery of waterbirds as well.
5.2
M aterials and methods
Study area
2
The Waza-Logone area covers approximately 8000 km in the Far North province
of Cameroon (Fig. 5.1). It receives an average annual rainfall of about 650 mm.
The rainy season lasts from June to September, followed by flooding of the Logone River, in parts of the area, from August to November. Rainfall and flooding are
complementary in their impact on flood depth.
134
Part II – Impact of ref looding
After the floods ceased in 1979, annual grasses invaded the desiccated floodplain.
The reflooding, started in 1994, initiated an annual conversion of 7% of the annu2
ally reflooded 180 km back into productive perennial grasslands (Scholte et al.
2000b). Each year at the end of December, immediately after the drying up, large
stretches of the floodplain are burnt by pastoralists, leaving a bare landscape that
soon turns green again with perennial grasses’ regrowth, attracting livestock, antelopes and birds.
Environmental factors
The Ndjamena station, although located at a distance of c. 100 km, is representative of rainfall in the Waza-Logone area. Discharge of the Logone river, at the entry
of the floodplain (Fig. 5.1), was measured during the flooding season. Discharge
is considered to be a function of rainfall in the upstream catchment area and
determines the extent of flooding in Waza-Logone (Sighomnou 2000). Vegetation
composition and cover were monitored in a 20 x 16 km grid (Scholte et al. 2000b).
Dry season w aterbird counts in Waza-Logone
Waterbird counts correspond to the 1992 and 1994-2000 rainy and flooding seasons. Counts took place in January, halfway through the dry season, except for the
1994 and 2000 flooding seasons when counts were conducted in February. Two
teams, of 2-3 observers each, covered, by car and foot, the entire Logone floodplain
in two weeks for a total waterbird count (Van Wetten & Spierenburg 1998; Dijkstra et al. 2002; Bredenbeek 2004). Each team consisted of at least one observer
who participated in all counts. Special attention was paid to the remaining wet parts
of the floodplain, such as depressions and streams. Observation intensity, expressed
as number of observation days, was not correlated with the number of observed
2
birds (R = 0.05, p = 0.69).
Waterbird species were categorised following the main taxonomic groups, Anseriformes; Charadriiformes, with Laridae and Sternidae kept separate; Ciconiiformes,
plus Pelecaniformes (i.e. Phalocrocorax africana and Pelecanus spp.) and Balearica
pavonina. Ciconiiformes were subdivided according to status and annual lifestyle,
i.e. Palearctic migrants versus Afrotropical residents or inter-Africa migrants (Scholte et al. 1999a), to examine the possible dependency on breeding possibilities in
the area. Trophic characteristics correspond largely to habitat use during the count
period, i.e. depressions for piscivorous birds and damp grasslands for omnivorous
birds.
Ardea melanocephala colony description and counting methods
The Ardea melanocephala colony is situated in Sclerocarya birrea woodland in the
southeastern corner of Waza National Park, 10 km west of the reflooded area (Fig.
5.1). The colony lies next to Andirni (Fig. 5.1), a village home to several Waza National Park guides, who regularly survey the colony and protect it from heron rob-
135
5 – Wat erbirds
bers. The area has witnessed an influx of seasonal labour from Chad, bringing in
cultures with other eating habits that pose a threat to large birds. In 1998, for instance, some of them were caught at night trying to trap young herons.
In 1993, the colony had existed over a decade and was restricted to the woodland
west of the village. In 1994 it expanded into the area south of the village, exposing
it to heron robbers. In 1998 the colony moved again into the western woodland
and in 2002 to the north of the village, where the most effective protection could
be expected. Till 1995, three small Ardea melanocephala colonies (<50 nests) were
found in the direct vicinity of Waza National Park.
Nest counts were carried out during 1993-2000 and again in 2003 in June-August,
after the first heavy rains (Table 5.1). Organisational problems forced us to bring
forward our visits of 1995 and 1998 when the main rains had not yet started. In
1997 the colony was counted both prior to and during the main rains. The heronry was subdivided in five parts, based on position relative to the village. From 7.30
till 12 am, we counted the number of nests and adult herons in each nest tree.
During the August counts in 1993 and 1997, the number of juveniles per nest tree
was counted to evaluate the colony’s success-rate. This number should be considered a minimum, as only the larger juveniles were detectable.
Table 5.1 – Breeding paramet ers of Black-headed Heron Ardea melanocephala colony
1993
dat e of
presence of
Number
Adult /nest
Young/nest
nest t rees
nest s/t ree
count
Bubulcus ibis
of nest s1
(median)2
(median)4
(number)
(median)2
7 Aug
breeding
742
nd 3
76
9.0bc
0.60a
3
1994
23 June
533 present
1160
nd
139
nd
1995
5 June
no
718
2.33a
102
5.8e
1996
2 June
>100 present
1604
1.71c
201
7.0cd
1997
17 M ay
no
1229
2.00ab
157
7.0de
1997
1998
16 Aug
12 June
breeding
no
2012
1372
f
180
10.0b
b
150
8.0c
c
0.94
0.55a
1.80
1999
1 July
1052 present
2418
1.60
167
14.0a
2000
15 July
breeding
2016
1.43d
197
9.0bc
2479
e
180
13.0a
2003
12 July
breeding
t ot al
1.00
1.50
0.56
1
see Fig. 5.4 f or correct ed numbers (explained in Result s)
2
diff erent let t er (e.g. a-b), p < 0.05, if at least one let t er is t he same p > 0.05 (e.g. a-ab)
3
nd: not det ermined, only t ot al know n
4
p = 0.91
9.0
Part II – Impact of ref looding
Other Ciconiiformes colonies
Our year-round presence in the study area in 1992-1997 and frequent contacts
with local communities allowed the survey of colonies of other Ciconiiformes. The
species that were involved were egrets, in particular Cattle Egret Bubulcus ibis and
the large piscivorous species, Marabou Stork Leptoptilos crumeniferus, Yellow-billed
Stork Mycteria ibis and Pink-backed Pelican Pelecanus rufescens.
Data analysis
Categories of the dry season counts were mutually compared with GLM analysis.
Developments in waterbird numbers, rainfall, river discharge and perennial vegetation cover were assessed by linear regression. Developments of the waterbird
counts and colony were described with regression analysis with best-fitted curve.
Median values were compared with Mann-Whitney tests. Correlation between number of nests and adults per nest-trees was investigated with Spearman rank correlation (SPSS 1999).
Laridae & Sternidae
Ruff
Shorebirds (-ruff)
Anatidae
Afrotropical omnivorous
Palearctic omnivorous
Afrotropical piscivorous
Palearctic piscivorous
120000
number of waterbirds
136
100000
R2
Total: 0.67
}
0.24
Ciconiiformes, Pelecanus, Balearica
0.01
80000
0.41
60000
0.63
40000
0.61
0.01
0.35
0.29
20000
0
1992
1994
1995
1996
1997
reflooding
1998
1999
2000
flooding season
Figure 5.2 – Dry season w at erbird numbers in Waza-Logone
Legend: R2 of linear regressions, in bold if signif icant (p < 0.05)
137
5 – Wat erbirds
Results
Waterbird dry season numbers
Total waterbird numbers increased from 60 000 to 105 000 during 1992-2000
(Fig. 5.2). Palearctic migrants made up 53% of the observed birds. Their numbers
2
fluctuated strongly from year to year and did not show an increase (R = 0.12, p =
0.40). The Afrotropical waterbirds on the other hand averaged 47% of the obser2
vations and showed a marked increase during the observation period (R =0.60,
p = 0.025).
In the period 1992-2000, 19 water bird species were present in numbers surpassing 1% of their known population size (Wetlands International 2002). The number of species passing this criterion found in a single year doubled between 1992
and 2000 (Fig. 5.3). Species that surpassed the 1% criterion in all years were Whitefaced Whistling Duck Dendrocygna viduata, with numbers up to 6% of known
population size, Balearica pavonina pavonina (up to 16%) and Collared Pratincole
Glareola pratincola (up to 39%).
The total number of waterbirds per category increased significantly in the years
between 1992 and 2000 (F = 19.3, p < 0.0001, df = 1,55) for all categories. The relative increase in bird numbers over time did, however, not differ significantly
amongst the categories (F = 1.4, p = 0.23, df = 7,48). The absolute increase in num-
14
number of species present with > 1%
of population size
5.3
y = 3.56Ln(year) + 4.28
12
2
R = 0.63 p = 0.019
10
8
6
4
2
0
1992
1994
reflooding
1995
1996
1997
1998
1999
2000
flooding season
Figure 5.3 – Number of w at erbird species in Waza-Logone surpassing 1% of know n
populat ion size, 1992-2000.
138
Part II – Impact of ref looding
bers of waterbirds was mainly due to the Afrotropical Anatidae such as Sarkidiornis melanota and to Afrotropical omnivorous Ciconiiformes, in particular Ardea
melanocephala and Bubulcus ibis (Fig. 5.2). Other species that showed at least a fivefold increase were Squacco Heron Ardeola ralloides (Palearctic piscivorous Ciconiiformes) and the shore birds Wood Sandpiper Tringa glareola and Black-winged Stilt
Himantopus himantopus (Palearctic), Jacana Actophilornis africana and Spur-winged Plover Vanellus spinosus (Afrotropical). No increase was observed for Ruff Philomachus pugnax nor for migratory omnivorous Ciconiiformes, whereas numbers of
the large resident piscivorous Ciconiiformes Leptopillos crumeniferus, Mycteria ibis
2
and Pelecanus rufescens fluctuated over time (R = 0.11).
The preceding season’s rainfall was not correlated with numbers observed for any
of the waterbird species or categories counted. Logone discharge, correlated with
the flooding extent, was only correlated with total number of waterbirds observed
2
(R = 0.61, p = 0.038).
Ardea melanocephala
Dry season numbers of Ardea melanocephala increased only from 1997 onwards
reaching 5800 in 2000 (Fig. 5.4), largely surpassing 1% of the known population.
In the early 1995, 1997 and 1998 colony counts, Ardea melanocephala was present
in large numbers but not yet breeding, as shown by the higher number of adults
per nest compared to later counts (p < 0.0001) (Table 5.1). The number of adult
herons per nest in the early 1997 count was intermediate to the early 1995 and
1998 counts (Table 5.1), which have been corrected with the 1997 late/ early season ratio of 1.64.
The size of the colony increased steadily from 742 nests in 1993 to 2418 nests in
1999 and subsequently leveled off (Fig. 5.4). There was no correlation between the
2
number of Ardea melanocephala nests and rainfall (R = 0.27, p = 0.19), nor with
2
Logone discharge (R = 0.14, p = 0.40). Perennial vegetation cover in the reflood2
ing impact zone was correlated to the number of nests (R = 0.92, p = 0.001). The
relation with preceding year’s perennial vegetation, expected to have supported
2
the population during the harsh dry season, was however less pronounced (R =
0.63, p = 0.032), suggesting non-causality.
Other Ciconiiformes colonies
Bubulcus ibis nested scatteredly in villages in the area, with no indications of disturbance. Traces of other egret colonies were found in the northern part of the
area that showed signs of destruction. Marabou stork Leptoptilos crumeniferus colonies have been subject to frequent disturbances, often targeting the young for consumption. After the destruction of the Zina town colony in 1994 (±250 nests), colonies with about 50 nests were found in the towns of Pouss and Guirvidig (Fig. 5.1).
139
5 – Wat erbirds
number
6000
nests A.melanocephala
dry season numbers A. melanocephala
5000
dry season numbers large piscivorous
Ciconiiformes.
4000
3000
y = 1073Ln(x) - 50.6
R2 = 0.91
2000
1000
0
1992 1993 1994 1995 1996 1997 1998 1999 2000
2003
flooding
season
reflooding
Figure 5.4 – Breeding and dry season Ardea melanocephala and large piscivorous w at erbirds,
Waza-Logone 1992-2003.
A Leptoptilos crumeniferus colony was found in 1996 near Halé on the border of
Waza National Park, where at least 210 nests were destroyed in March 1997 (Fig.
5.1). No colonies were found in 1998. In 1999, Leptoptilos crumeniferus was found
breeding in the center of Waza NP (80 nests), where the colony grew steadily to
over 500 nests. Young were robbed again in 2002 after which the site was abandoned. Till 1997, a small colony of Mycteria ibis (±60 nests) and pink-backed pelican Pelecanus rufescens (±30 nests) found relative protection inside Waza National
Park (Fig. 5.1). Villagers explained that its subsequent disappearance was caused
by destruction by fishermen. Only once we found a larger Mycteria ibis colony with
550 nest-building individuals near Mahé village in October 1997 (Fig. 5.1).
5.4
Discussion
General developments in w aterbird numbers
The standardised itinerary and stable personnel assured a constant counting intensity, not achieved in most African waterbird counts (Dodman & Diagana 2003).
We found, however, large inter-annual fluctuations, especially of Palearctic birds,
also observed in Niger (Brouwer & Mullié 2001). Initially the reflooding resulted
in an increase in large piscivorous species, i.e. Great White Pelican Pelecanus onocrotalus, subsequently followed by an increase in smaller piscivorous egrets. I attrib-
140
Part II – Impact of ref looding
ute these developments to increasing fishing intensity that targets large fish (Scholte
et al. 1998; Scholte 2003a).
Increased ornithological importance since the reflooding
Published waterbird population estimates seldom cover Central Africa (Wetlands
International 2002). A choice for (North-) East African instead of the applied WestAfrica population estimates would have given 17 species surpassing the 1% criterion. For species such as White Stork Ciconia ciconia, a choice between West and
East African populations ignores the position of Central Africa at the crossroad of
these populations (Brouwer et al. 2003).
Not only the number of species surpassing the 1% criterion has increased but also
the observed numbers of each of these species important for conservation. The
endangered Balearica pavonina pavonina, for example, showed an upward trend
2
(R = 0.48, p = 0.057).
Comparison w ith other West African w aterbird counts: impact of
generally improved rainfall conditions
The Sahel region from Senegal to Chad, including the Waza-Logone area, was distinguished as a region with a strongly correlated rainfall pattern (Gommes & Petrassi 1996). In the 1950s and early 1960s, rainfall was well above long-term-average. After this, a gradual decline took pace which included the Sahelian droughts
of 1969 and 1971-74. This ‘first Sahelian drought’, was followed by a relatively
humid period that ended with the severe 1983-85 drought, after which a gradual,
though highly fluctuating, increase in rainfall has taken place (Beauvilain 1995;
Gommes & Petrassi 1996).
To discern the impact of generally increased rainfall from the impact of floodplain
rehabilitation, developments in waterfowl numbers in Waza-Logone were compared with January counts from other West African countries (Dodman & Diagana 2003). We applied the same taxonomic and functional categories as distinguished above. The only data for which observation effort was more or less constant is from the Senegalese part of the Senegal Delta (Schricke et al. 2001; Triplet
pers.comm. 2004) and, looking at Anatidae only, from Hadjia-Nguru in neighboring Nigeria (Polet 2000; Dodman & Diagana 2003). Although only covering
the period 1992-1997, the Niger country-wide data was used because of the prominent presence of Ciconiiformes, as in Waza-Logone (Brouwer & Mullié 2001).
The total numbers of birds per distinguished category (Fig. 5.2) in the Senegal
Delta did not show an increase (F = 1.95, p = 0.17, df = 1,62), but the categories differed significantly in their relative increase (F = 7.05, p < 0.001, df = 7,55). This
was due to the remarkable increase in Afrotropical and Palearctic Anatidae in Sene2
gal (R = 0.73, p = 0.004). This was paralleled in 1993-1994 in Niger and in 1994-
5 – Wat erbirds
141
1997 in Nigeria. Anatidae numbers in Niger and Nigeria dropped since 1994 and
1998 respectively, possibly related to their concurrent increase in Waza-Logone.
Shorebirds that increased in Waza-Logone, did not show any trend in Niger nor
in Senegal. The lack of trend in numbers of Philomachus pugnax in Waza-Logone
(Fig. 5.2) was paralleled in other West African wetlands (Trolliet & Girard 2001).
Omnivorous Ciconiiformes, including herons, showed an increase from 1992-2000
in Waza-Logone but not elsewhere in West Africa. This was partly due to the sudden increase of Bubulcus ibis in 1994 in Niger, the Senegal Delta, albeit in low
numbers, and up to 260 000 in the Inner Niger Delta in Mali, attributed to the
exceptional 1994 flooding (van der Kamp et al. 2002a). A comparable increase of
Bubulcus ibis in Waza-Logone took place in 1998 with high numbers still present
in 1999. A comparison of the population trend of the Ciconiiformes Ardea melanocephala, Mycteria ibis and Leptoptilus crumeniformes between Waza-Logone and
other West African sites is hampered by the rarity of these species in the latter
sites. The smaller piscivorous Ciconiiformes, i.e. egrets, showed widely fluctuating numbers in all areas, with exception of Ardeola ralloides with its upward trend
in Waza-Logone. Of the omnivorous Ciconiiformes, the abundant Bubulcus ibis and
the rare Ardea melanocephala showed a 15% decrease (Van der Kamp 2002b).
Amongst the few increasing resident piscivorous Ciconiiformes in the Inner Niger
Delta one should note Phalocrocorax africanus (van der Kamp 2002b).
Improved rainfall conditions may largely explain the increase of Anatidae in WazaLogone. The increase of especially Ciconiiformes relied on other factors, of which
floodplain rehabilitation and colony protection will be discussed below.
Impact of floodplain rehabilitation on w aterbird numbers
The impact of floodplain rehabilitation, induced by the 1994 reflooding, was twofold: a change of annual into perennial vegetation in an area of approximately 180
2
km (Scholte et al. 2000b) and an increase in vegetation production and water
2
availability in an area of 600 km (Chapter 4).
In 1994, the first reflooding year, we compared waterbird diversity in always-flooded perennial grassland, reflooded grassland with still a predominantly annual
grassland vegetation that would convert in later years into perennial grassland and
degraded annual grassland vegetation, not subject to reflooding (Scholte et al.
2000a). Recently reflooded and especially undisturbed flooded habitats with a
water depth of 20-35 cm showed the highest density of passerines and waterbirds.
These observations correspond to the observed range for small herons and large
herons in the Everglades (USA) with a water depth of 10 to 40 cm respectively
(Bancroft, 1989) and in flooded grasslands in the Inner Niger Delta, Mali (van der
Kamp et al. 2005). With the progressing floodplain rehabilitation, differences be-
142
Part II – Impact of ref looding
tween reflooded and undisturbed habitats in Waza-Logone have gradually disappeared (Scholte et al. 2000b).
The impact of the floodplain rehabilitation is limited to about a third of the yearly-inundated floodplain. The floodplain rehabilitation targeted the area adjacent to
Waza National Park, where wildlife, including (resident) waterbirds receive protection. This may explain the relatively large impact of reflooding, further enhanced
by colony protection as discussed below.
Ardea melanocephala versus other Ciconiiformes: linking colony protection w ith floodplain rehabilitation
The number of Ardea melanocephala counted in the Waza-Logone floodplain from
1992-2000 equaled 25-80% of the numbers counted in 23-36 African countries
2
(excluding Chad and Sudan). Only in the 10 000 km Sudd area in Sudan, surveys
in the early 1980s revealed Ardea melanocephala numbers comparable to those in
Waza-Logone (Howell et al. 1988).
From 1993 to 1999, total dry season Ardea melanocephala numbers were lower
than numbers counted at the colony (Fig. 5.4). A. melanocephala also showed a time
lag of three – four years between the reflooding and the increase in dry season numbers, longer than other waterbirds (Fig. 5.2). These observations may be attributed
to dispersal of juveniles and young adults that was reported in Southern Africa to
cover distances of up to 800 km (Tarboton 1977). Presented results showed that
the direct impact of rainfall during the study period on the increase of the Ardea
melanocephala population was limited. Ardea melanocephala dry season increase
was not related with the Logone river discharges that influence the flooding in the
entire Waza-Logone floodplain. I suspect that the large changes due to reflooding
have masked the influence of rainfall and flooding.
With 742 nests (Fig. 5.3), the Ardea melanocephala heronry was in 1992 already
three times the size of the largest known colonies in Kenya (North 1963; Parsons
1976; Taylor 1972) and Somalia (Ash & Miskell 1998) and a multitude of the maximum size of 100-150 reported from South Africa (Skead 1964), Tanzania (Beesley
1972) and West-Sudan (Wilson 1982). In West Africa, reported colony sizes are
smaller than 100 nests (Elgood et al. 1994; Cheke & Walsh 1996; van der Kamp
et al. 2002b).
The number of breeding pairs in the Ardea melanocephala colony increased from
the first reflooding year onwards. The constant size of the colony since 1999, suggests that availability of feeding in the colony surroundings limits further increases (Fasola & Barbieri 1978; Gibbs et al. 1987). Low breeding success, less than one
young per nest (Table 5.1), was also reported from E.Africa (Brown et al. 1982).
The increase in colony size in especially the first years can not be explained by the
5 – Wat erbirds
143
output of the colony only and must have been supplemented by immigration. This
reasoning is supported by the above-mentioned reduction in number of smaller
colonies in the area, probably due to human pressure.
Hafner (2000) argued that exceptionally large nesting assemblages are, on a regional scale, caused by a shortage of suitable nesting places. The increased Logone colony
occupied an increasing number of nesting trees that did not lead to an increasing
number of nests per tree (Table 5.1). This suggests ample tree availability. The
colony’s location, in the eastern woodland area on the edge of the floodplain, predicts ample tree availability in the near future as well. The availability of nest trees
close to the village, source of protection, may be of crucial importance however.
The development of the large Ardea melanocephala colony in Waza-Logone is reminiscent of the grey heron Ardea cinerea population in France. The latter recovered,
after severe prosecution, from almost nil in 1900 to 19 000 pairs in 1990. In
1958, more than half of the French population was found in one single well-protected colony of 1 300 pairs, subsequently splitting into several new colonies (Marion et al. 2000). In the Netherlands, the Ardea cinerea population increased from
about 6 500 pairs in 1926 to 10 000 pairs at present, while the number of colonies
increased from 130 to 500 due to increased protection. Large colonies with over
500-1000 nests, in well-protected privately owned woodlands as occurred during
and just after periods of prosecution, are no longer found (Braaksma et al. 1950;
Van der Weide 2002). The Ardea melanocephala colony in Waza-Logone reminds
of the earlier European situation, although the tendency, towards increased prosecution, is the reverse.
The lack of increase of large piscivorous Leptoptilos crumeniferus, Mycteria ibis and
Pelecanus rufescens in 1992-2000 in Waza-Logone is attributed to the repeated destruction of their colonies. These species have become increasingly rare in the
Inner Niger Delta and the Senegal Delta, despite their rich fish resources (Van der
Kamp et al. 2002a). Consumption of chicks by immigrants has often been the direct
cause of colony disturbance in Waza-Logone. I expect that increasing competition
for fish resources has made local communities indifferent to the destruction of
the colonies in their territories. In the Inner Niger Delta in Mali, where larger piscivorous birds have already become rare, reproduction sites of the remaining smaller piscivorous Egretta species have now come under increasing pressure (Van der
Kamp et al. 2002b, 2005).
144
Part II – Impact of ref looding
5.5
Conclusions
Developments in the waterbird populations of Waza-Logone show the link between
long-term rainfall conditions and floodplain rehabilitation. To quantify the relative
contributions of these factors a more thorough knowledge of regional waterbird
population dynamics is necessary to reveal possible exchanges and compare developments between West-central African wetland areas. Long-term monitoring, with
constant or at least quantified effort, is vital in this context.
The observations in Waza-Logone show that in addition to improved rainfall conditions and floodplain rehabilitation, colony protection has played a crucial role for
at least some of the Afrotropical waterbirds. The role of colony- and habitat protection is likely to become more prominent in future because of the increasing
human pressure on floodplain resources (Scholte 2003a). Initiatives such as the
colony protection at Andirni merit further support to ensure their continuity, possibly through a modest touristic exploitation that elsewhere has proven its success
(Bouton & Frederick 2003).
6
Antelope Populations in Waza National
Park (Cameroon) from 1960 till 2001:
Impact of changes in rainfall, hydrology
and human pressure
Paul Scholte, Saleh Adam and Bobo Kadiri
Abstract
Antelopes are amongst the most prominent wildlife in Waza National Park
which, situated in the Sahelo-Sudanian savanna of Cameroon, has witnessed
changes in rainfall, hydrology and human encroachment. To assess their impacts,
we reviewed 26 surveys, comprising total, transect and localised counts, both
aerial and terrestrial. Estimated Kob numbers dropped from 20 000 in the 1960s
and 1970s to 2000 in the mid-1980s, recovering to 5000 in the 1990s. Estimated
Korrigum (‘Topi’) and Roan numbers dropped already in the early 1970s, for reasons largely unknown, and slightly recovered to respectively 2000 and 1000 in
the 1990s. The diversity of counting methods notwithstanding, the drop in Kob
numbers and the extinction of Waterbuck can be attributed to the construction
of the upstream Maga dam in 1979 and subsequent low rainfall. Yet population
structure data showed that Kob’s reproduction capacity was not severely hit and
Kob recovered in the late 1980 and early-mid 1990s, but did not persist, despite
increased flooding. Although few signs of poaching have been reported, it
might have been important in periods of drought, when Kob and Korrigum left
the protected area.
147
6.1
Introduction
Waza National Park (NP), situated in the Sahelo-Sudanian zone in North Cameroon,
has been the subject of ecological studies such as those on Giraffe Giraffa camelopardalis L. (Ngog Nje 1983, 1984), Elephant Loxodonta africana Blumenbach
(Tchamba 1995, 1996) and Lion (Panthera leo L.) (Bauer 2003). Waza NP also harbours large populations of Kob (Kobus kob Erxleben), Korrigum (Damaliscus lunatus korrigum Burchell), Roan (Hippotragus equinus Desmarest) and Red-fronted
Gazelle (Gazelle rufifrons Gray), as well as minor populations of Bohor Reedbuck
(Redunca redunca Pallas) and Common Duiker (Sylvicapra grimmia L.). These antelopes are important resources, from a conservation as well as tourist perspective
(Flizot 1962; East 1999). Previously published antelope population assessments
in Waza NP (Esser & Lavieren 1979; Tchamba & Elkan 1995) described the distribution of antelopes based on a single survey in December, at the end of the
flooding season, underestimating the role of the floodplain during the subsequent
dry season. Their limited time-span also hampered an assessment of the changes
triggered by droughts, natural as well as man-made.
The Waza-Logone floodplain was severely degraded when it was cut off from its
main water supply through the construction of a dam upstream. Built in 1979,
this dam has caused the disappearance downstream of the perennial floodplain
2
vegetation in an area of 1500 km , including approximately a third of Waza NP,
with a dramatic impact on livestock and fisheries. To counter this ongoing degradation, the IUCN Waza-Logone project was started. In 1994 an embankment that
had closed off a watercourse since 1979 was breached, bringing back the annual
2
floods in an area of 180 km . This triggered an annual 7% restoration of perennial floodplain vegetation in the reflooding impact zone, including a part of Waza
NP and a tripling of livestock densities (Scholte et al. 2000b; Scholte 2003a).
It was argued that antelope populations also suffered from the decreased flooding
and should recover after the 1994 reflooding. To assess the antelope populations
during the floodplain rehabilitation, annual counts were initiated at the waterholes
in Waza NP still holding water at the end of the dry season. However, changes in
flooding cannot be interpreted in isolation from the inter-annual fluctuating rainfall. We therefore reviewed all 26, mostly unpublished wildlife counts, conducted
from 1960 till 2001 with varying target species and methodology. The number of
counts and their repartition in time allowed an assessment of the impact on antelope populations of the dramatic changes in hydrology and rainfall that Waza NP
witnessed in the second part of the 20th century.
148
Part II – Impact of reflooding
6.2
Materials and methods
Study area
Waza NP covers 160,000 ha and lies in the transition zone between the Sudan
and Sahel savannas. The sandy soils in the slightly elevated south-western and
southern part of Waza NP support a wooded Sclerocarya birrea and Anogeissus leiocarpus savanna with Hyperthelia dissoluta and other Andropogon-like perennial
grasses. On the transition towards the lower parts of the area, this wooded savanna is replaced by Lannea humilis open shrubland on planosols with a sparse cover
of annual grasses and herbs (eg. Schoenfeldia gracilis, Blepharis linariifolia). Tree
cover has been steadily decreasing since the early-1970s (Esser & Lavieren 1979;
Piet Wit pers. comm.), probably due to lower rainfall and the late fire regime
applied. Most of Waza NP is dominated by vertisols, which are or used to be subject to flooding. Vegetation types differ in relation to the depth of present and past
flooding. In general, the clayey soils above the 1979 flood line, mostly in the central part of Waza NP, are dominated by Acacia seyal shrublands. The areas formerly flooded, which used to be productive perennial grasslands, are now covered
by annual herbs and grasses especially Sorghum arundinaceum. Flooding arrives
as a slowly progressing sheetflow from the east and south-east, the over-flow of the
Logone and its branches, from over 10 km of the Waza NP border. In the flooded
(north)-eastern part of Waza NP, covering ca. 30.000 ha, perennial grasses such
as Echinochloa pyramidalis, Hyparrhenia rufa and Oryza longistaminata and Vetiveria
nigritana dominate and on the occasionally flooded parts further west the biannuals Ischaemum afrum and Panicum anabaptistum. These rather monotonous grasslands are broken by (manmade) dwelling mounds, previously inhabited by fishing
communities, wooded with Balanites aegyptiaca and Tamarindus indica and often
with a waterhole in their vicinity. Each year, immediately after the area dries up at
the end of December, many parts are burned, leaving a bare landscape. Only some
local humid spots remain with green regrowth and concentrations of wildlife.
Rainfall and flooding
The rainy season is from June to September, followed in the eastern parts by flooding from the Logone river and its branches from August to November. During the
0
subsequent dry season (December-May) temperatures rise as high as 45 C, and
the only remaining water sources are a number of waterholes. With a long-term
average annual rainfall of about 650 mm, large inter-annual differences have been
recorded, as measured at nearby Ndjamena. Four rainfall periods have been identified throughout the West-Central Sahel, amongst others based on the N’djamena
data-set (Gommes & Petrassi 1996). In the 1950s and early 1960s, rainfall was
well above long-term-average after which a gradual decline has taken place resulting in the Sahelian droughts of (1969-) 1971-74. This ‘first Sahelian drought’ since
memory of the present generation, was followed by a relatively humid period that
6 – Antelopes
149
ended with the severe 1983-85 drought, after which a gradual, yet highly fluctuating, increase in rainfall has taken place (Fig. 2.3).
Antelope population numbers were compared with developments in annual rainfall that has been averaged for the preceding three rainy seasons. This choice was
based on the rate of conception of most adult female antelopes with a time-lag of
at least 10-12 months, i.e. two preceding rainy seasons, standardised into the mobile
average over three years (Beauvilain 1995).
Flooding of the Logone River ceased during the 1971-74 and the 1983-85 droughts.
The first cession of flooding coincided with the construction of the Waza-Kousseri
road that has cut several watercourses feeding the northern and northwestern part
of Waza NP. It was, however, the 1979 Maga dam construction that has had a profound impact on the floodplain, both in and outside Waza NP, by reducing flooding from the Logone River and secondary water courses.
Review of 1960-2000 count methodology
To facilitate comparison, we categorised counts into total, transect and local counts
conducted by either terrestrial or aerial surveys (Scholte 2000a), Table 6.1. For
reasons of standardisation, indicated years correspond to the preceding rainy- and
flooding season and not necessarily the survey year.
•
Total terrestrial counts
- Waterhole counts. Two observers, a park guard and an observer, were placed
from 6 am till 6 pm, during two consecutive days, at each of the 14 waterholes still holding water at the end of the dry season (April, May), The ex0
treme heat (45 C), low air humidity (<20%) and limited shade, made all but
the wariest species (Giraffe and Ostrich Struthio camelus L.) drink in the
observers’ presence. Numbers, time of drinking and species have been
recorded. In the open floodplain mainly, the number of animals that did
not drink was counted as well. Animals outside Waza NP, mainly Kob and
Korrigum, were counted by car during the following days.
- Counts by car and from strategic observation points focussed on the areas
around waterholes where antelopes concentrated at the end of the dry season. Flizot, park director during the 1960s and early 1970s, probably also
followed this method.
•
Total aerial counts. For Kob, concentrated around water holes, this method has
been used as reinterpretation of the 1991 transect aerial count that targeted elephants and was biased towards water holes (Tchamba & Elkan 1995), Table 6.1.
The 1974 counts of Vanpraet (1977), by helicopter, lacking a methodology description, were presented as total counts as well.
150
Table 6.1 – Background information on antelope counts in Waza NP
Month
Species
targeted
Total /
Transect2
Aerial/Terrestrial3
(sampling intensity)
Floodplain outside
NP included?
Author
1960
1967
1974
1974
1975
1976
1977
1980
1982
1982
1986
1987
1988
1990
1991
1991
1993
1994
1995
1996
1997
1997
1999
1999
2000
2001
General
General
April
May, Sept.
April
January
Dec.
April
Dry season
April
March - May
Idem
April
April
Dec.
Feb. - May
End April
End April
End April
End April
March, April
April
April
End April
April
End April
all
all
Kob
all
Kob
all
all
Kob
Kob
Kob
Kob
Kob
Kob
Kob
Elephant
all
all
all
all
all
Kob + Korrigum
all
all
all
all
all
TO
TO
TOP
TO/TR
TOP
TR
TR
TOP
TO
TOP
TO
TO
TOP
TOP
TR1
TR
TO
TO
TO
TO
TO
TR
TR
TO
TR
TO
T
T
T, NE corner only4
A (7%)
T, NE corner only4
A (13%)
A (13%)
T, NE corner only4
T
T
T
T
T, NE corner only4
T, NE corner only4
A (14%)
T, road
T, waterhole
T, waterhole
T, waterhole
T, waterhole
T, road
T, foot (31%)
T, foot (?)
T, waterhole
T, foot (31%)
T, waterhole
No
No
Flizot (1962)
Flizot (annual report 1968)
Ecole de Faune (internal report
Vanpraet (1977)
Ecole de Faune (internal report
Esser & Van Lavieren (1979)
Esser & Van Lavieren (1979)
Ecole de Faune (internal report
Thal (unpubl)
Ecole de Faune (internal report
Korthof (unpubl)
Njiforti (unpubl)
Ecole de Faune (internal report
Ecole de Faune (internal report
Tchamba & Elkan (1995)
Bos & Bus (unpubl)
Scholte (unpubl)
Scholte (unpubl)
Scholte (unpubl)
Bobo (unpubl)
Hebou (unpubl)
Bobo (unpubl)
Saleh & Njiforti (unpubl)
Saleh (unpubl)
Saleh (unpubl)
Saleh (unpubl)
1
No (?)
No
No
?
?
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
2
Kob data have been recalculated as of total count (see text); TO: Total count; TR: Transect count, TOP: Total count of part of the area;
3
4
2
A: Aerial; T: Terrestrial; In an area of 16 km in the NE corner of Waza NP, where Kob concentrates in the dry season.
‘75)
‘76)
‘81)
‘83)
‘89)
‘91)
Part II – Impact of reflooding
Year
151
6 – Antelopes
•
Transect counts, terrestrial
- Road count. While driving every accessible track in the park, observed animals were counted and their perpendicular distance to the road estimated.
Observations were stratified into the four main landscapes (see above) and
corrected with a detection curve (Burnham et al. 1980). Roads have, however, been designed for tourist purposes and pass the areas with water holes,
where wildlife concentrates. Tracks also tend to follow the Lannea humilis
shrubland, with its smooth soil surface, an area preferred by Red-fronted
Gazelle, Korrigum and Roan (Esser & van Lavieren 1979; Vanpraet 1977).
To avoid the waterhole bias, an alternative calculation was made, excluding
observed animals at a distance of less than 1 km of the waterholes.
- Counts on foot. The park and neighbouring floodplain were subdivided in
78 5-15 km long transects. Garoua Wildlife College students, assisted by park
tourist guides, counted in a fixed band of 200 m on each side of the observer. Difficulty existed with the assessment of the bandwidth, especially in the
floodplain grasslands with little reference points.
•
Aerial Transect counts. Counts were conducted with a four-seated Cessna plane
at an altitude of 100 m in fixed bands of 100 or 200 m (Esser & van Lavieren
1979; Tchamba & Elkan 1995) and with an Alouette helicopter (van Lavieren
pers. comm.; Esser & van Lavieren 1979).
•
Local counts. A 16 km triangle in the NE floodplain part of Waza NP, where
kob antelope concentrates, was counted by 40-50 observers, ‘sweeping’ the
area in bands of 200-300 m (van Lavieren 1976).
2
Kob population structure
During total counts in 1997, Kob’s sex and age class was assessed to detect possible trends in population, allowing a comparison with assessments in 1972 and
1986 during total counts as well. Methodology followed Wanzie (1988).
6.3
Results
Importance of the floodplain
Throughout the dry-season, almost the entire Kob population was found in the
floodplain (Fig. 6.1). Korrigum and Roan moved from the Sclerocarya woodland
and A.seyal shrublands into the floodplain during the second part of the dry season. A major part of the Korrigum population left the floodplain at the onset of the
rains, when rainwater stagnated in small pools in the Sclerocarya woodland and
neighbouring Lannea humilis shrubland and were not observed in the waterhole
counts. Korrigum that remained in the floodplain continued drinking at the waterholes (Fig. 6.1), but their numbers were limited (Fig. 6.2: 1994, 1996 and 2001).
Part II – Impact of reflooding
percentage of all observations
100
80
60
40
20
Kob
Korrigum
Roan
Red-fronted Gazelle
Progressing dry season
M
ay
.0
1
M
ay
.9
4
99
3
Ap
ri l
.1
.1
99
1
91
Fe
b
D
ec
.1
9
77
0
D
ec
.1
9
152
First rains
Figure 6.1 – Percentage of antelope observations in the floodplain. Details of count, see
Table 6.1
At the onset of the rains, Roan also left the floodplain (Fig. 6.1) but contrary to korrigum, Roan were included in the counts when drinking at the waterholes in the
A.seyal shrubland (Fig. 6.3). Only a limited, stable part of the Red-fronted Gazelle
population used the floodplain (Fig. 6.1). Whereas most antelopes came from some
distance drinking at the waterholes during day-time, an important part of the Kob
population was already present at dawn. After drinking in the morning, Kob generally did not leave the area around the waterhole and sometimes also drank in the
afternoon.
Population numbers
Korrigum population numbers were strongly correlated with numbers of Roan
and to a lesser extent with Bohor Reedbuck and Kob. The population developments
of Bohor Reedbuck and Kob, both floodplain species, were not correlated (Figs. 6.2,
6.3; Table 6.2). Only Kob population numbers showed a positive correlation with
rainfall, whereas numbers of Red-fronted Gazelle were negatively correlated with
rainfall (Table 6.2).
A drop in Kob population numbers occurred in the early 1970s, coinciding with
the 1969-74 drought and the lack of flooding (Fig. 6.2; Table 6.2). In 1971, 1972
and 1982, Kob losses, extrapolated from respectively 198, an unknown and 400
153
Number
6 – Antelopes
30000
KOB
KORRIGUM
25000
1
overestimation
20000
underestimation
2
15000
10000
5000
6
5
3
4
3
7
4
4
0
'60
'67
'74
drought
'76 '77
'82
dam
'86 '87
'91
drought
'93 '94 '95 '96 '97
reflooding
'99
'01
Year
1 Probably biased by experiences of the early 1960s, see text
2 A concentration of Kob outside the survey area (i.e. NE of the park) was observed (van Lavieren, pers.
comm.)
3 Targeting Kob around waterholes only (Table 6.1)
4 Early rains caused Korrigum leaving the floodplain, no longer drinking at waterholes (see text)
5 Estimations by road counts were 10730 Kob and 2800 Korrigum, corrected for waterhole bias (see methods)
2300 and 898 respectively,
6 Estimations of terrestrial transects were 7665 kob and 2138 Korrigum
7 idem, 9548 Kob and 4284 Korrigum; in 2000, 8929 Kob and 1512 Korrigum.
Figure 6.2 – Estimated numbers of Kob and Korrigum, 1960-2001
carcasses found, were estimated at about 1000 (Table 6.3; Vanpraet 1977; Thal
unpubl.). In the same period, Korrigum and Bohor Reedbuck numbers seemed to
have dropped more dramatically (Fig. 6.2, 6.3). Yet in 1974, Flizot only remarked
‘Korrigum seems to be less abundant and seems to have migrated to the north’
and ‘few Bohor Reedbuck have been observed, contrasting the numbers before the
droughts’.
The Waterbuck (Kobus elypsiprymmus Ogilby) population was estimated at more
than 200 throughout the 1960s, but not detected by the aerial surveys of 1976 and
1977 (Esser & Van Lavieren 1979). Waterbuck was no longer observed from the
mid-1980s onwards till one and two individuals were once seen again in the floodplain part of Waza National Park in 1998 and 2003 respectively. Bushbuck and
Red-flanked Duiker (Cephalophus rufilatus Gray) still present in the 1970s (Wit
Number
154
Part II – Impact of reflooding
4000
4000
1
2
2000
ROAN
1500
RED-FRONTED GAZELLE
BOHOR REEDBUCK
overestimation
underestimation
4
4
1000
8
9
500
7
3
3
3
5
6
0
'60
'67
'74
drought
'76 '77
'91
dam / drought
'93 '94 '95 '96 '97 '99 '01
reflooding
Year
1
2
3
4
5
6
7
Presumed overestimation, explanation not know.
Probably based on experiences of the early 1960s, see text
Aerial counts, underestimated smaller antelope species, see text
Waterhole count, overestimating Roan, probably drinking twice
Waterhole count, underestimation for unknown reasons
Waterhole count, underestimation because of early rains (see also Fig.6.3 and Text)
Estimations by road counts were 1446 Roan and 1574 Gazelles, corrected for waterhole bias (see methods)
935 and 1036 respectively .
8 Estimations of terrestrial transects were 1331 Roan and 525 Gazelles
9 idem, 3868 Roan and 593 Gazelles; in 2000, 496 Roan and 228 Gazelles.
Figure 6.3 – Estimated numbers of Roan, Red-fronted Gazelle and Bohor Reedbuck, 1960-2001.
pers. comm.) have no longer been observed in Waza NP. Common Duiker still
occurs in limited numbers.
The 1977-1982 drop in Kob population coincided with the construction of the Maga
dam and subsequent lack of flooding (Fig. 6.2; Table 6.2). The further drop from
1982 to 1986 in population numbers of Kob and possibly Korrigum populations
were accentuated by the 1985-86 drought (Figs. 6.2, 6.3; Table 6.2). Observed Kob
densities in the area with the largest concentrations dropped from 356 (1974) and
-2
-2
-2
210 km (1975) to 131 (1980) and 91 km (1982) to 45 (1988) and 23 kob km
(1990). Population numbers of Kob and Korrigum recovered between 1986-1987
and 1991-1993 (Fig. 6.2). This recovery continued from 1994 till 1997, possibly
enhanced by the increased (re-)flooding, but numbers stabilised subsequently.
155
6 – Antelopes
Table 6.2 – Correlation coefficients amongst antelope populations and key environmental parameters
Kob
Korrigum
Roan
Bohor
Reedbuck
r1
Kob
Red-fronted Rainfall3
Gazelle
1
p
Korrigum
Roan
Bohor Reedbuck
Red-fronted Gazelle
Rainfall3
Presence of Maga dam
1
r
0.732
p
0.003
1
r
0.66
0.99
p
0.013
< 0.001
1
r
0.47
0.81
0.79
p
0.15
0.004
0.004
r
-0.21
0.14
0.26
0.26
p
0.54
0.70
0.44
0.47
1
1
r
0.54
0.48
0.39
0.49
-0.64
p
0.032
0.084
0.19
0.13
0.032
S4
-0.80
-0.32
-0.17
-0.24
0.32
-0.45
p
< 0.001
0.27
0.58
0.48
0.33
0.078
2
1
3
Pearson Correlation Coefficient (SPSS 1999); Significant (p < 0.05) in bold; Cumulative rainfall of the preced4
ing three rainy seasons; Spearmann Correlation Coefficient (SPSS 1999).
Kob population structure
In 1986, half of the Kob population was observed outside Waza NP and had a sex
ratio (male/female) of 0.47, that was 0.60, 0.70 in the East and North-East of
Waza NP respectively; overall sex ratio was 0.53 (Table 6.3). The number of young
per female varied from 0.31 (E.Waza NP) to 0.39 (outside) to 0.50 (NE Waza NP).
Amongst carcasses found, juvenile Kobs were over-represented (Table 6.3), due to
their vulnerability to muddy water and warble flies (van Lavieren pers. comm.). In
February and April 1997, sex ratio was 0.51 and 0.53 and young/female ratio 1.04
and 1.18 respectively (Table 6.3).
6.4
Discussion
Count biases
To some extent differences in counting methods masked changing environmental conditions. This held especially for the 150-200% higher Kob numbers estimated in transect than in total counts (Fig. 6.2), due to their congregation around
water holes in all but the rainy and flooding season (Esser & van Lavieren 1979),
see also Fischer & Lindenmair (2001a). The terrestrial transect counts resulted in
deviating results for other species as well (Fig. 6.3), attributed to the difficulty of
156
Table 6.3 – Population structure of Kob 1
Part of
Waza NP
Month
1972
North East
April
1972
North East
April
Author3
Adult male
(Sub) Adult
Juvenile
Number
155 (31%)a4
177 (35%)a
176 (35%) 5a
508
16 (8%)b
carcasses
77 (39%)b
carcasses
105 (53%) 5b
carcasses
198
EFG 1973
Subadult male
female
110 (9%)c
262 (22%)c
534 (46%)c
265 (23%)c
11716
24 (15%)d
28 (17%)d
176 (53%)d
54 (16%)d
3356
Febr.
96 (7%)e
234 (18%)e
704 (54%)e
273 (21%)e
13076
All
Febr.
230 (8%)f
524 (19%)f
1414 (50%)f
592 (21%)f
28136
1997
Park and
Febr.
551 (12%)g
334 (7%)g
1741 (39%)g
1817 (41%)g
4472
1997
surroundings
April
563 (12%)h
366 (8%)h
1761 (36%)h
2082 (43%)h
4832
1986
North East
Febr.
1986
East
Febr.
1986
East of Waza NP
1986
1
Korthof (unpubl.)
Hebou (unpubl.)
Classification following Wanzie 1988.
Year corresponds to the preceeding rainy- and flooding season and not necessarily the survey year.
3
See also Table 6.1.
4
Different letter: p < 0.0001, except g-h: p = 0.045 (chi-square test).
5
Includes some subadult males, the 1972 counts can only be mutually compared.
6
Some double-counting (50%) has taken place (compare with Fig. 6.2).
2
Part II – Impact of reflooding
Year 2
6 – Antelopes
157
estimating the bandwidth by inexperienced observers, of which their large number (>40) and frequent changes did not allow any correction. The aerial surveys
were biased towards the larger species neglecting Red-fronted Gazelle and Bohor
Reedbuck (Van Esser & Lavieren 1979; Tchamba & Elkan 1995; Dejace et al. 2000).
During the waterhole total counts, frequent, i.e. more than once a day drinking,
was observed of Kob, leading to a potential overestimation of its population numbers. None of the other antelope species showed comparable drinking behaviour.
Other imaginable biases of the waterhole counts could only have led to an underestimation of population numbers. Red-fronted Gazelle is less dependant on daily
drinking than the other antelope species in Waza NP (Estes 1992). It is further
likely that some antelopes drank before or after the observers had been installed.
In addition to these methodological biases, it cannot be excluded that ‘counter psychology’ has influenced survey results. Flizot probably overestimated the numbers
of Korrigum, Roan and Bohor Reedbuck in 1967, neglecting a decline that is expected to have taken place after the above-average rainfall and flooding in the 1950s
and early 1960s when he was frequently in the field. Flizot, as well as his successors, further could have had an interest in showing a healthy animal population
in the park they managed.
Kob
Despite the above-mentioned count biases, major impacts of (man-made) droughts
have been detected, especially on Kob. The Kob population dynamics show the intrinsic link between stress from periods of low rainfall and man-made droughts,
induced by the Maga dam construction (Fig. 6.2; Table 6.2). This also holds for
the rehabilitation process, as the 1994 reflooding started when the Kob population
had already partly recovered from the mid-1980s drought.
Kob’s sex ratio (male/female) in Comoé NP (Ivory Coast) was only 0.34, which
Fischer & Lindsenmair (2002) attributed to high poaching pressure. Here, Geerling & Bokdam (1971) observed a sex ratio of 0.74 in a time when poaching pressure was much lower. The somewhat lower Kob sex ratio outside Waza NP in
1986 (0.47), compared to inside Waza NP (0.60), may possibly be attributed to
poaching. Sex ratio of the entire Kob population in 1986 and 1997 was comparable (0.51-0.53), suggesting no trend in poaching. Predation pressure has remained
relatively constant during the last decades (Bauer 2003) and does not explain
observed differences in Kob nor any of the other antelope species.
Kob year classes of Fischer & Lindsenmair (2002) did not entirely match Wanzie
(1988), yet with the broad class ‘young’ (calves and juveniles) this source of differences was limited. Fischer & Lindsenmair (2002) observed a low 0.49 ± 0.05
young per adult female, which they attributed to the high poaching pressure, as
argued for the above-mentioned low sex ratio. The comparably low young per
158
Part II – Impact of reflooding
adult female ratio observed in Waza in 1986 (0.42) can be attributed to drought
induced food stress as the juvenile percentage in 1997, preceded by three years of
almost average rainfall, was twice as high (Table 6.3). Although some differences
in interpretation cannot be excluded, drought seems to have affected especially the
youngest and eldest age classes. Although the 1985-1986 drought had limited the
reproduction output, it does not seem to have hit the (future) reproduction capacity, as confirmed in the Kob population increase in subsequent years (Fig. 6.2).
Korrigum, Roan and Bohor Reedbuck
The poorly documented late 1960s-early 1970s drop of Korrigum, Roan and Bohor
Reedbuck populations (Figs. 6.2, 6.3) remains difficult to explain. Contrary to
common belief, presented results show that lack of flooding due to the Maga dam
construction, that took place only after 1979, was not the reason for their decline
(Figs. 6.2, 6.3; Table 6.2). The lack of flooding since 1979 may, however, have
been a constraint for their full recovery. Elsewhere in West-Central Africa these
antelopes species have also been under increasing pressure, due to poaching, competition and disease transmission by livestock (Esser & van Lavieren 1979; Sayer
1982; East 1999; Fischer & Lindsenmair 2001b; Sinsin et al. 2002). With a continuing drop in the Arly-Singou-Pendjari area (Sinsin et al. 2002), Waza NP holds
the only remaining viable Korrigum population, probably due to the variation in
protected habitat, including both floodplain and upland savanna (Sayer 1982).
Although the floodplain habitat has been improved with the recent reflooding
(Scholte et al. 2000b), livestock competition and potential disease transmission in
the immediate surroundings of the park, have also increased (Scholte 2003a).
Extinction and human impact
The disappearance of Waterbuck that used to occur in the floodplain, yet in small
numbers since the early 1970s, appears to be linked to the hydrological changes.
Its timid come-back may possibly be associated with the reflooding from 1994
onwards. In addition to these changes in the floodplain, Vanpraet (1976) stressed
the negative impact of habitat changes in other parts of Waza NP such as increased
fire in the preferred Lannea humilis habitat, a practice that, though addressed in
its management plan remains difficult to change (Scholte 2000b). The extinct
Bushbuck and Red-flanked Duiker used to occur in the wooded parts of Waza NP
that experienced a loss of tree cover due to lack of flooding from upland water courses, the low rainfall in the 1970s and early 1980s (Fig. 6.2) as well as an increasing
impact of Elephant (Tchamba 1996).
Direct human impact on these extinctions can not be excluded as reported for
neighbouring Kalamaloué NP where the once common Waterbuck disappeared in
the same period, due to human encroachment (Scholte 2003a). Red-fronted Gazelle, with a predominant Sahelian distribution (East 1999; Scholte & Hashim
2005), seems to have increased in both national parks. Few direct signs of poach-
6 – Antelopes
159
ing have, however, been reported in and around Waza NP (Tchamba & Elkan 1995;
Scholte et al. 1999b). This may be due to the difficult relations that existed for most
of the time between park authorities and local communities and the inaccessibility of the area during the rainy and flooding season (Scholte et al. 1999b). Kob and
Korrigum populations that leave Waza NP for considerable part of the year (Flizot
1974; Esser & van Lavieren 1979) are especially vulnerable to poaching.
It has been difficult to assess with one single method the population size of all
antelope species, each with a particular distribution and habits. A combination of
standardised counts, such as annual waterhole counts and occasional aerial counts,
as proposed in the Waza management plan, seems appropriate and cost effective.
This will allow monitoring the antelope populations after the recently increased
confidence between park authorities and local communities and allow an assessment of the impact of more extensive reflooding (Scholte 2003a).
7
Pastoralist Responses to Floodplain
Rehabilitation in North Cameroon
Paul Scholt e, Saïdou Kari, M ark M orit z and Herbert Prins
Human Ecology, in press
Abstract
This chapt er examines t he responses of mobile past oralist s t o a f loodplain rehabilit at ion program in Nort h Cameroon. From 1993 t o 1999, w e measured
changes in number of camps and herds, and t he t ime t hey spent in t he 600 km
2
of t he Logone f loodplain t hat w as ref looded in 1994. The f irst year, f ew past oralist s ant icipat ed t he ref looding or it s impact , and t he increase in grazing
int ensit y w as caused by a prolonged st ay of past oralist s w ho already used t he
area f or t ransit . The f ollow ing t hree years show ed a sharp increase in t he number of camps and herds, w hich st abilised f rom 1997 onw ards. Overall, grazing
int ensit y increased t hree-f old, f ollow ing t he gradually recovering perennial
grasslands, w it h no signs of over-exploit at ion of t he area. These development s
closely mat ch t he Ideal Preempt ive Dist ribut ion model. We also examined how
ref looding aff ect ed past oral incursions in Waza Nat ional Park locat ed in t he
f loodplain.
163
7.1
Introduction
Mobile pastoralists are remarkably well adapted to the drylands of Africa, much
better than sedentary pastoralists or ranchers (Breman & Wit 1983; Niamir-Fuller
& Turner 1999). Floodplains play a key role in the yearly migration of many pastoralists, due to the availability of nutritious grass regrowth and surface water far
into the dry season, when surrounding grazing lands have dried up (Hiernaux &
Diarra 1983). However, human pressure on these floodplains is increasing and restricting pastoralists’ access to them. One of the more devastating impacts on floodplains is the construction of dams for electricity and irrigation purposes (Drijver
& Marchand 1986). Increasing awareness of the importance of wetlands for rural
economies in semi-arid Africa has led to a series of initiatives to reduce the impacts
of existing dams (Acreman & Hollis 1996). Most initiatives aim at improving water
management, so that ‘excess’ water can be released and restore former floodplain
functions, including that of dry season pastures for transhumant pastoralists. In
several places in Africa experiments with such releases have started. But despite
the importance of African floodplains for mobile pastoralists, there had not yet
been an assessment of their responses to floodplain rehabilitation. This paper examines the responses of mobile pastoralists to the reflooding of the Logone floodplain,
focusing on changes in transhumance patterns and grazing intensity from 1993
until 1999. It offered a unique chance to study pastoralists’ reactions to improving
grazing conditions in a Sahelian environment.
In 1979, the Waza-Logone area in the Far North Province of Cameroon met the
fate of other African floodplains, when a dam and embankment along the Logone
River were constructed forming a reservoir for a parastatal irrigation scheme (Fig.
7.1). In combination with lower than average rainfall during the last two decades
2
(Beauvilain 1995), the dams reduced flooding in an area of about 1,500 km , which
included Waza National Park, an important refuge for wildlife such as elephant,
lion, antelopes and waterbirds (Scholte et al. 1996a). Annual grasses invaded previously productive perennial grasslands, limiting regrowth in the dry season and
reducing the floodplain’s carrying capacity for wildlife and livestock (Scholte et al.
1996a). Studies that documented the detrimental effects of the dam on ecology
and local economies (Drijver & Marchand 1986) were the impetus for a reflooding program, which took shape in the Waza-Logone Project. An appraisal study
showed the potential for reflooding (Wesseling et al. 1996), which was supported
by local people, authorities and wildlife managers. Reflooding was started in May
1994 with the re-opening of a watercourse that had been closed off by the embankment along the Logone River, starting the recovery of perennial grassland vegetation (Scholte et al. 2000b).
164
Part II – Impact of ref looding
Fig. 7.1 – The Far Nort h Province w it h t he Waza-Logone area
7 – Past oralist s responses
165
One of the main goals of the reflooding was to stop the incursions of pastoralists
into Waza National Park in order to reduce the competition between wildlife and
livestock for dry season grazing (Scholte 2003a). Livestock is not allowed in national parks and this has been the subject of numerous confrontations between
park authorities and pastoralists (Scholte et al. 1996b, 1999b). The project anticipated that the floodplain rehabilitation would motivate pastoralists to find grazing
lands adjacent to the park, where the reflooding was expected to have its main
impact.
The reflooding feasibility study predicted that the productivity of the floodplain
grasslands would increase two- or three fold (Wesseling et al. 1996). One of the
questions was how mobile pastoralists would respond to this increase in rangeland productivity and whether a potential influx of ‘new’ pastoralists would jeopardize the projected rehabilitation of the Logone floodplain. Another question was
whether pastoralists would indeed refrain from entering Waza National Park with
increasing productivity of the floodplain.
Predicting pastoralists’ responses
Visions from pastoralists, researchers, and authorities diverged on the reactions of
mobile pastoralists to the floodplain rehabilitation and its subsequent impact on
Waza National Park. The visions can be grouped in three different scenarios:
1 Overshoot scenario
The first scenario assumed a ‘free-for-all’ open access situation in which pastoralists would flock to the improved grazing opportunities. This would result
in grazing intensities largely exceeding the improved carrying capacity, i.e., an
‘overshoot’ of cattle. The fear was that this would lead to overgrazing and conflicts with other pastoralists, thereby undoing the rehabilitation of the floodplain.
In this scenario, pastoralists would most likely continue to take their cattle into
Waza National Park because there would be no alternative grazing opportunities.
2 Territorial scenario
The second scenario assumed that pastoralists that used the area before reflooding would somehow defend their grazing lands, i.e., display territorial behavior (Casimir 1992). Territoriality among mobile pastoralists has been observed
in neighboring Chad (pers. obs.) and in the Niger floodplain in Mali (Legrosse
1999; Turner 1999). Although there seemed to be a system in the floodplain
in which pastoralists had customary rights to specific campsites and surrounding pastures, it was unclear whether and how pastoralists would defend these
rights if there were an influx of ‘newcomers’. In this scenario, there would be
most likely a slow increase in grazing intensity and a low exploitation of perennial grass cover in the rehabilitated floodplain. Potentially this would allow
166
Part II – Impact of ref looding
pastoralists with claims in the floodplain to return to their old campsites and
refrain from taking their cattle inside Waza National Park.
3 Ideal free distribution scenario
The third scenario predicted that any improvement in grazing land condition,
i.e. perennial grass cover, would lead to a corresponding increase in the number of cattle. The assumption underlying this scenario was that pastoralists
would adjust the number of herds in the reflooded area and/or the time they
spent there to make optimal use of the available resources. The result would be
a gradual increase in grazing intensity following a similar increase in perennial grass cover. In this scenario, grazing resources outside the park would gradually come available to pastoralists that were taking their cattle into the park.
The data presented in this article allows for an evaluation of which scenario best
matches pastoralist responses to reflooding. We will also consider how other factors, such as insecurity, cattle theft and predation, and conflicts with fishers and
farmers affected transhumance patterns.
7.2
Background
The Waza-Logone floodplain
The Waza-Logone floodplain receives a mean annual rainfall of c. 650 mm with
between-year fluctuations of up to 50%, a typical Sahelo-Sudanian climate (Beauvilain 1995). The rainy season is from June to September. In parts of the area, the
Logone River and its branches flood during August to November. In April and
May, temperatures rise as high as 45°C. In areas subject to annual flooding, perennial grasses such as Echinochloa pyramidalis, Oryza longistaminata and, on the
slightly elevated parts, Vetiveria nigritana and Hyparrhenia rufa dominate. When
the floods recede at the end of December, most parts are burned by pastoralists
and hunters, leaving a bare landscape and only some humid spots remain with
vivid green vegetation and concentrations of fishermen, livestock and birdlife.
Until the Maga dam construction in 1979, the study area, the ‘reflooded area’
located in the southern floodplain, used to be flooded and was composed of productive perennial grasslands (Fig. 7.1). When flooding ceased, perennial grasses
were replaced by annual grasses, especially the reed-like unpalatable Sorghum
arundinaceum. The reflooding in 1994 brought back the annual floods in an area
2
2
of about 180 km and raised the water level in 600 km . Following the reflooding
an annual 7% increase in perennial grass cover was measured (Scholte et al.
2000b). The improvements were dramatic: Sorghum arundinaceum has disappeared and one can again see for miles and miles fertile perennial grasslands.
7 – Past oralist s responses
167
This period coincided with average rainfall and flooding conditions, and without
droughts like those that plagued the area in the 1970s and 1980s (Beauvilain 1995).
Pastoralists in the Waza-Logone floodplain
The Logone floodplain historically has been a Key Resource Area for pastoralists
in the Lake Chad Basin (Scholte & Brouwer 2005). When rangelands in the surrounding plains and mountains are drying up and sources of surface water are
disappearing, pastoralists trek to the Logone floodplain to find forage and easily
accessible surface water for their animals. Each year, thousands of pastoralists
with about 200,000 cattle from Cameroon, Chad, Nigeria, and Niger enter the
1
floodplain for a period of six to eight months (Beauvilain 1981).
The mobile pastoralists that use the floodplain pastures belong to two ethnic
groups: (Shuwa) Arabs (Braukamper 1996) and FulBe (e.g. Azarya et al. 1999),
(Table 7.1). The Arabs do not sell milk, which they use for household consumption or leave for the calves. This is one of the reasons why their herds grow at a
higher rate than those of FulBe pastoralists. Other reasons are their involvement
in livestock trade, the profits of which are invested in cattle, and the fact that they
care for entrusted animals from relatives in Nigeria. The Arab pastoralists that use
the reflooded area are nomadic and practice transhumance from the Diamaré plains,
south of the floodplain, where also most FulBe pastoralists come from (Fig. 7.1).
The FulBe can also be divided in nomadic and agro-pastoral groups. This distinction refers to the fact that the former do not practice agriculture and are without
permanent settlements. Agro-pastoral FulBe, on the other hand, have families
that live in villages where they cultivate. Agro-pastoral herds may be permanently
on transhumance and come back to the village only once or twice a year for a few
weeks. There are two distinct groups of agro-pastoral FulBe that practice transhumance to the reflooded area. One group comes from the area south of Waza
National Park and is referred to as the FulBe Ngara’en, which is the name of the
ruling clan of the towns of Pétté and Fadaré (Fig. 7.1). Their transhumance takes
them along the borders of Waza National Park, and not surprisingly, the FulBe
Ngara’en are the ones responsible for incursions into the park. The other agropastoral FulBe group comes from further south and is often referred to as the
FulBe Yillaga’en, which is the ruling clan in the towns of Mindif and Moulvoudaye (Fig. 7.1). Most agro-pastoralists from both groups are descendents of FulBe
that conquered the area two centuries ago (Mohammadou 1988).
1 Use of floodplains in the Chad basin is facilitated by international agreements (CIRAD-CTA, 1996).
In principle, pastoralists have to pay a small fee per head of cattle and show certificates of herd vaccinations when they cross the border, but in practice border controls are not very strict.
168
Part II – Impact of ref looding
Table 7.1 – Charact erist ics of Past oral Groups at t he end of t he St udy Period, 1999
Nomadic Past oralist s
Agro-Past oralist s
FulBe
FulBe
FulBe
Adanko’en
Alijam’en
M are’en
25
89
114
135
Number of households per camp
2.5
9.8
3.8
9.6
NA
NA
Number of camps
10
9
30
14
26
16
Number of herds per camp
7.2
9.4
8.4
13.9
10.9
12.0
Number of herds per household
2.9
1
2.2
1.5
NA
NA
Tot al number of herds
72
85
252
195
283
192
4,593
5,338
21,344
23,653
23,829
16,320
20
30
60
30
>100
>100
1
Number of households
Arabs
FulBe
FulBe
Ngara’en
Yillaga’en
89
(or herding groups)
Est imat ed number of cat t le
Presence in t he f loodplain
(in years)
1
Unlike the nomadic groups, the FulBe Ngara’en and FulBe Yillaga’en are mostly herders on transhumance
without their families. The term household cannot be used for this group. It is better to think of these units as
herding groups in which a number of herders of one village cooperate. Large herding groups often constitute
one camp; smaller herding groups camp together with herding groups from neighboring villages. The herds in
these herding groups, in turn, are often composed of animals from multiple households in the village.
The nomadic FulBe group can be divided into three distinct sub-ethnic groups
2
Adanko’en, Alijam’en and Mare’en. The majority of nomadic FulBe originally
came from Nigeria, but have been in Cameroon for the last twenty to sixty years.
They have Cameroonian identity cards and identify themselves as Cameroonians.
Their kin network extends across all borders, which reflects their readiness to
move, responding to ecological as well as socio-political conditions.
Although cattle account for more than 90% of the total Tropical Livestock Units in
the floodplain, there are also FulBe Uuda’en shepherds who transhumance from
Nigeria and Niger to the Logone floodplain. But since they do not exploit the reflooded area we have excluded them from this study. We also excluded the sedentary,
village-based Musgum, who possess less than 10% of the total number of cattle in
the floodplain.
2 A sub-ethnic group refers to a group consisting of different clan fractions that have in common a
migratory history and certain cultural traits (Burnham 1996). Other, numerically smaller nomadic
FulBe groups in the Logone floodplain have been grouped with one of the three main nomadic FulBe
groups with whom they associate.
7 – Past oralist s responses
169
Transhumance patterns
One can distinguish two phases of pastoral exploitation of the floodplain. During
the first phase from November through January, after the recession of flooding,
cattle feed on young grass shoots. At the beginning of this phase, when pastoralists enter the floodplain, they often split their herds. Older and lactating animals
stay with the families, while young herders take the stronger, healthy animals on
a separate transhumance, called luci, following the retreat of the water in search
of the best pastures. During this period camps and herds move frequently. In
January-February, the herds are re-united again and pastoralists often settle in one
or two campsites for the rest of the dry season. During this second phase, cattle
feed on the regrowth of perennial grasses that result from fires. At the onset of the
rainy season in May-June, pastoralists direct their animals to those places where
the first rains have fallen and leave the floodplain till the following dry season.
One can also distinguish different ways in how the reflooded area fit into the transhumance of different pastoral groups before it was reflooded in 1994. A first
group of pastoralists, mainly FulBe Alijam’en and Arabs, used the reflooded area
as a transition zone from the rainy season pastures in the south to their dry season pastures further north and east (Fig. 7.1). A second group, FulBe Mare’en and
FulBe Adanko’en, used the reflooded area only in the cold dry season (NovemberJanuary), before they went to their dry season pastures at the borders of Lake
Maga, further south. A third group of pastoralists spent the entire dry season in
the reflooded area. This last group consisted predominantly of the agro-pastoral
FulBe Ngara’en and FulBe Yillaga’en, and the FulBe Mare’en. We will see below
that most pastoral groups continued these transhumance patterns after reflooding, while spending increasing time in the reflooded area.
7.3
M ethods
Early 1994, before the reflooding, we visited all pastoral camps in the southern part
of the floodplain (Fig. 7.1). The main objective was to start a dialogue with a wary
group that had suffered from previous governmental and project interventions
(Moritz et al. 2002; Scholte et al. 1996b). Most of the time was spent on discussing
ongoing developments in the floodplain and pastoralists’ concerns, in particular
insecurity. During the visits in following years, we obtained information on ‘new’
pastoralists in the study area, which were included in our annual monitoring of
pastoralists’ responses to the reflooding. We also obtained information on pastoralists that we missed in our earlier campaigns. Generally, these ‘missing’ pastoralists used the reflooded area only as a transition zone from rainy season pastures
in the south to dry season pastures further north outside our study area. At the
end of each campaign, we organized a session in a centrally located market town
to discuss results and upcoming issues, such as further reflooding initiatives.
170
Part II – Impact of ref looding
Assessing grazing intensity
The visits in 1994 and subsequent years also served to establish a database of all
pastoral groups that spent time in the reflooded area (Table 7.1). Groups were
entered in the database by the name of the camp’s leader and for each group we
listed the campsites visited in the floodplain and the number of weeks spent there.
Early in our work, we compiled a list with all campsites and marked them on a topographic map. We used this list to record transhumance patterns and determine
the number of weeks that pastoralists spent in the reflooded area. The interview
method was an effective and reliable way to assess changes in grazing intensities;
pastoralists could recall without difficulty how many days they had camped in
3
each site over the last year. In subsequent campaigns, we also collected data on
the number of herds and herd size (i.e. the numbers of cattle per herd) in infor4
mal discussions with group leaders. The trust we had built up through the projects’ commitment to resolving pastoralists’ primary concern of insecurity (Scholte
et al. 1996b), allowed the estimation of these otherwise very sensitive topics. We
checked self-reported herd numbers and size with direct counts during overnight
stays, crosschecking with pastoralists in other camps, and data from vaccination
5
campaigns (Table 7.2). Because there were significant differences between selfreported herd size from FulBe Mare’en and Arabs and our own direct counts and
vaccination campaigns, we have disregarded the former. In this study of changes
in grazing intensity, we consider herd size to be a constant value, although grassland productivity increased after the reflooding and assumedly led to greater reproductive rates and thus larger herds. Moritz (2003), for example, found that nomadic
herds in the Logone floodplain grew on average 4.5% per year, slightly higher than
the 3.4% postulated for a ‘normal’ herd by Djahl & Hjort (1976). Increased reproductive rates and herd growth is reflected in an increase in the number of herds
per pastoral group when pastoralists ‘split’ their herds.
We have operationalised grazing intensity as the number of herds multiplied by the
number of weeks each camp spent in the reflooded zone and that multiplied by
the number of cattle per herd of the concerning pastoral group. The resulting graz2
ing intensity is expressed as cattle densities, i.e., the number of cattle per km aver-
3 Pastoralists used moon cycles and weekly market days as reference points to recall how long they
had camped in particular sites. We cross-checked this information with other pastoralists.
4 We have considered herds as management units, i.e., the animals that graze together during daytime under the guidance of a herder, and not as property units (Dahl and Hjort 1976: 134).
5 Initial analysis using stem-and leaf plots in SPSS (1999) showed that herd size data followed a normal distribution. Herd size data from different sources have subsequently been analyzed with t-tests,
variances of herd sizes with Levene’s Test for equality of variances (SPSS 1999) to enable possible
lumping. Subsequently, herd size data from different pastoral groups were lumped if results of
ANOVA and Tamhane Post Hoc tests (SPSS 1999) allowed for it.
171
7 – Past oralist s responses
Table 7.2 – Average Herd Size by Source of Inf ormat ion
Self -report
Vaccinat ion
Direct count
records
Average herd size used in
t his st udy (vaccinat ion records
and direct count s lumped)
Nomadic Pastoralists
FulBe Adanko’en
FulBe Alijam’en
FulBe M are’en
Arabs
65.8 a1
63.8 a
(25)
(5)
64.4 b 2
(36)
–
63.8i
64.5 b
50.0 b
62.8i
(15)
(2)
75.3 e
84.0 f
92.3 f
(72)
(38)
(3)
98.7 c
120.8 d
122.3d
(15)
(13)
(6)
80.8 g
84.2 g
–
84.2k
(62)
(20)
79.1 h
85.4 h
82.7 h
85.0k
(51)
(29)
(6)
84.7k
121.3j
Agro-Pastoralists
FulBe Ngara’en
FulBe Yillaga’en
Number of report s or count s is given in parent heses.
1
2
same let t er: p (means) > 0.05 and p (variance) > 0.05; same let t er in it alics: p (means) > 0.05 (t -t est , SPSS 1999) and p
(variance) < 0.05 (Levene’s Test f or equalit y of variances, SPSS 1999).
aged over the 6 months of the dry season. This also allows us to compare our data
6
with livestock surveys in other floodplains.
7.4
Results
Changes in transhumance patterns
In individual and group discussions, pastoralists invariably expressed their appreciation of the reflooding initiatives. At a meeting in 1997 with about twenty representatives of different pastoral groups, a proposed opening of a second watercourse, in addition to the 1994 reflooding, was applauded with ‘the more water
the better’. How pastoralists responded in terms of changing transhumance patterns will be discussed below.
6 The 27 to 69 cattle km2 increase in the reflooded area in the Logone floodplain falls within the
range of aerial surveys of other African floodplains, which assessed average cattle densities from 25 to
80 cattle per km2 (Scholte & Brouwer 2005). Obviously, care should be taken to compare surveys of
different scales and methodologies.
Part II – Impact of ref looding
Changes in the number of pastoral camps
In the first year of reflooding, few pastoralists had heard about the reflooding or
had anticipated its impact, reflected by the limited increase in the number of pastoral camps (Fig. 7.2). When the positive impact of the reflooding became apparent, it was subject of much discussion on market places and other pastoralists’
fora. The number of pastoral camps increased dramatically in the second year,
1995, and continued until 1997 after which it stabilised (Fig. 7.2). Unlike the other
pastoral groups, the number of Arabs camps did not increase, although some ‘newcomer’ Arabs pastoralists are said to have joined the existing camps, contributing to
the highest ratio of herds per camp (Table 7.1).
120
Y(total) = 119 - 64.94/ t
( p = 0.003)
100
number of camps
172
t = years after reflooding
80
Fulbe Mare'en
Arabs
Fulbe Ngara'en
Fulbe Yillaga'en
Fulbe Alijam'en
Fulbe Adanko'en
60
40
20
0
1993
1994
1995
1996
1997
1998
1999
reflooding
Figure 7.2 – Development of number of camps in t he ref looded area, 1993-1999.
Y
R2
p
Total
119.2-64.9/ t
0.96
0.003
FulBe Adanko’en
12.7 - 11.4/ t
0.94
0.007
FulBe Alijam’en
LnY = 2.54 - 1.06/t
0.87
0.020
Arabs
LnY = 2.64-0.14/t
0.52
0.169
FulBe M are’en
35.1-15/ t
0.99
0.000
FulBe Ngara’en
LnY = 3.44-1.36/ t
0.93
0.008
FulBe Yillaga’en
18.9-10.0/t
0.86
0.023
Bold: t able-w ide signif icant (α = 0.05). Regression equat ions have been det ermined bet w een t he
number of camps and grazing int ensit y w it h increasing t ime, using best -f it t ed curve est imat ion
(SPSS, 1999). Signif icance levels have been adjust ed on a t able-w ide level, using t he Bonf erroni t est
f or mult iple t est s at a 0.05 signif icance level (Rice 1989).
173
7 – Past oralist s responses
1200
Y(total)= 261ln t +601
number of herds
1000
(p =0.002)
t = years after reflooding
800
Fulbe Mare'en
Arabs
Fulbe Ngara'en
Fulbe Yillaga'en
Fulbe Alijam'en
Fulbe Adanko'en
600
400
200
0
reflooding
1994
1995
1996
1997
1998
1999
Figure 7.3 – Development of number of herds in t he ref looded area, 1993-1999.
Y
R2
p
Total
Y=261.3 ln t + 601.4
0.97
0.002
FulBe Adanko’en
LnY = 5.02- 3.12/ t
0.96
0.004
FulBe Alijam’en
2.49X + 68.2
0.69
0.082
Arabs
6.73X + 152.9
0.74
0.061
FulBe M are’en
LnY = 5.56- 0.55/ t
0.93
0.009
FulBe Ngara’en
LnY = 5.97 - 1.71/ t
0.93
0.008
FulBe Ngara’en
13.15X + 111.3
0.99
0.000
Bold: t able-w ide signif icant (α = 0.05), see also remark Fig. 7.2
Changes in the number of herds
Initially, the number of herds followed a relatively low increase rate as newcomers had relatively few herds compared to the increase in the number of camps (Fig.
7.3). The number of herds continued to increase until the last monitoring year.
Part of the increase is due to normal herd growth and subsequent splitting of herds
that were already present. The 4.5% annual growth rate of nomadic herds (Moritz
2003) would translate in an increase of 30% in the number of cattle from 1993 to
1999. If we assume a similar increase in the number of herds, then approximately
one-third of the increase in the number of herds in the study area during that period is due to growth of existing herds, while two-thirds is due to the arrival of new
7
herds.
7 This assumption seems reasonable given the large number of animals and herds in our study,
respectively 95,077 and 1,079 in 1999.
Part II – Impact of ref looding
Changes in grazing duration
We also measured how long pastoralists stayed in the reflooded zone, which we refer
here to as grazing duration, i.e. the number of weeks pastoralists stayed in the study
area. Before reflooding in 1994, FulBe Mare’en stayed in the reflooded zone almost
twice as long as other groups. After reflooding the difference between the FulBe
Mare’en and the other pastoral groups has become less distinct as contrary to most
other pastoral groups, FulBe Mare’en stayed only slightly longer than before. From
1993 until 1999, FulBe Alijam’en and Arabs groups have stayed increasingly
longer in the reflooded zone on their yearly migration to pastures further north
8
and east into Chad.
70
Y(total) = 7.1 t + 17.9
(R2=0.92
grazing
intensity
(cattle km-2)
174
p =0.003)
t(1993) = 1
60
50
40
30
20
10
0
1993 1994
1995 1996
1997 1998
reflooding
1999
Total
Fulbe Mare'en
Arabs
Fulbe Ngara'en
Fulbe Yillaga'en
Fulbe Alijam'en
Fulbe Adanko'en
Figure 7.4 – Development of grazing int ensit ies in t he ref looded area, 1993-1999.
Y
Total
7.1X + 17.9
R2
p
0.91
0.003
Fulbe Adanko’en
0.38X -0.017
0.54
0.096
Fulbe Alijam’en
0.59X - 0.40
0.93
0.002
Arab
1.7X + 2.16
0.88
0.006
Fulbe M are’en
LnY = 2.9 - 0.63/t
0.67
0.046
Fulbe Ngara’en
2.33X + 3.4
0.71
0.035
Fulbe Yillaga’en
0.99X - 0.26
0.85
0.009
Bold: t able-w ide signif icant (α = 0.05), see also remark Fig. 7.2
8 Respectively Y = 0.00344(Year)-6.9, R2 = 0.91, p = 0.003 and Y = 0.0039(Year)-7.8, R2 = 0.87, p =
0.006.
175
7 – Past oralist s responses
Changes in grazing intensity
With the data on the changes in the number of herds and the time that they spent
in the reflooded area, we can estimate the changes in overall grazing intensity for
the impact zone. Grazing intensity in the reflooded area increased from 27 to 69
2
cattle km for the six-months dry season between 1993 and 1999 (Fig. 7.4). Three
pastoral groups, FulBe Mare’en, Arabs and agro-pastoral FulBe Ngara’en accounted for the largest increase in total grazing intensity. The FulBe Ngara’en and the
Arabs pastoralists have been the main beneficiaries of the reflooding, accounting
for respectively 33% and 24% of the increase in grazing intensity (Table 7.3).
Table 7.3 – Cont ribut ions of Past oral Groups t o Increase in Relat ive Grazing Int ensit y, 1993-1999
Nomadic Past oralist s
Agro-Past oralist s
FulBe
FulBe
FulBe
Arabs
FulBe
FulBe
Adanko’en
Alijam’en
M are’en
Ngara’en
Yillaga’en
5.0%
8.4%
15.0%
24.3%
33.2%
14.1%
0.2%
0.09%
0.13%
0.18%
0.53%
1.1
1.5
0.7
1.0
1.1
Increase in grazing
int ensit y (Fig. 7.4)
Increase in grazing int ensit y
per herd
Increase in grazing int ensit y
-3
per head of cat t le (10 )
In terms of increase in grazing intensity per individual household, agro-pastoral
Ngara’en and Yillaga’en FulBe have benefited most of the reflooding, although this
was biased by the low number of households accompanying the herds (Table 7.3).
FulBe Mare’en, who used to be numerically the most prominent pastoralists in
the reflooded area before reflooding, were the only group for whom grazing intensity did not increase, meaning that their cattle did not increase and did not stay
longer in the study area. Overall, FulBe Mare’en have benefited the least from the
floodplain rehabilitation (Table 7.3).
In the first year the increase in grazing intensity was for a large part due to a prolonged stay of pastoralists in the reflooded zone, reducing their stay elsewhere in
the floodplain. This was most evident in the case of the FulBe Alijam’en, who on
their yearly migration to the northern floodplain, have increased their stay in the
reflooded area. When grazing duration in the reflooded area increased, this did
not lead to lower grazing intensities in other parts of the floodplain. We observed
that other groups, some coming from Nigeria, have replaced groups in the northern and eastern parts of the floodplain that are now visiting the reflooding area.
176
Part II – Impact of ref looding
Other factors affecting transhumance patterns
It is important to keep in mind that pastoralists’ transhumance patterns are not
only determined by ecological factors; socio-political factors may be important as
well (Niamir-Fuller 1999b; Stenning 1957). We expected that this might also have
been here the case, because the topic of cattle losses due to theft, diseases or predation, was nearly always raised in our discussions with pastoralists. Moreover,
pastoralists argued that the insecurity threatened their access to the rehabilitated
floodplain (Scholte et al. 1996b). The start of the floodplain rehabilitation coincided with a period of insecurity, unparalleled in recent history (Issa 2004), in which
several pastoralists lost their lives defending their herds. We anticipated that cattle losses might have influenced pastoralists’ transhumance patterns, thereby potentially reducing the impact of the reflooding.
In 1997 we assessed the effects of cattle losses on transhumance patterns and the
exploitation of the rehabilitated floodplain. In our annual visits to camps we asked
about the number of cattle lost and categorised them as diseases, accidents, predation and theft. We calculated cattle losses as a function of grazing duration in a
particular part of the floodplain, following the equation below:
Report ed losses
Loss rat e (f ract ion per w eek) =
(Number of herds) x (herd size) x (grazing durat ion)
This survey shows the spatial distribution of risks in the floodplain (Fig. 7.5). We
were unable to make a causal link between data on cattle losses with grazing intensities in particular areas of the floodplain and thus limit our discussion here to the
following general observations. When we recorded cattle losses in 1997, pastoralists reported the theft of 600 cattle. This was considerably less than they had experienced in the early 1990s before the Waza-Logone Project had undertaken action
(Scholte et al. 1996b). A year later, cattle losses were even lower, as special governmental forces had been stationed in the area to counter the insecurity. In subsequent years losses to cattle theft continued to decline. Pastoralists frequently
argued that the reflooded area was the area with the highest risk of cattle raids as
confirmed by the reported loss rate in 1997 (Fig. 7.5). This may have inhibited
some pastoralists visiting this area in the early years of reflooding. The fact that
there was no sharp increase in the number of camps and herds after 1998, the
year that the special forces practically halted cattle thefts, is an indication that insecurity did not influence the transhumance into the reflooded area.
Later discussions with nomadic pastoralists focused on conflicts with fishers and
agriculturalists (Kouokam et al. 2004). These conflicts concerned the blocking of
transhumance routes into and within the Logone floodplain by agricultural fields
and fish canals (Moritz et al. 2002). The Waza-Logone Project later assisted in the
177
a
La
ke
Ma
g
a
Ar
e
Zin
a
ed
Flo
od
pla
in
Ea
ste
rn
od
flo
Re
No
rth
er
nF
loo
dp
Ar
e
a
lai
n
7 – Past oralist s responses
Figure 7.5 – Number of cat t le lost in diff erent f loodplain areas in 1997
For areas, see Fig. 7.1
resolution of these conflicts through the demarcation of livestock corridors into
the floodplain and establishment of conflict resolution mechanisms (Kari & Scholte
2001). These conflicts did not seem to have had a lasting impact on transhumance
patterns neither.
Another factor that might have affected transhumance patterns in the reflooded
area was livestock predation. Agro-pastoral FulBe Ngara’en frequently raised this
problem in meetings with the Waza-Logone Project (Scholte et al. 1996b). Although
predation by lions was a great concern for individual FulBe Ngara’en, the 155 cattle lost to lions in 1997 represent only a fraction of the total losses in the floodplain. Lion predation only occurred in areas adjacent to Waza National Park, and
even there accounted for a relatively small percentage of the total losses (Fig. 7.5).
Hyena predation was reported in areas far from the park. Overall, predation accounted for only a small percentage of total livestock losses. It is unclear what exactly
the effect of predation was on transhumance patterns, but the fact that FulBe Ngara’en continued to use the pastures in and adjacent to Waza National Park suggests that it affected their transhumance patterns only minimally.
Pastoralists suffered the largest cattle losses because of diseases, such as trypanosomiases, foot-and-mouth disease and a variety of parasites. These diseases are
endemic, but outbreaks more often affect areas with concentrations of herds, like
178
Part II – Impact of ref looding
the Logone floodplain. The impact of diseases on transhumance patterns is unclear,
save in the case of serious outbreaks. In 1997, for example, an epidemic of footand-mouth disease struck the borders of Lake Maga, where many pastoralists stayed
towards the end of the dry season. The pastoralists there suffered high losses;
1,512 of their cattle died (Fig. 7.5). Other pastoralists, who used the pastures of Lake
Maga as a transition zone to their rainy season pastures, waited as long as possible in the reflooded area before moving south, thereby increasing the grazing
intensity in the reflooded area that year.
Pastoralists’ incursions into Waza National Park
One of the goals of the floodplain rehabilitation was to reduce the competition between livestock and wildlife in Waza National Park by stopping pastoralists’ incursions into the park. The assumption was that reflooding would offer the pastoralists responsible for illegal grazing in the park alternative grasslands outside the
park.
After the first year of reflooding, the Waza-Logone Project started a dialogue with
informal leaders of the FulBe Ngara’en from the Pétté-Fadaré area, the pastoralists who most frequently were caught in the park. This resulted in a verbal commitment in October 1995 of one of the most influential and respected leaders that
his herds and those of his followers would no longer enter the park. In December
1997, this commitment was formalized into a written contract between his group,
the Waza National Park authorities and the Waza-Logone Project.
In the period 1995-1997, the FulBe Ngara’en group showed a striking reduction
in grazing duration in the reflooded area, following the decision of the (absentee)
owners to send their herds further north during the early dry season to avoid any
risk of entering Waza National Park. Other FulBe Ngara’en pastoralists followed this
example, in part because their leader no longer intervened on their behalf in conflicts with the national park’s authorities. But FulBe Ngara’en pastoralists from
Pétté and Fadaré also argued that a major consideration for their commitment to
refrain from grazing inside Waza National Park was the improved condition of
grazing lands due to the reflooding. This left sufficient grasses in the reflooded
zone for the hot dry season when herd mobility is limited. Nonetheless, pastoralists noted that they suffered major losses of newborn calves during the first year
of changing transhumance routes and pastures, a situation that improved in subsequent years when herders acquainted themselves with the site-specific circumstances. The reduction in pastoralists’ incursions into the park during 1995-1997
was confirmed by park authorities who only occasionally reported herds within
park boundaries.
However, from 1997 onwards, pastoralists complained about the increasing number of cattle in the reflooded area and elsewhere in the floodplain due to an influx
7 – Past oralist s responses
179
of ‘new’ pastoralists, which left little forage for them towards the end of the dry
season. These developments undermined the agreement between the FulBe
Ngara’en and the Waza-Logone Project, which was already under pressure because
of illness of the FulBe leader and personnel changes at the Waza-Logone Project.
In the end, despite the increase in perennial grassland productivity, FulBe Ngara’en
pastoralists returned to graze their animals in the park.
7.5
Discussion
How did mobile pastoralists respond to the rehabilitation of the Logone floodplain? Did territorial behavior or the lack thereof lead to a lower or higher exploitation of the gradual increases in perennial grassland productivity? Did the rehabilitated floodplain limit the incursions of pastoralists into Waza National Park? The
data presented above allows to answer these questions.
Our data shows that risks of cattle losses are spatially distributed (Fig. 7.5). Pastoralists are aware of this and make decisions, if relevant, accordingly. Cattle losses due to predation were limited in both relative and absolute terms, and probably
did not influence transhumance patterns. Cattle losses to theft were much higher, particularly in the reflooded area. But the effect on transhumance patterns
seemed minimal; grazing intensities did not increase sharply when security improved. Outbreaks of contagious diseases, however, did affect transhumance patterns in 1997, but only temporarily.
Our data on pastoralist responses can be summarised as follows. During the first
six years of reflooding, grazing intensity in the reflooded area has increased li2
nearly, from 27 to 69 cattle km , contrasting the density dependent increase in the
number of camps that after a rapid increase in 1995 levelled off. The continuing
increase in the number of herds, albeit slowing down to 7% in 1999, can be partly
attributed to natural growth by splitting of the herds present, but the influx of new
herds account for most of the increase. The increase in grazing intensity in 1994
and from 1997 onwards is for a large part due to the increasing grazing duration
in the reflooded area, because pastoralists spent there more time before they moved
further north. The question is what this means for the three scenarios of pastoralists’ responses to floodplain rehabilitation that we earlier postulated.
The overshoot scenario predicted social and ecological ‘chaos’, triggered by a rapidly increasing number of pastoralists and their cattle. In 1995 the area indeed witnessed a rapid increase in the number of camps (Fig. 7.2). This was followed by violent clashes between Arabs and Musgum fishers. However, these conflicts did not
continue in the following years, in part due to mediation by the Waza-Logone Project. The number of camps continued to increase after 1995, but at a much slower
180
Part II – Impact of ref looding
rate. Between 1993 and 1999 overall grazing intensity increased steadily, in part
because of the influx of newcomers, in part because of natural growth of existing
herds, and in part because some pastoralists spent more time in the reflooded area
before they moved further north to their dry season pastures. There are no indications that the increase in grazing intensity has led to overgrazing as from 1995
onwards perennial vegetation has increased with an annual 7% of the reflooded
area (Scholte et al. 2000b). We conclude therefore that the overshoot scenario has
not taken place.
The territorial scenario predicted that pastoralists who used to exploit the area
before reflooding would prevent newcomers from using the reflooded area. In the
Logone floodplain, there is not a system in which pastoralists ask formal authorisation to access the floodplain as in the Inner Niger Delta in Mali (Legrosse 1999).
There used to be, however, a system in which pastoralists pay taxes and tributes
to traditional chiefs in exchange for protection of access to pastures and personal
safety (Moritz 2002). This ‘nomadic contract’ has come under considerable pressure in the last decades. Contrary to what could be expected in a territorial situation,
most pastoral groups have received proportional benefits from the reflooding, calculated as grazing intensity increase per household (Table 7.3). FulBe Mare’en, who
could have expected to behave territorially given their dominant presence prior to
the reflooding, even have had lower benefits, calculated per head of livestock, than
the other pastoral groups (Table 7.3). The limited increase in number of camps in
the first reflooding year could be attributed to territoriality, but in view of the
sharp increase in number of camps in subsequent years, it is more likely that lack
of information was the main cause. The stabilisation in number of camps in 1996,
the low increase in number of herds and the prolonged stay in the reflooded area
might suggest a territory effect. In 1997, however, the number of camps increased
again, while the number of herds stabilised and grazing duration declined. In 1999,
the 1996 pattern repeated. This fluctuating pattern is difficult to explain with a territorial scenario.
The fluctuating pattern seems to indicate the existence of a feedback mechanism
between the number of camps, herds and grazing duration, the product of which, grazing intensity, fluctuates around the saturation level of a linearly increasing grazing
capacity. We imagine the following mechanism. In a given year with a high number of camps and herds, saturation of grazing intensity is relatively quickly reached,
leading to reduced grazing duration, i.e., early departure of pastoralists to other
parts of the floodplain. This negative experience will dissuade potential newcomers from visiting the following year the reflooded area. In the following year, the
reflooded area will have an increased carrying capacity because of its further developing perennial vegetation. With an equal number of pastoralists, the reflooded
area offers more forage, meaning that they stay there longer. This leads to positive
evaluations and an increasing number of herds in the subsequent year, etc.
7 – Past oralist s responses
181
This fluctuating pattern supports an Ideal Free Distribution Scenario, which assumes
that any increase in perennial vegetation would be subject of consumption by cattle. Apart from the lag in vegetation development in the first year (Scholte et al.
2000b), perennial vegetation and grazing intensity both increased linearly during
the study period, suggesting optimal exploitation of grassland resources. Models
of the ideal distribution posit different ways in which ideal distribution and thus
optimal exploitation is achieved. The Ideal Free Preemptive distribution model assumes that the first individual to use a territory can preempt it. Individuals first
occupy the territories with highest quality and only the lower quality ones when
9
the population density increases (Pulliam & Danielson 1991). The preemptive
distribution gives distinct advantages to someone who arrives first with his herds,
over his competitors in exploiting and depleting the vegetation around his campsite before moving on. The question is why our data on changes in grazing intensity fits the ideal free distribution scenario and whether there was a pre-emption
effect, i.e., how did pastoralists adjust their transhumance patterns to achieve an
optimal exploitation of the rehabilitating grassland resources.
Pastoralist responses
The Logone floodplain is best described as an ‘annual grazing area’ for mobile
pastoralists (Niamir-Fuller 1999a), i.e., an area used by one or more ethnic groups
in which land is not held in common and no action is undertaken against intruders (Casimir 1992). However, in the case of mobile pastoralists in the Logone floodplain this does not mean that access to grazing lands is regulated by the principle
10
of first-come-first-served, as Casimir suggests (1992:160). Mobile pastoralists in
the Far North of Cameroon have negotiated access to grazing lands through higher-level institutions, in what we have called the ‘nomadic contract’ (Moritz et al.
2002). Compared to the rainy season transhumance area of the Diamaré, there is
limited contact between mobile pastoralists and higher-level institutions in the Logone floodplain, although pastoralists pay taxes and tributes to the various authorities in the floodplain.
The overall result is that no pastoralists are denied access, as long as they pay dues
to the authorities. In the Logone floodplain, negotiations and coordination occur
therefore primarily amongst pastoralists. Niamir-Fuller (1990) calls this ‘passive
coordination’ in which no formal agreements are made between pastoralists but
where coordinated movements result from individual decision-making. When spatio-temporal distribution of cattle is determined by herders on the basis of social
9 The relatively uniform floodplain vegetation masks possible quality distinctions, an issue that has
not been further studied.
10 Although access to grazing lands for main camps is not regulated by the principle of first-comefirst-served; it is the principle by which the luci herds gain access to pastures. There is increasing competition among luciji herders to enter the floodplain as soon as possible, even though the partial flooding carries health risks.
182
Part II – Impact of ref looding
and ecological characteristics (Turner 1999: 104, 105), this does not automatically
lead to competition over the same campsites or grazing lands, since individual
herders weigh costs and benefits differently and consequently make different
choices. Passive coordination in the floodplain is likely also achieved through formal and informal arrangements or mechanisms. Preemptive use, for example,
might be reflected in the customary rights that pastoralists have to campsites in
‘sojourn pastures’ that they have occupied for a long time (i.e., a first-come firstserve basis at a different time-scale). However, pastoralists have only limited usufruct rights over the grazing lands radiating from their campsite, the boundaries
of which ‘expand and contract in response to the local availability of fodder’ (Turner 1999: 108). Because the radii of different campsites overlap, pastoralists from
a number of different camps use the same grazing lands. When reflooding led to
greater grassland productivity, this allowed newcomers to occupy new sites in
proximity to existing customary campsites while avoiding conflicts over campsites
and grazing lands. This explains how the number of camps in the study area could
more than double without leading to significant conflicts among mobile pastoral11
ists.
The other major contributors to the increase in grazing intensity were pastoralists
who use the reflooded area as a transition zone to pastures further north. These
pastoralists use ‘transit pastures’ and occupy ‘travel campsites’ along the transhumance route that takes them further north. We did not find any timing conventions on these transit pastures, meaning that pastoralists can adjust the time they
spend in the reflooded area depending on the resources available. The data presented above shows that pastoralists indeed adjusted their transhumance patterns
primarily by shortening or prolonging the time they spent in the rehabilitated part
of the Logone floodplain.
The development of pastoral exploitation of the perennial grasslands in the
reflooded area confirms that mobility and flexibility are key adaptations to the
African drylands (Behnke et al. 1993; Breman & de Wit 1983; Niamir-Fuller 1999b).
More importantly, there is no evidence that the reflooding led to a ‘tragedy of the
commons’ (Hardin 1968). On the contrary, pastoralists adjusted their numbers
and the time spent in the reflooded area quickly and effectively in response to
changing productivity of perennial grasslands through ‘passive coordination’ (Niamir 1990). Our studies thus not only show the resilience of the Logone floodplain
ecosystem (Scholte et al. 2000b) but also the continued flexibility of mobile pastoralism.
11 In all the years of working with mobile pastoralists in the Logone floodplain, we have not come
across conflicts between pastoralists over campsites or grazing ranges.
PART III
Enhancing Conservation –
Development Integration by
Management Planning and
Training
8
9
10
11
Risks: the overshoot of success
Immigration: A potential time bomb under the integration of
conservation and development
Grip on the Whole: management planning
At the Interface of Legislation and Wildlife Management: A decade of
experience in consensual protected area management planning
in Cameroon
Foundation: development of human capacities
Curriculum Development at the African Regional Wildlife Colleges,
with special reference to the Ecole de Faune (Cameroon)
Wildlife Managers’ Perceptions of Community Conservation Training
in West and Central Africa
8
Immigration: A potential time-bomb
under the integration of conservation
and development
Paul Scholte
Published in Ambio 32: 58-64
Abstract
Integrated Conservation and Development Projects (ICDPs) aim to stimulate conservation without the previous negative experiences for local people, but pay little attention to their long-term impact such as immigration. The rehabilitation
of the Logone floodplain in North Cameroon, the core activity of the WazaLogone ICDP, has led to a 34% increase of sedentary fishermen and a multiple
number of temporary fishermen. Whereas livestock pressure tripled, Kob antelopes, a key floodplain species, have not increased, reducing their competitiveness. The virtual disappearance of wildlife in nearby Kalamaloué NP, due to
advanced human encroachment forms, is therefore a bleak perspective for Waza
NP. Examples from the Central African Republic, Galapagos, Nigeria and Zimbabwe also showed that in open-access systems, improvement in living standards
(development) may stimulate immigration, jeopardising the stability necessary in
protected areas (conservation). Most ICDPs lack demographic monitoring, masking its possible immigration risk. To counter the immigration risk in Waza, a policy was formulated based on local stakeholder categorisation and subsequent
privileges, resulting in the voluntarily displacement of a village out of Waza NP.
It is further recommended that ICDPs should be involved in regional land-use
planning and discourage development activities that stimulate immigration.
187
8.1
Introduction
Decades of conservation efforts have left a global network of protected areas, rich
in wildlife, forming a true heritage for mankind (Bruner et al. 2001). However, in
many instances, the eviction of villagers and nomadic groups associated with the
creation of these areas has left deep scars. Subsequent lack of participation in the
profits of mass tourism or other park revenues has nourished an almost universal
local hostility to protected areas (Anderson & Grove 1989; IIED 1994). In the mid1980s this led to a change in conservation paradigm, whereby development was
incorporated to counter the negative aspects of conservation for local people (Hulme
& Murphree 1999), in particular their loss of access to exploitable resources, generally called people-oriented conservation (Jeanrenaud 1999) or community conservation (Adams & Hulme 2001). Further, it was felt that the efficiency of conservation could be improved by linking conservation and development (Brown &
Wyckoff-Baird 1992), hence the dominant project form of Integrated Conservation and Development Projects (ICDPs) referred to in this paper. In an idealised
form ICDPs cover three subjects: i) protected area management, ii) management
of buffer zones, and iii) local social and economic development (Brown & WyckoffBaird 1992). Emphasis has been placed, at least on paper, on stimulating the participation of local communities in the formulation and implementation of these
projects.
The decade after the start of the first ICDPs, a variety of reports on their activities
have appeared (IIED 1994; Brown & Wyckoff-Baird 1992; Wells et al. 1992; Brandon 1997; Brandon & Wells 1992; Caldecott 1996). Initially the discussion centred
on the justification of people-oriented conservation (IIED, 1994; Wells et al. 1992)
or, alternatively, its reject (e.g. Spinage 1998). Later reports discussed preliminary
results of ICDPs and the need for adjustments. Brandon (1997), for example, identified questionable assumptions on which many ICDPs have been based which led
to disappointing results, especially with regard to their conservation objectives. She
argued that many ICDPs have been set up as aid projects where typical governmental responsibilities, such as law reinforcement, have been neglected. Lately,
Adams & Hulme (2001) argued to move beyond rhetoric and proposed more diversified community conservation, tuned towards the specific situation.
The long-term role of ICDPs in the future of protected areas, most notably the
danger of immigration encouraged by its own activities, has thus far received only
scant attention. In a review of ICDP case studies Salafsky (1994: 455) noted that
‘none of the case studies mentions developing strategies to deal with the influx of
poor migrants that can be expected if the project does indeed succeed in raising
standards of living relative to surrounding areas’. Brandon & Wells (1992) and to
some extent Caldecott (1996) also discussed the potential risk of people being attracted by the activities of an ICDP. First-hand experiences with the phenomenon
188
Part III – Enhancing conservation – development integration
have been reported, in narrative terms only, from the forests of West Africa (Oates
1999) and the Central African Republic (Noss 1997).
One can well imagine the following, highly simplified, immigration-risk scenario
for open-access systems (Ostrom 1990). A protected area is surrounded by a few
poor communities, who depend on the (illegal) exploitation of the protected area’s
natural resources. However, this low-intensity exploitation has hardly any impact
on the area’s ecology. For a variety of reasons, an ICDP is started in the area, improving local living conditions, thereby attracting (poor) newcomers, who want to
share in the increased resources. In the absence of barriers, immigration will continue until everybody has once more reached the same (low) level of development.
The result is that the protected area is now surrounded by many, instead of a few,
poor communities, who still have to make use of the same resources, thereby jeopardising its ecology due to their increased numbers. Is this development or conservation or neither one?
This article provides case studies of an ICDP in the Waza-Logone area of Cameroon
with which to examine the validity of this scenario for open-access and possibly
other property systems as well. These experiences are compared with those from
other ICDPs, and strategies to deal with immigration problems associated with
such projects are illustrated with a recent example from Waza-Logone.
8.2
Waza-Logone case studies
The Waza-Logone area
2
The Waza-Logone project area covers about 8000 km in the Sahelo-Sudanian zone
of Cameroon (Fig. 8.1). The area is bordered in the west by Nigeria and in the east
and north by the Logone River, which forms the border with Chad. The region is
characterised by floodplains, which are intensively used for fishing (Drijver et al.
1995) and dry-season grazing (Scholte et al. 1996b). The area includes two national parks, Waza and Kalamaloué, containing a diverse population of large mammals: Elephant (Loxodonta africana), Giraffe (Giraffa camelopardalis), Hippopotamus
(Hippopotamus amphibius), seven antelope species, three primate species, Warthog (Phacochoerus africanus) and predators such as Lion (Panthera leo), Spotted
and Striped Hyena (Hyaena hyaena, Crocuta crocuta). The Waza-Logone area is
also considered as one of the principal waterbird areas in West-Central Africa
(Scholte et al. 1999a).
Since 1979, the natural hydrological regime of the Logone floodplain has been
affected by the construction upstream of a dam near Lake Maga and an embankment along the Logone river, as part of the parastatal irrigated rice scheme (see Fig.
8.1). These structures have reduced both the depth and extent of flooding in an
8 – Immigration, a potential time-bomb
Figure 8.1 – Map of the Waza-Logone area.
189
190
Part III – Enhancing conservation – development integration
2
area of 1500 km . Subsequently, productive perennial grasslands have been invaded by annual grasses, limiting dry-season regrowth and reducing the carrying
capacity of the area for both livestock and wildlife (Scholte et al. 1996b). Fishing
resources have declined dramatically as well (Drijver et al. 1995). Since 1979,
about 40% of the human population then settled on the floodplain has left the
area, while a large part of the remaining population has shifted from fisheries and
animal husbandry to agriculture (Scholte et al. 1996b). Lack of flooding is perceived as the major problem in villages around Waza National Park (NP), followed
by the denial of access to fishing and grazing grounds in the park, and intimidation by park authorities (Scholte et al. 1999b; Chapter 9).
Impact of human encroachment on Kalamaloué NP
In contrast to Waza NP, which has a size of 170 000 ha, Kalamaloué NP is a small
national park of 4500 ha situated within a densely inhabited area (Fig. 8.1). As will
be argued below, developments that have taken place around Kalamaloué NP and
subsequently had their impact on the park, might well take place around Waza NP
if the ICDP, which was only started in 1992 and therefore has only had a limited
impact as yet, was to become too successful. Kavanagh (1978) concluded, based on
his experiences in 1974 and 1975, that Kalamaloué NP had tremendous potential
for tourism with its wide variety of mammals and birds. The ten park guards,
since 1974, and their reasonable working conditions should have been sufficient
to maintain its condition (Bruner et al. 2001). Nonetheless, 20 years later dramat1
ic changes in wildlife numbers have taken place: Waterbuck (Kobus ellipsiprymnus) became extinct, and the reduction of Kob (Kobus kob) and Warthog overshadow variation in numbers due to counting bias (Table 8.1).
Recent studies have indicated an enormous pressure on the park from surround2
ing communities. Firewood extraction has been estimated at more than ten tons
(wet mass) a day, considered to be unsustainable as indicated by the declining cover
of woody species. Each day an average of five herds of 60-120 cattle and ten smallruminant herds from neighbouring villages, complemented by several herds from
further away, are found inside the park. Annually approximately 300 fishermen
1 Wildlife numbers in Kalamaloué NP (based on unpublished reports by the Ecole de Faune Garoua
and Scholte and Saleh unpublished data): Line transect counts, covered by foot with three observers
per line, have been carried out in 1979, 1981 and 1984. All animals observed in a 100 m band on both
sides of the 15 transects, 1 km apart, were counted, corresponding to a 20% overall sample. In 1994
and 1996 these counts were repeated by park guards. The latter also confirmed the disappearance of
species that were no longer found.
2 Natural resource exploitation in Kalamaloué NP, based on Saleh (1994): A general survey was carried out by a forestry student, originally from the area, in each of the 26 villages bordering Kalamaloué
NP, obtaining information on village history and demography. This student assessed the number of
fishermen within the park during a three-month study. Firewood use extraction from the park was estimated in 1996 during three sessions each of 4 days on 15 main transit points around the park in different seasons. Information on the presence of livestock in the park was obtained during these studies.
191
8 – Immigration, a potential time-bomb
Table 8.1 – Wildlife counts in Kalamaloué NP, 1979-19961
Species
Warthog
Red-fronted Gazelle
Kob Antelope
Waterbuck
1979
1981
1984
1994
1996
563
618
141
175
50
38
132
0
410
195
2733
1837
112
920
95
481
1159
extinct
extinct
extinct
Bushbuck
68
73
9
60
present
Common Duiker
88
0
0
5
10
1
Based on line transect counts, see footnote 1.
make use of the fish resources of the park. Despite the official ban on exploiting
park resources, many people have deliberately chosen to live on its borders. Of the
26 villages in or on the border of the park (within a distance of less than 1 km),
eight were created after its delimitation as a forest reserve in 1946 followed by
another six after its designation as a national park in 1973. This substantial immigration was caused by the availability of fish as well as pastures and wood in the
park, which are easy to market as a result of the proximity of the ever-growing
markets of Kousseri and N’djamena.
The Waza-Logone ICDP’s reflooding’s impact on Waza National Park
The Waza-Logone project started in 1992 with the goals of planning and implementing interventions to restore the area’s biological diversity, and to sustain the
development of its inhabitants: a genuine ICDP. After consultation with local communities and (national park) authorities, a water course blocked by the embankment
2
along the Logone river, was re-opened in May 1994, reflooding an area of 180 km
of desiccated floodplain.
Monitoring (Scholte et al. 2000b) showed that in 1994, the first year of reflooding, vegetation production was high and perennial grasses set seed abundantly,
but no changes in vegetation composition were observed. In subsequent years an
annual 7% increase in perennial grasses, most notably Echinochloa pyramidalis
and Oryza longistaminata, has been observed, and these are expected to fully cover
the area once again in 2003 (Chapter 3). Vetiveria nigritana, dominant before the
dam construction, has not shown any increase as yet. Waza NP harbours an important large-mammal population, with fluctuating numbers during the last 34
3
years of counts based on various methods. Numbers of Kob antelope have reduced almost ten-fold between 1977 and 1993, a variation that can not be explained
3 Wildlife numbers in Waza NP (based on Esser & van Lavieren 1979, Scholte et al. 1996a, Scholte
unpublished data, see also Chapter 6).
Part III – Enhancing conservation – development integration
by count bias (Fig. 8.2). The decline of this typical floodplain species (Kingdon
1997) may be primarily attributed to the reduction in dry-season forage availability due to the Maga dam construction in 1979. The impact of poaching on antelope’ numbers was probably of secondary importance. With the rehabilitation of
the floodplain, the Kob population has initially increased, but numbers stabilised
from 1995 onwards (Fig. 8.2).
40000
Kob antelope
in Waza NP and
reflooded zone
number
192
Cattle
number in reflooded zone
(dry season)
30000
20000
10000
0
1962
1976 1977
dam
1993 1994 1995 1996 1997
reflooding
1999
Legend: Methodology Kobs (Chapter 6), cattle (Chapter 7). No cattle surveys have been conducted
prior to 1994. Indicated years reflect the relevant flooding season (= actual survey year -1).
Figure 8.2 – Number of Kob antelopes and cattle in Waza National Park and the
reflooded area.
In spite of 40 years of protection, with 20-30 reasonably well equipped park guards,
almost all surrounding communities continued to make intensive use of Waza
NP. Or, as they themselves put it: ‘we have always fished in the water holes our
parents dug; the park only came afterwards’. When the park is flooded (from June
to December), people are not hindered by the presence of game guards, who cannot easily access the flooded area. ‘It is with the grader opening the roads that the
park boundaries return’.
In 1994-5, a 25% increase was recorded in the numbers of sedentary fishermen
fishing within and near Waza NP, increasing to 34% in 1995-6, the second year of
4
-1
reflooding (Table 8.2). In 1994-5 the harvest was estimated at over 10-kg fish ha
4 Fishermen population in Waza NP, based on Saleh (1995): The number of sedentary fishermen
exploiting Waza national park and its direct surroundings (<5 km of boundary) in the previous and
193
8 – Immigration, a potential time-bomb
Table 8.2 – Changes in number of sedentary fishermen in Waza NP villages, 1993-19951
villages
1993-4
1994-5
1995-6
% increase
location of origin2
Tchédé
22
26
28
27
Lagdo
Zouang
21
21
21
0
–
Lougouma
14
26
30
114
Lagdo, Maga, Garlé
Mahé
15
19
21
40
Zina, Tila, Guidiba
Baram
16
18
18
13
Lagdo
Total
88
110
118
34
1
2
Methodology, see footnote 4; Lagdo, situated 300 km to the south; Maga, situated 50 km to the
south; others situated nearby in floodplain.
with a wet-season income per fisherman of US$ 400, much more than a farmer
can make. The number of people fishing in and near Waza NP has been estimated to have tripled with the arrival of seasonal fishermen from neighbouring villages. Their highly variable presence and the inaccessibility of the floodplain during flooding prevented monitoring their numbers in any detail.
Livestock intrusion from nearby grazing lands is considered to be one of the most
urgent problems by the park authorities (Scholte et al. 1999b). In the first three
years after the reflooding, the number of pastoral camp groups in the area near
Waza NP increased from 47 to 106. Livestock pressure in the dry season increased
5
in the period 1993-1999 from 14 000 to 39 000 cattle (Fig. 8.2). The cattle: Kob
ratio increased from 5 till 8, illustrating the marginalisation of the antelope populations that, towards the end of the critical harsh dry season, compete with the abundant cattle for the same, scarce grasses (Noupa 1995).
In the first years after the reflooding, the sedentary population of the villages on
the Waza NP edges was increasing at an annual overall rate of 5%, composed of
3% immigration and 2% natural growth, representing a doubling within a period
6
of 14 years. Only one new village is planned to be established, at Doudou Ndiyam
near Goulou (Table 8.3; Fig. 8.1). Five households who left in 1985 have written to
the administrative authorities asking to be allowed to return to their natal site.
current year was assessed in 1994-5. Fishing revenues were estimated based on direct observations
and interviews. Repeated visits complemented data for 1995-6.
5 Grazing intensity of nomadic and transhumance pastoral population in the Waza area (Scholte et
al. 2005; Chapter 7).
6 Waza park village demography, based on Tambekou et al. (1997): During PRA studies (Scholte et
al. 1999b), a baseline demography study was executed, nine villages were covered in 1994-5, the
remaining five in 1995-6. In 1996-7, the number of households was re-examined and information
was obtained on motivation of settlement and village of origin. Households were used as sampling
unit, being more reliable than individuals.
194
Part III – Enhancing conservation – development integration
Waza NP complements the fishing and grazing grounds elsewhere in the plain,
due to its position downstream in the floodplain (Fig. 8.1). Depressions outside
the park dry up several weeks earlier than those inside and fishermen shift their
fishing activities into the park when catches have fallen elsewhere. With the reflooding, water holes deep in the interior of the park have become well stocked with fish
and nutritious vegetation until the end of the dry season, and therefore attractive
to fishermen and pastoralists. The reflooding has had major benefits for the park’s
natural resources as well as for the surrounding human communities, at first
sight a ‘win-win situation’. As suggested by the deteriorating cattle-antelope ratio,
it seems unlikely that the increased benefits for the park can keep pace with the
demands of the increasing human population and, without interventions to restrict access to park resources and immigration, the situation at Waza may ultimately resemble that at Kalamaloué NP.
Table 8.3 – Changes in number of households in Waza NP villages, 1994-19961
Number of households
migration balance
household/year
1996-7
1995-6
Amkodje
14
11
Andirni
68
70
-1
Badadaye
29
27
1
Baram
17
17
0
Diéguéré
23
23
Goulou
52
Lougouma
22
Ndiguina
91
Niwadji
69
Mahé
22
Mbili
79
78
Tagawa 1,2
29
26
Tchede
24
Zwang
242
TOTAL
515
1
2
1994-5
3
3
20
1
64
2.5
19
1.5
90
1
1
3
22
23
208
0
1
285
1
0
15 (3.0%)
Methodology, see footnote 6; Includes one household ‘naturally’ constituted (not included
in migration calculations).
8 – Immigration, a potential time-bomb
8.3
195
Other immigration examples
The Campfire program in Zimbabwe is more a rural development program than
an ICDP (Metcalfe 1994). Nevertheless, owing to its success it is internationally
seen as a prime example of a community conservation project. In and around the
various districts where it is active, ‘land hunger’ is the major (push) factor for
immigrant installation (Dziginrai 1996). Immigration in wildlife areas causes
fragmentation of wildlife habitat and competes with Campfire revenues and projects (Bird & Metcalfe 1995). In Hurungwe district (West Mashonaland), the achievements of the Campfire project itself, visible in the form of revenue distribution,
rural clinics and schools, also attract people from other, non-Campfire, villages in
the district as well as from outside the district. This is considered ‘a very serious
problem’ by the Hurungwe district officials (Mr. Banda and Hurungwe district
councillor, pers. com. 1996) and Campfire officials alike (Anon. 1996). In a neighbouring district (Omay), mechanisms have been developed to control immigration
by an adapted form of land-use planning and its, till recently, (rigorous) application (Taylor, pers. comm.).
Lately, the impact of immigration on the forests of West and Central Africa has
received particular attention (Noss 1997; Oates 1999). Low initial population densities, a lack of land tenure and sudden large-scale (conservation) interventions in
an environment with few employment opportunities, are some of its principal reasons. Passionately, Oates (1999) described how various prime wildlife areas have
been destroyed through the bias of development projects and ICDPs alike. He
concluded that in the West-African rainforest context any mutual benefit between
conservation and development is an illusion.
The Galapagos Islands are a prime example to show that conservation and development are intrinsically linked. Wildlife tourism, with an estimated annual value
of US$ 100 million, and fisheries, has made the islands the richest province of the
country and triggered an influx of 12 000 people or three-quarter of the
Galapagos’ present population (Abbott 1996; Int. Am. Dev. Bank 2000). Despite
the restricted area open to human habitation, 3%, this influx threatens the islands’
long-term conservation, an issue that has received a fair amount of attention in the
popular press (e.g. Benchly & Franklin 2001). For years, conservationists have
pleaded for restrictions on immigration, arguing that any loss in biodiversity
would have major consequences for the booming tourist industry as well (Abbott
1996).
Immigration has also been described as a major threat for protected areas by
Sherbenin and Freudenbergen (1998), who analysed various push and pull factors
that lie at its origin. Surprisingly, they rejected the link between the activities of
ICDPs, influencing several of these push and pull factors, and immigration based
196
Part III – Enhancing conservation – development integration
on ‘incomplete evidence’ (Sherbenin & Freudenberger 1998: 51). Indeed almost
no ICDP monitors demographic trends in its project area, making it particularly
difficult to assess the impact on immigration. This is at odds with ICDPs’ development objectives and complements the concern of Kremen et al. (1994), who earlier noted a comparable lack of wildlife monitoring in ICDPs.
8.4
Discussion
Justification for ICDPs
The doling-out of development activities such as wells and roads is widely used to
motivate people to refrain from exploiting protected area resources, although there
is little experimental evidence for the success of this approach. It is also often seen
as a means of compensating for lost resources (Sayer 1991) and destined to all
people living at the edge of the protected area. It becomes somewhat theoretical to
compensate already evicted people for resources lost in the past. Instead of conceptualising solutions in terms of compensation, it seems more appropriate to
approach the problem by ‘learning by doing’, with initial activities such as the
Waza reflooding constituting a first step in a process of stimulating an atmosphere of trust and dialogue in which possible strategies to control the immigration risks can be negotiated.
But why should people in Waza receive such a ‘development compensation’ if they
still use park resources? The studies referred to above (Saleh 1994, 1995; Scholte
et al. 1999b), showed that hundreds of people exploit fish, wood and pastures
from inside Kalamaloué NP and Waza NP on a very regular basis. Their habitation on the park boundary would no longer be possible if the parks were really closed
to them. In a number of cases, the exclusion of local people has even privileged
others, like the inhabitants of a remaining village in Waza NP who fish intensively in water holes on the territory of one of the evicted villages.
Risks of ICDPs
Compensation for lost resources to all people in and around protected areas may
provide incentives leading to an intensification of exploitation of park resources,
based on increased migration into the area around the park. Even if exploitation can
be prevented in the short term with an improved park control system, the sharing
of improved resources by a larger number of people around the park, to the extent
that everybody is as poor as before, will only result in increased long-term risks to
the park’s survival. The occupation of forest reserves in Nigeria by immigrants
attracted by rural development opportunities (Oates 1999), showed the necessity
of establishing a link between conservation and development objectives (Brown &
Wyckoff-Baird 1992). The underlying problems are even more far-reaching: how
can an ICDP work in such frequently occurring open-access systems, where an
8 – Immigration, a potential time-bomb
197
increase in the standard of living attracts people and thereby increases pressure
on the protected area?
It is proposed to emphasise the long-term objectives of the ICDP that must be
realistic, even under changing conditions. Conservation risks should be explicitly
distinguished, if possible, from the start of project implementation, and a system
of adaptive management able to deal with them should be developed during the
execution of short-term activities. In cases of collapse of the state, or more common its advancing erosion, dependency on governmental agencies has shown to
be far less successful compared to a system based on people with traditional ties
to the area supported by international non-governmental organisations (Hart &
Hart 1997). Experiences from the Waza-Logone area have shown the unpredictability of changes, such as the insecurity in the area from 1990 onwards and the presence of key persons brooking deals between authorities and local communities.
Sinclair & Arcese (1995), Hart & Hart (1997) and especially Oates (1999) have previously questioned the efficiency of ICDPs. They argued that ICDPs fail to provide
success vis-à-vis conservation and enhance dependency on funding agencies, risking over-exploitation if the support system collapses. Without doubt the general
complexity of ICDPs and conflicting interests will lead to these problems, which
are well known in the development-project world (Brandon & Wells 1992). One
can question whether it is realistic to expect ICDPs to function properly in terms
of all their objectives, even though this is essential for their output (Brown &
Wyckoff-Baird 1992).
8.5
A way forward
Who is local?
Not everybody found in or around a protected area can be considered ‘local’. It is
therefore surprising that little effort has been made to identify local user/stakeholder groups in and around protected areas (but see Jeanrenaud 1999). In a publication on community wildlife management (IIED 1994), local resource users were
not further categorised, masking diversity in interests and, consequently, management potential. Around Waza NP, communities have been categorised based on
i) territory/space, ii) activities undertaken and iii) kinship (Fig. 8.3). Other, predominantly non-local, stakeholders can be categorised as park authorities, (governmental) agencies, (inter) national visitors, donors, researchers, business people
etc. (IIED 1994).
198
Part III – Enhancing conservation – development integration
WAZA C OMMUNITIES
Territory (i)
Kinship (iii)
• people returning to
their home village
8
• people of direct descent,
but not born in or in the
vicinity of the protected
area
9
protected area
village on
boundary
2
village
inside
village territory on
boundary
3
1
• agropastoralists
4
• seasonal fishermen with
activities outside their own
village territory
4
Activities undertaken (II)
transhumant pastoralists 7
villages/towns
further away
• village or town
based pastoralists 5
• seasonal
fishermen
6
Legend: Categories based on i) territory ii) activity undertaken iii) kinship, see also text.
Figure 8.3 – Categorisation of Waza communities
An appropriate categorisation of stakeholders is a crucial start for immigration
control. A solution formulated by Caldecott (1996: 52) for the open-access system
around the national parks in Cross River State, Nigeria: ‘an indigene was therefore anyone whose claim to be one was endorsed by an appropriate village council’
remains too open and may still lead to an unmanageable situation. National park
and/or ICDP authorities should also have a say in excluding some categories. ‘Advising villagers that registration of non-indigenes would be against their own
interests because it would cause benefits to be diluted’ (Caldecott 1996: 52) will
work only if the committees who decide on the criteria of who is local are small
enough, without the danger of being dominated by leaders.
Barriers to immigration: a (remaining) role for buffer zones?
Buffer zones have long been regarded as a tool to integrate protected areas in their
regional setting (Batisse 1982). They are assumed to extend the habitat for wildlife
and offer products for surrounding communities (MacKinnon et al. 1986). They
have also been proposed to halt further encroachment arising from the increasing
8 – Immigration, a potential time-bomb
199
demand for agricultural land by subsistence farmers (Nepal & Weber 1994). Implementation of the buffer zone concept is difficult in Waza-Logone and elsewhere,
as its legal status, as a type of protected area, may result in the perception of an
encroachment of the national park (Neumann 1997).
But even if such constraints are lifted, administrative regulations are by no means
sufficient for influencing people’s behaviour. Traditional leaders play a major role in
the decision-making process regarding the creation of new villages (Dzingirai
1996). In many areas the authority of traditional leaders is eroding, making these
leaders no longer the single deciding authority. Worse, this ‘power vacuum’ is one
of the main causes of the development of open-access situations, the single-most
important condition for immigration (Ostrom 1990). The traditional leader of
Kousseri, who controls part of the area covering Kalamaloué NP, explained that he
decides on the creation of new villages. Asked about the villages recently established, he was not aware of them: ‘Apparently my say in this is not always necessary’. It is, further, in the interest of local leaders to increase population numbers
in their territory, an economic incentive that cannot easily be reversed (Sultan of
Logone-Birni, pers. comm. 1996).
Analysing the human encroachment around Kalamaloué NP, one can conclude
that parameters that are regularly the focus of improvement by ICDPs, such as
accessibility, year-round water availability, availability of land for cultivation, grazing and wood collection and a nearby market, were all present. To avoid a similar
encroachment near Waza NP, the access to these resources should be regulated,
for example by strengthening the control of the park (Bruner et al. 2001). Experiences from Kalamaloué NP, with its relatively high park guard-size ratio (1: 500
ha), showed the limits of such an approach, however.
A more powerful tool would be a policy of active discouragement of development
activities in the surroundings of protected areas. In a buffer zone adjacent to the
protected area, no rural development activities should be practised, or only such
activities that have no appeal to human communities. In the transition or support
zones (Wells et al. 1992), sustained rural development as part of a general strategy of regional development may be promoted, in a manner that limits the risks of
human attraction. These support zones should provide alternatives to those communities that exploit the buffer zones and protected areas and will not receive
development assistance in this framework. General guidelines are difficult to provide and depend also on pragmatic choices that often must be made. However, the
utmost caution should be adopted vis-à-vis road construction (Wilkie et al. 2000)
and increasing pastoral and domestic water availability. This contradicts present
ICDP practices and buffer zone development as described by Sayer (1991), whereby various extractive uses originally proposed for the support zones only, are promoted right in the buffer zones.
200
Part III – Enhancing conservation – development integration
Some preliminary results of a new immigration policy in Waza-Logone
In September 1997, a management plan for Waza NP and its periphery zone was
approved by the government of Cameroon to improve relations between park
authorities and its neighbors (Scholte 2000b; Chapter 9). Another objective was
to prepare the park for the immigration of people to its borders with the floodplain
rehabilitation efforts by the Waza-Logone Project. In this management plan, a committee was created, with representatives of selected communities and park authorities (both with voting rights) and administrative and traditional authorities (without voting rights). The Waza management committee, formalised by a ministerial decision, has representatives of all villages either situated on the park boundary,
2, or with their territories bordering it, 3 (Fig. 8.3). Agropastoralists as well as transhumant pastoralists 5, 7, are represented as long as they exploit at present the
periphery zone of Waza NP. Seasonal fishermen, 6, were excluded. Baram, the
only village inside the park boundary, 1, was also excluded to participate in the
committee with the idea that this could ultimately persuade them to settle outside
the park boundaries, in accordance with the official conservation policy. In 1997
and 1998, village water pumps, small-scale irrigated rice cultivation and local
tourist camps, were promoted by the Waza-Logone Project in the park villages, 2, 3,
but excluded Baram, 1. People returning to their home or parents’ village, 8 and 9,
have been considered for the time being as belonging to one of the categories 1-7.
Characteristically, all Waza villages originate from well before the establishment
of the national park (Scholte et al. 1999b). In other situations it might be useful to
distinguish those villages and inhabitants present prior to the establishment of the
protected area from those which arrived more recently with, obviously, less rights.
The list of distinguished activities inside the park boundary may be extended to
include other natural resources than grazing lands and fishing waters as well as
ceremonial or religious attractions, as is the case of Niokolo Koba NP in Senegal
(pers.comm. students Ecole de Faune).
Early 1998, one of the families in Baram persuaded the other 14 families to move
out of the village (and out of the park). A letter was sent to various authorities and
to the Waza-Logone Project asking for assistance. In a subsequent correspondence,
the village was promised that, once moved beyond the park boundary, they would
benefit from the same advantages as the villages outside the park. In October
1998, the initiator settled with his family in Halé on the edge of the park (Fig. 8.1)
followed by eight Baram families between March and June 1999. Whereas in
1996 and 1997 the initiator received fines of about USD 150 for fishing inside the
park, in 1998 his relationship with the park warden improved considerably. In
June 1999, the Waza-Logone Project installed a highly appreciated water pump in
the new village. Whereas decades of oppression towards Baram by virtually all
park wardens did not change their continued presence, the new more balanced policy of ‘the carrot and stick’ has yielded a tangible result. As the initiator of the
8 – Immigration, a potential time-bomb
201
move told me (June 1999), ‘as long as they pressed us to leave we were determined
to stay, but as soon as there was a perspective of a more prosperous stay outside
the park and people were no longer daily harassed, we changed our opinion’. This
experience should guide the further development of the policy with regard to the
newcomers in a regional context.
8.6
Conclusions and recommendations
The Waza-Logone area is unusual in having experienced a degradation process
with subsequent emigration and having (partly) been successfully rehabilitated,
provoking a population build-up on the park’s borders. The risk of ICDPs becoming poles of attraction for people has also been noticed in the Central African
Republic (Noss 1997), Nigeria (Oates 1999) and Zimbabwe (Anon. 1996). These
areas have in common that, for a variety of reasons, access is not well regulated
and surrounding areas exercise a high (emigration) pressure. In other open-access
areas it is expected that the lack of real impact of ICDPs, often running for a limited period only, has masked the threat of immigration. Anyhow without a proper
demographic monitoring system only massive immigration can be detected. Such
demographic information is prerequisite to further analyse the risk of ICDPs, in
other than open-access property systems possibly as well.
These negative experiences do not detract from the necessity of integrating protected areas in their regional setting, for which the ICDP concept, if more realistically designed and executed, remains the best alternative. Taking into account
the following suggestions can already make some major improvements:
• More emphasis should be put on conservation objectives as experience shows
that they are often marginalised in daily ICDP routine (Brandon 1997). Improved
park control can reduce some of the impacts of immigration.
• ICDPs will have to devote more effort to targeting local resource users. Selected groups can play a key role in the management of the areas surrounding
the protected areas and may form a social fence for further influxes. Special
privileges will have to be given to these people and in some cases exclusive land
rights (Neumann 1997), without risking a loss of control over the protected
area. This could reduce the ‘openness’ of the area.
• To influence the development of activities around protected areas, ICDPs should
devote more attention to regional planning, an activity demanding the presence
of project personnel in the lobbying worlds in district, provincial and national
capitals. This will have to be accompanied by further commitments of the various authorities involved, most importantly an appropriate legal context and its
subsequent implementation.
202
Part III – Enhancing conservation – development integration
•
Generally, smaller budgets spread over longer periods will allow ICDPs to concentrate on their role as facilitator in conservation-development planning instead
of becoming a financing party with all its ‘spending obligations’ that often are
easier funnelled into development than conservation.
If these recommendations result in an ICDP becoming so complex that its conservation and development objectives are not realistic, while immigration remains
a threat for the protected area, its continuation is no longer justified. If continuing, the ICDP may otherwise appear to be a Trojan Horse.
9
At the Interface of Legislation and
Wildlife Management: A decade of
experience with consensual protected
area management planning in Cameroon
Paul Scholte
In: The Future of Conservation in Africa: Law, Biodiversity, Livelihoods and
Development (2005).
Abstract
Management plans are considered to be indispensable for the daily management of protected areas. Historically, their focus was on relatively straightforward technical or ecological issues. Since the 1990s management plans have
increasingly been based on the consensual interpretation of pressing management issues, often related to the legal context of protected areas. The most
important product of the modern management plan is a consensus building
process based on negotiation among a large number of stakeholders. This chapter asks whether the high expectations for management planning as an interface between protected area management and wildlife law and policy are realistic. The analysis first sketches protected area management policy in Cameroon,
particularly the limitations of the legal context in which it occurs. This is followed by an account of the formulation process for the Waza NP management
plan, a process that has led to a consensus on sensitive key issues among local
communities, local and national authorities, and international environmental
NGOs. This particular exercise in management planning was a learning process
for all concerned. The analysis questioned, however, whether the resulting management plan was sufficiently realistic to lead to its successful implementation.
The Waza case is then compared with management planning experiences in
other protected areas in Cameroon and in other African countries.
205
9.1
Introduction
Recent developments in African wildlife conservation
In the late 1980s and 1990s, African wildlife conservation underwent profound
changes. It became increasingly evident, for example, that wildlife conservation
could no longer be considered the sole responsibility of government. Local communities and, in many regions, the private sector would have to play a vital role.
Conservation has now come to be seen as part of an overall sustainable development
strategy. Although the precise meaning and consequences of this new conservation (Hulme & Murphree 1999) will continue to be debated by some (Spinage 1998)
and needs to be conceptualised more fully (Adams & Hulme 2001) protected area
managers can no longer ignore the pressure to come to grips with the new expectations imposed on them by the outside world.
In an increasing number of African countries, elements of the paradigm shift are
taken into account in newly drafted national wildlife legislation. But in most countries, including Cameroon, the situation on the ground in local communities
seems not to have changed very much (Bomba & Bigombo Logo 2001). This complicates things for protected area managers who are trapped between the ‘unrealistic’ aspirations of local people, ‘rigid’ legislation, ‘distant’ superiors and ‘impossible’ demands of donors and pressure groups.’ (Scholte 2000b). African protected area managers also have to face conflict situations (Hart & Hart 1997), the
intrinsic limits of conservation and development integration (Scholte 2003a) and
a lack of essential equipment and personnel. In the face of these constraints it is
hard to be sure what is actually happening in the field. Evidence from various parts
of Africa suggests, however, that there is a ‘silent’ wildlife crisis. Some explain this
by pointing to the continued dominance of fortress conservation, which limits the
viability of wildlife resources on common land (Brockington 2002; Gibson 1999).
Others blame the new conservation for the lack of success (Oates 1999).
Here, I focus on the long implementation trajectory for changing wildlife law and
policy. Now that the rules are shifting, is wildlife conservation still being pursued
as business as usual? Or are some of the offshoots of the new conservation paradigm really having an impact on the ground? Management plans are increasingly
seen as valuable tools for redirecting protected area management, especially when
they are part of donor assisted integrated community development projects
(ICDPs) (Thomas & Middleton 2003). Given the controversies surrounding many
African protected areas, it is no surprise that external support is often conditioned
by consensus building on the need for management planning. But does it work?
How does it work?
206
Part III – Enhancing conservation – development integration
Management plans
The ‘fashion’ for protected area management planning
Management plans are said to be indispensable tools for the daily management
of protected areas around the world (MacKinnon et al. 1986; Ortega-Rubio & Arguelles-Méndez 1999; Parr 1998; Rettie 1995). Management plans have multiple
functions, the most frequently cited of which are the identification of management needs for a protected area, the setting of its management priorities, and organising an approach to its future (MacKinnon et al. 1986), see also Fig. 9.1. Historically, management plans were focused on relatively straightforward technical
or ecological issues, for which ecologists, together with protected area managers
themselves provided the information base. Now, the ‘fashion’ for management
plans brings a broader mix of scientists, decision makers, and financing agencies
into the picture, each with their own motivations and expectations.
For scientists, management plans seem to play a role as baskets for research
results. A literature search of peer-reviewed science journals returned 172 conser1
vation related hits on the key words management plan. Sixteen per cent of these
papers analysed (parts of) existing management plans, 28 per cent described research that, according to the author, potentially supports the design of elements of
(future) management plans, and 50 per cent described more fundamental research
and superficially concluded that it could be useful for management planning.
In a number of countries, the preparation and implementation of management
plans is now legally prescribed for national parks. They include Australia (ANZECC
2000), Mexico (Ortega-Rubio & Arguelles-Méndez 1999), England and Wales (Thomas & Middleton 2003) and Cameroon, where
2
A management plan shall be drawn up for the State forests defining, in accordance with the conditions laid down by decree, the management objectives and
rules for each forest, the means needed to achieve the said objectives, as well as
the conditions under which the local population may exercise their logging
3
4
rights in accordance with the provisions of the classification instrument.
1 Based on a search through the abstracts of Institute of Scientific Information (ISI) journals from
1995-2003, 26% of the hits concerned papers dealing with conservation, the others had a medical
(46%), hydrological or geological (11%), technical (10%), fisheries/agronomic (7%) or management
(1%) orientation.
2 State forests include wildlife protected areas, such as national parks, game and hunting reserves,
state game ranches, wildlife sanctuaries, buffer zones and state zoological gardens as well as several
forest reserves categories (MINEF 1994: 9-10).
3 In the French law text the broader term ‘droits d’usage’ is used (traditional user rights) (MINEF
1994: 11).
4 MINEF 1994: 11.
207
9 – Management planning
Management plans are also required for natural sites being considered for inclusion
on the World Heritage List (Thomas & Middleton 2003) and ‘...provisions should
be made for a management policy or plan’ for Biosphere Reserves (UNESCOMAB 2002). NGOs and donor agencies have arguably become the strongest supporters of management planning. They see them as indispensable instruments
for the integration of conservation and development (Margolius & Salafsky 1998).
Legislation
Agency, policies,
strategies,
business plan
Regional plan, broadscale land
management planning
MANAGEMENT PLANS
FOR INDIVIDUAL OR
GROUPS OF RESERVES
Subsidiary, e.g. fire plans, site plans etc.
Operational/action plans, works programs
Implementations
Figure 9.1 – Planning hierarchy (adapted from ANZECC 2000)
In bold: relevant strata for Cameroon
Recent development of management plans
Parallel to the change in the wildlife conservation paradigm in the 1980s and 1990s,
the aim of management plans also changed. They developed from being essentially technical documents into more formal presentations of the outcomes of negotiation. Prior to the 1980s, a management plan was often a technical document,
informally prepared by a dynamic protected area warden. It had no formal significance and was only for internal distribution, often in photocopy form or even as
a carbon copy. Consultants were often asked to co-ordinate the management plan
formulation process and describe in detail the biological environment and pro-
208
Table 9.1 – Management Planning in Protected Areas in Cameroon1
Name
International
Status
Waza
National Park
(1934-1968)
Benoué
National Park
(1968)
Management Plan formulation
Main Issues
New policy/law
Interpretation?
Period
Agency
Coordination
Biosphere
Reserve
1995-1997
IUCN
Project team +
Warden
User rights
Yes
Biosphere
Reserve
1998-2002
WWF
Project team
Zoning in relation
to immigration,
new internal organisation
No
Park infrastructure
Developments
No
SAVANNA
Faro
National Park
(198?)
–
2001 - ongoing
WWF
Project team
LOWLAND FOREST
Dja
Game Reserve
(1950)
Biosphere
Reserve
World Heritage Site
1994 – 20042
EU-ECOFAC
Consultants +
project team
+warden
User rights in
relation to zoning
but status
Korup
National Park
(1986)
World Heritage
status pursued
1996-20023
WWF
Consultants
mainly
Park exploitations,
re-settlement
of park villages
Yes
Lobéké
National Park
(2001)
–
Ongoing
WWF
Consultants?
Zoning – park
exploitation
(village hunting)
Yes
(as Waza)
CampoMa’an
National Park
(2000)
–
2001-20044
Tropenbos
Int.
MINEF,
Project team
Poaching, tourism
development,
co-management local
communities
1
2
Comparable to
Waza, but status
as game reserve
gives more flexibility
No
This list is limited to protected areas where management planning is ongoing; Last version submitted for approval of the minister, two earlier versions rejected
3
4
(1999 and 2000); An earlier ‘development plan’, serving as project document, had no formal status; Submitted for approval of the minister.
Part III – Enhancing conservation – development integration
National
Status
(since)
9 – Management planning
209
posed management measures for its conservation. The internally discussed plan
sometimes achieved a more formal status when, in addition to instructing the park
warden and his/her colleagues, it was used to attract financing (as was the case in
Korup NP, see Table 9.1). More recently, management plans have become bulky
documents clearly aimed at external constituencies.
Lengthy descriptions in a management plan can mask the nature of its formulation process. Are the proposed management options and subsequent actions the
choices of the authors and their superiors, or do they reflect a widespread consensus? Consensus could bring wider support for sensitive management issues.
And recognising this, management plans have increasingly been based since the
1990s on the consensual interpretation of pressing management issues, where
there is often a conflict of interest among stakeholders. Indeed, the most important product of the modern management plan is a consensus building process
5
based on negotiation among a large number of stakeholders. In the case of Waza
National Park (NP) (Fig. 9.2), elaborated below, the key conflicts were between
local people and park authorities over the illegal exploitation of natural resources
inside protected areas.
Structure of the analysis
This chapter asks whether the high expectations for management planning as an
interface between protected area management and wildlife law and policy are realistic. The analysis first sketches protected area management policy in Cameroon,
particularly the limitations of the legal context in which it occurs. This is followed
by an account of the formulation process for the Waza NP management plan, a
process that has led to a consensus on sensitive key issues among local communities, local and national authorities, and international environmental NGOs.
Everyone agrees that this particular exercise in management planning was a learning process for all concerned (MacKinnon et al. 1986.) The analysis goes on to
question, however, whether the resulting management plan was sufficiently realistic to lead to its successful implementation. The Waza case is then compared
with management planning experiences in other protected areas in Cameroon
and in other African countries.
5 In the case of Kruger N.P., sensitive and highly controversial international issues involving elephant culling were involved (Braack 1997). In the case of Hwange NP in Zimbabwe, the key frictions
were between tour operators and park authorities (UNDP 1998).
210
Part III – Enhancing conservation – development integration
Figure 9.2 – Map of the Waza-Logone area, Cameroon.
9 – Management planning
9.2
211
Background of protected area management in Cameroon
Introduction
Cameroon has lowland and mountainous rainforest in the south and mountains
and lowland savannas in the north. Its wide-variety of ethnic groups and cultures
lead people to call it a ‘miniature Africa.’ Biogeographic diversity makes Cameroon one of the African countries with the highest number of mammal, bird and
plant species (Alpert 1993; Fotso et al. 2001). In the 1960s and 1970s, a strong
conservation policy led to the creation of some of the first national parks in West
Africa and to the setting up of a widely respected regional wildlife training centre
(Scholte 2003b). National parks were well staffed and had reasonable working
conditions. Although in the mid-1980s governmental support for conservation
diminished in a context of stagnating national economic growth, a Ministry of
Environment and Forestry was created in 1993, partly in response to the energy
generated by the Rio Summit. This was followed, in 1994, by a new law on the
environment designed to facilitate a more participatory style of conservation and,
in 1995, by its application decree.
The need for local participation in the development process has been underscored
by traditional development organisations, which have spent about ten million US
dollars a year on Integrated Conservation and Development Projects (ICDPs) in
Cameroon since the early 1990s. The Government of Cameroon has increased the
number of protected areas – a favored focus for ICDPs – especially in the rainforests and montane forests of southern Cameroon (Alpert 1993; Fotso et al.
2001). These changes have substantially complicated protected area management
in Cameroon, with a multitude of actors involved in an increasing number of lo6
calities. The results of all this activity have been mixed. The recently created rainforest protected areas, for example, have had only limited attraction for the tourism
and safari hunting businesses, even though their biodiversity is remarkable (Blom
2001; Wilkie & Carpenter 1999ab).
Ambiguous wildlife law
The 1978 law on the protection of national parks, though still valid, has been largely forgotten with the introduction of a new series of legal texts in the mid-1990s
(Kamto 1996). In the 1994 forestry and wildlife law broad, inclusive purposes
were suggested in the first section:
This law and the implementing instruments thereof lay down forestry, wildlife
and fisheries regulations with a view to attaining the general objectives of the
6 Cameroon strives to bring 20 per cent of its territory under some form of protection. There are currently ten national parks and seven faunal reserves, including three international Biosphere reserves
and one World Heritage Site (see Table 9.1). National parks situated in the savanna zones originated
from colonial times and have played an important role as catalysers for tourism (Fotso et al. 2001).
212
Part III – Enhancing conservation – development integration
forestry, wildlife and fisheries policy, within the framework of an integrated
management ensuring sustainable conservation and use of the said resources
7
and of the various ecosystems.
In subsequent sections, however, the 1994 law also appeared to lay the basis for a
more rigid and centralised approach. In this and other respects the law contained
significant ambiguities. ‘The state shall ensure,’ it said, ‘the protection of the
8
forestry, wildlife and fishery heritage’ and ‘The genetic resources of the national
9
heritage shall belong to the State of Cameroon.’ One analysis of the process by
which the 1994 law was drafted, however, stressed the dominant role of the World
Bank, the lack of leadership from the Government of Cameroon, especially from
the newly created Ministry of Environment and Forestry (MINEF), and the lack of
involvement of other stakeholders (Ekoko 2000).
The 1995 implementation decree, in describing categories of protected areas to be
designated and managed under the 1994 law, did not depart substantially from a
10
strict protection regime, and was in some respects stricter than the IUCN pro11
tected area designations. The decree article on resource user rights in protected
areas lacked any reference to the grazing of domestic animals, a major land use in
Northern Cameroon, and further highlighted a legal bias towards the economi-
7 MINEF 1994: 5.
8 MINEF 1994: 7.
9 MINEF 1994; 7.
10 The decree recognized the following types of areas: Integral ecological reserves (no area has been
designed as such); Game Reserves (see Table 9.1, infra); Areas set aside for the conservation, management and propagation of wildlife and for the protection and management of its habitat, where
hunting is forbidden, except by authorisation of the minister in charge of wildlife, as part of duly
approved management operations, and where dwelling and other human activities are regulated or
forbidden; National Parks (see Table 9.1, infra), defined as uninterrupted areas whose fauna, flora, soil,
subsoil, atmosphere, waters and natural environment as a whole are of special interest and should be
preserved from any natural deterioration and protected against any human interference likely to alter
their outlook, composition and evolution [in which respects there shall be taken into consideration the
preservation of endangered animals and plant species, as well as habitats, on all or parts of the national territory; the preservation or gradation of major migration routes of wildlife; the scientific or technical studies indispensable for the development of human knowledge; and where there are prohibitions on hunting and fishing, except as part of a park management operation; industrial activities;
extraction of materials; pollution of any nature; farming, grazing and forestry activities; stray domestic animals; aircraft flying at an altitude of less than 200m; the introduction of local or imported animal and plant species, except for scientific purposes or as part of management operation authorised
by the minister in charge of wildlife]; Sanctuary, zoological garden, game-ranch, and game-farming
areas; and Buffer zones (infra, section 3.2.4) (MINEF 1995: 120-121).
11 In comparison with the IUCN (1994), categories of protected areas, for example, only the strictly
protected areas are represented in the Cameroon decree. And in the case of national parks the prohibition on subsistence resource use is stricter than suggested by IUCN criteria. On the other hand,
human habitation is not explicitly excluded in Cameroon national parks.
213
9 – Management planning
12
cally important forest zone of Southern Cameroon. Wildlife in Cameroon is not
confined, however, to protected areas and, because of increasing pressures on the
remaining habitat, serious resource conflicts can arise in and around human settlements and pastoral camps. The 1990s legal texts are based on the assumption
that local wildlife authorities will have a strong presence and adequate intervention capacity, but this is unrealistic given the low number of personnel and limited transport and other equipment. This has made it particularly difficult for farmers and pastoralists to react to frequent crop and livestock damage, provoking bitter reactions (Scholte et al. 1999b). The 1995 decree says that ‘Any battue [culling]
13
shall be undertaken by officials of the services in charge of wildlife.’ But it also
says that ‘...no person may be charged with hunting a protected animal if his act
was dictated by the urgent need to defend himself, his livestock or his crops. Proof
of lawful defence must be given within 72 hours to the nearest official in charge
14
of wildlife.’
Some of the criticism of the 1994 law and 1995 decree has been directed at their
fragmentation, loopholes and inconsistencies (Mayaka 2002). More problematic,
however, are the implementation difficulties they pose, because they lack clear guidance on the meaning of crucial concepts, such as community, and because they do
not reflect consensus amongst the main stakeholders (Egbe 2001). The more general problem is that these legal texts were drafted in isolation, too far removed
from the field and lacking the feedback from field experiences.
Management planning: a way forward?
The implementation of a specific ICDP project, like Waza-Logone, cannot correct
ambiguities and other weaknesses in national legislation. Carrying out such a project does, however, provide a window of opportunity for making decentralised interpretations and policy decisions under the law. The making and adoption of
management plans are especially important sites for this activity, because the law
states that ‘Any activity in a state forest shall, in all cases, be carried out in accor15
dance with the management plan.’
Although two types of plans, for management and development, are recognised in
the 1995 decree, the application portion of the decree focuses only on the management plan, which it defines as ‘A technical document drawn up by the service
12 The decree defined user rights as ‘the exploitation of forestry, wildlife and fishery produce by the
local population for personal use. However except for game reserves, sanctuaries and buffer zones
where they may be authorised, such right shall apply neither to integral ecological reserves, national
parks, zoological gardens nor to game-ranches’ (MINEF 1995: 123).
13 MINEF 1995: 127.
14 MINEF 1995: 128. Not surprisingly the difficulty of presenting this proof has frequently created
conflict between wildlife authorities and farmers and pastoralists.
15 MINEF 1994: 11.
214
Part III – Enhancing conservation – development integration
in charge of wildlife, or by any other natural person or corporate body commissioned
by the service, for the purpose of planning in space and time all strategies to be
16
implemented for the sustainable use of one or several given wildlife resources.’
The room for manoeuvre in the process of making and adopting a management
plan is not entirely clear, however, especially with regard to user rights inside
national parks. In what follows I try to answer this question. The first step is to
describe the process whereby the first formally approved management plan in
Cameroon, for Waza NP, was drafted and adopted. This is then compared with
management planning experiences in other protected areas in Cameroon. A further brief comparison with management planning in other African protected areas
provides some insight into the extent to which experience in Cameroon offers
more general lessons for improving the management of protected areas.
9.3
The Waza-Logone case
Introduction
Waza NP is probably the best-known protected area in Cameroon. It receives as
-1
many tourists (6000 yr ) as all other Cameroonian protected areas combined and
more than any other protected area in Central Africa. It is situated in the SaheloSudanian Far North Province and covers large stretches of wooded grassland and
bushland (Fig. 9.2). The eastern part is characterised by tree-less grasslands and
is highly productive, thanks to the annual flooding of the Logone River. Waza’s
main attractions are the large Elephant, Giraffe, and Lion populations, as well as
high concentrations of antelopes and waterbirds during the dry season. Waza NP
was created in 1934 as a game reserve. It became a national park in 1968 and an
international biosphere reserve in 1979. Its history is closely linked with the various villages that surround the park, all dating from before its creation. They are
composed of ethnic Kotoko fisheries communities, Musgum and Bornouas agriculturists, and Fulani and Arab pastoralists (Drijver et al. 1995; van Est 1997;
Scholte et al. 1996b). At present only one village, Baram, remains inside the park.
The process of developing a management plan for Waza NP was based on previous work, which beginning in 1984 had identified four major problems. The first
was that the construction of a dam upstream, on the Logone River, had degraded
the environment in the eastern part of the park, causing dramatic reductions in
16 By contrast, a development plan is defined as ‘a technical document drawn up by the service in
charge of wildlife, or any natural person of corporate body commissioned by it, which determines in
space and time the nature and programme of works and studies to be carried out in a protected area
and which shall apply to the said area.’ However, development plans for protected areas managed by
individuals may be drawn up by the individuals themselves and approved by the service in charge of
wildlife (MINEF 1995: 120).
9 – Management planning
215
wildlife populations and serious declines in fisheries and livestock production (Drijver et al. 1995; Scholte et al. 1996b). There were, secondly, severe tensions between
national park authorities and neighbouring populations leading to an almost complete lack of collaboration facilitating access to poachers (Scholte et al. 1999b). A
third problem stemmed from a reduction in government commitment to park management, which led to deteriorations in vehicle condition, a lack of maintenance
for park infrastructure, and low salaries and bleak career prospects for park personnel. The park was also plagued by a legacy of poor decisions involving uncontrolled bush fires and poor road building and maintenance, which also resulted in
unnecessary habitat degradation. This interrelated set of problems was gradually
reducing wildlife numbers and yielding low economic benefits from the park to
neighbouring populations.
Management plan formulation (1995-1997)
Introduction
In 1995, the government of Cameroon delegated the formulation of a manage17
ment plan for Waza NP to the Waza-Logone Project. A core team to co-ordinate
the process was formed, comprising the warden of Waza NP, a short-term consultant who had worked in the area 25 years ago, the director of the Waza-Logone
Project, who had extensive experience in government service, and the author.
Additional project teams played important roles in carrying out various surveys, as
well as in stimulating discussions with local populations and between the Project
and its partners. Local consultancies were arranged to assist with financial, institutional, and legal issues.
To address the four problems noted earlier four main intervention strategies were
developed. One was the creation of a park protection and anti-poaching policy. A
second sought to improve the technical and ecological basis for park management
through better infrastructure and resource monitoring. A third strategy aimed at
better integration of development and protection goals, through improved community relations and tourism initiatives. And fourthly, there was a strategy to
improve the internal environment of park management by ensuring that the necessary human and material resources would be available. Some of these strategies
are relatively straightforward and require no major changes in policy. They essentially depend for their success on the availability and commitment of qualified
17 The decree requires that ‘Development plans,... , shall be rendered enforceable by order of the minister in charge of wildlife... All development plans shall be prepared in accordance with directives
issued by the ministry in charge of wildlife, and shall stipulate in particular: a general description of
the protected area; the basic objectives to be achieved, taking into consideration the interests of the
local populations and the necessity for the conservation of biodiversity; the works to be carried out and
the schedule thereof; the cost of the project; and indicators for its follow-up and assessment.’ (MINEF
1995: 127).
216
Part III – Enhancing conservation – development integration
people. The one strategy that clearly challenged existing policy for the park was the
third one, analysed more closely below.
A new start
As result of the dam and embankment built along the Logone River in 1979 to
enable irrigated rice cultivation in the Maga area, upstream of Waza NP (Fig. 9.2),
the lack of flooding has had major impacts on the Waza-Logone floodplain ecosystem. Fisheries production has dropped and livestock and wildlife can no longer eat
the perennial grasses which used to produce palatable regrowth during the dry
season. Floodplain rehabilitation by reflooding has, therefore, been the central
activity of the Waza-Logone Project since its inception. The first pilot release of
2
water in 1994 reflooded about 180 km and coincided with the start of the management planning process. When the release rehabilitated perennial grasslands
and produced more fish (Scholte et al. 1996a), it also created an atmosphere of
renewed confidence among local people in their collaboration with outsiders.
Increased resource availability, especially on the rangelands, enabled initiatives to
exploit alternative resources outside the national park.
Coincident with the floodplain rehabilitation there was also a change in park management. The former warden, who had been in place for over twelve years and had
tense relations especially with the Kotoko fishermen communities, had tried several times to remove a village only to be stopped by the traditional authorities. His
ethnic background and marriages made him much more aware of the situation in
Fulbé and Arab villages than elsewhere. The new warden, by contrast, had an academic background, was young, and was a Kotoko. Local people took this change
in manager as a sign of increased tolerance towards their presence, and they immediately asked the new warden for permission to exploit natural resources, such as
fish and firewood, within park boundaries.
Intelligence and information (Scholte et al. 1999b)
To supplement available information on village-park relations (Njiforti et al. 1991)
a systematic survey was undertaken in 1995 in each village, following the methods
of Participatory Rural Appraisal (Gueye & Schoonmaker Freudenberger 1991).
Complementary studies analysed fishery and pastoral exploitation inside the park,
as well as transhumant and nomadic groups originating outside the park villages.
The results showed that among a wide variety of problems several were directly
related to the existence of the park (Table 9.2). In addition to the lack of flooding,
previously mentioned and substantially alleviated by the pilot water release, concern was raised amongst the management plan co-ordination team about the high
degree of the dependence of local communities on park resources, especially for
fishing and grazing. Were the park really closed for exploitation, villages would not
be able to stay where they were. The new information strongly suggested that
217
9 – Management planning
Table 9.2 – Ranking of the ten main problems perceived by Waza NP village communities1
Problem
Times the problem has been ranked as
Lack of flooding
1st
2nd
6
1
1
1
3rd
4th
5th
6th
7th
8th
1
Lack of formal authorisation to
fish in the park
Animal theft
1
Lack of grinding mill
1
1
1
1
Crop damage
2
1
Lack of pastoral water in dry season
2
1
Intimidation by the Waza NP authorities
1
Well without water
1
1
1
2
1
3
Predation on domestic animals
2
Lack of grazing
2
1
1
3
1
Not all figures add up because of the omission of less important problems.
resource exploitation within the park, although formally prohibited, was being tolerated informally by local park personnel in exchange for an illegal ‘tax.’
When the management plan was being put together, Waza NP had about 20 guards.
Most of them were recruited in the 1970s and based in the park for years (Fig. 9.3).
The warden, who was the only staff member on the core team, felt that the guards
were to some extent unaware of the feelings villagers had about them. They did
understand, however, the ingenious methods local people used to circumvent the
rules, and by not addressing these they were missing an opportunity to create better relationships with local communities.
There are also in the park about 35 tourist guides, who originate from the park villages and are paid by their clients. They are registered at the park office and, apart
from their tourist work, are supposed to support actively the surveillance performed
by the game guards. In some ways, they have been the true guardians of Waza NP
over the years. They have excellent local knowledge and have shown their commitment to the park during confrontations with poachers. They are pursuing, in effect,
a participatory conservation approach to managing the park. The survey work
showed that their main problem was lack of support from their fellow villagers.
The guides, in other words, were the ones most directly confronted by tensions
between park and people.
Part III – Enhancing conservation – development integration
50
number of park personnel
bar: park guards (armed)
40
+ village
guards
(unarmed)
line: park guides (unarmed)
30
20
10
2002
2000
1998
1996
1994
1990
1988
1985
1982
1980
1978
1974
1970
1965
1962
0
1953
218
Figuur 9.3 – Development of personnel numbers of Waza NP
Preparing for consensus
The process of trying to reach a new consensus on how to manage Waza NP began
in February 1996, with a three-day workshop in which selected invitees met to discuss the 1995 survey work. The outcome was a set of preliminary proposals to improve both resource exploitation inside the park and the management of surrounding areas. They formed the basis for further discussions in subsequent months with
authorities and local populations.
The workshop agreed that existing park borders should be regarded as inviolable and
definitive. They were needed to guarantee conservation values and to reassure local
18
villages that they did not have to fear any further extension of the park. The idea of
authorising specific resource exploitations inside the park also found agreement, on
the basis that it would counter an increase in park exploitations. This would require
18 Park boundaries cannot be changed through a management plan, but widespread suspicion on this
subject reigned during the planning process. People’s attitudes towards Waza NP are rooted in the
past. The removal of villages, through frequently changing park boundaries, has left deep scars and is
still feared. In 1966, one village inside the park, Zeila, was burned down for political as well as park
management reasons. Its inhabitants, dispersed over neighbouring villages often on the park borders,
continued exploiting its former fishing grounds in the park. During the late 1940s, two other villages,
Andirni and Badadaye, shifted a few kilometres to escape being trapped in the park. All villages have
seen the park approaching or as someone put it ‘as soon as there are animals you make sure that you
leave.’ Fear of the park has been especially acute in the eastern villages, which found themselves right
on the park border in the 1950s. In the late 1980s, the park warden caused havoc when he used a road
grader to redraw the border of the park, putting many villages inside the park that had not been there
before. Although the management plan corrected this situation, local uncertainty can only be alleviated by giving park boundaries a more obvious material presence on the ground.
9 – Management planning
219
finding a way around the prohibition in the 1995 decree on resource exploitation
within national parks, and legal advice was sought for this purpose. The third no19
table outcome of the workshop involved buffer zones. Although a buffer zone is
itself a protected area and falls under the jurisdiction of the Minister of Environment
and Forestry, there is a risk that proposals to create buffer zones will be viewed by
local people as a reduction in their ability to control what they think of as their area.
The workshop, therefore, proposed that all existing and legal resource exploitation
be guaranteed for those living in buffer zones, and for their descendants.
Negotiations
In January 1997, after weeks of preparation to help select preliminary representatives from the local population, a meeting was held to set up a management committee. Additional meetings were organised in March 1997 in the provincial capital to decide on the main components of the management plan. After a day on
which ministry officials dealt with a number of procedural planning issues, the second day of these meetings constituted in effect the first session of the management committee. The workshop proposals on park resource exploitation proved
uncontroversial. The proposals on buffer zones, however, provoked heated discussion and clearly made ministry officials uncomfortable. They did not want to confirm that present users of resources in potential buffer zones would have rights of
access to those resources. It was eventually agreed that the management plan would
refer to a periphery zone, a concept not defined in law and, therefore, open to
future elaboration. The core team of the Waza-Logone Project, together with a senior staff member of the Ministry of Environment and Forestry, then wrote a draft
of a five-year management plan for Waza NP. After some additional discussions
at the ministry, the draft was finalised without fundamental changes and was
20
approved by the Minister of Environment and Forestry in September 1997.
Implementation (1998-2003) and the 2002 evaluation
Although the implementation of the management plan, for various reasons, did
not get off to a fast start, within a year of its approval a new warden was appointed for Waza NP. He warmly endorsed the plan and took particular steps to see that
an outside expert would evaluate it (Amougou 2002).
In line with the intervention strategy for improving internal management, the management plan called for the creation of a sub-unit to pull together and enhance the
19 According to the 1995 decree a buffer zone is ‘A protected belt around each national park, natural
reserve or game reserve designed to mark a transition between the above-mentioned areas and the
zones where cynegetic, agricultural and other activities are freely carried out. Nevertheless, certain
human activities may be regulated therein following a management plan duly approved by the minister in charge of wildlife. The instrument setting up a protected area shall lay down the limits of its
buffer zone’ (MINEF 1995: 121).
20 Arrêté No. 732/A/MINEF/DFAP/DAJ du 16 Septembre 1997, rendant exécutoire le Plan Directeur d’Aménagement du Parc National de Waza.
220
Part III – Enhancing conservation – development integration
protection, ecological, development integration, and administrative and financial
aspects of park management. But circumstances worked against this goal. The
recruitment of park guards was frozen following the imposition of International
21
Monetary Fund restrictions on the government. The number of staff functions
at the central ministry increased during the same period, however. Clearly, field
positions did not have high priority. Initiatives to put middle-level technicians at
the head of various park units were not followed up, despite financial support for
this promised by the Waza-Logone project. So, the park warden remains the only
higher-level person on the park staff. And the fact that he is surrounded by an ageing group of park guards seriously compromises the intervention capacity of park
management. Since 2001, the recruitment of 600 park guards has been approved
22
in principle, but there has been no implementation. In fact, the number of park
23
guards has declined by 50 per cent (Fig. 9.3).
The management plan also tries to improve the technical and ecological basis for
park management decision making. Although elementary ecological and wildlife
research and monitoring, set up in the early 1990s, has continued during the
implementation phase (Table 9.3), this strategy has also not made much progress.
While the various institutions that can and could play a role in research have been
identified, their several responsibilities and budget allocations have not been settled.
Long term monitoring of park conditions has been neglected and limited follow
up means that research and monitoring have had almost no impact on the management of the park (Table 9.3). Zoning, a subject the plan left to be developed
during the implementation phase, has had no priority, probably because of a lack
of ideas and directives (Table 9.3). A scientific committee, composed of nine Came24
roon-based experts, advises the plan’s management committee on ecological issues. Its recommendations on the controlled exploitation of park resources attracted some attention. But the scientific committee has met only five times in the
management plan implementation period, and this is simply not sufficient for
either effective supervision or advice.
25
The management committee has a consultative role on subjects dealing with activities inside the park and a management role in the park’s periphery. It approves
the annual plan for the park and brings together representatives of local populations and park authorities, who have, respectively, 15 and 6 voting members. Ad21 In 1992, only 5 per cent of MINEF’s total budget was allocated to non-staff expenditures (Brunner
& Ekoko 2000).
22 Personal communication from the Director of Wildlife, March 2001 and January 2002.
23 The lack of park guards has led to speeding up the recruitment and training of village guards. But
they were never meant to replace park guards. Village guards are not allowed to carry arms and, therefore, can play only a secondary role in anti-poaching campaigns.
24 Including the author.
25 Officially approved by a ministerial decision in April 1998. Decision No. 0728/D/MINEF/DFAP/
DAJ portant organisation et fonctionnement du Comité de concertation/gestion du Parc National de
Waza et de sa zone péripherique.
221
9 – Management planning
Table 9.3 – Evaluation of Waza Management Plan’s implementation (Amougou 2002)
Actions prescribed in
Management Plan
Realisations %
(own impression)1
Comments (adapted from Amougou 2002)
1 Protection and anti-poaching control
1.1 Recruitment of park guards
0% (-50%)
During the implementation period a 50% drop,
see Fig. 9.3 and text
1.2 Recruitment of village guards
100%
Initiative of the park warden, yet village guards
cannot replace park guards
1.3 Training of guards
100% (30%)
1.4 NP border demarcation
10%
Limited, as no new park guards had to be trained
A proposal has been drafted
2 (Ecological) management
2.1 Research2
2.2 Monitoring2
60%
20% (40%)
Generally ecological research
Ecological monitoring mainly, responsibilities of
partners involved not clear
2.3 Follow-up Research & Monitoring 10 %
Only one study has had an impact on management
2.4 Zoning
Not considered a priority
0%
2.5 Fire policy developed
0 % (10%)
Some field knowledge being applied
2.6 Infrastructures developed
40% (30%)
Especially maintenance of water holes
3 Integration into local development
3.1 Tourism
Tourist promotion
Tourist management
70%
Brochures, local and international press cover
60% (30%)
Some ecotourism activities started but failed
3.2 Relations with local communities
Committee
90%
Functioning, several conventions signed, as yet
external investments only
Access to park resources
50% (25%)
Only partly executed as strongly dependant on
presence of park guards (1.1)
Local development
70%
See Table 9.4
The Baram case
50%
A part of the village has voluntarily moved out of
(village inside Waza)
Waza NP (Chapter 8)
3.3 Awareness creation amongst:
Neighboring communities
80% (50%)
See Table 9.4
Women
100% (50%)
In all structures there is a representation of women
Youth
Private sector
Decision makers
3.4 Other forms of Integration
50%
Waza NP visited by pupils
0% (10%)
0%
0% (10%)
Considerable publicity in scientific fora
1
Note that a score of 50% should be considered as reasonable
2
Scores and comments ‘averaged’ over six categories, including vegetation, fish, birds and large mammals.
222
Part III – Enhancing conservation – development integration
ministrative and traditional authorities are present as consultative members.
Clusters of 2-3 villages, as well as the transhumant pastoralists, also select one man
and one woman as committee representatives. Agropastoralists, whose families are
outside the area, select one representative, who was subsequently designated as
the first president of the committee, with the park warden as vice-president. Baram,
the only village inside the park boundary, was excluded from participation in the
committee, on the theory that could ultimately persuade the people there to settle
outside the boundaries, in accordance with the official conservation policy.
The creation of this committee has yielded some of the most remarkable impacts
that are attributable to the management plan. The committee meets at least once
a year and has shown itself on several occasions to be concerned not only with
development issues important to the communities it represents but also with the
conservation of Waza NP. In 1999, for example, the committee decided to send a
delegation to the ministry in Yaoundé, the national capital, to plea for additional
park guards. On the other hand, park communities are so diverse that their commitment to the park planning and management has been difficult to secure through
their representation by only a limited number of committee members, whose
feedback to local communities is necessarily limited (Table 9.4). The most visible
impacts of the Waza-Logone Project have clearly come from the pilot reflooding,
beginning in 1994. Although these efforts were not extended for financial and technical reasons, the positive correlation between reflooding and the presence and
intensity of development activities is clear (Table 9.4). In other places, even small
26
changes in material conditions have yielded results.
9.4
Management planning in other protected areas in Cameroon
Why management planning?
In all the major state protected areas in Cameroon where an ICDP project is being
carried out, management planning has started (Table 9.1). In contrast, there are
no management planning initiatives in the mountain forest areas in western
Cameroon (at Mt. Cameroon, Mt. Kupé, and Mt. Kilum-Ijum), where protection
is a matter not for the state but for the community, with ICDP support. At first
glance, this suggests that the planning under way in the state protected areas is
just part of a management fashion, a nominal response to the legal requirement
that each protected area have a management plan. Or do these management planning initiatives, on closer scrutiny, turn out to address real and pressing management needs, as is arguably the case in Waza NP ? The principal issues addressed
26 In the village of Baram, where no development interventions were initiated by design (Table 9.4),
one of the families persuaded the others to move out of the village, and out of the park, in 1998 (see
Chapter 8).
Table 9.4 – Impact of Waza-Logone ICDP activities on park communities (1995-2002)
Impact of
reflooding1
Ndiguina
Goulou
Mbili
Baram
Halé4
Mahé
Lougouma
––
––
–
+
+
+
+
Zwang
++
Tchede
+
none
Diégueré
Andirni
Amkodje
Badadaye
Niwadji
Tagawa
Waza town
–
+/–
+/–
+/–
––
––
––
Agro
pastoralists5
Transhumant
pastoralists5
++
1
Development interventions
+
failed
continuing
Apiculture
X
X
X3
Water pump
Improved rice
culture
Tourist camp
Water pump
Improved rice
culture
Water pump
well
X
X
X
Tourist camp
Pastoral
well + pump
X
Reception hall +
Tourist camp
(x)
Convention on
non-grazing Insecurity issue
(Scholte et al.
1996b)
2
Representation
in Waza
committee1
Park resources
presently exploited
(illegally)
Proposed
conditional
access to park
resources2
Arabic Gum
Arabic Gum, fish
All
Fish
Fish
Fish
S, A
S
S, F
None
S, F
S, F
S, F
+/–
Fish
S, F
+/-
Fish
S,F
+/–
+
+/–
+/+/–
+/–
+
Grazing
Grazing
Grazing
Grazing
Grazing
Grazing
Various
S, A
S, A
S
S
S
S
S, A
+/–
Grazing
None
––
none
none
–
–
+/–
– –3
––→+
–
–
9 – Management planning
Village/
Community
(Fig. 9.2)
3
223
– –: none, –: slight, +/–: moderate, +: reasonably high, ++: very high; S: straw, A: Arabic Gum, F: Fish; in bold already conditionally active; As prescribed in the manage4
ment plan, no development activities nor any representation in the committee were pursued in order to dissuade this village to leave the national park; Village created in
5
1999 with descendants from Baram, subsequently integrated in development activities as well in the committee; Agro-pastoralists have a village base, transhumant pastoralists not.
224
Part III – Enhancing conservation – development integration
by management planning in Cameroon are zoning and resource exploitation within
protected areas (Table 9.1). These are also subjects that generally require a consensual interpretation of legislation and policy. Zoning has received little attention in
the defined categories of protected areas in Cameroon, because even a buffer zone
is considered a protected area offering few, if any, rights for local communities.
How does management planning in Cameroon take place?
International conservation NGOs, such as the World Wildlife Fund (WWF) and the
World Conservation Union (IUCN) play an important role in the organisation and
facilitation of management planning in Cameroon. They help provide and find
project staffs, who are often supported by short-term consultants and who invariably constitute the planning co-ordination teams. The warden of the respective protected area is generally closely associated with these teams but seldom plays a more
active co-ordination role, as he did at Waza. In Cameroon, the co-ordination teams
are usually constituted informally, as with Waza and Benoué NPs, but in some
cases (Korup NP, for example) ministerial approval of the team is pursued. Most
planning activities started in the mid-1990s, with three management plans that
have been formally approved, Waza, Korup and Benoué NPs. In the Dja game
reserve, management planning has dragged on for years, in part due to the multitude of actors involved and also because of the complexity of the area and issues
addressed. The Dja case shows that planning can be overly ambitious in scope in
relation to the quantity and quality of the human and other resources available.
Participatory studies with local communities (Gueye & Schoonmaker Freudenberger 1991) are generally being used as preparation for management planning (Wild
& Mutebi 1997; Vabie 2000). Unfortunately, these studies are seldom carried out
or supervised by the management planning co-ordination team. The opportunity to
establish a dialogue with local communities is, therefore, missed. Furthermore,
these studies are not undertaken, overall, on a systematic basis and they neglect
evasive communities, such as the nomadic pastoralists in the Benoué and Faro
areas.
Review of the management planning outcomes
Apart from Waza NP, other protected areas in Cameroon have insufficient experience with management planning to be the basis of detailed analysis. Some useful
impressions can be gleaned, however, from Benoué NP, where the second management plan in Cameroon was approved in 2002 and the Dja World Heritage and
27
Biosphere Reserve, which plan was submitted for the minister’s approval in 2004
(Table 9.1). The Benoué plan generally resembles that for Waza, especially in its layout and internal organisation, no doubt because several ministry officials had ear27 This section is inspired by a review of the Benoué and Dja management plans by participants (park
wardens of Central Africa) of the bushmeat training courses by the Garoua Wildlife College in July
2003 and September 2004 respectively.
9 – Management planning
225
lier participated in the Waza planning. More emphasis is put on zoning, however,
which is the main theme in the Benoué plan, and a detailed work plan and budget
are included. Proposals for the internal organisation of the park include a doubling
of staff, which is even more ambitious than for Waza. But in the monitoring and
research section no collaboration mechanism is proposed, even though at Waza
this has proven to be a main limitation on implementation. There is to be a committee representing the main stakeholders at Benoué, and a scientific committee is
announced. So, overall the main features of management planning at Benoué suggest that there has been a fairly direct transfer of experience at Waza. The Dja management plan is part of a generation of plans in which the central ministry, after
the failures of previous versions, has taken the lead during its formulation (Table
9.1). Although previous versions were inspired by the Waza experience, the role of
local communities in the consultation fora seems to have been limited to a representative role only. Much of the sensitive zoning issues have, as at Waza, been postponed for further investigations during the implementation of the first plan.
9.5
Management planning in other African countries
Why management planning?
On the basis of an unsystematic sample of seven management planning exercises from a range of other African countries some admittedly limited comparisons
with the Cameroon experience can be made. Based on an analysis only of the management plan text, none of these other seven management plans is an explicit
response to detailed legal requirements, although several interpret the law in specific local contexts on issues such as the exploitation of resources from within the
protected area and zoning (Table 9.5). Perhaps the most interesting comparison
is with the Réserve Spéciale de Forêt Dense Dzanga Sangha in the Central African
28
Republic, which was created by a specific law, signed by the head of state in 1990,
29
and which describes local user rights in a site-specific 1992 ministerial arrêté,
in much the same way as in management plans in Cameroon.
How did the management planning take place?
All seven management plans in one way or another have involved local communities, a distinct difference with earlier practices. This holds for the long established
protected areas of Kruger, Hwange, and Serengeti where the involvement of local
communities is limited to a consultative role, as well as to some newly created protected areas in which a stronger involvement is sometimes pursued (Table 9.5).
28 Loi No. 90.017 du 29 Décembre 1990. Portant création d’une Réserve Spéciale de Forêt Dense
Dzanga-Sangha.
29 Arrêté No. 007 du 25 Mars 1992. Portant règlement Intérieur de la Réserve Spéciale de Forêt Dense
Dzanga-Sangha.
Protected
area
Formulation
Approval by
(year, previous
Duration plan)
Involvement local
Principal themes
Communities /
Restitution of mgt plan
New policy
interpretations
Proposed
Implementation
consultation experiences
structures
Yes / yes
Zoning, Exploitation
of NTFP-resources
from the park
Few
Yes
Negative publicity
around eviction of
pygmy communities
Bwindi NP
(Uganda)
WH
Project team,
consultants
warden
2,5 year
Board
(1995, none)
Serengeti
(Tanzania)
IUCN-project
team
5 year
Minister (4
2 seminars with
earlier versions representatives of
never approved) communities / no
Livestock / agriculture in the area
Likely
No
NA, as not
approved
Hwange NP
(Zimbabwe)
Consultants +
park team
1 year
Minister
(1999, 1992)
Yes, several
seminars / yes
Relations between
tourist agencies and
park authorities
No
No
Deteriorating political situation has
overruled any impact
Kruger NP
(S.Africa)
Park scientists
3 year
Board
(1998-1986)
55 workshop days,
internal mainly,
a few with others /
more or less.
Elephant culling
Fire management
Tourist reception
No, park
exploitation
not treated
No
Elephant culling
under increased
attention, justifying
undertaken planning
Djoudj NP
(Senegal) WH
IUCN-project
team
5
Minister
(1995, none)
Yes / yes
Collaboration with
local communities
?
Yes
Widely claimed
as positive
Boucle du
Baoulé NP
(Mali) BR
Consultants
?
Minister
(1998, none)
Zoning and newly
drawn borders
No, follow-up
of new law on
park borders
No
?
?
Minister
(1995, none)
Framework for
more detailed plans
and programmes
No
No
Claimed as positive
Banc d’Arguin WWF with
(Mauritania)
park team
WH
and scientists
BR: Biosphere Reserve; WH: World Heritage Site
?
Part III – Enhancing conservation – development integration
Coordination
226
Table 9.5 – Management Planning in selected African protected areas
9 – Management planning
227
Review of the outcome
As in Cameroon, all of the management plans reviewed cover a period of five years,
except for Banc d’Arguin NP where the period is ten years. Since the making of a
plan can take a year or more (Table 9.5) and took more than three years in Cameroon (Table 9.1), a longer time span for the plan seems realistic. This is underscored by the experience at Kruger NP, where the management plan was formed
as an annex to a plan prepared twelve years earlier but still took three years to complete. As in Cameroon, most management plans emphasise zoning, which can be
a sensitive issue; so sensitive at Waza that zoning was omitted from the final plan
document.
None of the management plans reviewed anticipates independent review and only
two, Bwindi and the Banc d’Arguin, provide for internal review. This will make it
unusually difficult to assess and evaluate plan implementation and is probably a
hangover from the time when plans were regarded as essentially technical documents. Anecdotal information suggests that actual experience with implementation of the management plans ranges along the entire spectrum from good to bad.
9.6
Discussion and conclusions
If meaningful and effective, management planning is assumed for purposes of
analysis to require success at every phase of the process, the lessons of the experiences reviewed, here, are usefully considered phase by phase. It might be useful
to consider management planning as a process in three phases: (1) reaching a consensus on main management issues; (2) ‘writing’ the management plan, based on
this consensus and an analysis on less controversial management issues and (3)
ensuring the continued commitment of stakeholders during the management
plan’s implementation. The success of management planning depends on all
three stages.
The process of management plan formulation
At Waza, which was already the best known protected area in Cameroon, the additional studies and surveys deemed necessary for the management plan, and the
negotiation surrounding these, still took two years. This was a shorter time than
was needed elsewhere in Cameroon but longer than in most other African countries (Table 9.5). While additional information and intelligence were being gathered, at least one staff member was occupied full-time in co-ordinating project
teams and consultants. Regular meetings brought together up to ten governmental staff members and more than fifty local representatives. The costs of making
the Waza plan, including studies, meetings, and tours but not salaries and consultancy fees were nearly US$ 80,000. This is at the upper end of cost estimates
for similar management plans in Australia and the US, which generally did not
228
Part III – Enhancing conservation – development integration
include any base-line studies (ANZECC 2000) but lower than the US$ 250,000
estimated cost of management plans in Africa as a whole (Clarke 1999).
The planning at Waza occurred in quite exceptional circumstances, created by the
successful floodplain rehabilitation and the arrival of a new park warden. No
changes in associated governmental staff and key contact persons occurred. This
was unusual but it did facilitate the learning process and made it productive for
plan staff to spend time on study tours for central government officials, who were
then encouraged to begin other planning exercises elsewhere in Cameroon. Under
less favourable circumstances, a plan preparation period of one to two years would
be much too optimistic (Table 9.1). And when planning is under pressure from
time, budget, and donor-imposed constraints, third parties may come to dominate
the process to the extent that local people and park authorities feel no attachment
to either the process or its outcomes.
The first five-year management plan for Waza has, to some extent, been dominated by the Waza-Logone project staff. One important reason for this is imbalance in power relations between authorities and local populations. It is not clear
how participatory management should be approached in conditions where tension
is prevalent and people are reluctant to mention park exploitation. The stronger parties, namely the park authorities and those who have good relations with them,
may fear a loss of power. Weaker parties may be reluctant to begin an open process
that will change the status quo when the outcome is uncertain. This dilemma played
itself out during the buffer zone discussions where, in principle, the Waza-Logone
project’s proposal might have facilitated land use control and could have been of
mutual benefit. The central authorities, however, would not provide guarantees
against misuse, thus making it too risky to rely on project promises. Local people
have been able to play a somewhat more active role in park decisions through the
management committee, but a co-ordinating third-party agency, like the project,
with a combined role of financing and facilitation (Groot 2002) is essential in making this work.
Generally speaking, a park warden is too much of an interested party to be an effective facilitator of the management planning process (Groot 2002). However, park
wardens could be more involved than is now the case in Cameroon (Table 9.1),
30
especially if they were taught the requisite skills. More could also be done, as suggested by Clarke (1999) to review the successes and failures of park management
planning and build these experiences into the wildlife-training curriculum.
30 I attribute their present limited role to a lack of necessary skills and not, as often assumed, to a
lack of interest (Chapter 11). The regional wildlife colleges have taught elements of park management
planning since the 1980s and recently updated their courses, although facilitation skills may not have
received sufficient attention as yet. The Garoua wildlife college recently developed a park management planning refresher course for former students who occupy park warden posts (Scholte 2003b).
9 – Management planning
229
Overall, the Waza management planning process led to a break-through in protected area management in Cameroon. For the first time, park authorities committed
themselves to explaining a plan in a committee setting dominated by local populations. Despite legal constraints, procedures for access to park resources were formulated, bringing transparency to an old practice of individual negotiations with
park authorities. This has had major implications for other protected areas in Cameroon as well.
Management plan approval authority
Although the management plan approval authority in Cameroon is the minister
of Environment and Forestry, this has been subject of discussion. In protected areas
where the planning process was controversial, such as Dja and Benoué, interventions made by ministries for Livestock, Agriculture, and Mines, for example, contradicted positions taken by the Ministry of Environment and Forestry, prompting
the Dja management plan co-ordinator to suggest that approvals should come from
the Office of the Prime Minister. This seems certain, however, to further complicate what is already a difficult political process. If the view is that management
plans are primarily technical documents, approval can more readily occur at lower
levels in the ministerial hierarchy, and this may be an option for Cameroon in the
future, when approvals deal with second versions of management plans.
Form, presentation and distribution of management plans
Management plans still consist mostly of technical and ecological descriptions (Claridge 1999). Opinion is divided about why this is and whether it is a good thing.
Description is favoured by external consultants, who would rather not deal with
complex policy issues. It also gives plans a scientific aura and, thus, some claim
to neutrality as between conflicting interests. But the value of lengthy descriptions
in plans has been questioned on the grounds that that protected area managers
are more interested in conservation values and the management choices that
might help achieve them than in compendia of descriptive information (Claridge
1999). One simple way to accommodate both would be to split planning documents into an annex and a main volume, with the former containing the descriptions and the latter addressing the management and policy issues.
Only the Hwange and Kruger management plans include either a detailed account
of how the planning process actually unfolded or a list of who participated in various workshops, stakeholder meetings, and other relevant deliberations. Given the
current tendency to emphasise that the quality of management planning is better
gauged by looking at the process rather than at the final document, in order to
‘create ownership’, this is a strange omission. Certainly, a wide variety of methods
exist to contact and reach out to various publics, all the way from the global constituency that cares about Kruger NP, for example, to local communities (Pretty et
al. 1995). Participants in management planning exercises ought to take better ad-
230
Part III – Enhancing conservation – development integration
vantage of these methods, whether they involve the internet or more conventional printed materials.
Management plan implementation
Before starting the Waza management plan implementation, the prediction was
that ‘Essential for the success of the implementation of the Waza management
plan will be that at least some of the results will be put quickly into practice, justifying a continued commitment of local people’ (Scholte 2000b). In fact, there
was genuine commitment to the plan by local communities and protected area
authorities, and this contrasted with the disappointing roles played by central government and by the international donor community (Table 9.3). But for the governmental partners and other officials involved, management planning is a job, a
part of their normal work, for which they are paid. For local people the situation
is very different. To participate, they have to sacrifice time and energy that could
be spent doing other more immediately productive things, and for them the expectation that planning will lead to improvement in their daily lives is acute. One way
to hold people accountable for their adherence to a plan would be to make the plan
contractual (Kirkpatrick 2001; Pollack & Horty 1999). But even this could not
guarantee success.
The tendency to predict and to try to guarantee the success or failure of management planning by focusing on the form and content of the plans is understandable (Clarke 1999). But there are some larger contextual variables at work, too.
Management planning is one of the weakest links in the conservation chain and
the conservation community puts too little emphasis on it. None of the seven management planning formulation processes in Cameroon, for example, has costed
more than five per cent of the total budget of conservation projects. This limits the
financial impact of their failure, but it also causes major frustration for the local
people and private companies who invest their time and energy in the expectation
that the conservation community will get behind the implementation of plans in
a serious and immediate way.
Clearly, management planning has proven in many cases to be too ambitious, given
the availability of human resources, the complexity of problems, and the multiple
and to some extent conflicting expectations of local communities, authorities and
donors. It might make more sense to focus on removing the most salient obstacles to the day-to-day management of protected areas and only then begin a more
comprehensive management planning process (Thomas & Middleton 2003). Certainly, if no obstacles for management exist, or no consensus can be envisaged,
management planning should be limited to a simple (internally) formulated working document.
9 – Management planning
231
The mixed results of management planning shows that the implementation trajectory for changed wildlife law and policy is often longer than many anticipated.
Offshoots of the new conservation paradigm do have an impact on the ground, but
they are often too limited to counter the increasing conflicting interests of conservation and development. This should not deter one from the need to continue with
management planning, as this may provide much needed feed-back to national
policy and legislation that will be discussed below.
Implications for national policy and legislation
‘Empowering poor people to conduct their own appraisal and analysis and to
present their realities is one thing. Whether their voices are heard, understood
and acted on is another. There are two weak links: from voice to policy change
(policy-in-principle); and from policy change to practice (policy-in-practice)’
(Chambers 1998).
Although most of this chapter deals with policy-in-practice, the question of how
local experience with management planning can lead to changes in policy and legislation also deserves some comment.
Policy
It is important to remember, despite all the previous criticism of ministry personnel and their poor use of human resources, that a radical change in ideas and perceptions has taken place at the Ministry of Environment and Forestry in Cameroon
since the appearance of the 1995 decree (Egbe 2001) Although there was speculation that the 1995 decree would bring about a turn towards a new community
wildlife policy in Cameroon, even going beyond the obvious implications of the
1994 law, there was at the time very little practical understanding among ministry
officials of what this might mean. The idea that policy should involve and encourage the participation of local communities has now become mainstream thinking,
in part because of the exposure ministry officials had to the day-to-day problems
of park communities as evidenced during the Waza management planning process.
There has even been discussion of creating a distinct unit for the management of
protected areas, a sort of parastatal wildlife service that could give local people and
officials more independence in day-to-day management decision making than they
have, now.
Legislation
Experience with the Waza-Lagone Project showed that new national legislation is
needed in Cameroon to diversify protected area categories and, thus, to reflect different resource exploitation regimes, especially where there is substantial human
presence, as with IUCN categories IV and V (IUCN 1994). This would be helpful
in national parks, which now lump together areas that serve various objectives,
232
Part III – Enhancing conservation – development integration
some of them incompatible with those laid down in the 1995 decree. This diversification would not yield quick results, but without this legal change only limited
further progress with management planning can be made. The legal definition of
buffer zones, now seen as extensions of protected areas, also needs to be amend31
ed (Neumann 1997; Scholte 2003a)
The Cameroon experience also has lessons for other African countries. In Gabon,
for example, where a network of 13 national parks was created in 2002 (Quammen
2003), problems identified for Cameroon’s protected areas, such as above-mentioned lack of differentiation in status of national parks, are also present. A wildlife
decree for Gabon is expected to give management plans a prominent role in decentralising national park decision making and include terms of reference for plans
32
that call for consultation with local communities.
The final point to be made about legislation concerns the incompatibility between
the protected area management regime now legally in place in Cameroon and the
management scheme envisioned for international biosphere reserves. As previously noted, management planning for Waza, Benoue and Dja biosphere reserves
in Cameroon began more than ten years after their designation. Yet, even though
33
biosphere reserves are meant to serve multiple functions and are, therefore,
unsuitable for management under a single authority striving to meet the require34
ments of a single type of protected area, the 1994 law and the 1995 decree in
Cameroon proceeded as if this were not the case. The ambiguous legal status of
Cameroon’s national parks vis-à-vis biosphere reserves needs to be resolved. The
decennial evaluation, proposed by UNESCO-MAB for all biosphere reserves, but
never conducted in Cameroon, could play a role in stimulating this reflection. The
experiences with the World Heritage Sites’ periodic review (World Heritage
Centre 2002), that encompass the Dja Reserve in Cameroon, shows that feed-back
from periodic reporting into national policy and legislation cannot be taken for
granted and could be more actively pursued by UNESCO.
31 These changes in national policy need to be accompanied by change at the grass roots level, too, if
they are to have tangible impacts. Grass roots initiatives are presently limited, however, by the unwillingness of the Ministry of Environment and Forestry to assign more field personnel, one of the principal indicators of the central government’s lack of commitment to management plan implementation.
32 Personal Communication, René Hilaire Adiaheno, secrétaire permanent Conseil National des
Parcs Nationaux, Gabon. December 2003.
33 ‘A biosphere reserve should perform three complementary functions: a biodiversity conservation
function (with a focus on conserving a representative sample of major ecosystems); a development
function (with a focus on humans in the biosphere, emphasising an integrative role for local communities) and a logistical function (combining conservation research, education, training and monitoring). Biosphere Reserves are a special kind of conservation area-traditionally a nested series of
zones with differing management intensities (core area, buffer zone and transition area)’ (Bridgewater et al. 1996).
34 ‘It should be noted that a Biosphere Reserve which is equivalent to a single IUCN [protected area]
category does not conform fully to the Biosphere Reserve concept. Such cases should promote a reexamination of the Biosphere Reserve to test its effectiveness’ (Bridgewater et al. 1996).
10
Curriculum development at the African
Regional Wildlife Colleges, with special
reference to the Ecole de Faune
(Cameroon)
Paul Scholte
Environmental Conservation 30: 249-258
Abstract
Regional colleges in Tanzania, Cameroon and, recently, South Africa have
trained some 4000 wildlife managers. Training need assessments called for major
curriculum reforms, which were developed and implemented in the late 1990s.
This is an analysis of the factors that influenced this curriculum reform in the colleges’ endeavour to respond to new developments in African wildlife management. Since 1979, the curriculum of Garoua Wildlife College, Cameroon, has
changed only gradually, whereas work placement subjects, selected by students
and their employers, have quickly responded to developments in wildlife management, with an increase in the number of people-oriented subjects amongst
other things. In the new curriculum, Garoua’s mid-career students appreciated
biology and inventory disciplines for their relevance, as well as courses in other
disciplines tailored to conservation practice. The curriculum reforms implemented at Garoua depended on the presence of interested lecturers with an appropriate background, often obtained by additional training. The curricula of the
regional wildlife colleges at Garoua (Cameroon), Mweka (Tanzania) and SAWC
(South Africa) showed important differences, as a result of regional differences
and differing visions of the wildlife management profession. All three colleges
have struggled to establish a feed-back system for continuous curriculum review.
Increasing the exchanges between the colleges could further develop the curricula. While pursuing necessary changes in curriculum and institutions, care should
be taken not to reduce the colleges’ sustainability.
235
10.1
Introduction
European managers ran African protected areas long after independence and the
international conservation community was generally reluctant to see the custodianship of wildlife transferred into African hands (Robins 1970). The subsequent
lack of qualified national wildlife managers in independent Africa drove the establishment of the College of African Wildlife Management in Mweka (Tanzania) in
1963 and the Ecole pour la Formation des Spécialistes de la Faune in Garoua (Cameroon) in 1970. Both colleges had a regional orientation because of the limited size
of national markets. In total, the colleges of Mweka and Garoua have trained 2500
and 800 students, respectively, from some 20 African anglophone and francophone countries. They focus on mid-career, medium-level wildlife professionals, and
many of their graduates have become protected area or wildlife service managers
(Gamassa 1995; Ngog Nje 1995a; Njoya 2001). The Southern African Wildlife College (SAWC) was founded in 1997, focussing on Southern Africa, including lusophone Mozambique and Angola (Venter 1995).
National training centres in Botswana, the Central African Republic, Kenya, Nigeria
and Zimbabwe (Child 1976) aimed at the low and, occasionally, medium professional levels (i.e. park guards and rangers), but some collapsed due to their dependence on foreign assistance (Ngog Nje 1995b). Although universities also provided
courses in ecology and wildlife biology, Mweka, Garoua and SAWC have remained
virtually without competition for mid-career professional wildlife management
training (Snelson & Lanjouw 1997).
Since the creation of the regional wildlife colleges, major changes have taken place
in wildlife management that, apart from ecological and technical management,
now also includes communication and community development (Cannon et al.
1996; Saberwal & Kothari 1998; Hulme & Murphree 1999). This new orientation
has resulted in increased awareness of the role of women in environmental management (de Bruijn et al. 1997). The private sector has taken over governmental
responsibilities especially in Southern Africa (Hulme & Murphree 1999). Armed
conflicts in Central and West Africa, the student catchment for Garoua, have resulted in loss of state power, without a private sector or other institutions to take over.
New models for conservation and relevant training need to be implemented during such critical periods (Hart & Hart 1997; Shambaugh et al. 2001). As in other
professions (Van den Bor 1989), the information technology revolution has helped
generate powerful tools such as geographical information technology (GIS) and
remote sensing, which few wildlife managers have mastered.
In 1993, Mweka and Garoua were criticised, among others, for being slow in adapting their curricula to meet evolving needs, providing insufficient field practice,
having curricula insufficiently responsive to identified needs, failing to evaluate
236
Part III – Enhancing conservation – development integration
training effectiveness and not including a greater representation of ecosystems in
training programs (Snelson & Lanjouw 1997). In response to such criticism, clients
and experts became involved in curricula reforms that hitherto had been conducted internally. None of the colleges had a feedback system to evaluate and review
the curriculum. Participatory curriculum development (Rogers & Taylor 1998) can
only be effective with appropriate facilities and qualified and motivated staff to
implement reform (Tuntivanich 1989); this required institutional changes when
external support for the colleges stagnated.
In this chapter I review developments in African wildlife management and the
resulting training requirements and organisation of the colleges. I assess how Garoua has adapted its curriculum over the last 20 years and compare its 1996 reform
with those of the other wildlife colleges. An analysis of the factors that have influenced curriculum reform and review will show the importance of the organisational and institutional settings of the colleges.
10.2
Methods
I began this study during a two-year assignment at Garoua, holding informal discussions with students and staff, which were presented at feedback sessions. Information on SAWC and Mweka was obtained from internal documents and discussed during the SADAC Wildlife Management Training Workshop in Harare,
Zimbabwe and at SAWC in November 1999 and Mweka in May 2002.
Training needs and developments in African wildlife management
I used the two wildlife training needs assessments conducted in the mid-1990s that
covered central, eastern and southern Africa (Pitkin 1995; ULG 1998). Because of
the lack of earlier systematic training needs assessments, I analysed 492 work placement subjects. Since 1979, Garoua students have spent six weeks in the field between their first and second year, often at their previous position of employment.
Generally, the students’ employers or financing agencies and, to a lesser extent,
the students themselves selected these work placements or research projects (Fig.
10.1). The role of the college at Garoua has usually been limited to academic approval. Work placement subjects were categorised into four disciplines, namely
biology and ecology, protected area management, tourism and other forms of
exploitation and people-oriented subjects, and the Pearson Correlation (SPSS 1999)
coefficients were calculated. It was not possible to perform a comparable analysis
at Mweka and the recently-created SAWC, because of the limited role of work
placements or research projects at these colleges (see later).
237
10 – Curriculum development
50
Percentage of students
40
30
20
10
0
EmployerEmployer +
student
Student Student +
college
College
All
Figure 10.1 – Who chose students’ work placement subjects in 1998 (as disclosed by
students, n=48)
Background to students and colleges
Garoua students indicated their prior work and training experience as part of the
evaluation. No systematic information was available from Mweka or SAWC. Tests
of normality were carried out with Kolmogorov-Smirnov tests (with Lillefors significance correlation).
Garoua’s curriculum development and students’ perception
To study Garoua’s curriculum development, its >40 courses were categorised into
the ten disciplines, including work placements, commonly used at Garoua.
An anonymous evaluation of students’ perceptions was introduced to provide
feedback on Garoua’s 1997 curriculum. The 31 diploma and 18 certificate students
judged the relevance of the courses’ contents, their presentation and the balance
between theory and practice on a scale of 1-5. Scores were averaged and ranked for
each course. I also allocated the scores to the disciplines and subsequently ranked
them. I was unable to obtain data on the perceptions of Mweka or SAWC students
for comparison (see later).
Curriculum comparison between the regional colleges
Information on the curricula of Mweka and SAWC was obtained from the 20002002 course prospectuses and syllabi and categorised according to the disciplines
applied at Garoua. The number of hours each subject occupied in the curriculum,
the main criterion in weighing student results at the three colleges, was used as
comparison.
238
Part III – Enhancing conservation – development integration
10.3
Results
African wildlife managers’ job requirements and training needs
Graduation from the wildlife colleges has become an important criterion for promotion to medium level staff functions. Whereas in the past certificate students
accounted for most enrolments, their numbers at Garoua and Mweka, and recently at SAWC, have dropped (Table 10.1), suggesting an increase in job requirements.
In Cameroon, Niger, Tanzania, Uganda and elsewhere, some of the more prestigious national parks have recruited M.Sc. graduates as park wardens, thus bypassing the wildlife colleges. Only Mweka delivers advanced diploma and post-graduate courses, but struggles with their accreditation.
Relatively few wildlife college graduates continued advanced training. Out of a sample of 221 who received training at Mweka from 1989 to 1999, 29 continued their
training after graduation. However, more than 50% of the 221 expressed a desire
to continue to study for a B.Sc. or M.Sc. in order to improve career advancement
options (Lipya 2000).
Table 10.1 – Student numbers at the African regional wildlife colleges.
Year
Garoua
1998
Certificate
Diploma
Male Female
Male Female
14
1
Mweka
4
27
Certificate
Male Female
492
25
2
0
0
2
522
22
2
12
1
552
0
22
2000
11
4
21
5
432
0
2
0
Male Female
432
0
0
Diploma
4
0
0
Certificate
Male Female
0
2001
Diploma
Male Female
1999
1
SAWC
54
27
15
1
18
1
6
nd
nd
nd
nd
2002
8
4
26
6
37
4
34
2
18
8
17
5
%
73
27
83
17
90
10
88
12
86
14
87
13
1
2
Garoua had a biannual intake: Male and female numbers not specified.
nd not determined.
Developments in wildlife management reflected by work
placement subjects
Amongst the work placements, there has been a steady increase in people-oriented conservation subjects between 1980 to 2000, their frequency being negatively
correlated (p = 0.006) with the number of biology and ecology subjects (Fig. 10.2).
The number of protected area management subjects received a largely constant
level of attention (Fig. 10.2).
239
10 – Curriculum development
Biology & ecology [1,2]
People-oriented [8]
80
Protected area management [3,4,6,9]
Percentage of work placements
Tourism & exploitation [5,7]
60
40
20
0
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
Year
Figure 10.2 – Subject areas of Garoua work placements, 1980-2000. Numbers in parentheses correspond to the disciplines on the x-axes of Figures 3 and 4.
Garoua’s work placement subjects have been responsive to topical developments
in wildlife management. An example is the rinderpest outbreak in 1983 that received
considerable attention amongst the 1984 subjects, contributing to the peak of interest in biology and ecology (Fig. 10.2). People-oriented subjects peaked in 1994 (after
the United Nations Conference on Environment and Development) and subsequently decreased. On average 47%, and never less than 30%, of the work placement
subjects dealt with tourism and exploitation, and people-oriented conservation.
Background of colleges
Students
With on average approximately 10 years prior work experience, Garoua students
have been professionals, working in protected areas or forestry or agricultural services (Table 10.2). Mweka and SAWC have also attracted an increasing number of
students from community organisations, and hunting and tour operators, while
some students have been self-financed.
Since the late 1980s, there has been a steady increase in the number of female students. The preferential scholarship policy at Garoua has raised the number of
female students to 25%, double that of Mweka and SAWC (Table 10.1).
240
Part III – Enhancing conservation – development integration
Table 10.2 – Prior education and work experience (years ± 95% confidence interval) of
Garoua students between 1997 and 1999 (BAC roughly equivalent to English A-levels, BEPC is
roughly equivalent to English O-levels/GCSE).
Experience
Average
Diploma
Certificate
Prior education
Median
BAC + 2.0a
BEPC+ 2.0b
Prior field work
Mean
6.4e ± 1.7
6.7e ± 2.4
Prior office work
Median
1.0c
5.0d
Prior total work
Mean
9.2f ± 1.7
12.3g ± 2.4
Comparisons between diploma and certificate students are given by a different superscript:
p < 0.05 by Mann-Whitney (a-d) or independent t-test (e-g) (SPPS 1999).
Trainers
Five of Garoua’s nine permanent training staff had university training in biology
or agronomy; the others were Garoua graduates. Increasingly, temporary staff are
hired, while expatriate staff have declined from an initial majority to none. Most
of the staff used to have a management or scientific background, whereas the current staff members have spent their entire career at the college. At its establishment, most Mweka staff had a B.Sc. and there were few diploma holders. At present, all but one of the 14 staff at Mweka hold M.Sc. degrees. SAWC has four permanent staff to ensure the functioning of the college, drawing on a variety of subject specialists for training.
Facilities and financing
All students have access to computers, but it has been difficult to keep track of the
latest developments, such as in the Internet, at Garoua. Garoua and Mweka spend
periods of up to two weeks in the field, requiring important investments in cars,
camping gear and other field equipment. Day trips dominate at the SAWC, which
is surrounded by Kruger NP. Garoua and Mweka manage nearby savannah and
mountain rainforest reserves for fieldwork.
Garoua has received significant Cameroonian government financing, albeit on a
lower level than in the 1980s, ensuring the salaries of its personnel. Tuition fees
cover most of the running costs, including the fees for temporary trainers, whereas donors have provided investments and staff training. The annual costs of a
Garoua student is estimated to be € 8800, including scholarships (€ 4600) and
annual tuition fees (€ 2100) (Njoya 2001).
Although a government structure, Mweka does not receive government financing
-1
for recurrent activities and depends on tuition fees (US$ 3125 yr and US$ 6000
-1
yr for Tanzanian and foreign students, respectively) and project support. The
241
10 – Curriculum development
-1
SAWC is a private institution and tuition fees (US$ 4500 yr ) and company and
donor grants assure its continued functioning and cover investments.
Curriculum reform and review
Three internal curriculum reforms have taken place since Garoua’s establishment,
namely in 1977, when certificate and diploma courses increased from one to two
years in duration (Allo 1978), 1981 and 1989. The 1996-1997 reform, which led to
a modular programme, was based on a five-day workshop with a variety of participants, including three non-Cameroonian graduates, followed by a workshop with
donors and ministerial personnel (Table 10.3 and 10.4). Garoua staff produced syllabi and began curriculum implementation in September 1997.
Table 10.3 – Curriculum reform and ongoing review at the African wildlife colleges
in the 1990s
Mweka
Garoua
SAWC
White blanket
Existing curriculum
White blanket
Workshops,
Technical workshop
Consultancies
consultancies
PR follow-up
Workshop
PROCESS
• Starting base
for reform
• Reform methods
workshop
• Period of reform
1993–1997
1996-1997
1994-1997
• Frequency of review
Annually
Bi-annually
?
Ineffective (lack
Yes
FEEDBACK
• Course evaluation:
Trainee
of confidence)
Initially by each
trainer, since
recently centrally
Faculty
Yes
Yes
Limited
Outsiders
Yes
No
Yes
2 limited studies
Ongoing (n>>100)
• Tracer studies
(n=55, 18)
Results not yet
available
Mweka has gone through six curriculum reforms (Snelson 1993), driven mainly
by consultants. The 1990s reform took five years as it deliberately ignored the
existing curriculum (Table 10.3). Although facilitated by Mweka staff, substantial
technical assistance was delivered through workshops and the production of syllabi in 2001. The production of course manuals is, as at Garoua, still underway.
242
Part III – Enhancing conservation – development integration
Table 10.4 – Participants of curriculum reform and review at the African wildlife colleges.
Mweka
(Potential) trainees
Garoua
SAWC
Reform
Review
Reform
Review
Envisaged1
Initiation2
Review
+
––
––
+
+
––
––
Protected Area managers
++
+
+
–
+
?
+
Local communities /
––
––
––
––
––
–
–
Development NGOs
Clients:
Governmental
++
+
+
––
+
?
+
Private sector
+
+
–
––
–
+
+
Conservation NGOs
Trainers, inside college
+
–
+
––
+
+
++
++
++
++
++
+
–
–
Trainers, outside college
+
+
+
––
+
+
+
Scientists
+
+
+
––
+
+
+
Presence of participants: ++ = majority, + = one-several, – = occasionally one, – – = none. ?= unknown.
1
2
Pending a presidential decree, (re-) institutionalising the college; due to the recent start of the college
The SAWC leaves much initiative to its visiting expert staff that, not bound to syllabi, have had the ability to develop their own courses for which they generally produce manuals.
Recent curriculum review at Garoua has been driven mainly by student feedback
(see below), reported to a commission of training staff (Table 10.3). A similar commission deals with course development at Mweka, which has recently installed an
external curriculum review commission like that at SAWC (Table 10.3).
Developments in Garoua’s curriculum
Changes from 1979 to 1997 in Garoua’s curriculum have been gradual (Fig. 10.3).
One of the few changes was the 1981 introduction of ornithology, a biology subject that has remained a major course, possibly because of the enthusiasm of the
trainer. In 1997, veterinary techniques (biology), a prominent course due to the
presence of French veterinary assistance until 1996, was reduced. In 1997, computer science (administration) and GIS (inventories) were introduced, as well as
participatory approaches and environmental education, thus tripling the importance of socio-economics.
243
10 – Curriculum development
Percentage of curriculum
40
1979
1981
1989
1997
30
20
10
m
is
ur
to
6.
E
ng
i
ne
er
in
g
&
ar
k
s
&
ab
5.
P
m
ec
ha
7.
ni
cs
Ar
m
s
&
hu
8.
nt
So
in
g
ci
oec
on
9.
om
Ad
ic
m
s
in
&
le
gi
sl
10
at
.W
io
n
or
k
pl
ac
em
en
t
tm
gt
ita
or
ie
s
4.
H
3.
I
nv
en
t
ol
Ec
2
1.
B
io
l
og
y
og
y
0
Figure 10.3 – Curriculum subject composition of the diploma course at Garoua Wildlife College,
1979-1997.
Feedback on the new Garoua curriculum
The perceived relevance of courses at Garoua has been remarkably consistent but
varies between disciplines (Table 10.5). Courses on biology, inventories, parks and
tourism, and arms and hunting were considered to have above-average relevance
(Table 10.5). Surprisingly, courses in the field of ecology and habitat management
were ranked lower. New courses ranked from very high (for example participatory approaches: socio-economics), to very low (for example remote sensing and GIS:
inventories). The relevance of the extensive ornithology course (biology), a topic that
received far less attention at Mweka and SAWC, was considered above average.
Students and training staff alike expressed their interest in field trips and exercises as the most effective type of training, addressing skills and complementary to
the predominantly knowledge-based classroom training. Moreover, fieldwork (several weeks of bush experience) was considered unsurpassed in developing professional attitudes by students at all three colleges.
Curriculum comparison between the three African wildlife colleges
The curricula of the two-year certificate and diploma courses were similar at Garoua
(Fig. 10.4). Mweka and the SAWC focussed their one-year certificate course on traditional wildlife management (inventories, mechanics) and their diploma courses
of two years and one year, respectively, on socio-economics and administrative subjects. This contrasted with the substantial prior office experience of certificate stu-
244
Part III – Enhancing conservation – development integration
Table 10.5 – Relevance ranking of disciplines by Garoua students (1997-1999).
Rank
Diploma students
Mean1,2
Certificate students
Mean2,3
1
Parks & tourism
4.49a
Inventories & wildlife
4.37a
2
Biology
4.41a
Arms & hunting
4.28a
3
Inventories & wildlife
4.39a
Biology
4.26a
4
Arms & hunting
4.38a
Parks & tourism
4.25a
management
management
5
Habitat management
4.20
Administration & legislation
4.18a
6
Socio-economics
4.19ab
7
8
9
Ecology
Administration & legislation
Engineering & mechanics
ab
Socio-economics
4.07ab
bc
Ecology
4.09b
c
Habitat management
3.93b
cd
Engineering & mechanics
3.54c
4.14
4.08
3.91
2
1
Median of rank 1 and 2 is 5; rank 3 is 4.25, rank 4-9 is 4; Comparison of medians amongst disciplines, different
3
letters in superscript indicate p < 0.05 (Mann-Whitney); For all ranks median is 4.
dents at Garoua (Table 10.2). Differentiation between certificate and diploma courses enabled students to start at certificate level and continue the following year with
the diploma course. Although possible, this has created considerable overlap at
Garoua.
Biology and ecology received ample attention at Garoua and at Mweka (Fig. 10.4).
The lack of these disciplines at diploma level at SAWC was striking, given the preference for biology exhibited by Garoua students (Table 10.5). Inventories and wildlife
management and habitat management received attention in all curricula. Parks
and tourism were dealt with at all colleges and contributed to a third of the diploma curriculum at the SAWC. Garoua and Mweka scarcely covered engineering
and mechanics, nor did the SAWC diploma course, a development consistent with
the low appreciation of these subjects by Garoua students (Table 10.5). Socio-economics received attention in all curricula, with courses at SAWC and the diploma
course at Mweka spending up to 20% of the time on this area. The SAWC further
spent an important part of its curriculum on administration.
10.4
Discussion
The wildlife colleges have made considerable internal efforts to keep pace with the
new developments in wildlife management. Do these efforts result in a curriculum
that meets the demands of wildlife practice?
245
10 – Curriculum development
Garoua/C
Garoua/D
Mweka/C
Mweka/D
SAWC/C
SAWC/D
Percentage of curriculum
40
30
20
10
&
s
ng
i
ne
er
in
g
&
ar
k
5.
P
4.
H
m
m
ec
ha
7.
ni
cs
Ar
m
s
&
hu
8.
nt
So
in
g
ci
oec
on
9.
om
Ad
m
ic
in
s
&
le
gi
10
sl
at
.W
io
n
or
k
pl
ac
em
en
t
to
ita
ur
is
tm
gt
gt
m
e
lif
ab
6.
E
3.
I
nv
en
t
or
ie
s
&
w
ild
2.
E
1.
B
io
l
co
lo
g
og
y
y
0
Garoua/C=Garoua, certificate course. Garoua/D = Garoua, diploma course. Mweka/C= Mweka, certificate course.
Mweka/D = Mweka, diploma course. SAWC/C= SAWC, certificate course. SAWC/D= SAWC, diploma course.
Figure 10.4 – Curriculum subject comparison of the three African wildlife colleges.
Continuing discrepancy between curricula and professional requirements
There are different requirements for a two-year curriculum and a six-week work
placement and these cannot be translated automatically into developments of job
requirements. For example, protected area management subjects have received less
attention in work placements than in Garoua’s curriculum, yet are expected to occupy a major share of a wildlife manager’s time. Tourism and exploitation and people-oriented conservation have received more than twice as much attention in work
placements as in the curriculum. More important than this absolute difference
was the observed time lag in attention. Whereas in 1994 people-oriented subjects
peaked amongst work placements, their subsequent decline coincided with an
increase in attention in the curriculum, a lag of five years.
Three hundred interviews with protected area managers across eastern, central
and southern Africa, conducted in the early 1990s, identified training needs in three
job responsibilities, namely (1) implementing intervention programs, (2) ensuring
visitor satisfaction and (3) promoting conservation in local communities. Managers
further expressed the need for (4) knowledge-oriented training in policies and procedures, planning and administration, as well as (5) skills-oriented training in creativity, problem analysis and evaluation techniques (Pitkin 1995). A training needs
246
Part III – Enhancing conservation – development integration
assessment in Southern Africa concluded that the wildlife colleges remain ‘by
their nature’ weak in providing skills and attitudes especially in administration,
project and business management (need 4) and community wildlife management
(need 3) (ULG 1998).
These training needs were already addressed in the curricula, needs (1) and (2) in
inventories, habitat management, and parks and tourism (Fig. 10.3), introduced or
reinforced during the recent reform, such as needs (3) and (4) in socio-economics,
and administration and legislation (Figs. 10.3, 10.4), or have remained largely uncovered (need 5), with the possible exception of the SAWC. In the early and mid
1990s, a reported general lack of skills amongst graduates of Garoua and Mweka
(Pitkin 1995; Snelson & Lanjouw 1997; ULG 1998), explained the continuing need
for training in responsibilities that were already addressed by the curricula (needs
1 and 2), but probably too theoretically. This was in line with the evaluation by
Garoua students in 1998-2000, showing that training after the curriculum reform
was still too knowledge-oriented.
Curriculum reform and review
The differences between the curricula (Fig. 10.4) can be partly explained by the
reform processes. The short intensive reform at Garoua with a limited number of
participants (Table 10.4) who were mostly familiar with the old curriculum, led to
gradual changes imprinted by the training staff, as reflected by the reported course
introduction. The elaborate reform process at Mweka (Table 10.3) led to a break
with the previous curriculum, as illustrated by the new one–year only certificate
course. The newly-developed focus on the wildlife profession probably played a
critical role.
None of the colleges had a feedback system in place prior to the curriculum reform.
At Garoua, feedback from students with an average of 9-12 years’ work experience
(Table 10.2) was introduced in 1998 and considered an important indication of
quality. Despite sometimes-harsh judgements that could have provoked resistance
(Tuntivanich 1989), trainers used the subsequent evaluations to monitor their
functioning and impact. This has also led to a reduction in the number of temporary trainers and adjustments in overlapping courses, but not to changes in the
course syllabi. The external review commissions at Mweka and the SAWC have
not yet had an impact on the curriculum.
All three colleges recognised the potential of communicating with their alumni
and started tracer studies (Table 10.3); however, communication and methodological problems prevented their feedback into the curriculum.
10 – Curriculum development
247
Training and learning
The three colleges have the distinct advantage of attracting experienced students,
who are motivated to learn and, with the exception of Cameroonian students, will
be rewarded with promotion once back home. Trainers with relatively little field
exposure and little experience with participatory training (Moss 1991) may have
difficulties using students’ prior knowledge (Dochy et al. 1997).
Socio-economic and administrative courses should especially be tailored to wildlife
practice. These courses are not automatically in the interest of students, who are
after all wildlife managers and not human-resource or business managers, and
therefore demand specialised trainers. The student feedback highlighted the discrepancy in socio-economics where the course participatory approaches was ranked
fourth in relevance, while the more general economics course ranked only fortythird.
Curriculum reform has been hampered by the difficulty of introducing new ideas
in an environment where students and trainers will not recognise them and hardly ever receive incentives for somewhat-controversial knowledge. An example is
the rejection of the transfer of technology paradigm (Chambers & Jiggins 1987) in
the Garoua participatory approaches course, whereas extension services, considered to be subject specialists, were still working with it. The SAWC has the distinct advantage of being located in South Africa, where it is in the forefront of developments in wildlife management.
Field work and exercises
Assignments and field exercises, appreciated for their effectiveness (see also Galindo-Leal 2001) were programmed to occupy a quarter of the time at Garoua and
slightly more at Mweka (Table 10.6). In practice they hardly ever attained this as
Garoua students noted, mainly due to planning problems, and not for financial
reasons as Garoua staff sometimes claimed. An increasing problem was the disintegrating protected area system, especially in Cameroon, depriving Garoua of
prime wildlife conservation field examples (Alpert 1993; Scholte 2003a).
Modularisation and short courses
Modularisation has been presented as an instrument to facilitate a transition from
a classical towards a more flexible education system (Meel 1997) and the introduction of a modular system brought advantages such as more focused learning
and training at the three colleges (Banham 1999). Some expected results, such as
increased exchanges between lecturers were, however, a function of changes in the
colleges’ management. At all three colleges, little use was made of the ability to
combine long and short courses. Modularization did, however, allow trainers to be
involved in additional research, consultancies and short-course training.
248
Part III – Enhancing conservation – development integration
Table 10.6 – Type of education at the three wildlife colleges.
Course length (hrs) (years)
Garoua
Mweka
SAWC
Certificate Diploma
Certificate Diploma
Certificate Diploma
2200 (2)
2700 (2)
1152 (1)
2352 (2)
1272 (1)
1314 (1)
Long field trips (%)
18
15
27
20
0
3
Exercises/ Assignments (%)
25
23
291
291
?3
?3
Lectures (%)
39
47
442
442
?3
?3
18
15
0
6
0
19
Work placement /
Research period (%)
1
2
3
Includes student presentations and directed study; Includes seminars and self-directed study; Not centrally
fixed, dependant on trainer.
Since the mid-1980s, Garoua has organised short courses on wetland management
that were initially externally financed and attended by both outsiders and students.
These stimulated the 1996 modularization based on ecosystems and were subsequently integrated as modules in the curriculum. Mweka and the SAWC organised short courses more regularly than Garoua, providing the well-situated SAWC
in particular with regular income. Moreover, short courses allowed the colleges to
keep in touch with senior wildlife managers (Pitkin 1995; Stone 1997).
Science bias
The comparison between work-placement subjects and curricula indicated a bias
at Garoua towards biological sciences, whereas employers were of more diverse
backgrounds. Science bias is widespread in conservation training (Cannon et al.
1996) and an argument for including people from various disciplines and interest groups in curriculum reform (Rogers & Taylor 1998). Stakeholders absent
from the Garoua 1996 reform were development organisations and local people
(Taylor 1997), whereas Garoua trainers dominated in numbers (Table 10.4). With
a comparable background Mweka staff probably also had a science bias, but they
involved more outsiders in curriculum reform and review (Tables 10.3 and 10.4).
The SAWC with its temporary staff has had the flexibility to draw trainers from
other disciplines.
Personnel aspects
Given the small size of the colleges, the staff have inevitably placed a personal imprint on the curriculum and this is to be applauded when the result of initiatives
of dedicated staff. At Garoua, trainers with what at first sight appeared to be unsuitable professional backgrounds developed several new courses. Examples were
a veterinary trainer who attended a six-week participatory rural appraisal (PRA)
10 – Curriculum development
249
training course and a wildlife utilisation trainer who attended a 9-month remote
sensing/GIS course; both subsequently introduced these topics at Garoua. Trainers at Garoua attended courses and workshops organised by international NGOs
and there has also been specific funding to Garoua for such training. Expatriate
staff at Garoua have sometimes set important directions, such as on rainforest
and wetlands management, which have been continued by counterpart staff after
their departure. Subject-specialist trainers have had the freedom to develop their
own courses with their own personal imprint at the SAWC.
Some pronounced personnel impacts on the curricula, however, were due to institutional deficiencies. In Garoua’s curriculum, training personnel could be held
responsible for several deplored developments, and the English language course
was abandoned with the departure of the anglophone director. That choice of specific trainer may have an impact beyond presentation was illustrated by the differences in perceived relevance between diploma and certificate courses that could
sometimes be related to the trainer concerned (for example, habitat management,
Table 10.5). Garoua and Mweka have been linked to their wildlife ministries and
these should have allowed frequent rotation of personnel, maintaining contacts
with wildlife management practice. Unfortunately, two-thirds of the training staff
have been at Garoua and Mweka for more than 15 years, probably because of the
personal preferences of trainers, low status of the colleges and the lack of appropriate human resource policies at the ministries and colleges. Given their regional
orientation, it is regrettable that none of the colleges has employed trainers from
within their regions.
10.5
Conclusions
Curriculum reform, especially at Garoua, has been related with personnel management, in particular the limited recruitment of trainers with a practical wildlife
management background and interest, and subsequent training and follow-up.
The 1996 Garoua curriculum reform workshop did not institute an external follow-up, nor has Mweka’s reform been effective in this regard. Curriculum reform
should, however, be a regular phenomenon and include representatives of development agencies and local people (Taylor 1997). Students’ feedback on the curriculum has already played an important role at Garoua.
Obviously there is no ‘best’ curriculum and opinions will always differ on the ideal
composition of any curriculum. Nevertheless, the differences between the curricula with similar target groups are remarkable. To some extent this may reflect
regional differences as indicated by different views of the wildlife profession, such
as the ‘wildlife MBA’ diploma course at the SAWC. It would therefore be interesting to compare the reactions of students to their respective curricula among the
250
Part III – Enhancing conservation – development integration
three colleges. Interest has been shown in stimulating such exchanges (Kanyamibwa 1999), although as yet not on a structural base. This may also result in student exchanges for a specific module such as private wildlife management at SAWC
and lowland rainforest wildlife management at Garoua.
The regional wildlife colleges have been pressured to orient training towards prestigious university level, as in their short courses. Collaboration with universities
could avoid the accreditation difficulties experienced by Mweka and enable the
small colleges to concentrate on their practical training niche.
Despite the discussed drawbacks, Garoua and Mweka have shown by their continued existence to be sustainable, an achievement given the harsh institutional environment (Ebohon et al. 1997) that has led to the virtual collapse of the university
system in Cameroon and other countries. While pursuing changes in curriculum
and the colleges’ structures, care should be taken not to jeopardise their sustainability. Donor fatigue in supporting the colleges is not entirely justified given their
regional role and, in the case of Garoua, continuous support from the host government. Outside support has been crucial, not only as a source of investment,
but also through staff training as an impetus for the implementation of curriculum reform. This has led to the education of several generations of African wildlife
managers, including women, at a third of the costs in Europe (Van den Bor 1989).
Future conservation success in Africa will depend to a large extent on the capacity and dedication of wildlife college students, as it has in the past.
11
Wildlife Managers’ Perceptions of
Community Conservation Training in
West and Central Africa
Paul Scholte, Wouter T. de Groot, Zacharie Mayna and Talla
Environmental Conservation, in press
Abstract
Training need assessments for African protected area managers have revealed
the need for people-oriented training to increase their job performance. The
Garoua Wildlife College, Cameroon, developed long (diploma and certificate)
and refresher courses in community conservation for mid-career protected area
managers and park guards from West and Central Africa in the late 1990s.
Through lectures, case studies and Participatory Rural Appraisal exercises, the
courses emphasised the development of skills for tuning principles of people
participation to the conservation objectives of protected areas. The present
study reviews the trainees’ evaluations of these courses to appreciate their relevance and support their further development. Diploma students judged the
courses as highly relevant because of the acquired analytical skills, whereas certificate students considered them only of medium relevance. The reaction on
short refresher courses varied as a function of the use of cases from either the
trainees’ professional experience or from the fieldwork location. The reactions of
trainees to this learning opportunity suggest that protected area personnel are
not ‘attitude-limited’ as often suggested. Their constraints to develop more people-oriented work style lie largely in the areas of knowledge and skills. These
findings motivate increased efforts to implement training of protected area personnel in community conservation, preferably early in their careers.
253
11.1
Introduction
Community conservation, i.e. the notion that conservation cannot and should not
be pursued against the interests and wishes of local people, has received increasing attention from researchers, authorities, development and conservation agencies
and local community organisations (Adams & Hulme 2001; Hulme & Murphree
1999). Bruner et al. (2001) and Vanclay (2001) showed the importance of the
number of park guards and their awareness building activities in local communities for the effectiveness of protected areas. Protected area personnel, i.e. wardens
and park guards, have seldom been involved in the development of community
conservation, however, in spite of its large international funding (El-Ashry 2001).
Training needs assessments for Eastern and Southern Africa (ULG 1998) and the
African continent (Pitkin 1995) have highlighted the need to train protected area
managers in human-related subjects, see also Western (2003). Also in India and
the USA the need to include the human dimension in conservation education has
been stressed (Saberwal & Kothari 1996; Jacobson & McDuff 1998). The African
training needs assessments have remained, unfortunately, without further specification, e.g. stressing a ‘need for all aspects of Community Based Natural Resource
Management training’ (ULG 1998). In some publications, emphasis is put on
required changes in attitudes of protected area personnel (IIED 1994).
Existing training courses in participatory natural resource management (BorriniFeyerabend et al. 2000; Nguinguiri 2001) are based on insights from participatory rural development (Pretty et al. 1995) that do not take potential frictions between conservation and short-term development objectives into account (Oates
1999; Scholte 2003a). For example, protected area personnel also have a policing
role, rendering their position towards local communities sometimes ambiguous,
which requires special skills and attitudes. Another difference with participatory
rural development is that protected area personnel are evaluated on overall conservation results. From their perspective therefore, community conservation tends
to be seen as a means rather than a goal.
This chapter reviews the evaluation by the trainees of the first long and refresher
courses in community conservation for West and Central African protected area
personnel given by the basic training college at Garoua (Cameroon). A comparison will be made with courses at the College of African Wildlife Management at
Mweka (Tanzania) and the Southern African Wildlife College, SAWC (SouthAfrica), that train mid-career medium level wildlife professionals from especially
Eastern and Southern Africa respectively. This study should enhance an appreciation of the relevance of training in community conservation and support its further development.
254
Part III – Enhancing conservation – development integration
11.2
Garoua college and the community conversation courses
The Garoua Wildlife College and its students
The Garoua Wildlife College or simply ‘Garoua’ is one of the three regional African wildlife colleges that train mid-career protected area managers (Scholte 2003b).
Founded in 1970, Garoua draws its students from twenty francophone, mainly
West and Central African countries. Garoua provides courses of two years on Certificate and Diploma levels, with entry requirements BEPC (‘O-levels’) and BAC
(‘A-levels’) respectively, and at least three years of working experience (Table 11.1).
Many park wardens in West-Central Africa are Garoua diploma graduates, whereas certificate graduates occupy medium-management positions such as assistant
warden.
Garoua’s curriculum has been dominated by courses on species identification and
the functioning of ecosystems, while some attention has been paid to tourism and
sport hunting. Prior to 1996, hardly any attention was paid to the role of local people in wildlife exploitation and protected area management, apart from incidental
theoretical courses on sociology and extension science (Scholte 2003b). Garoua is
not an exception; regional colleges in India (Saberwal & Kothari 1996) and Tanzania (Snelson 1993) shared, till recently, this almost complete focus on bio-ecological wildlife management. An external review carried out in 1996, stressed the
introduction of a community conservation course and an update of courses on
park management planning and environmental education. A total of about 300
hours, or 10 percent of Garoua’s curriculum, has now become people-oriented. In
addition, 40% of the students presently choose a people-oriented subject for their
practical period (Scholte 2003b). With the curriculum reform also the need was
identified to upgrade skills of former Garoua students who are active as protected
area managers.
Community conservation courses
Protected area personnel are often considered insensitive to the needs of local communities (IIED 1994), resulting in a lack of involvement and difficult implementation of Integrated Conservation and Development Projects (ICDPs; Scholte et al.
1999b). The goal for the short and long courses was therefore to raise awareness and
increase capacities for an atmosphere of transparency and mutual trust between
protected area managers, local communities and conservation-development project personnel (Table 11.2).
Long courses for mid-career protected area managers
The average professional experience of Garoua students is over ten years (Table
11.1). Such prior professional experience has been argued to be the most important
variable influencing learning results (Dochy et al. 1997), and calls for the active
participation of students that often have more field experience than the trainers.
Table 11.1 – Background of trainees and community conservation courses at the Garoua Wildlife College (1996 –2000)
Type of courses
Sessions
Number of
participants
Prior Education1
Prior work
experience1
Prior experience
in community
conservation2
O-level + 2 additional
years (average)
4.3 to 9.1 years
fieldwork + 5 years
office (average)
None: 50%
Passive: 30%
Active: 0%
In other profession: 20%
A-level + 2 years
(average)
4.7 to 8.1 years
fieldwork + 1 year
office (average)
None: 27%
Passive: 45%
Active: 9%
In other profession: 18%
< 5 years
primary education +
about 3 months initial
professional training
10–30 years, with
little change of posts
None
LONG COURSES
Certificate course3
1997-99
1999-01
Diploma course3
18
20 days /5 days
Waza and Benoué,
NP, Cameroon
14
1997-99
33
1999-01
25
11 – Community Conservation Training
Course length /
Length of
village exercises
Protected area
SHORT COURSES
Park guards
(Far North
Cameroon)
1996
2 days /none
2 x 13
Park guards
(North Cameroon)
2000
3 days /
0.5 day role play
2 x 30
Park wardens
(C. Africa4)
2000
10 days / 2 days
Dja Reserve,
Cameroon
15
Park wardens
(W. Africa5)
2000
10 days / 1 day
Niokolo NP, Senegal
16
1
2
Predominantly Diploma 5-20 years, often a
course at Garoua with
variety of posts and
additional training
responsibilities
Majority with little to
no experience
3
Information on long courses based on 1997-1999; Information on long courses based on 1999-2001; From 17 francophone West-Central African countries;
5
CAR, Cameroon, Chad, Congo-B. and Gabon; Burkina Faso, Guinée C., Mali, Mauritania, Senegal
4
255
256
Part III – Enhancing conservation – development integration
In view of the present predominant lecturing style Garoua faces the challenge to
include insights from participatory adult education (van der Bor et al. 1995; Crowder et al. 1998).
Both long and short courses started out with discussing the objectives of community conservation, addressing knowledge and attitudes. These objectives were specified by the tuning of objectives of a protected area (IUCN 1994) with possible participation level such as described by Pretty et al. (1995) and Barrow & Murphree
(2001), essentially addressing analytical skills. For the remainder of especially the
long courses, a ‘community conservation cycle’ was put at central stage, in analogy to the project cycle (Cracknell 2000), to stress the interrelationship between
undertaken activities (Table 11.2). The diagnostic stages of Participatory Rural
Appraisal (PRA) and problem- and stakeholder analysis (de Groot 1998) were followed by the facilitation of stakeholder representation. Confidence building, addressing attitude and communication skills, was part of the course as well because
of its relevance for protected areas with tense relations between protected area
authorities and local communities. Problem analysis and categorisation allowed
trainees to get a grip on development problems with conservation requirements
and their inclusion into the community conservation cycle.
Long courses were designed with a focus on analytical and communication skills,
addressed by lectures, fieldwork and examples from protected areas either visited
or from students’ home areas (Table 11.2). A one-day village PRA exercise was
taken up in the college’s regular 1-2 week field trips to one of the national parks
in northern Cameroon. Unfortunately, fieldwork planning did not allow for overnight stays in a village, which could have stimulated more informal contacts.
Short courses for park guards
Both short courses (Table 11.1) aimed at initiating park guards to collaborate with
local communities (Table 11.2) and the development of a dialogue between ICDP
projects and guards. The courses were structured around the main phases of the
community conservation cycle, with examples from the local protected area and its
ICDP project, and based on the concept that involvement of local communities
should not imply loss of authority over the protected area.
Short courses for park wardens
The refresher course was limited to 10 days because of financial and absencefrom-post constraints, motivating a specific course theme: ‘The role of management planning in the involvement of local people in protected area management’.
This choice was based on the interest that Cameroonian park wardens had shown
in the park management planning of Waza National Park and the frequent discussions on community conservation (Chapter 9). These refresher courses focused
participants from several West and Central Africa countries on a selected protect-
Table 11.2 – Contents of community conservation courses at the three African Wildlife Colleges
College
Level
Long
Unspec.
Short
Guard
Warden
A,Sc
Sc
Sa
K, S, A
K, Sf
Sc
K
+
+
+
++
++
++
++
++
+
–
–
+
–
–
+
–
––
+
–
––
–
––
+
+
++
+
–
–
–
–
–
–
–
Sa
Sa, K
+
–3
–
––
S
S
S
+3
–
–
–
––
––
Mweka1
SAWC2
Long
Long
Certificate
Diploma
unspec
–
+
–2
–
–
+
–
––
––
–
–
+/–
+
–2
–
+
+
–
––
––
–
–
–
+
–
–
–
+
–
––
––
–
–
–
–
–
–
+
+
+
++
–
––
–
+
––
+
++
+
+
++
–
+
Type of training3
Description
Community conservation: goal or means?
Synonyms of community conservation
Tuning participation level to objectives
The community conservation cycle, including:
• Confidence building
• Diagnosis: PRA
• Diagnosis: Problem / stakeholder analysis
• Representation of stakeholders
• Negotiation of contracts,
• Implementation of contracts
• Monitoring and evaluation
Skills addressed in:
• Analysing park – people interactions
• Formulating conservation & development
strategies
• Communication
• Facilitation
• Conflict management
K,A
K
Sa
11 – Community Conservation Training
Course length
Garoua
++: dominant; +: much attention; –: fair attention; – –: no attention:
1
2
3
4
2001 syllabi; 1998 course manuel; K: Knowledge – A: Attitude – Sa: Skills, analytical; – Sc: Skills in Communication – Sf: Skills in Facilitation; mostly limited to participa5
tion typology (Pretty et al. 1995); also subject of follow-up long courses Protected Area Planning and Extension
257
258
Part III – Enhancing conservation – development integration
ed area where management planning was ongoing. In preparation of the short
courses, park wardens prepared 10 minutes presentations on management planning in relation to local community involvement in their own protected area, allowing a focus of the introductory lectures on management planning and community conservation, for both trainers and trainees. In the session with Central
African park wardens the hosting park warden presented the ongoing park planning process, subject of subsequent fieldwork. In the training session with West
African park wardens, the host park warden was not able to carry out his presentations and emphasis of the course was shifted to an exchange of experiences
amongst participants, with fieldwork limited to a one-day visit of development
realisations. Both sessions included lectures and discussions on PRA and communication skills, with PRA field exercises in the Central African session only.
Comparison with other African Regional Wildlife Colleges
With the Garoua courses as starting point, we compared the 2001 syllabi of the
one-year certificate and two-year diploma courses at the College of African Wildlife Management (Mweka, Tanzania) and the 1998 course manual of the one-year
course at the Southern African Wildlife College (SAWC, S.Africa).
11.3
Evaluation methods
Training is a systematic process, which helps individuals develop knowledge (facts
and information), skills (the ability to carry out an action) and attitudes (feelings
expressed to others through behaviour or language) to increase their job performance (Rothwell & Sredl 1992; Stone 1997). These three training dimensions are
subject of the review in this paper. We followed Kirkpatrick’s hierarchy (Kirkpatrick 1994; Rothwell & Sredl 1992), based on trainees’ (1) reaction, (2) learning,
(3) behaviour change during job performance. No course evaluations from Mweka
and SAWC were available for comparison.
Trainees’ reactions
At the start of the 1999-2001 long course, students were invited to write a statement on people’s involvement in conservation, its justification and their earlier
experiences. These statements were categorised in four classes, ranging from development to conservation focus. Long course students judged the ‘relevance’ of each
course of the curriculum for their (future) profession, as part of the curriculum
evaluation. Students also judged the ‘presentation’ of each course, in which they
addressed whether they ‘liked’ the course and its training methods. Both long and
short course participants were asked to judge the relevance of course themes and
training methods on a scale of 1 (non-relevant) to 5 (highly relevant) and explain
their choice in few words.
11 – Community Conservation Training
259
Trainees’ learning
At the end of all but the 1996 short course, trainees’ reactions on their perceived
learning experiences were anonymously assessed with two multiple-choice questions. Half way through and at the end of each long course, Garoua organised a
written assessment, addressing acquired knowledge mainly, constituting half the
students’ mark. An oral assessment at the end of the semester constituted the
other part, together with, in 1999-2001, an appreciation of the PRA field exercise’s
report written by groups of four diploma students. Oral assessments generally
emphasised students’ analytical skills and attitude and often referred to work placement experiences. These assessments fulfilled regular educational requirements,
and were a check on learning progress.
Behaviour change during job performance
Experience with tracer studies of Garoua has shown that post-course assessments
based on sent-out questionnaires only do not yield valid results and need more
cost intensive interviews conducted in the respective countries of former students.
For the present courses, post-training observations could be made for park guards
based in North Cameroon only.
Statistical analysis
Correlations of formal assessments were analysed with Pearson correlation, comparison of their means with paired t-test (SPSS 1999). The five point rating scale
used in the curriculum and course contents and -methods evaluation was analysed
with a Mann-Whitney test (SPSS 1999).
11.4
Results
General reactions to relevance and presentation of the courses
Three-quarters of the diploma students and half of the certificate students indicated they had some experience in community approaches prior to the course,
especially where they had a background in agriculture or forestry (Table 11.1). Asked
to justify community conservation at the start of the course, certificate students
emphasised conservation objectives, whereas diploma students tended to emphasise development objectives.
In the curriculum evaluation of the diploma long course, the relevance of both sessions of the community conservation course was evaluated as above average
amongst the well-established courses in wildlife and vegetation inventories (p <
0.01). Diploma students judged the presentation of the 1997-99 session as average amongst these, improving in 99-2001 (p < 0.0001), without correlation with
perceived course relevance. Certificate course students perceived the relevance as
260
Part III – Enhancing conservation – development integration
well as the presentation of both sessions of the community conservation course as
average, outperformed by courses on mammals and legislation.
The theory-practice balance of the community conservation course was judged
equal to the other courses that, with the exception of courses on mechanics, computer science and vegetation inventories, were all considered too theoretical.
Park guards judged the course as useful (4.3 and 3.7 on a 1 to 5 scale for the 1996
and 2000 session respectively). The park warden refresher course was considered
very useful by half and two-third of the trainees of the West and Central African
sessions respectively. The remaining wardens considered the course merely useful.
Specific reactions to course contents and training methods
Courses were held in a stimulating atmosphere, with trainees eager to know what
this community conservation approach that so many people refer to really is about.
Discussions on fieldwork sessions were, especially at certificate level, sometimes
considered overwhelming. Frequent structuring and summarising course sessions
allowed students to take notes that they could consult and ask questions about in
a next session.
The introduction of both long and short courses emphasised the responsibility of
park personnel towards conservation objectives, stressing that community conservation is a means and not a goal as such. This was a response to some trainees
who felt threatened by the mere idea of community conservation, which they initially perceived as loss of authority.
Certificate students considered problem analysis, problem categorisation and communities’ representation as very relevant compared to the introductory subjects (p
< 0.05). The former had received much attention in the course, stimulated by the
large number of questions, exercises and discussions with examples from the visited protected areas, whereas the latter had remained on a somewhat more abstract
level. Diploma students singled out the topic of ‘Who is local?’ as more relevant
than the introduction (p = 0.04).
Diploma and certificate students expressed a strong preference for cases that were
familiar through their own fieldwork. The discussion of the Waza National Park
committee and especially the visit of a local community member who had signed
a contract with the Waza National Park authorities were thus considered above
average relevant (p < 0.05). Certificate students considered cases presented by fellow students on protected area committees very relevant (p < 0.05), contrasting
with a merely average appreciation by diploma students. This difference may be
attributed to a highly skilled certificate student who presented an innovative case
11 – Community Conservation Training
261
from Dzangha-Sangha (CAR). Contrasting with the high relevance of these direct
experiences were the discussions on an article on Campfire in Zimbabwe (Feron
1997) and a video on buffer zone management in Uganda (Brown & Singer 1991)
(p < 0.05).
In the park guards’ short course, stimulating question and answer sessions developed following each introductory lecture. These discussions were often centred on
the ambiguous legislation with regard to community involvement. Attention was
paid to development problems of the region and the role park guards could play
to gain confidence of local communities. The participation of wardens, the guards’
supervisors, in the course limited the frankness of the discussions, but assured a
consistent approach towards local communities. The ICDPs that financed the short
courses were happily surprised by the discussions as they had expected a somewhat hostile attitude of the park guards towards them.
In the refresher course for Central African park wardens, park management planning received a lot of attention with a high appreciation for the course theme (4.8
on a scale of 1 to 5) and perceived relevance (4.4). In the West African park wardens course, with the poorly developed practical case, exchanges of experiences (4.4)
were considered above average (p = 0.014), whereas community conservation
approach ranked second (4.2). Contrary to the long-course students, park wardens
expressed a lot of interest in the somewhat abstract subject of tuning participation
level to protected area objectives.
Trainees’ learning
Student assessments were (partly) based on fieldwork and associated classroom
discussions, and presentations of fellow students. Assessments showed generally
large differences amongst students that Garoua trainers generally attribute to the
students’ educational background that vary between their countries of origin.
Repeated assessments yielded little individual variation. With exception of the
1999 certificate long course, oral assessments targeting skills and attitudes, had
higher results than written tests (p < 0.05). Results of the diploma students’ exercise on PRA fieldwork, that stimulated working till late at night were marked even
higher.
Half to three-quarters of the diploma and certificate students declared that the
course had changed their attitude on the need to involve local people (Table 11.3).
All diploma students and a majority of the certificate students declared that the
course had changed their (analytical) skills on how to involve local people in conservation (Table 11.4).
262
Part III – Enhancing conservation – development integration
Table 11.3 – . Community conservation course evaluation, attitude dimension, assessment of change of
opinion by course participants on the need to involve, in one way or another, local people in nature
conservation (% of respondents).
Long courses
Short Course
1997-1999
2nd year
Yes, a lot
Yes, a little bit
No, I was already aware
No, the course was not of
any use
1999-2001
1st year
2000
Diploma
n=31
Certificate
n=16
Diploma
n=24
Certificate
n=14
Guards
n=55
39
23
39
0
44
38
13
6
75
17
8
0
64
14
7
14
62
33
5
0
Table 11.4 – Community conservation course evaluation, skills dimension, assessment by course participants on the change of their capacity to involve local people in nature conservation (% of respondents).
Long courses
Short Course
1997-1999
2nd year
Yes, a lot
Yes, a little bit
No, I already mastered
the subject
No, the course was not of
any news
1999-2001
1st year
2000
Diploma
n=30
Certificate
n=15
Diploma
n=24
Certificate
n=12
Guards
n=51
77
23
0
47
40
13
75
21
0
67
33
0
29
65
6
0
0
4
0
0
Virtually all park guards declared having changed their attitude on the need to
involve local people (Table 11.3), which was the primary objective of the park guard
course. This contrasted with the limited perceived changes in skills (Table 11.4),
not specifically addressed by this introductory course (Table 11.2).
Post-course behaviour change during job performance
In the months following the 1996 training, at least five (out of 26) guards undertook, stimulated by follow-up instructions of their wardens, initiatives such as the
organisation of village discussion sessions, the assistance of villagers against problem animals and a bushfire campaign. In addition, the warden of one of the northern province national parks initiated with his guards a public awareness campaign
11 – Community Conservation Training
263
in all park villages, the first time during their presence in the area. These initiatives much improved, at least temporarily, their hitherto negative relation with
local communities. The fact that half of the Far North Province guards went on
retirement between 1996 and 2004 limited the long-term impact of the course.
Community conservation courses at the other African Regional Wildlife
Colleges
In the 2001 curricula, Mweka has developed a two-week certificate course ‘People
and Conservation’ and, at diploma level, a two-week communication skills- and a
three-week Community Conservation course. A three-week course on participatory planning and community conservation is given at post-graduate level for senior
wildlife managers without prior training in participatory approaches, like the
Garoua warden refresher course. Long course training methods at Mweka and
Garoua were comparable, with predominantly lectures (50%), case studies by students (15%) and only 15% of the course time consisting of field exercises. Although
course objectives differed amongst the wildlife colleges, the contents were largely
comparable (Table 11.2). SAWC paid more attention to Community Based Natural
Resource Management (conservation outside protected areas), whereas Mweka and
Garoua centred their courses on protected areas.
11.5
Discussion and conclusions
Relevance of courses
Reaction to training is considered as one of several parameters to study the impact
of training (Kirkpatrick 1994; Rothwell & Sredl 1992; Stone 1997). Long and short
course participants had an average of over ten years of professional experience and
we are therefore inclined to consider their reaction on the relevance of the courses and their perceived learning experience as an indication of the impact of training for their future job performance.
Are park personnel ‘attitude-limited’ towards community conservation as generally assumed? (IIED 1994). In other words, do they reject learning and changing?
Rather our hypotheses is that, as with all people, it is difficult to have a positive
attitude towards something that you are insecure about and that you cannot put
into action. The trainees’ reactions show, however, that once roles and needs are
clarified and basic skills acquired, attitudes change quickly (Tables 11.3 and 11.4).
The evaluation suggests that even the three-day introductory courses were already
successful in changing attitudes on the importance of involving, in one way or another, local people in conservation. The high appreciation of the relevance of the
course by diploma students is consistent with the number that declared that the
course had increased their analytical skills (Table 11.4). Prior to the course, these
264
Part III – Enhancing conservation – development integration
diploma students associated community conservation with economic development
objectives mainly. The eye-opener was that communities can also be involved in
conservation. The merely average perception of certificate students and lower numbers who had acquired necessary skills, suggest that the long course was more
tuned to diploma students’ level or interest, probably because of its largely analytical character.
The course evaluations indicated the need for continuous reference to practice.
Training on the spot, as practised with the park guards, thus almost becomes mentoring, requiring trainers who are familiar with the working situation. For park
wardens, experiences from colleagues played an important role, which need to be
streamlined with the course theme, putting high demands on the trainers as well.
Experiences with long course students showed that sharing the experiences strongly depends on the availability of interesting cases and showed the need to assist
students in presenting them. For long course students, experiences from fieldwork locations have played an essential role. Cases presented on paper or video
should be limited to only the truly necessary, e.g. for broader comparison.
A more participatory training should be pursued to allow effective adult learning.
Didactic courses and follow-up mentoring may assist trainers, who generally use
the lecturing training style they themselves have been trained in, to develop the
necessary skills. Long course students with their hitherto passive training experiences should also be introduced to this participatory training, e.g. at the start of
the first year as at the other regional African wildlife colleges.
Perspectives
Experiences from Gabon suggest a low perseverance of training experiences of governmental protected area personnel, often based in the capital cities (Thibault &
Blaney 2001). Participants of the Garoua courses have field-based posts however
and may therefore be expected either to use the training results themselves or to
stimulate their application by others.
Garoua proposed to train a selection of park guards in communication skills, once
the policy described in the management plan would be clarified and clear instructions could be given. This would prepare them as community liaison agents, reducing some of the difficulties in combining policing and extension tasks (Van den
Ban & Hawkins 1988). Protected area personnel should receive training in community conservation attuned to their education level and working experience, thus
developing not only necessary knowledge and attitudes but skills as well. Apart from
the necessity to organise refresher courses for existing protected area personnel, a
need exists to develop community conservation courses as part of pre-recruitment
training for future personnel.
11 – Community Conservation Training
265
Training in community conservation allows protected area managers to participate in public outreach (Barrow & Murphree 2001; Gilbert 1971) and play an active
role in community involvement. This, in turn, may stimulate protected area personnel to influence authorities and projects towards a balanced integration of conservation and development objectives (Scholte 2003a) and the institutionalisation
of community conservation (Bergin 2001).
PART IV
Synthesis
12
Floodplain Rehabilitation and the Future
of Conservation & Development:
Synthesis and concluding remarks
In this concluding chapter, I review the outcome of the floodplain rehabilitation by
comparing the reflooding responses that, because of their different response times,
caused human supported resources to become dominant. I continue analysing the contributions of management planning and capacity building to correct these undesired
effects of the otherwise successful reflooding.
I subsequently review different concepts of ecosystem change and management that
may explain the observed changes. I first apply panarchy (Gunderson & Holling 2002), a
holistic empirical concept, that allows observed ecological as well socio-economic developments and interventions to be linked in a temporal and spatial hierarchy. It provides
limited causal explanations however. Secondly, I use the empirical concepts hysteresis
and resilience that show that the rehabilitation of Waza-Logone, despite its success in
the field, can be considered as a classic case of hysteresis in which both end results and
the rehabilitation path are different from the pre-dam situation and the degradation
process. I introduce the concept of ‘evolving states’ to explain that not all observed
states are static, but may undergo cyclic succession, such as has taken place with the
reflooding. I subsequently applied the holistic-normative ecosystem approach to present
the key findings of this study in a wider context by suggesting some further development of these guidelines for analysis.
At the end of this chapter I analyse the adaptive management approach pursued, that I
use as overarching concept including elements of the above discussed empirical and
normative concepts. I revisit the monitoring observations and insights to reformulate
the floodplain rehabilitation expectations and hypotheses postulated in the beginning
of this chapter and discussed in the respective chapters. The consequences for the overall floodplain rehabilitation targets as programmed in Waza-Logone (IUCN 1999) and
possibly in other (African) floodplains are discussed. I call especially attention to the
role of a ‘vision’ , including not only targets for rehabilitation but also understanding
the required enabling environment. Empirical concepts of change and exchanges with
other experiences played an important role in the development of this vision. One may
conclude that adaptive management, based on monitoring of separate ecosystem components and a regular review of targets within an overall vision, forms a useful framework for the rehabilitation and management of poorly known tropical (floodplain)
environments.
271
12.1
Synthesis of the monitoring results on reflooding responses
In this section I synthesise the changes in vegetation, wildlife and pastoralism in
Waza-Logone observed during the last two decades. I will start, insofar available
information allows, with a description of the degradation processes induced by the
1
construction of the upstream Maga dam in 1979. This is followed by a summary
of the changes observed from 1993-2000(-2003) and their interpretation. I conclude with the reformulated hypotheses for future reflooding. A particular point
of attention is whether observed changes can be attributed to the reflooding or to
alternative explanations such as the relatively favourable rainfall during the study
period (Chapter 2), (see Table 12.1 for an overview of their relative impacts).
Table 12.1 – Impact of reflooding and the importance of other factors that explain observed
developments (for temporal and spatial scale see Fig. 12.2)
Resource
Reflooding
Increased
Extraction
Protection
rainfall +
(grazing,
(by humans)
related
poaching
flooding
Vegetation
Composition
++
1
Diseases
Developments
elsewhere
or theft)
+/– 1994-1998
–
–
––
––
2
––
+ 1998-2000
Production
+
Waterbirds
Anatidae
(dry season
Omnivorous
numbers)
Ciconiiformes
Piscivorous large
1
+
–
–
––
+/–
++
+/–
+/–
–
?
+
+/–
+/–
+
–
+
+
+/–
++
–
–
–
+
+/–
+/–
++
–
+ (?)
+
+/–
+
+
+
4
–
+/– (?)
+
–?
+?
+?
–
Ciconiiformes
3
Waterbird
Black–headed
colonies
heron (nests)
Floodplain
Kob (number)
antelopes
Korrigum (number)
Pastoralism
Camps (number)
+
–
+/–
+/–
+
Grazing intensity
+
+
–
+/–
+
-2
(no. cattle km )
Impact : ++: major; +: important; +/–: existing; –: marginal; – –: none; ?: unknown; nothing mentioned: not relevant.
1
2
3
Includes the in the long–term impact of increased physical protection; no grazing during flood season; potentially
4
important; rinderpest
1 For earlier changes, see Chapter 2.
272
Part IV – Synthesis
Vegetation composition
Impact of the Maga dam
• Fragmentary information suggests that, in the early 1980s, in extended parts
of the area perennial grasses died. This can be largely attributed to the dam construction, although there were indications that the condition of especially Vetiveria nigritana was already poor due to the mid-1970s drought, the first since
the 1930s (Fig. 3.2). Till 1985 and in some occasions 1993, remaining individuals of the rhizomatous Echinochloa pyramidalis were found in the desiccated
parts of the floodplain (Table 3.2). In the central part of Waza NP, out of reach of
the reflooding but with local rainfall ponding, a few individuals of the tussock
grass Vetiveria nigritana survived for over twenty years.
Recent changes
• The 1993-2000 monitoring, supported by field observations in 2002, indicates
that in 2002 the perennial rhizomatous grasses Oryza longistaminata and Echinochoa pyramidalis occupied almost the totality of the reflooded area once again.
• In terms of species composition, this rehabilitation was only partial however:
of the previously dominant perennial floodplain grasses, Vetiveria nigritana did
not show any sign of re-establishment despite its good condition elsewhere in
the floodplain.
Impact of the reflooding
• The initial hypotheses of a ‘full recovery, in five years, to the pre-Maga state from
the 1970s’ was therefore rejected.
• The rehabilitation has taken place under (long-term) average rainfall and flooding conditions, but above average if considering rainfall of the last 30 years
only. The lay-out of our study, with plots both in and outside the area impacted by the reflooding, showed the large impact of reflooding compared to a low
impact of merely average rainfall.
Interpretations
• The gradual but steady recovery of Oryza longistaminata and Echinochloa pyramidalis grasses can be explained by lateral growth through their rhizomes.
• The tussock grass Vetiveria nigritana depends for its multiplication on seed dispersal, a less secure strategy that may depend on (temporarily) entirely different environmental conditions. This holds in particular for the direction of flooding water, and therefore the sediments it carries, that has remained entirely
different from the pre-dam situation.
12 – Synthesis
273
Hypotheses for future reflooding
• With future reflooding one should take into account a period of at least ten
years for the recovery of rhizomatous grasses. If rainfall is lower than average,
this period may be much longer.
• Full recovery that includes again a predominant presence of Vetiveria nigritana,
may last much longer than 10 years and may depend on siltation processes.
Vegetation production
Impact of the reflooding
• Above-ground biomass production showed a 2-year time lag in response to the
start of the reflooding.
• Assessments in Waza-Logone, supported by data from three other seasonally
flooded grasslands from the three main African geographic regions, showed
an increasing above-ground biomass production with maximum flood depth in
a 0-1 m range.
• Data from Waza-Logone did not support an established linear relation between
dry season regrowth production and wet season above-ground biomass from
the Inner Niger Delta (Mali), and thus with maximum flood depth. A comparison with other African floodplains hints at an exponential relation of regrowth
production if maximum flood depth exceeds ca. 100 cm. Below this maximum
flood depth, the timing of burning has an overwhelming influence on the (low)
regrowth production.
Interpretations
• We applied the relationship of maximum flood depth and above-ground biomass to the 20 cm water level rise, due to the reflooding, in an area of 60 000
ha. This way it may be estimated that wet season above-ground biomass has in-1
creased by approximately 800 kg DM ha in the first reflooding year (≈ 10%
production increase) increasing to an additional 2800 kg DM ha1 (≈ 37% increase) in later years.
Hypotheses for future reflooding
• I postulate a two year time-lag and the maximum flood depth - above-ground
biomass equation, for future reflooding in the plain grasslands (with a maximum flood depth < 1m).
• I hypothesise that the time lag of the response of dry season regrowth production to reflooding is a multitude of the two years time-lag of wet season aboveground biomass. Regrowth production is expected to start only after several
years, when sufficient rhizome biomass has been accumulated.
274
Part IV – Synthesis
Waterbirds
Impact of the Maga dam
• The list of bird species no longer observed since 1980 suggests changes amongst
especially the woody savanna species, more influenced by long-term rainfall
change than by the flooding (Scholte et al. 1999).
• I attribute the drop in numbers of the area’s flagship species, the resident
Black-crowned Crane, from 10 000 in 1973 to 2000- 2500 individuals in 1992
to the Maga dam construction. No quantitative information was available about
other waterbird species.
Recent changes
• The numbers of waterbirds, observed during the dry season, increased from
60 000 to 105 000 between 1992 and 2000.
• The numbers of waterbirds surpassing the international 1% criterion of their
population size, doubled from 6 to 12 per year. Also the population numbers
of each of these species important for conservation increased, as was shown for
Black-headed Heron and Black-crowned Crane.
• The already exceptional size of the Black-headed Heron colony at the start of
the reflooding tripled during the study period to a total of 2500 nests.
Impact of the reflooding and interpretations
• The increase in numbers of Anatidae (ducks and geese) corresponds to the generally observed trend in their recovery from the mid-1980s droughts as observed
elsewhere in West Africa.
• The change in numbers of Ciconiiformes (storks, herons and egrets) exceeded
their developments elsewhere in West Africa.
• The spectacular development of the Black-headed Heron colony was attributed
to the reflooding, and the incorporation of smaller colonies previously subject
of destruction.
• Amongst the water birds that showed a trend less than the general two-fold increase, were the large predominantly piscivorous resident Marabou and Yellow-billed Storks. This is attributed to the repeated destruction of their colonies
by (fisher) men, a recent development in Waza-Logone that has elsewhere in
West Africa already been under way for years.
Hypotheses for future reflooding
• Hypotheses for the recovery of waterbird by future reflooding should take into
account their long-term dynamics based on regional changes (e.g. rainfall) and
local changes especially human protection and prosecution. The categorisation
in functional groups, with a comparable reaction to reflooding, should therefore not only be based on status (Palearctic migrant – Afrotropical resident)
12 – Synthesis
275
and feeding ecology (piscivorous – omnivorous), but also on vulnerability to persecution.
Floodplain antelopes
Impact of Maga dam
• Numbers of Kob antelopes crashed from 20 000 to 5 000 in the period 19791983 due to the man-induced drought caused by the Maga dam construction.
The subsequent drying up of the area coincided with a period of rinder pest.
Numbers further declined to 2 000 at the end of the 1985 drought, showing
the antelope’s vulnerability to natural droughts as well.
• The disappearance of Waterbuck, soon after the construction of the Maga dam,
has become a symbol of the area’s degradation.
Recent changes
• The Kob population numbers increased between the late 1980s and the earlymid 1990s from 2 000 till 5 000 individuals, after which the population size
stabilised.
• Korrigum (‘Topi’) antelope numbers have increased steadily since the earlymid 1980s as well.
• In 1998 and again in 2003, two Waterbuck individuals were observed once
again.
Impact of the reflooding and interpretations
• Kob antelope is the only large mammal whose population dynamics in the period 1960-2000 could be linked to water management, i.e. the 1979 Maga dam
construction and the reflooding from 1994 onwards.
• The initial recovery in Kob numbers has not continued beyond 1997. The main
explanation is the increased competition with livestock due to the rapidly increasing cattle grazing intensity, also inside Waza NP (see below). Direct persecution
has almost certainly played an important, yet poorly known, role as well. The
dramatic drop in number of park guards (see below) is an indication of the
reduced conservation effectiveness.
Hypotheses for future reflooding
• Floodplain rehabilitation hypotheses should not only be based on water management but include feedback mechanisms of conservation effectiveness as well.
Pastoralist Responses
Impact of Maga dam
• Only scarce qualitative information is available on the reduction in grazing intensity following the Maga dam construction. One may assume a rapid drop in
276
Part IV – Synthesis
cattle grazing intensity by a reduction in length of stay in the area impacted by
the Maga dam. If production of the perennial vegetation did indeed cease abruptly, many pastoralists may have had to decide already at the end of the rainy season to direct their herds South (as during the 1985 drought, Chapter 7). In any
of these cases, the grazing intensity decline will have taken place faster than
the recovery process.
Recent changes and impact of reflooding
• From 1992 to 2000, we observed a doubling of the number of cattle herds, twothirds of which was due to newcomers and one-third to ‘natural’ increases based
on increased floodplain productivity due to reflooding and favourable rainfall.
• The average time cattle herds were present in the reflooded area increased by
60%.
• The resulting grazing intensity, expressed as cattle density, almost tripled
between 1993 and 1999.
• The increase in the number of herds and camps slowed down after 1996. Grazing intensity continued to increase linearly however, showing density independence, following the gradually recovering perennial grasslands. It is expect-2
ed that this increase has levelled off at a maximum of about 100 cattle km
(≈ 70 TLU) around 2003.
• I attribute the quasi-totality of the observed changes to the reflooding, the only
factor that differed between the studied floodplain and surrounding areas from
where pastoralists moved away.
Interpretations
• No ‘territorial blocking’ nor ‘chaotic overshoot’ scenario has taken place with
the reflooding. The Ideal Free Preemptive Distribution scenario, assuming that
any increase in perennial vegetation is consumed by cattle with the first individuals using the territories pre-empting them, best explained pastoralist responses to floodplain rehabilitation. Pre-emptive use might be reflected in the
customary rights that pastoralists have to campsites, which they have occupied
for a long time.
• One of implications of the Ideal Free Pre-emptive distribution is that with reflooding, apart from a short lag in pastoralist reaction time, all additionally produced forage is consumed, leaving few incentives for pastoralists to refrain from
grazing inside Waza National Park.
Hyptheses for future reflooding
• I postulate the Ideal Free Preemptive Distribution scenario for future reflooding of the area. Unfortunately, this scenario predicts continued grazing of livestock inside Waza National Park, unless strict measures are taken. This scenario is conditional on reflooding being carried out in steps with comparable
size of impact area. If the impact size is much larger, as proposed by IUCN
12 – Synthesis
277
(1999), the time-lag of the arrival of new groups may be somewhat longer than
observed during the 1994 reflooding.
12.2
Comparison of reflooding respronses
Floodplain rehabilitation may lead to unexpected changes in land use. Quickly
recovering species will have a distinct advantage over slowly recovering competitors, driving the latter into a secondary role or even towards their local extinction.
In the Waza-Logone floodplain with competing land uses, speed of response has
therefore a profound influence on their relations. One may argue that area’s continuing resilience (see 12.4) will lead to corrections in the long run. With the virtually irreversible changes in hydrology and landuse, exchanges between wildlife
populations in the Lake Chad Basin have become very limited, however, hampering any recolonisation.
To allow a comparison of the 1993-2000 (-2003) developments of perennial grass
cover, numbers of Black-headed Heron nests, dry season numbers of waterbirds,
Kob antelope numbers and cattle grazing intensity, I indexed these parameters to
their 1993 pre-reflooding situation (Fig. 12.1). I consider the development of perennial grass cover and vegetation production as reference lines for the floodplain
ecosystem recovery.
This comparison shows that the recovery of human-supported resources such as
cattle grazing intensity and villagers’ protected heron nests, exceed the increase in
resources with which humans compete, i.e. Kob antelopes and large piscivorous
waterbirds. Human populations have been quick to respond to the improved conditions, out-competing wildlife (Fig. 12.1). The isolated Waza-Logone population
of 4000 Kob antelopes had a limited capacity to increase, at most at an annual rate
of 50% i.e. 2000 antelopes. The ‘reservoir’ of 4.7 million cattle in the Convention2
al Lake Chad Basin (Martin et al. 1996) of which 200 000 cattle visit the floodplain has, however, a potential of increasing cattle grazing intensity in the floodplain many times over, by redirecting migration routes. A number of species that
compete with people for the same resources (e.g. piscivorous birds) have also
come under increased pressure, not only because of the increased competition (as
indicated by the number of fishermen, Chapter 8) but also because of direct persecution (Chapter 5).
2 Comprising Nigeria, NE of Damaturu, Cameroon north of Guider, Chad in a line of (Figuil) PalaLai-Melfi-Moussoro-Mao, and extreme SE Niger.
278
Part IV – Synthesis
400
Increase (100=1993)
heron nests
300
cattle
density
perennial grass cover
200
waterbirds
Vegetation production
Kob
100
large piscivorous
birds
0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
start reflooding
(0--> 20 m3/s)
increase reflooding
(20 --> 30 m3/s)
flooding season
Notes:
1. From 1999 onwards, development of perennial grass cover is based on transect data
2. The total number of these parameters increased significantly over time (GLM, F = 31.1 p < 0.0001,
Df=1,49), their relative increase over time differed as well (groups*time F = 6.083, p > 0.001, Df =
2
2
6,43). One should take into account that the impact areas cover with 180 km and 600 km respectively, only parts of the area where waterbirds have been counted. A comparison of the ecosystem
components that were monitored on approximately the same scale gives, however, comparable
-2
results. Cattle grazing intensity (cattle km ) with Kob antelope numbers: significant increase in
time (F = 6.8 p = 0.026, df = 1,10), and significant difference amongst components: (F = 12.7, p =
0.006, df = 1,9). Numbers of waterbirds with heron nests and numbers of piscivorous birds: significant increase in time (F = 8.7, p = 0.008, df = 1,21) and significant differences between components
(F = 4.4, p = 0.027, df = 2,19).
Figure 12.1 – Development of Waza-Logone Resources, 1993-2003.
12.3
Enhancing the integration of conservation and development
Human population and land use dynamics
The Waza-Logone floodplain rehabilitation has led to the partial, yet substantial
3
recovery of its main land use types (fisheries, pastoralism and conservation), truly
the success story of the Waza-Logone project and worth the investments (Chapter
3-7; IUCN 1999; Emerton 2004). However, already strained relations between
pastoralists, fishermen and farmers have come under further pressure, leading to
violent clashes during the first reflooding years (Kari & Scholte 2001). For changing relations between farmers and fishermen one is referred to Kouakam et al.
3 As explained in the preface, not treated in this thesis. See, however, Loth 2004.
12 – Synthesis
279
(2004). Also within the same category of land use, different groups have benefited
differently. For example Mare’en pastoralists, most prominent prior to the reflooding, benefited less than Arabs and agro-pastoralists (Chapter 7). This has, as far as
we are aware of, not led to tensions between these groups, probably because of
their long existing and as yet little disputed power relations.
More problematically, the rise of human population numbers at the borders of
Waza National Park has put the conservation – development balance under pressure (Chapter 8) leading for instance to increasing domestic grazing inside the
national park. One might argue that Waza-Logone is an exception because it has
experienced a degradation process with subsequent emigration. Experiences with
development activities around conservation areas becoming poles of attraction for
people have, however, been reported in an increasing number of cases (Chapter
8; ABCG 2004).
In the long term, people may become convinced of the need to refrain from exploiting resources from inside the protected areas. A population build-up may potentially even have advantages in awareness building. There does not seem to be an
analogy with cases of positive influences of population density on land management (‘more people, less erosion’ Tiffen et al. 1994) unless one would accept to
follow the path of the developed countries with the loss of species which may be
reintroduced subsequently.
Wildlife policies: developing instruments for Conservation & Development
At present, no politically or ethically viable alternatives for conservation & development approaches exist (Adams & Hulme 2001; Chapter 8). Most of the interest
in conservation by authorities and politicians in Cameroon and elsewhere, even
though still limited, is due to the activities of conservation & development projects. Continuing on the path of integrating conservation with development is further justified because present failures can be largely attributed to the poorly developed instruments in the field. It is as one says ‘If your only tool is a hammer everything looks like a nail’.
One of the most promising tools explored in Waza-Logone was social fencing,
where local communities exclude ‘outsiders’ from exploiting the surroundings of
a protected area (Chapter 8), an approach that is no longer taboo (Anonymous
2004). This was part of management agreements between protected area authorities and local communities, where a controlling and information duty of local communities was rewarded by exclusive exploitation rights as well as other privileges
such as a small-scale development assistance (Chapter 9). The pursued ‘carrot and
stick’ approach towards local communities was another promising instrument
applied for conservation with development. New opportunities (‘carrots’) were created for park border villages that were targeted for small-scale development activi-
280
Part IV – Synthesis
ties, whereas Baram, the village inside Waza National Park, was actively deprived
of such opportunities (‘stick’). A local initiative resulted in Baram being moved
thus breaking a deadlock of tense relations with the park authorities and assuring
a new future for the village (Chapter 8).
The convention with the pastoralists from Fadaré, at 40 km from Waza National
Park, showed the conservation potential of development activities at a certain distance from the protected area (Chapter 7). For two years this convention reduced
livestock pressure on Waza NP, but the rapidly increasing grazing intensity in the
floodplain and difficulties to monitor the convention with a limited number of
guards (see below) made this tool less successful in the longer term.
The implementation of the management plan has been successful on a local level
as the greatly improved relationship between park authorities and local communities testified. Lack of commitment of especially the Cameroonian government
has led, however, to a decline in the number of park guards that compromised the
momentum of gained confidence between protected area authorities and local communities (see next section). These experiences should be taken into account in other
protected areas in Cameroon and Central Africa, where consensual management
planning has just started. More emphasis should be put on the design and implementation of locally adapted arrangements in which not only local people but also
local and central governments should keep their side of the bargain.
Capacity building of wildlife managers
Protected area personnel such as wardens and guards generally have little input in
protected area planning and community conservation. Their lack of capacity in these
disciplines was not unique to Waza Logone, but also identified as a significant constraint in a WWF-portfolio of seven ICDPs in Gabon, Ethiopia, Ecuador, Honduras,
Pakistan and the Philippines (McShane & Newby 2004). The need for capacity
building in these disciplines motivated the development of a training programme
at the Garoua Wildlife College, the first of its kind in West and Central Africa. Developed courses aimed at raising knowledge and skills of protected area personnel
to support the dialogue with local communities enhancing the conservation of
Waza-Logone and other protected areas in Cameroon and West-Central Africa.
Experience with these courses showed the potential to upgrade the knowledge and
skills of present personnel that are not attitude-limited as often supposed.
The study on the curriculum development at the Garoua wildlife college in Cameroon showed only gradual changes contrasting with the rapidly changing market
demand that calls for more attention to community conservation and other subjects that reflect present park management objectives. Garoua’s curriculum differs
from the other African regional wildlife colleges at Mweka (E.Africa) and SAWC
(S.Africa) in its limited attention to community conservation and business skills.
12 – Synthesis
281
Erosion of the Cameroonian wildlife sector
Poor human resource management in the Cameroonian wildlife sector has overshadowed many of the reported results of management planning and training.
Within four years following the first refresher courses the number of Waza park
guards had dropped from 20 to 10, due to retirement, illness or change of posts
(Fig. 2.4). The number of guards in the northern province dropped less dramatically but also experienced ageing and lack of post rotation. Personnel management has captured our attention since the first discussions with the officials of the
Human Resources department of the Ministry of Environment and Forestry (Chapter 9), but can be largely attributed to a lack of government commitment (van Est
& Scholte 2001). The proposed institutional changes of the wildlife department
into a parastatal (Director of Wildlife, pers. comm. 2002), inspired by the Kenya
Wildlife Service model (Leakey & Morell 2001), has the potential to change the
precarious personnel situation. The forthcoming recruitment of 600 park guards
(Director of Wildlife, pers. comm. 2001 repeated in subsequent years) may be an
outstanding opportunity for preparatory training in new management techniques.
In contrast with this lack of government commitment towards biodiversity conservation was the protection of the area against armed robbery and cattle theft. At
the start of the reflooding in 1994, pastoralists expressed their concern about the
problematic security situation that threatened to block access to the newly reflooded area (Chapter 7; Scholte et al. 1996b). This situation improved dramatically after
1998 when the government sent special troops to control the situation (Chapter
8). This shows that the difficulties with conservation did not only result from government weakness, but ultimately from a lack of government commitment.
12.4
Empirical concepts of ecosystem change
In the sections 12.1 and 12.2, I synthesised and compared the changes of the individual Waza-Logone ecosystem components, followed by an overall comparison.
In this section I will discuss approaches that aim to explain ecosystem changes
from a whole-system perspective. Experiences from the Waza-Logone rehabilitation suggest that the time-dimension should play a prominent role in these approaches. In addition I argue that some of these ‘holistic’ approaches do not help
identify causal relations between change agents and the subsequent ecosystem
responses. They should therefore not be considered as models of change. Their role,
as metaphor, lies largely in facilitating communication on theories and insights.
Exploring the holistic panarchy concept for Waza-Logone
Panarchy is a holistic concept that aims to help explain developments of ecosystems as well as societies (Gunderson & Holling 2002). Two features distinguish
panarchy from traditional hierarchical models (Holling et al. 2002). The first is its
282
Part IV – Synthesis
elementary unit, the adaptive cycle. The second feature of panarchy concerns the
manner in which elements of complex adaptive cycles nest to one another in a
hierarchy (Holling et al. 2002).
I applied the hierarchical concept to illustrate the different recovery speeds and
scales at which the floodplain rehabilitation has taken place, see Fig. 12.2. In addition, Panarchy allows putting the importance of the main historic events, introduced in Chapter 2, in a wider perspective, thus highlighting the evolving reference image (Gunderson et al. 2002). Periods of collapsing ‘panarchies’ encompassed the late 1800s that were hit by a cascade of events (rinderpest, elephant
slaughter, war and colonisation) that still determine the present state of the area
(Chapter 2). A second downward spiral occurred after the construction in 1979 of
the Maga dam, immediately followed by the 1980s droughts, a drop of investments in public services and, as some would argue, a collapsing state. These developments took place against a background of an increasing population pressure,
not only in the area itself but with surrounding regions exercising a push of
human migration as well.
= adaptive cycle
3
1000
= floodplain rehabilitation monitoring
= intervention
Elephant/Woodland
cycle
wet dry spells
Time (year)
institutional cycle
2
100
1
10
0
(log)
political
Waza mgt.
veg.
comp.
recovery
impact of
Maga-dam
pastoral
recovery
mgt.
veg. pro- planning
duction
recovery
waterbird
recovery
curr.
development
long
courses
1
veg.flood
pilot
release
refreshment
course
1
100
1000
10.000
100.000
0
2
3
4
5
Figure 12.2 – The Waza-Logone Panarchy
(log)
Area (km2)
283
12 – Synthesis
One may expand this conceptual framework by adding to the factors time (T)
-2
[years] and spatial (S)- scale [km ], a ‘system level’ (L) that represents the delay in
response to reflooding, or its impact on ecosystem components and activities,
based on levels of production (e.g. 1 = primary, 2 = secondary), communication,
etc. in a hierarchy (see Table 12.2).
The recovery of perennial vegetation, waterbirds, wildlife and pastoral use, characterised by its direct link with the reflooding (L ≤ 2) has taken place in a period
of 3 to 10 years (Fig. 12.2; Table 12.2). Although not without surprises (Chapter 3),
it has been possible to forecast most of its impact. The short training courses were
of limited impact in a relatively short period only (< 1 year), due to the rapidly
degrading personnel situation. The effects of the long-course training, let alone of
curriculum development, took more time before being felt (Fig. 12.2). The many
steps in its implementation (L ≥ 4, Table 12.2), has made it rather complicated to
forecast their outcome.
Table 12.2 – Hierarchical levels of ecosystem component’s response to reflooding
System level
Abiotic
Hydrology
Grassland
production
1
2
3
4
commitmentimplementation
primary
production
composition
change
Waterbirds
waterbird
responses
Antelopes
antelope
responses
Pastoral use
pastoralist
responses
Management
planning
diagnostic,
dialogue
negotiations
↓
trainer
design
Curriculum
6
Change agent
Grassland
composition
Training course
5
implementation
↓
design
development
department
wildlife
college community
284
Part IV – Synthesis
One of the lessons learnt is that training people the moment one is confronted
with problems is too late. This may seem ‘an open door’, yet it is common practice with conservation & development interventions (Chapter 11). Refresher courses should build on proper pre-recruitment training as earlier argued by Lindley
(1999) for extension professionals. This is an additional argument for the need for
sound pre-recruitment training and the presence of institutions that can take care
of this (Chapter 10).
Box 12.1 – Vegetation states and (cyclic) succession in Waza-Logone
Three states (sensu Westoby et al. 1989) can be distinguished for the present WazaLogone floodplain vegetation. State A, the ‘pre-dam’ floodplain state, has a diverse
perennial grass vegetation, including rhizomatous grasses (Echinochloa pyramidalis
and Oryza longistaminata) and tussock grasses (Vetiveria nigritana). This state can
still be found in the area that has continued receiving floodwater after the Maga
dam construction (Chapter 3). State A has a somewhat perpetuating nature (Rietkerk
et al. 2002), the numbers and the condition of Vetiveria nigritana have for example
remained remarkably constant throughout a period of relative drought (mid-1980)
and more abundant rainfall (study period), Fig. 3.2. This does not exclude long-term
changes such as those that occurred in the density of Vetiveria nigritana tussocks in
the late 1970s when decades of relatively humid conditions were halted by a
drought (12.1).
State B, derived from State A after the dam construction, has different appearances,
due to the cyclic (Gleasonian) succession that takes place (Fig. 12.3). With the reflooding, Sorghum arundinaceum annual grasses were replaced by short living perennial
grasses and subsequently by rhizomatous perennial grasses. Such cyclic succession
also occurs in Dinder National Park (Central Sudan) where the same range of annual
to perennial grass communities were found in neighbouring backswamps (pers.obs.
Feb. 2005). The flooding conditions of these backswamps vary every few years,
depending on the shape and therefore discharge capacity of the feeders from the
river. Van der Valk (1992) analysed this Gleasonian succession for environments with
more predictable cyclic circumstances.
State C, on the edge of the floodplain, is dominated by annuals herbs. It can be considered a further ‘degeneration’ of State B, where even in the presence of (late)
flood water no further changes in vegetation were observed.
The variation in plant species within the three states, most notably B, is larger than
differences between the three states. This is particularly important for the interpretation of ecosystem functions, such as providing forage for grazers.
285
Above-ground Biomass – (Nutrient Status?)
12 – Synthesis
Predam State A
Rhizo.grasses +
Vetiveria
nigitana
State B. Cyclic succession amongst:
Echinochloa /
Oryza rhizo.
grasses
Panicum / Ischaemum
per. grasses
no changes
observed
Dam catastrophe
Sorghum
annual grasses
State C annual
herb vegetation
Max. flood depth
Figure 12.3 – Vegetation States and cyclic Succession in Waza-Logone (1993-2003)
Hysteresis and resilience
Ecological restoration was defined as the complete and functional return to a predisturbance state, whereas rehabilitation has often been defined as the only partial structural and functional return to a pre-disturbance state (Tapsell 1995; Middleton 1999). The 2000-2003 reflooding ‘end’ result was different from the 1979
pre-dam situation with regard to the area’s hydrology, vegetation composition, the
ratio of numbers of livestock and kob antelopes and, probably, the numbers of
large piscivorous birds compared to the number of fishermen. As shown in section 12.2, land use has shifted into a direction in which conservation has lost some
of its (relative) pre-dam importance. I therefore conclude that the 1994 reflooding
induced the rehabilitation and not the restoration of the floodplain. The WazaLogone results are not less profound, however, than the results of many ‘restoration’ projects, such as those envisaged by the Everglades Restoration Plan (Sklar
et al. 2001).
The end results of the reflooding, discussed in section 12.1, are obviously of importance, but we should not neglect the rehabilitation process itself. A central concept
in restoration ecology is hysteresis. Hysteresis, etymologically ‘coming behind’, has
its background in physics, where it is the ‘retardation of the effect when the forces
acting upon a body are changed. Hysteresis represent the history dependence of
(physical) systems: If you push on something, it will yield: when you release does
286
Part IV – Synthesis
it spring back completely? If it does not it is hysteresis in some broad sense’ (Setna
2004). In ecology hysteresis is used in a more holistic sense, emphasising the
word ‘completely’ in above description. This results in the concept that ‘that forward and backward switches [of the changes into a state] occur at different critical
conditions’ (Scheffers 2001).
With the reinstatement of flooding, tussock grasses and floodplain antelopes have
experienced a slower, if any, recovery than the dramatic drop in their population
numbers that took place following the Maga dam construction (12.1). One may
speculate on the entirely different conditions that should reign to cause, if possible, a switch of the vegetation into its genuine pre-dam state (Box 12.1), such as
increased sediment deposits, triggered by water flows feeding the western part of
the floodplain (Chapter 3). For the antelopes this may hold for a collapse of livestock grazing intensities or protection measures not only more effective than at
present but also more than when the population drop occurred. The hysteresis of
these vital resources in the floodplain system recovery has led to a different land
use state outcome of which the consequences were discussed in section 12.2.
Resilience, in its broadest sense the ability of a system to persist in an uncertain
world, has intrigued ecologists and increasingly social scientists (Perrings 1998).
Resilience was initially defined as ‘the ability to experience change and disturbance without catastrophic qualitative change in the basic functional organisation’
(Holling 1973), and more recently as ‘the long-term adaptive property of the system to withstand external perturbations without changing its basic structure’
(Holling 1995). Resilience also refers to ‘...the maximum perturbation that can be
taken without causing a shift to another ecosystem state’ (Scheffer 2001). Resilience
is not necessarily constant during the adaptive cycle (Holling & Gunderson 2002).
Populations are more vulnerable when under stress, as became clear with the cascade of events (drought, rinderpest) following the Maga dam construction (12.1).
Pimm (1984) follows an alternative approach in defining resilience as ‘the time
taken for a system to return to its state prior to disturbance’. The concept of resilience appears thus to have different meanings (Perrings 1998; Van den Berth
et al. 2004), of which Pimm resilience may be easier to measure (Batabyal 1998)
and is closer to the original meaning of hysteresis as simple time lag between
cause and effect (see above). Holling resilience is the most frequently being used,
reflecting more lasting changes (e.g. Levin et al. 1998). Below I discuss both concepts of resilience for the Waza-Logone floodplain in the light of the changes induced by the construction in 1979 of the Maga dam and, since 1994, the reflooding.
Local people argued that in 1988, a year of abundant rainfall, the flooding in WazaLogone resembled the floods in the years before the construction of the Maga dam
12 – Synthesis
287
in 1979 (Chapter 3). Also in 1994, the first reflooding year, the flooding appeared
once more as its pre-dam stage. The dam has ‘only’ changed the frequency of
flooding in the study area, well within the area’s (hydrological) resilience. Yet following the Maga dam construction ‘catastrophic changes’ (Holling 1973; Rietkerk
1998) of the ecosystem have taken place indicating that the area’s resilience has
been exceeded, switching from a productive floodplain into a desiccated unproductive state (Chapter 12.1).
In the ten years following the reflooding, the perennial grassland vegetation rehabilitation in the reflooding impact zone was virtually complete, albeit with a different species composition than before the dam construction (Box 12.1, Chapter 3).
This rehabilitation took place in a period with relatively favourable rainfall, when
perennial grasslands generally expand into the annually grass dominated drier
parts of the floodplain (r-phase in the adaptive cycle). In the areas further west,
only partly reached by the reflooding, a vegetation occurred that showed only an
increasing dominance of annual herbs (Figure 12.3). Because the pre-dam state (A
in Fig. 12.3) may not have been reached, I conclude that Pimm resilience is (much)
more than ten years. If neglecting the species composition and only considering
perennial grass cover as target for floodplain rehabilitation as we initially did
(Wesseling et al. 1994), one may conclude that Pimm resilience is about ten year
(Fig. 12.4: target image range). Pimm resilience allows a comparison with other
areas where rehabilitation trials were conducted, indicating the high resilience of
the Waza-Logone floodplain compared to other areas (Chapter 3).
Characteristics that explain the resilience of the seasonally flooded grasslands
include the high below-ground biomass (± 70% of total biomass) that is inaccessible to grazing animals because of the firm vertisols and the six months of seasonal protection due to rainfall ponding and subsequent flooding. Drent & Prins
(1987) also argued that, because the herbivore is prisoner of its food supply, vegetation under ‘natural conditions’ is free from disturbance due to the herbivore.
Ideal Free Distribution is one the mechanisms that makes it rather unlikely that
herbivores exercise long-term disturbance (Scholte & Brouwer 2005).
The Lake Chad Basin showed earlier in the 1900s a remarkable resilience with the
recovery of its elephant populations (Chapter 2). A return to the ‘basic pre-disturbance state’ was obviously impossible with the changes in environment that have
taken place in the 70-year time-span. The initial increase in numbers of Kob
antelopes in response to the reflooding, but also the recent return of some individual waterbuck antelopes, showed the resilience of the Waza-Logone floodplain
area in maintaining an appropriate habitat. Pimm resilience, certainly much more
than ten years, shows that this ‘antelope resilience’ is limited however. The best
example of the area’s resilience became visible with the development of pastoralism in the reflooded area. Grazing intensities recovered in ten years to a pre-dam
288
Part IV – Synthesis
After
Range of reference image
with vs-without dam
Before
Range of target image
Holling resilience
Recovery time
(=Pimm resilience)
Dam
Reflooding
Note: The range of reference and target image represent their development during the study period (see text).
Figure 12.4 – Evolving States and with-vs-without comparison
cattle density as suggested by a comparison with cattle densities in other African
floodplains (Scholte & Brouwer 2005).
In analogy to the discussion on restoration versus rehabilitation (12.4.2), we may
conclude that the concept of a static or basic pre-disturbance state as used in both
the Holling (1973, 1995) and Pimm (1983) resilience concepts is difficult to apply.
A more realistic concept for resilience takes into account that changes in the predisturbance state have taken place, in analogy to what economists call ‘with-versus-without comparison’ in Cost Benefit Analysis (Gittinger 1982). Van den Berg
et al. (2004) developed a ‘with-without scenario’ for a simple linear succession in
marshland in Western Netherlands in case of a single instantaneous disturbance.
‘Evolving states’ may be a more appropriate concept that one may conceive as the
pre-disturbance state that has continued developing as could have been expected
in the absence of the disturbance. Imagine, for example, that the Maga dam was
not built and the area had developed with the continuing impact of factors such as
drought and poaching (Table 12.1). Almost certainly the cover of Vetiveria nigritana
would also have dropped from 1979 onwards (Fig. 12.4). This also holds for the
size of the floodplain antelope populations. Both Vetiveria nigritana and antelopes
would of course not have dropped to the same extent as with the presence of the
dam. Reference areas play an important role to assess these background changes,
in analogy to the monitoring of the reflooding (Chapters 3 and 6).
12 – Synthesis
289
Empirical concepts of ecosystem change, concluding remarks
Holling & Gunderson (2002) stated that ‘the adaptive cycle should be considered
as metaphor to help interpreting events and their gross causes, and not as a [causal]
theory’. This did not, however, stop them as well as others using panarchy as a theoretical framework for ecosystem and societal change (e.g. Gallopin 2002). Laplante
(2004) rightly questioned such straightforward interpretations towards a unique
ecological theory. The straightforward discussion on results of the reflooding in
section 12.1 has a relatively high causal content, contrasting the discussions on
panarchy. Not by coincidence I started this chapter by presenting conclusions
based on field data and brought them subsequently together in an overall comparison. Only afterwards a comparison was made with empirical concepts of ecosystem change that would have been of little value if not fed with field experiences.
The vegetation rehabilitation experiences, for example, showed that the concept of
target communities (‘pre-dam perennial grasslands’) has limits because of their
dynamics (Fig. 12.4). Ultimately we deal with individual species and their traits
and not vegetation communities that are important for rehabilitation (Bakker et al.
2000; Chapter 3). The concept of ‘State & Transition’ plays an important role in the
understanding of the dynamics of the system. At least in Waza-Logone, it has less
relevance to understand the services its provides.
To understand functional relations one has to study individual ecosystem components often targeting individual species. The somewhat ‘dull’ description and interpretation of the floodplain rehabilitation case becomes however much more
attractive when put in a hierarchical time and space dimension such as panarchy.
These holistic metaphors play a complementary role in communication and exchange of experiences, contributing to vision forming and ultimately to interpretations of ecological philosophy (Laplante 2004)
12.5
The key findings of this study and the ecosystem approach
A discussion on concepts of ecosystem change and management, should include
the normative ecosystem approach, formally endorsed by the fifth Conference of
Parties (COP-5) of the Convention of Biological Diversity in 2000 (www.biodiv.org/
decisions; Shepherd 2004). The 12 principles of the Ecosystem Approach reflect
the present state of ‘practical wisdom’ on ecosystem management, obviously an
improvement from the inflexible approaches that dominated till recently. Korn et
al. (2003) developed several criteria to create an enabling environment for the application of the ecosystem approach. Shepherd (2004) facilitated the implementation of the ecosystem approach by categorising its 12 principles into five implementation steps.
290
Part IV – Synthesis
The potential use of the ecosystem approach was recently voiced by the Director
General of IUCN:
‘The ecosystem approach has emerged as a core element that can bring the
widespread and diverse activities of this movement [IUCN] together, not by subsuming any individual activity, but by providing an orientation point. Anything
from the management of a local wetland to issues such as global climate change
can be aligned under the ecosystem approach. This allows us to speak in terms
and concepts that can influence an entire global debate... and indeed other sectors’ (Steiner 2004).
In restoration practice the principle ‘management must recognise that change is
inevitable’ is especially relevant. For conservation practice this holds for the principle ‘management should be decentralised to the lowest appropriate level’. As such
the ecosystem approach is an excellent framework for analysis, especially if reflections such as of de Groot (2001), Korn et al. (2003) and Shepherd (2004) are kept
in mind.
The Waza-Logone project was implemented well before the ecosystem approach
was conceived and operationalised. It remains speculative if its knowledge beforehand would have influenced the Waza-Logone approach. It is, however, the present lack of relevant examples that makes it difficult to interpret the present state
of the Ecosystem approach into practical recommendations. In 2002, the sixth
Conference of Parties therefore requested the submission of case studies and lessons-learnt on the development and implementation of the ecosystem approach
on national and regional levels (www.biodiv.org/decisions).
Below I will present the main lessons of this study, allowing the further development of the ecosystem approach, by first addressing the five key findings of the
present study, namely:
1 The impact of different response times of ecosystem components to the rehabilitation results (resulting in ‘rehabilitation instead of restoration’ and ‘win-loose
situations’).
2 The Ideal Free Pre-emptive distribution, probably reigning in other (tropical
floodplains) systems as well, makes it particularly hard to set resources aside
for conservation purposes (failure of convention with pastoralists on refraining
from grazing inside Waza NP to be compensated by reflooding benefits).
3 The risk of human population build-up at the border of protected areas, accelerated by the integration of conservation and evelopment (immigration timebomb).
4 Need to assure the commitment of all parties in consensual protected area management (lack of governmental commitment).
12 – Synthesis
291
5 Need to have adequate intervention capacities of all parties (training of protected area personnel in planning and community conservation)
Finding 1 has a clear link with principle 8 ‘Recognizing the varying temporal scales
and time-lags that characterise ecosystem processes, objectives for ecosystem
management should be set for the long-term’ (Table 12.2). Finding 2 shows the
difficulty to seek an appropriate balance between use and conservation of resources
(principle 10, Table 12.2). Finding 3 may be linked, if some adaptations are made,
to principle 3, ‘ecosystem managers should consider the short and long term effects
(actual or potential) of their activities including [on] adjacent and other ecosystems’. Findings 4 and 5 are somewhat related to principle 12 ‘The ecosystem approach should involve all relevant sectors of society and scientific disciplines’.
Findings 4 and 5 further provide input in the categories of measures proposed by
Korn et al. (2003), namely ‘institutional strengthening and co-operation’ and
‘Information – knowledge – capacity’.
In addition to above key findings, other comments linked to the 12 ecosystem
principles are made based on insights from this study (Table 12.2). Of these comments, I will in the last part of this chapter specifically address the ones on adaptive management, applied by Shepherd (2004) as implementation steps of the
ecosystem approach. Adaptive management is particularly relevant as overarching
framework for our experiences. It needs further development to become fully
operational, however.
12.6
Towards adaptive management of success
Introduction
As Lee (1993) puts it
‘Adaptive management is an approach to natural resource policy that embodies
a simple imperative: policies are experiments; learn from them... Adaptive management takes uncertainty seriously, treating human interventions in natural
ecosystems as experimental probes. Its practitioners take special care with information. First they are explicit about what they expect, so that they can design
methods and apparatus to make measurements. Second they collect and analyse
information so that expectations can be compared with actual... Finally, they
transform comparison into learning – they correct errors, improve their imperfect understanding, and change action and plans. Linking science and human
purpose, adaptive management serves as a compass for us to use in searching
for a sustainable future’.
292
Part IV – Synthesis
Table 12.2 – Findings of this study as elaboration upon the principles of the Ecosystem Approach
No Ecosystem Approach Principle
Findings of this study
1
The objectives of management of land, water
and living resources are a matter of societal
choice
True, although the term ‘societal’ leads to different interpretations (e.g. de Groot 2001)
2
Management should be decentralised to the
lowest appropriate level
‘Lowest appropriate’ may also mean higher-level
government (e.g. in land use planning and security
issues)
3
Ecosystem managers should consider the
effects (actual or potential) of their activities
on adjacent and other ecosystems
Human population build-up should be included in
this principle
4
Recognising potential gains from management
there is usually a need to understand and
manage the ecosystem in an economic context
Economic context should be broadly interpreted
and does, for example, not necessarily imply a
monetary context
5
Conservation of ecosystem structure and
functioning, in order to maintain ecosystem
services, should be a priority target of the
ecosystem approach
‘Conservation’ should be broadly interpreted and
include rehabilitation
6
Ecosystems must be managed within the limits
of their functioning
‘Ecosystem’ should be defined broadly such as to
enable removal of undesirables states (eg. annual
herbs State C Box.12.1)
7
The ecosystem approach should be undertaken
at the appropriate spatial and temporal scales
True, varying from village level to Lake Basin
levels. Care should be taken to be flexible to take
into account new insights on what appropriate
scales are (see 12.6)
8
Recognising the varying temporal scales and
lag-effects that characterise ecosystem
processes, objectives for ecosystem management should be set for the long term
Very true, different response times will make short
term ‘restoration’ objectives out of reach
9
Management must recognise that change is
inevitable
Very true, hence the need for adaptive management, guided by a long term vision (12.6).
10 The ecosystem approach should seek the
appropriate balance between and integration
of conservation and use of biological diversity
True, but Ideal Free Distribution disputes the ‘balance’ concept.
11 The ecosystem approach should consider all
forms of relevant information, including
scientific and indigenous and local knowledge,
innovation and practices
Note the importance of historic data that complements and sometimes even contradicts ‘scientific’
and local knowledge (case of elephants, Chapter
2).
12 The ecosystem approach should involve all
relevant sectors of society and scientific
disciplines
– Sometimes, relevant sectors should be excluded
(see case of park village)
– Need to assure the commitment of ‘all’ parties
in consensual protected area management
– Need to have adequate intervention capacities
of ‘all’ parties
12 – Synthesis
293
Following Kessler (2003), the above definition of adaptive management runs the
risk of leading to a sum of small adaptations that may end, finally, in an undesired
situation. Adaptive management therefore needs to be guided by a long-term vision.
Adaptive management then entails the use of monitoring for continuous adaptation plus inputs for reflection for the identification, re-identification and work
towards this long-term vision.
In this section, I review the management approach that was pursued in the reflooding experiment in the light of adaptive management principles. In particular
I want to discuss which insights obtained from monitoring have led to changes in
the field. For this purpose I distinguish between three levels of abstraction: concrete actions, planning, and targets. I will first address the difficulty to monitor in
a project context beyond a time-span of five years. I will subsequently illustrate the
impact of monitoring on concrete actions and policies and on planning. A revisit
of the earlier discussed hypotheses for future reflooding allows the development
of an overall vision of floodplain rehabilitation. Lessons from this approach allow
a further development of a planning framework for managed flood releases in
African river systems.
Monitoring, studies and concrete actions and policies
In 1993 we hypothesised that restoration of the floodplain environment and its land
use, to the pre-Maga dam situation, would take five years (Chapter 1; Wesseling et
al. 1994). A considerable part of the (financial) resources of the Waza-Logone project during the first two project phases (1992-1997) were allocated to the monitor4
ing of this scenario and to discuss and formulate larger-scale reflooding options.
The general appreciation of the reflooding in these years was overwhelmingly positive (Acreman & Pirot 2004). Difficulties to assure the necessary financial sources
for large scale reflooding caused the shift of attention of the Waza-Logone project
from 1997 onwards to more mainstream development interests such as smallscale projects for sanitation and agriculture (Table 9.4). Local communities as well
as project personnel lost interest in monitoring beyond a time span of 3-5 years in
which tangible results had to be presented and continuing financing had to be
assured.
Many obtained insights of monitoring were put into action, often even before
being formally reported. Decisions and subsequent actions taken on the management of the Waza-Logone floodplain can be summarised as follows:
4 Not restricted to hypotheses of resources analysed in this thesis. Fisheries and agriculture were
amongst the extensively monitored resources as well. Project monitoring also encompassed social
relations, including gender, see also Loth (2004).
294
Part IV – Synthesis
Wildlife policies
• Based on results from the developing grazing pressure, efforts were undertaken that resulted in the convention with pastoralists on refraining from grazing
inside Waza NP (Chapter 7)
• Based on the monitoring of impact of reflooding, the ‘social buffering policy’
was explored (Chapter 9).
• Based on results from village PLAs (Scholte et al. 1999b), the ‘carrot & stick
policy’ with regard to Baram village inside Waza National Park, was implemented (Chapters 8,9)
Capacity building
• Based on the observations on the poorly developed skills of park personnel, a
course in People-oriented Conservation and Planning was developed and
implemented, initially for Waza personnel (1996), later for personnel of other
protected areas as well.
• Experiences with these courses highlighted the need for curriculum and institutional reform of the Garoua wildlife training college.
Above-presented monitoring results of the reflooding years 1993-1997 were integrated in the management plan of Waza National Park (Chapter 9) and the progressing Cost-Benefit Analysis of large-scale reflooding options (Wesseling et al.
1994; IUCN 1999; Emerton 2004). Based on continued monitoring from 1998
till 2003, studies were undertaken outside the project framework. Reported in the
preceding chapters, the resulting progressing insights can be summarised as follows:
2
• In addition to the changed vegetation composition in an area of 180 km
2
(Chapter 3), the importance of water raising in an area of 600 km was clarified (Chapter 4). The increase in vegetation production explained the larger
than initially calculated impact of the reflooding on livestock grazing and fisheries.
• The importance of the generally increasing trend of waterbirds recovery from
the 1980s drought was highlighted by the comparison with developments in
waterbird populations elsewhere in West Africa (Chapter 5). Most probably, the
same holds for other wildlife, including floodplain antelopes, but this has been
difficult to show due to the lack of reference areas.
• The divergence of developments of floodplain resources from 1997 onwards
(Figure 12.1, see above; Chapter 8) reinforced the necessity of proposed actions
in the management plan.
These decisions and actions were undertaken to meet the overall goal of the floodplain rehabilitation in achieving both sustainable development and biodiversity
conservation. Progressing insight showed the need to pursue corrections to plan-
12 – Synthesis
295
ning, to the rehabilitation targets and to the overall vision on floodplain rehabilitation that will be discussed below.
Monitoring and planning
Probably the most far-reaching decision on floodplain rehabilitation, motivated by
the difficulty to translate initial survey results into actions in the field, was to implement the 1994 reflooding (Chapter 1). This ‘pilot release’ became the identity
of the Waza-Logone project and it is difficult to imagine that it had initially not
even been programmed. Based on the preliminary monitoring results of this first
reflooding, subsequent negotiations and the resulting consensual decision, a follow-up of the pilot reflooding was implemented by an opening of a second watercourse, the Areitekele, in 1997.
Important, yet expensive, adaptations were made of the larger-scale reflooding
proposals, the so-called ‘main release’ (IUCN 1999). Monitoring showed that
although flooding water reached the more central parts of Waza National Park, its
arrival was delayed till well into the dry season. To allow the recovery of the more
central parts of Waza National Park, possibly including Vetiveria nigritana stands,
it was deemed necessary to imitate the original hydrographic pattern where floodwater reaches already during the rainy season the (western) floodplain.
An extrapolation of the preliminary reported benefits of reflooding for vegetation
production (Chapter 3) and especially livestock grazing (Chapter 7) played an important role as justification of the proposed large scale reflooding options. Livestock grazing for example was estimated to represent about 50% of the incremental benefits of the various reflooding options (Emerton 2004).
Monitoring and floodplain rehabilitation targets
A target image for future (large-scale) reflooding can be regarded as the ‘embodiment of insights of feasible policy and management goals along temporal
and spatial scales’ (Lenders et al. 1998).
The goal of the floodplain rehabilitation, voiced at the start of the reflooding, was:
‘The hydrological and ecological rehabilitation of the floodplain that involves the
restoration, as far as possible, of the physical conditions on the floodplain that existed before the construction of the SEMRY infrastructure in 1979’ (IUCN 1999).
Although not explicitly formulated prior to the start of the 1994 reflooding, this
5
goal has, rather surprisingly, not evolved over the years. The reflooding goal was
based on a single hypothesis: ‘back to the natural hydrology’, resembling the target for the Everglades restoration (Davis & Ogden 1994). The reformulated flood5 It should be mentioned that I was part of the project team preparing these formulations!
296
Part IV – Synthesis
plain rehabilitation hypotheses allow the adaptation of the overall floodplain rehabilitation goal, the subject of this section.
Below, I present some thoughts of how this floodplain rehabilitation target should
look like taking into account above-presented hypotheses (12.1) categorised into
specific temporal and spatial scales and reigning distribution models.
Different temporal scales:
• The recovery of tussock grasses lags several years behind the recovery of rhizomatous perennial grasses, which in their turn lag several years behind changes
in annual grasses’ composition and cover
• Dry season regrowth production lags several years behind the increase in wet
season above-ground biomass that in its turn lags two years behind the installation of reflooding
• The recovery of ‘human-competing’ resources, if taking place at all, lags behind
‘human-supported’ resources
• The impact of management planning lags several years behind immigration
and other developments described above
• The impacts of training lag behind management planning
• Institutional adaptations lag behind curriculum requirements
Different spatial scales:
• Waterbird (sub-) continent changes related to environmental conditions versus
site-specific situations such as protection.
• The recovery of livestock whose recovery is regionally determined versus the
recovery of antelope populations that is generally based on locally determined
factors only.
Different distribution models:
• For pastoralism, but possibly for other land uses such as fisheries as well, the
Ideal Free Pre-emptive Distribution model was postulated. This model predicts
that animals and plants will distribute evenly according to availability of
resources. This is a strong counter force against the reservation of resources
for conservation purposes.
From reference to target image and back: developing a vision for the
rehabilitation and management of African floodplains
The reference image(s) described in Chapter 2 was obviously not within the target
of a rehabilitation program. Progressing knowledge, obtained by the conduct of
historical and other surveys and the monitoring of the reflooding, has led to the
reformulation of the reference and target images (respectively Chapter 2 and 12.1).
Only after this first stage of reflooding probing (1993-1999), have we been able to
formulate more precise floodplain rehabilitation hypotheses and targets. These
297
12 – Synthesis
hypotheses and targets should be adapted based on progressing insights in the
future. In situations with poorly known reference situations and insufficient insight to formulate a target for rehabilitation, one should consider floodplain rehabilitation as an iterative process. One should pass not only from trial & error into
passive and subsequently active adaptive management (Walter & Holling 1990),
but also continue along in this reiterative spiral, passing on to future trial & error,
passive adaptive management stages. As hysteresis is almost always present, therefore prediction, even based on events in the past, will often be poor.
Integrating these insights in a planning framework for managed flood releases,
instead of the (one) circle framework proposed by Acreman & Pirot (2004), I propose the spiral as shown in Figure 12.5.
CHANGING THEORIES (reflected in empirical concepts)
adaptive vision
ACTIONS
continuous monitoring
CHANGING OUTSIDE WORLD
Figure 12.5 – Planning framework for flood releases
A vision encompasses not only the aspiration of (developing) targets but also
envisages the enabling environment in which rehabilitation and management targets can move. I express my vision for (further) floodplain rehabilitation in WazaLogone as follows:
Floodplain rehabilitation should be conducted in a probing approach with simple
and long-term (‘continuous’) monitoring that allows interventions and regular
redefinitions of targets. Because of expected imbalances in Conservation versus
Development, appropriate mitigating measures should be prepared beforehand,
e.g. by the designation of distinct areas for conservation and other competing land
uses. In addition one may consider zoning with exclusive rights for present local
communities that form a social buffer against the installation of newcomers. Formal agreements on the exclusive use of resources may facilitate such measures.
To enhance the efficiency of conservation, it will be important to increase the intervention capacity of protected area personnel well before major changes such as
those induced by floodplain rehabilitation, are undertaken. This often implies that
institutional arrangements have to be taken up early. In the long term one aspires
towards governmental commitment for conservation and local development in the
way they have been mobilised in Cameroon for security in the late 1990s. In addition one should strive to more coherence between local, national and international conservation policies and interventions.
References
ABCG (2004). Meeting on Migration and Conservation: Issues and Interventions. www.ABCG.org
(accessed Feb. 2005).
Abbott, A. (1996). Conservationists seek compromise over new threat to Galapagos. Nature
383: 3.
Acreman, M. and J.Y. Pirot (2004). Tools and practices for Floodplain Restoration. Pp. 125144 in Loth, P. (ed.) The return of the water. Restoring the Waza Logone floodplain in Cameroon. IUCN, Gland, Switserland.
Acreman, M.C. and G.E. Hollis (1996). Water management and wetlands in sub-Saharan Africa.
IUCN, Gland, Switzerland.
Adams, W.M. (1996). Economics and hydrological management of floodplains. Pp. 21-33 in
Acreman, M.C. and G.E. Hollis (eds.) Water Management and Wetlands in Sub-saharan
Africa. IUCN, Gland, Switserland.
Adams, W.M. and T. McShane (1992). The Myth of Wild Africa. Conservation without Illusion.
Norton & Company, New York, USA.
Adams, W.M. and D.Hulme (2001). If community conservation is the answer in Africa, what
is the question? Oryx 35: 193-200.
Allegret, M. (1993). Carnets du Congo. Voyage avec André Gide. CNRS Litterature, Paris, France.
Allo, A. (1978). The Garoua College of Wildlife Management. Parks 3: 8-9.
Alpert, P. (1993). Conserving biodiversity in Cameroon. Ambio 22: 44-49.
Amougou (2002). Rapport de la mission d’évaluation du plan d’aménagement du PN Waza. Internal Report, Waza, Cameroon.
Anderson, D. and R. Grove (eds.) (1989). Conservation in Africa. People, policies and practice.
Cambridge University Press, Cambridge, UK.
Anon. (1996). Campfire News, December 1996, Harare, Zimbabwe.
Anonymus (2004). Peering at the Future. Conservationists are, reluctantly, starting to get real.
The Economist June, 19th 2004.
ANZECC (2000). Working Group on National Parks and Protected Areas. Best practice in
Protected Area Management Planning. (available at www.pandora.nla.gov.au/tep/24147).
Ash, J.S. and J.E. Miskell (1998). Birds of Somalia. Pica press, Sussex, UK.
Azarya, V., A. Breedveld, M. de Bruin and H. van Dijk (1999). Pastoralists under pressure? Fulbe
societies confronting change in West Africa. Brill, Leiden, The Netherlands.
Bakker, J.P., A.P. Grootjans, M. Hermy and P. Poschlod (2000). How to define targets for ecologial restoration? Introduction. Applied Vegetation Science 3: 3-6.
Bancroft, G.T. (1989). Status and Conservation of wading birds in the Everglades. American
Birds 43: 1258-1265.
Banham, W. (1999). Modularisation: will this affect the way training is done: the experience of
Mweka. Unpublished report, SADC Wildlife Training Programme Training workshop,
Harare, Zimbabwe.
Barbraud, C. and H. Hafner (2001). Variation des effectifs nicheurs de hérons pourprés Ardea
purpurea sur le littoral Méditerranéen Français en relation avec la pluviométrie sur les
quartiers d’hivernage. Alauda 69: 373-380.
Barrat-Segretain, M.H. and C. Amoros (1996). Recolonisation of cleared riverine macrophyte
patches: importance of the border effect. J. Veg. Sci. 7: 769-776.
Barrett, C.B. and P. Arcese (1995). Are Integrated Conservation-Development Projects (ICDPs)
Sustainable? On the Conservation of Large Mammals in Sub-Saharan Africa. World
Development 23: 1073-1084.
Barrow, E. and M. Murphree (2001). Community conservation, from Concept to Practice. Pp.
24-37 in Hulme, D. and M. Murphree (eds.) African Wildlife & Livelihoods. James Currey,
Oxford, UK.
Barth, H. (1857). Reisen und Entdeckungen in Nord-und Central-Afrika in den Jahren 1849 bis
1855. Tagebuch seiner im Auftrag der Brittschen Regierung unternommenen Reise. Volume 1-5.
Gotha Justus Perthes. Translations in English taken from the first American edition: Barth,
H 1857. Travels and discoveries in North and Central Africa,being a journal of an expedition
undertaken in the years 1849-1855. New York: Harper and brothers, Volume 1-3.
Batabyal, A.A. (1998). The Concept of Resilience: retrospect and prospect. Environment and
Development Economics 3: 234-239.
Bates G.L. (1927). Notes on some birds of Cameroon and the Lake Chad Region: their status
and breeding-times. Ibis 3: 1-65.
Batisse M. (1982). The biosphere reserve: A tool for environmental conservation and management. Environmental Conservation 9: 101-111.
Bauer, H. (2003). Lion Conservation in West and Central Africa: the case of Waza National Park,
Cameroon. Thesis Leiden University, The Netherlands.
Beauvilain, A. (1981). Élevage et éleveurs du grand yaéré (Nord Cameroun). Revue de geographie du Cameroun 2: 163-176.
Beauvilain, A. (1989). Nord-Cameroun. Crises et Peuplements. Volumes 1 and 2. Thesis University of Rouen, France.
Beauvilain, A. (1995). Tableau de la pluviometrie dans les bassins du Tchad et de la Benoué.
Tableaux et documents scientifiques du Tchad. Documents pour la recherche III. CNAR,
N’djamena, Chad.
Beesley J.S.S. (1972). A nesting colony of Black-headed Herons Ardea melanocephala in
Arusha, Tanzania. Bull. E. Afr. Nat. Hist. Soc. 1972: 47-48.
Behnke Jr., R.H., I. Scoones and C. Kerven (1993). Range ecology at disequilibrium: new models
of natural variability and pastoral adaptation in African Savannas. Overseas Development
Institute, London, UK.
Beilfuss, R., E. Williams, T. Dodman (2003). Status survey and Conservation Action Plan for the
Black Crowned crane Balearica pavonina. International Crane Foundation, Wetlands
International, Wageningen, The Netherlands.
Benchley, P. and S. Franklin (2001). Galapagos. Paradies in Gefahr. National Geographic
Deutschland, February 2001.
Bergin, P. (2001). Accommodating New Narratives in a Conservation Bureaucracy. TANAPA
& Community Conservation. Pp. 89-105 in Hulme, D. and M. Murphree (eds.) African
Wildlife & Livelihoods. James Currey, Oxford, UK.
Bird, C. and S. Metcalfe (1995). Two views from Campfire in Zimbabwe’s Hurungwe District.
Wildlife and Development Series No. 5. IIED, London, UK.
Blokhuis, W.A. (1993). Vertisols in the Central Clay Plain of the Sudan. PhD Thesis Wageningen University, The Netherlands.
Blom, A. (2001). The Impact of Tourism on Protected Area Management and the Local Economy in Dzanga-Sangha (Central African Republic). J. Sust. Tourism 8: 175-189.
Blom, C.W.P.M. and L.A.C.J. Voesenek. (1996). Flooding: the survival strategies of plants.
TREE 11: 290-295.
Bomba, C.M. and P. Bigombo Logo (2001). Les Droits des Populations dans la Legislation Forestiere Camerounaise: Quels Acquis? Quelles Insuffisances? Bulletin Arbres, Forests et Communautres Rurales 22: 47-62.
References
301
Borrini-Feyerabend, G., M., Taghi Farvar, J-C. Nguinguiri and V. Ndangang (2000). La gestion
participative des ressources naturelles: Organisation, Négotiation et Apprentisssage par l’Action.
Kasparek Verlag, Heidelberg, Germany.
Boudet, G. (1983). Manuel sur les pâturages tropicaux et les cultures fourragères. Manuels et précis
d’élevage. IEMVT, Ministère de la Coopération. La documentation française, Paris, France.
Bouton, S.N. and P.C. Frederick (2003). Stakeholders’ Perceptions of a Wading Bird Colony
as a Community Resource in the Brazilian Pantanal. Conservation Biology 17: 297-306.
Braack, L. (1997). A revision of parts of the management plan for Kruger N.P. Vols. VII. & VIII
(www.parks-sa.co.za).
Braaksma, S., M.F. Morzer Bruyns and G.A. Brouwer (1950). Overzicht van de broedkolonies
van de Blauwe Reiger Ardea cinerea L. in The Netherlands. Ardea 38: 135-162.
Brabant, P. and M. Gavaud (1985). Les sols et les ressources en terre du Nord-Cameroun. MESRES/IRA-ORSTOM, Yaoundé/Paris, France.
Brandon, K.E. and M. Wells (1992). Planning for people and parks: Design Dilemmas. World
Development 20: 557-70.
Brandon, K. (1997). Policy and Practical Considerations in Land-use Strategies for Biodiversity
Conservation. Pp. 90-114 in Kramer, R., C.van Schaik and J. Johnson (eds.) Last Stand.
Oxford University Press, New York, USA.
Braukamper, U. (1996). Strategies of environmental adaptation and patterns of transhumance of the Shuwa Arabs in the Nigerian Chad Basin. Nomadic Peoples 39: 53-68.
Bredenbeek, J. (ed.) (2004). Wader and Waterbirds in the floodplain of the Logone Cameroon and
Chad, February 2000. WIWO report, Zeist, The Netherlands.
Breen, C.M., K.H. Rogers and P.J. Ashton (1988). Vegetation processes in swamps and flooded plains. Pp. 223-247 in Symoens, J.J. (ed.) Vegetation of inland waters. Handbook of
Vegetation Science 15/1. Kluwer, Dordrecht, The Netherlands.
Breman, H. and C.T. de Wit. (1983). Rangeland productivity and exploitation in the Sahel.
Science 221: 1341-1347.
Breman, H. and N. de Ridder (1991). Manuel sur les pâturages des pays sahéliens. Karthala, Paris,
France.
Bridgewater, P., A. Philips, M. Green and B. Amos (1996). Biosphere Reserves and the IUCN
System of Protected Area Management Categories. IUCN, Gland, Switserland.
Broadbent, J.A. (1971). Additions to the avifaunas of Waza (Cameroun) and Lake Natu (Sokoto). Bull. Niger. Orn. Soc. 8: 58-61.
Brockington, D. (2002). Fortress Conservation: The Preservation of the Mkomazi Game Reserve,
Tanzania. James Currey, Oxford, UK.
Brouwer, J. and W.C. Mullié (2001). A method for making whole country waterbird population estimates, applied to annual waterbird census data from Niger. Ostrich supplement 15:
73-82.
Brouwer, J., W.C. Mullié and P. Scholte (2003). White Storks wintering in Chad, northern
Cameroon and Niger: a comment on Berthold et al. (2001). Ibis 145: 499-501.
Brown, M and A. Singer (1991). La gestion des zones tampons en Afrique. PVO NGO/NRMS.
Unpublished report + video
Brown, L.H., E.K. Urban and K. Newman (1982). The Birds of Africa, Vol I. Academic Press,
London, UK.
Brown, M. and B. Wyckoff-Baird (1992). Projets Integrés de Conservation de la Nature et de
Développement. World Wildlife Fund, The Nature Conservancy and World Resources Institute. WWF publications, Baltimore, USA.
Bruner, A.G., R.E Gullison, R.E., Rice and G.A.B da Fonseca (2001). Effectiveness of Parks in
Protecting Tropical Biodiversity. Science 291: 125-128.
302
Floodplain rehabilitation and the future of Conservation & Development
Brunner J. and F. Ekoko (2000). Cameroon in the Right conditions Pp. 59-80 in Seymour,
F.J. and N.K. Dubash (eds.) The World Bank, Structural Adjustment and Forest Policy Reform.
WRI, Washington DC, USA.
Bruijn, M. de, I. van Halsema and H. van den Hombergh (1997). Gender and Land Use. Diversity in Environmental Practices. Thesis publishers, Amsterdam, The Netherlands.
Bruwer, C., C. Poultney and Z. Nyathi (1996). Community based hydrological management
of the Phongolo floodplain. Pp. 199-211 in Acreman, M.C. and G.E. Hollis (eds.) Water
Management and Wetlands in Sub-saharan Africa. IUCN, Gland, Switserland.
Bunting, A.H. and J.D. Lea (1962). The Soils and Vegetation of the Fung. J. of Ecol. 50: 529558.
Burnham, K.P., D.R. Anderson and J.L. Laake (1980). Estimating of density from line transect
sampling of biological populations. Wildl. Mon. 72: 1-202.
Burnham, P. (1996). The politics of cultural difference in northern Cameroon. Smithsonian Institution Press, Washington DC, USA.
Butler, R.W., J.A. Kushlan and I.J. Davidson (2000). Herons in North America, Central America and the West Indies. Pp. 151-176 in Kushlan, J.A. and H. Hafner (eds.) Heron Conservation. Academic Press, London, UK.
Caldecott, J. (1996). Designing Conservation Projects. Cambridge University Press, Cambridge,
UK.
Cannon, J.R., J.M. Dietz and L.A. Dietz (1996). Training conservation biologists in human
interaction skills. Conservation Biology 10: 1277-1282.
Casimir, M. J. (1992). The determinants of rights to pasture: territorial organisation and ecological constraints. Pp. 153-204 in Casimir, M.J and A. Rao (eds.) Mobility and territoriality : social and spatial boundaries among foragers, fishers, pastoralists and peripatetics. Berg,
New York, USA.
Catling, D. (1992). Rice in deep water. IRRI. The Macmillan Press Ltd, London, UK.
CCTA. (1953). Commission for Technical Co-operation in Africa south of the Sahara. Proceedings of the Third International Conference for the protection of the Fauna and Flora of Africa.
Bukavu, October 1953. Bruxelles, Belgium.
Cellot, B., F. Mouillot and C.P. Henry (1998). Flood drift and propagule bank of aquatic macrophytes in a riverine wetland. J. Veg. Sci. 9: 631-640.
Chambers, R. (1998). Afterword. Pp. 197 in Holland, J. and J. Blackburn (eds.) Whose voice?
Participatory research and policy change. IT Publications, London, UK.
Chambers, R. and J. Jiggins (1987). Agricultural research for resource poor farmers part I. Transfer of Technology and farming system research. Agricultural Administration and Extension
27: 35-52.
Cheke, R.E. and J.F. Walsh (1996). The Birds of Togo. B.O.U. Checklist No. 14. Tring, UK.
Child, G.S. (1976). Training Personnel for National Parks. Parks 1: 14-16.
Child, G. (1995). Wildlife and People: the Zimbabwean Success. How the conflict between animals
and people became progress for both. Wisdom Foundation, Harare, New York.
CIRAD-EMVT (1996). Atlas d’élevage du Bassin du Lac Tchad/Livestock Atlas of the Lake Chad
Basin 1997. CTA, Wageningen, The Netherlands.
Claridge, G. (1999). Protected Areas Management Planning. Tigerpaper 26: 15-17.
Clarke, P. (1999). Park Management Planning in Africa. Oryx 33: 281-282.
Connah, G. (1981). Three thousand years in Africa. Man and his environment in the Lake Chad
region of Nigeria. New studies in archeology. Cambridge University Press, Cambridge, UK.
Cracknell, B.E. (2000). Evaluating Development Aid. Sage Publications, New Delhi, India.
References
303
Crivelli, A., P. Grillas, H. Jerrentrup and T. Nazirides (1995). Effects on fisheries and waterbirds of raising water levels at Kerkini reservoir, a Ramsar site in N. Greece. Environmental
Management 19: 431-443.
Cronk, J.K. and M.S. Fennessy (2001). Wetlands Plants. Biology and Ecology. CRC Press/Lewis
Publishers, Boca Raton, Florida, USA.
Crowder, L.K., W.I. Lindley, T.H. Bruening and N. Doron (1998). Agricultural Education for
Sustainable Rural Development: Challenges for Developing Countries in the 21st Century.
The Journal of Agricultural Education and Extension 5: 71-84.
Cunnison, I. (1960). The Social Role of Cattle. The Sudan Journal of Veterinary Science and Animal Husbandry 1: 8-25.
Dahl, G., and A. Hjort (1976). Having herds: pastoral herd growth and household economy. Stockholm (Sweden): Department of Social Anthropology, University of Stockholm, Norway.
Davis, S.M. and J.C. Ogden (1994). Everglades. The Ecosystem and Its Restoration. St. Lucie
Press, Delray Beach, Florida, USA.
De Bie, S. (1991). Wildlife Resources of the West African Savanna. Wageningen Agric. Papers.
91-2. Wageningen, The Netherlands.
Dejace, P., L. Gauthier and P. Bouché (2000). Les populations de grands mammiferes et
d’autruches du Parc National de Zakouma au Tchad: Statuts et Tendances Evolutives. Rev.
Ecol (Terre Vie) 55: 305-320.
Denny, P. (1985). The ecology and management of African wetland vegetation. Junk, Dordrecht,
The Netherlands.
Denham and Clapperton. 1826 (1985 reprint). Narrative of Travels and Discoveries in Northern
and Central Africa in the years 1822, 23 and 24. Darf publishers limited, London, UK.
Denny, P. (1993). Wetlands of Africa: Introduction. Pp. 1-31 in Whigham, D.F., D. Dykyjova
and S. Hejny (eds.) Wetlands of the World: Inventory, ecology and management I. Kluwer Academic Publishers, The Netherlands.
Denny, P. (1994). Biodiversity and wetlands. Wetlands Ecology and Management 3: 55-61.
Denny, P. (2004). Wetlands are our life-blood. Valedictory Address UNESCO-IHE.
Dochy, F.J.R.C., M.S.R. Segers and G. Moerkerke (1997). The importance of prior knowledge
and assessment for increasing efficiency of the learning processes, especially in ‘problembased’ powerful learning environments. Pp. 73-96 in Charrier, A. and R. van Haarlem
(eds.) Biodiversity, an issue in Higher Education. Proceedings of the third European Conference
on Higher Education for Agriculture. Wageningen University, Wageningen, The Netherlands.
Dodman, T. and C.H. Diagana (2003). African Waterbird Census/Les dénombrements des oiseaux
d’eau en Afrique 1999, 2000, 2001. Wetlands Global Series 16, Wageningen, The Netherlands.
Dragesco J. (1960). Notes biologiques sur quelques oiseaux d’Afrique Equatoriale. Alauda 28:
81-92, 262-273.
Dragesco J. (1961). Oiseaux des savanes d’Afrique équatoriale. Oiseau Rev. fr. Orn. 31: 179-192,
261-271.
Drent, R.H. and H.H.T. Prins (1987). The herbivore as prisoner of its food supply. Pp. 133-149
in van Andel, J., J. Bakker and R.W. Snaydon (eds.) Disturbance in grasslands; species and
population responses. Dr. W. Junk Publishing Company, Dordrecht, The Netherlands.
Driessen, P.M. and R.Dudal (1989). Lecture Notes on the major soils of the world. Agricultural
University Wageningen and Katholieke Universiteit Leuven, Wageningen, The Netherlands.
Drijver, C.A. and M. Marchand. (1986). Taming the floods: environmental aspects of floodplain development in Africa. Nature and Resources 22: 13-22.
304
Floodplain rehabilitation and the future of Conservation & Development
Drijver, C.A., J.C.T. van Wetten and W.T. de Groot (1995). Working with Nature: Local Fishery
Management on the Logone Floodplain in Chad and Cameroon. Pp 29-45 in van der Breemer, J.P.M., C.A. Drijver and L.B. Venema (eds.) Local Resource Management in Africa.
Wiley, Chichester, UK.
Drijver, C.A. and E. Kouahou (1995). What a small dam can do: local initiative in the Logone
Floodplain, Cameroon. Pp. 179-184 in Roggeri, H. (ed.) Tropical Freshwater Wetlands. A
guide to current knowledge and Sustainable Management. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Dupuy, A.R. and P. Dupont (1984). Les guardiens de la vie sauvage. WWF-UICN, Gland, Switserland.
Dijkstra, A.J., W. Ganzevles, G.J. Gerritsen and S. de Kort (2002). Waders and waterbirds in the
floodplains of the Logone, Cameroon and Chad, January-February 1999. WIWO report 75,
Zeist, The Netherlands.
Dzingirai, V. (1996). Every man must resettle where he wants: The Politics of settlement in
the context of community wildlife management programme in Binga, Zimbabwe. Zambezia 23: 19-30.
East, R. (ed.) (1999). African Antelope database 1998. IUCN/SSC Antelope Specialist Group.
IUCN, Gland, Switserland and Cambridge, UK.
Ebohon, O.J., B.G. Field and R. Ford (1997). Institutional deficiencies and capacity building
constraints: the dilemma for environmentally sustainable development in Africa. Int. J. of
Sust. Devel. and World Ecol. 4: 204-213.
Egbe, S.E. (2001). The law, communities and wildlife management in Cameroon. Rural Development Forestry Network Paper 25e 1-11 Overseas Development Institute, UK. Available
at www.odi.org.uk/fpeg/publications/rdfn/25/rdfn-25e-i.pdf
Ekoko, F. (2000). Balancing Politics, Economics and Conservation: The Case of the Cameroon
Forestry Law Reform. Pp. 129-152 in Doornbos, M., A. Saith and B. White (eds.) Forests:
Nature, People, Power. Blackwell, Oxford, UK.
El-Ashry, M.T. (2001). Foreword. Pp x in Fishpool, L.D.C. and M.I. Evans (eds.) Important Bird
Areas in Africa and Associated Islands: Priority Sites for Conservation. Pisces publications and
BirdLife International (Birdlife Conservation Series No. 11), Newbury and Cambridge, UK.
Elgood, J.H., J.B. Heigham, A.M. Moore, A.M. Nason, R.E. Sharland and N.J. Skinner (1994).
The Birds of Nigeria. British Ornithologists’ Union, UK.
Ellenbroek, G.A. (1987). Ecology and productivity of an African wetland vegetation. Junk publishers, Dordrecht, The Netherlands.
Ellery, K., W.N. Ellery, K.H. Rogers and B.H. Walker (1991). Water depth and biotic insulation: Major determinants of black-swamp plant community composition. Wetlands Ecology
and Management 1: 149-162.
Emerton, L. (2004). Economic Justification of Additional Floodplain Re-inundation. Pp. 105121 in Loth, P. (ed.) The return of the water. Restoring the Waza Logone floodplain in Cameroon.
IUCN, Gland, Switserland.
Esser, J.D. and L.P. van Lavieren (1979). Importance, repartition et tendance evolutive des
populations des grands herbivores et de l’autruche dans le parc national de Waza, Cameroun. Rev. Ecol (Terre Vie) 33: 3-26.
Est, D. van (1997). The changing use and management of the floodplain environment by
Mousgoum women in North Cameroon. Pp. 9-26 in de Bruyn, M., I. van Halsema and H.
van den Hombergh (eds.) Gender and Land Use; Diversity in Environmental Practises. Thela
Publishers, Amsterdam, The Netherlands.
Est, D.M.E. van (1999). Vissen in Andermans Vijver. PhD thesis Leiden University, Leiden, The
Netherlands.
References
305
Est, van D. and P. Scholte (2001). Linking social and ecological diversity: organisations for
adaptive natural resource management in the Waza-Logone floodplain (North Cameroon).
Pp. 55-65 in Ali, M., P. Loth, H. Bauer and H. de Iongh (eds.) Management of fragile ecosystems in the North of Cameroon: the need for an adaptive approach. CEDC/CML, Leiden,
The Netherlands.
Estes, R.D. (1992). The behavior guide to African Mammals. The University of California Press,
Berkeley, USA.
Evans-Pritchard (1940). The Nuer. A description of the modes of livelihood and political institutions
of a Nilotic people. Oxford University Press, New York.
FAO (1977). Guidelines for soil profile description. FAO, Rome, Italy.
FAO (1988). FAO/Unesco Soil Map of the World, revised legend. World Resources Report 60.
FAO, Rome, Italy.
Fasola, M. and F. Barbieri (1978). Factors affecting the distribution of heronries in Northern
Italy. Ibis 120: 337-340.
Fasola, M., H. Hafner, J. Prosper, H. van der Kooij and I.V. Schogolev (2000). Population
changes in European herons in relation to African climate. Ostrich 71: 52-55.
Feron, E. (1997). Les processus participatifs de gestion et de conservation de la faune en Afrique Australe, ou la recherche d’un nouvel Eden. Le Flamboyant 42: 20-25.
Fischer, F. and K.E. Linsenmair (2001a). Spatial and temporal habitat use of kob antelopes
(Kobus kob kob, Erxleben 1777) in the Comoé National Park, Ivory Coast as revealed by
radio tracking. Afr.J. Ecol. 39: 249-256.
Fischer, F. and K.E. Lindsenmair (2001b). Decreases in ungulate population densitites. Examples from the Comoe National Park, Ivory Coast. Biol. Conserv. 101: 131-135.
Fischer, F. and K.E. Linsenmair (2002). Demography of a West African kob (Kobus kob kob)
population. Afr. J. Ecol. 40: 130-137.
Fishpool, L.D.C. and M.I. Evans (eds.) (2001). Important Bird Areas in Africa and Associated
Islands: Priority Sites for Conservation. Pisces publications and BirdLife International (Birdlife Conservation Series No. 11), Newbury and Cambridge, UK.
Fjeldsa, J., N.B. Burgess, S. Blyth and H.M. de Klerk (2004). Where are the major gaps in the
reserve network for Africa’s mammals? Oryx 38: 17-25.
Flizot, P. (1948). Les élephants des régions du Nord Cameroun et de la Bénoué. Mammalia 4:
148-151.
Flizot (1962). The Waza National Park in Northern Cameroun. African Wildlife 16: 293-297.
Flizot, P. (1962, 1968, 1974). Annual Reports. Provincial delegation, Garoua, Cameroon.
Foppen, R.P.B., C.J.F. ter Braak, J. Verboom and R. Reijnen (1999). Dutch Sedge Warblers
Acrocephalus schoenobaenus and West-African rainfall: empirical data and simulation
modelling show low population resilience in fragmented marshlands. Ardea 87: 113-127.
Fotso, R., F. Dowsett-Lemaire, R.J. Dowsett, COC, P.Scholte, M. Languy and C. Bowden (2001).
Cameroon. Pp. 133-159 in Fishpool, L.D.C. and M.I. Evans (eds.) Important Bird Areas in
Africa and Associated Islands: Priority Sites for Conservation. Pisces publications and BirdLife
International, Newbury and Cambridge, UK.
Francois, J., A. Rivas and R. Compère (1989). Le pâturage semi-aquatique à Echinochloa stagnina (Retz) P. Beauv. Etude approfondie de la plante ‘bourgou’ et des bourgoutières situées
en zone lacustre du Mali. Bull. Rech. Agron. Gembloux 24: 145-189.
Fry, C.H. (1970). Birds in Waza National Park, Cameroon. Bull. Niger. Orn. Soc. 7: 1-5.
Galatowitsch, S.M. and A.G. van der Valk (1996). Vegetation and environmental conditions
in recently restored wetlands in the prairie pothole region of the USA. Vegetatio 126: 89-99.
306
Floodplain rehabilitation and the future of Conservation & Development
Galindo-Leal, C. (2001). Design and analysis of conservation projects in Latin-America: an integrative approach to training. Conservation Ecology 5 (2): 16. online: www.consecol.org/
vol5/iss2/art16.
Gallopin, G.C. (2002). Planning for Resilience: Scenario’s, Surprises, and Branch Point. Pp.
361-392 in Gunderson and C.S. Holling (eds.) Panarchy. Understanding transformations in
human and natural systems. Island Press, Washington DC, USA.
Gamassa, D.G. (1995). Mweka College of African Wildlife Management: Its role as a Regional
Training Institution. Pp. 126-127 in Robinson, R. (ed.) Africa Heritage 2000: The future of
Protected Areas in Africa. Proceedings of the IUCN Commission on National Parks and
Protected Areas African Regional Working Session. National Parks Board, Pretoria, South
Africa.
Gaston, A. and D. Dulieu (1976). Etude agrostologique pour l’amènagement hydraulique pastorale
des Yaérés. Etude agrostologique no. 46. IEMVT Maison Alfort.
Geerling, C. (1982). Guide de terrain des ligneux sahéliens et soudano-guinéens. Mededelingen
Landbouwhogeschool 82-3, Wageningen, The Netherlands.
Geerling, C. and J. Bokdam (1971). The Senegal Kob (Adenota kob kob Erxleben) in the Comoé National Park, Ivory Coast. Mammalia 35: 17-24.
Gibbs, J.P. (1991). Spatial relationships between nesting colonies and foraging areas of Great
Blue Herons. The Auk 108: 764-770.
Gibson, C.C. (1999). Politicians & Poachers: The Political Economy of Wildlife in Africa. Cambridge University Press, Cambridge, UK.
Gide, A. (1927, 1928). Voyage au Congo/Retour du Tchad. Editions Gallimard, reprinted in Folio
1995, France.
Gilbert, D. (1971). Natural Resources and Public Relations. The Wildlife Society, Washington,
USA.
Gittinger, J.P. (1982). Economic Analysis of Agricultural Projects. John Hopkins University Press,
Baltimore, USA.
Gommes, R. and F. Petrassi (1996). Rainfall variability and drought in Sub-saharan Africa. FAO
agrometeorology series working paper no.9. (www.fao.org/sd/EIdirect/Eian0004.htm).
Greling, C. de (1972a). Sur les migrations et mouvements migratoires de l’avifaune Ethiopienne, d’apres les fluctuations saisonnieres des densités de peuplements en savane
Soudanienne au Nord Cameroun. Oiseau Rev. fr. Orn. 42: 1-27.
Greling, C. de (1972b). New records from northern Cameroon. Bull. Brit. Orn. Club 92: 24-27.
Grenier, C. (2000). Conservation contre nature. Les Iles Galapagos. IRD editions, Paris, France.
Groot, A.E. (2002). Demystifying Facilitation of Multi-Actor Learning Processes. Ph.D. thesis,
Wageningen University, Wageningen, The Netherlands.
Groot, W.T. de (1998). Problem-In- Context. A framework for the analysis, explanation and
solution of environmental problems. Pp. 22-43 in Nath, B., L. Hens, P. Compton and D.
Devuyst (eds.) Environmental Management in Practice. Volume I. Instruments of Environmental Management. Routledge, London, UK.
Groot, W.T. de (2001). Two paradigms for drylands research: a personal overview of the CEDC
2000 conference results. Pp. 11-22 in Ali, M., P. Loth, H.Bauer and H. de Iongh (eds.)
Management of fragile ecosystems in the North of Cameroon: the need for an adaptive approach.
CEDC and CML, Maroua, Leiden, The Netherlands.
Grunwald, M. (2002). Everglades Revival Work Halted. Washington Post August 15, 2002.
Gueye, B. and K. Schoonmaker Freudenberger (1991). Introduction à la methode accelerée de
recherche participative (MARP). IIED, London, UK.
Gunderson, L.H. and C.S. Holling (2002). Panarchy. Understanding transformations in human
and natural systems. Island Press, Washington DC, USA.
References
307
Gunderson L.H., C.S. Holling and G.D. Peterson (2002). Surprises and Sustainability: Cycles
of Renewal in the Everglades. Pp. 315-332 in Gunderson, L.H. and C.S. Holling (eds.) Panarchy. Understanding transformations in human and natural systems. Island Press, Washington DC, USA.
Hafner, H. (2000). Heron Nest Site Conservation. Pp. 201-217 in Kushlan, J.A. and H. Hafner
(eds.) Heron Conservation. Academic Press, London, UK.
Hardin, G. (1968). The tragedy of the commons. Science 162: 1243-1248.
Hart, T. and J. Hart (1997). Zaire: new models for an emerging state. Conservation Biology 11:
308-314.
Held, J.J. den, (1981). Population changes of the Purple Heron in relation to drought in the
wintering area. Ardea 69: 185-191.
Hepper, F.N. (1954-1972). Flora of West Tropical Africa, 2nd edition. Crown Agents, London, UK.
Herzog zu Mecklenburg, A.F. (1912). Vom Kongo zum Niger und Nil. Berichte der Deutschen
Zentral Afrika-Expedition 1910-1911. Leipzig, Germany.
Hiernaux, P. and L. Diarra. (1983). Pâturages de la zone d’inondation du Niger. Pp. 42-48 in
Wilson, R.T., P.N. de Leeuw and C. de Haan (eds.) Recherches sur les systèmes des zones
arides du Mali: résultats préliminaires. Rapport de recherche 5. ILCA, Addis Abeba, Ethiopia.
Hiernaux, P. and L. Diarra (1984a). Is it possible to improve the traditional grazing in the flood
plain of the Niger river in Central Mali? Proceedings 2nd International Rangeland Congress, Adelaide, Australia.
Hiernaux, P. and L. Diarra. (1984b). Savanna burning, a controversial technique for rangeland
management in the Niger floodplains of Central Mali. Pp. 238-243 in Tothill, J.C. and J.C.
Mott (eds.) Ecology and Management of the World’s Savannas. International Savanna
Symposium. Australian Academy of Science, Canberra, Australia.
Holling (1973). Resilience and stability of ecological systems. Annual Review of Ecological Systems 4: 1-24.
Holling, C.S. (ed.) (1978). Adaptive environmental assessment and management. Wiley, Chichester, UK.
Holling, C.S. (1995). What barriers? What bridges? Pp. 3-34 in Gunderson, L.H., C.S. Holling
and S.S. Light (eds.) Barriers and Bridges to the Renewal of Ecosystems and Institutions.
Columbia University Press, New York, USA.
Holling, C.S. and L.H. Gunderson (2002). Resilience and Adaptive Cycles. Pp. 25-62 in Gunderson, L.H. and C.S. Holling (eds.) Paranarchy: Understanding transformations in Human
and Natural Systems. Island Press, Washington DC, USA.
Holling, C.S., L.H. Gunderson and G.D. Peterson (2002). Sustainability and Panarchies. Pp.
63-102 in Gunderson, L.H. and C.S. Holling (eds.) Paranarchy: Understanding transformations in Human and Natural Systems. Island Press, Washington DC, USA.
Hollis, G.E., W.M. Adams and M.Aminu-Kano (eds.) (1993). The Hadejia-Nguru Wetlands. Environment, Economy and Sustainable development of a Sahelian Floodplain Wetland. IUCN,
Gland, Switserland.
Hollis, G.E. (1996). Hydrological inputs to management policy for the Senegal River and its
floodplain. Pp. 155-184 in Acreman, M.C. and Hollis, G.E. (eds.) Water Management and
Wetlands in Sub-saharan Africa. IUCN, Gland, Switserland.
Holmes, D.A. (1972). Bird notes from the plains south of Lake Chad, winter 1971-1972. Bull.
Niger. Orn. Soc. 9: 47-55, 76-84.
Holmes, D.A. (1974). Bird notes from the plains south of Lake Chad, winter 1971-1972. Part
3. Bull. Niger. Orn. Soc. 10: 28-37.
Hopkins and N. Levtzion (1981). Corpus of early Arabic sources for West African History 2001.
Cambridge University Press, Cambridge, UK.
308
Floodplain rehabilitation and the future of Conservation & Development
Howell, P., M. Lock and S.Cobb (1988). The Jonglei Canal. Impact and Opportunity. Cambridge
University Press, Cambridge, UK.
Hughes, F.M.R. (1990). The influence of flooding regimes on forest distribution and composition in the Tana river floodplain, Kenya. J. Appl. Ecol. 27: 475-491.
Hulme, D. and M. Murphree (1999). Communities, Wildlife and the ‘New Conservation’ in
Africa. Journal of International Development 11: 277-285.
IIED (1994). Whose Eden? An Overview of Community Approaches to Wildlife Management. IIED,
London, UK.
Inamdar, A., H. de Jode, K. Lindsay and S. Cob (1999). Capitalizing on Nature: Protected Area
Management. Science 283: 1856.
INTECOL (2004). Conference Abstracts of the 7th Intecol Wetlands conference, Utrecht, The
Netherlands.
Inter-American Development Bank (2000). Ecuadorian Plan to protect Galapagos Islands’ ecosystems receives IDB loan for $10.4 million. Press release, November 1. (www.iadb.org/exr/
prensa/2000/cp19300e.htm).
Issa, Saibou (2004). L’embuscade sur les routes des abords sud du lac Tchad. Politique Africaine 94: 82-104.
IUCN (1994). Guidelines for Protected Area Management Categories. Gland, Switserland and Cambridge, UK: CNPPA-WCMC. www.unep-wcmc.org/protected_areas/categories/index.html.
IUCN (1999). Rehabilitation of the Waza-Logone Floodplain, Republic of Cameroon: Proposals for
the Reinundation Programme. IUCN, Gland, Switserland.
IUCN (2003). Red List of Threatened Species. www.redlist.org.
Jacobson, S.K. and M.D. McDuff (1998). Training idiot savants: the lack of human dimensions in conservation biology. Conservation Biology 12: 263-267.
Janssen, M.A. and M. Scheffer (2004). Overexploitation of Renewable resources by ancient
societes and the role of sunk-cost effects. Ecology and Society 9 (1): 6. www.ecologyandsociety.org/vol9/iss1/art6
Jeanin, A. (1936). Les Mammiferes sauvages du Cameroun. Paul Lechevalier, Paris, France.
Jeanrenaud, S. (1999). People-oriented conservation: progress to date. Pp. 126-134 in Stolton,
S. and N. Dudley (eds.) Partnerships for Protection. Earthscan, London, UK.
John, D.M., C. Leveque and L.E. Newton (1993). Western Africa Pp. 47-78 in Whigham, D.F.
D. Dykyjova and S. Hejny (eds.) Wetlands of the World: Inventory, ecology and management
I. Kluwer Academic Publishers, The Netherlands.
Johnson, M. (1978). By Ship or by Camel ? the Struggle for the Cameroons. Ivory trade in the
19th century. Journal of African history 17: 95-117.
Junk, W.J., P.B. Bayley and R.E. Sparks (1989). The Flood Pulse concept in River-Floodplain
Systems. Pp. 110-127 in Dodge, D.O. (ed.) Proceedings of the International Large River
Symposium (LARS). Honey Harbour, Canadian Special publication of Fisheries and Aquatic Sciences 106, Department of Fisheries and Oceans, Ottowa, Ontario, Canada.
Kamto, M. (1996). Rapport de Mission. Aspects juridiques du Plan d’Aménagement des Parcs Nationaux de Waza et de Kalamaloué. Report to the Waza-Logone Project, Maroua, Cameroon.
Kanyamibwa, S. (1999). Project Development Workshop: Integrating Biodiversity Information
Management into the Curricula of Wildlife Training Institutions. WCMC, Cambridge, UK.
Kari, S. and P. Scholte (2001). La réhabilitation pastorale de la plaine d’inondation WazaLogone (Cameroun): comment consolider sa réussite écologique? Pp. 315-316 in Tielkes,
E. Schlecht and P. Hiernaux (eds.) Atelier regional: les approches de la gestion des pâturages et
les projets de développement: quelles perspectives? Verlag Ulrich E. Grauer, Stuttgart, Germany.
Kavanagh, M. (1977). Some birds in Cameroon. Bull. Niger. Orn. Soc. 13: 133-138.
References
309
Kavanagh, M. (1978). National Park in the Sahel. Oryx 14: 241-245.
Kessler, J.J. (2003). Working towards SEAN-ERA. A framework and principles for integrating environmental sustainability into planning. Tropical Resource Management Papers 43. Wageningen University, Wageningen, The Netherlands.
Kiker, C.F., J.W. Miloton and A.W. Hodges (2001). Adaptive Learning for science-based policy: the Everglades restoration. Ecological Economics 37: 403-416.
Kingdon, J. (1997). The Kingdon Field Guide to African Mammals. Academic Press. San Diego,
USA.
Kirkpatrick, D. (1994). Evaluating Training Programs, the Four Levels. Barrett-Koehler Publishers San Francisco, USA.
Kirkpatrick, J.B. (2001). Ecotourism, Local and Indigenous People and the Conservation of the
Tasmanian Wilderness World Heritage Area. Journal of the Royal Society of New Zealand
31: 819-828.
Korn, H., R. Schliep and J. Stadler (eds.) (2003). Report of the International Workshop on the
‘Further Development of the Ecosystem Approach’. Bundesambt fur Naturschutz Skripten 78,
available on www.bfn.de.
Kouokam, R., M. Moritz, D.Ngantou and P. Loth (2004). Conflicts and conflict management
in the Waza Logone Floodplain. Pp. 69-76 in Loth, P. (ed.) The return of the water. IUCN,
Gland, Switzerland.
Kramer, R., C. van Schaik and J. Johnson (eds.) (1997). Last Stand. Protected Areas & the defense
of tropical biodiversity. Oxford University Press, New York, USA.
Kremen, C., A.M. Merenlender and D.D. Murphy (1994). Ecological monitoring: a vital need
for Integrated Conservation and Development Programs in the Tropics. Conservation Biology 8: 388-397.
Kushlan, J.A (2000). Heron Feeding Habitat Conservation. Pp. 219-236 in Kushlan, J.A. and
H. Hafner (eds.) Heron Conservation. Academic Press, London, UK.
Laplante, K. (2004). Environmental Alchemy: How to turn Ecological Science into Ecological
Philosophy. Environmental Ethics 26: 361-380.
Lavieren, L.P. van (1976). Methodes d’inventaire des populations des grands mammiferes d’Afrique.
FAO document de travail RAP/74/056, Rome, Italy.
Lebeuf, A.M.D. (1969). Les Principautés Kotokos: essai sur le caractère sacré de l’autorité. Paris.
CNRS, Paris, France.
Leakey, R. and V. Morell (2001). Wildlife Wars. My Fight to save Africa’s Natural Treasures.
MacMillan, London, UK.
Lee, K.N. (1993). Compass and Gyroscope: Integrating Science and Policy for the Environment.
Island Press, Washington DC, USA.
Leendertse, P.C., A.J.M. Roozen, and J. Rozema (1997). Long-term changes (1953-1990) in the
salt marsh vegetation at the Boschplaat on Terschelling in relation to sedimentation and
flooding. Plant Ecology 132: 49-58.
Legrosse, P. (1999). Les règles d’accès des troupeaux peuls aux pâturages du Delta central du
Niger (Mali). Pp. 76-96 in M. Niamir-Fuller (ed.) Managing mobility in African rangelands.
Intermediate Technology Publications, London, UK.
Le Houerou, H.N. (1989). The Grazing Land Ecosystems of the African Sahel. Ecological Studies 75. Springer-Verlag, Berlin, Germany.
Lenders, H.J.R., B.G.W. Arts, H. Strijbosch and G. Van der Velde (1998). The role of reference and target images in ecological recovery of river systems: lines of thought in the
Netherlands. Pp. 35-52 in Nienhuis, P.H., R.S.E.W. Leeuwen and A.M.J. Ragas (eds.) New
Concepts for sustainable Management of River Basins. Backhuys Publishers, Leiden, The
Netherlands.
310
Floodplain rehabilitation and the future of Conservation & Development
Letouzey, R. (1968). Etude phytogeographique du Cameroun. Lechevalier, Paris, France.
Levin, S.A., S. Barrett, S. Aniyat et al. (1998). Resilience in natural and socio-economic systems. Environment and Development Economics 3: 222-235.
Lindley, W.I. (1999). Constraints and Potentials of Training Mid-Career Extension Professionals
in Africa. International Workshop on Innovative Training Programmes for Mid-Career
Agricultural Extension Field Staff in Sub-Saharan Africa, 6-8 July 1999. Addis Ababa,
Ethiopia.
Lipya, A.K. (2000). Evaluation of the progress of students trained at CAWM-Mweka. Unpublished
report, SADC Wildlife Management Training Programme, Harare, Zimbabwe.
Loth, P., and H. de Iongh (2004). The Waza-Logone Project: History, Objectives and Main Results. Pp. 77-87 in Loth, P. (ed.) The return of the water. Restoring the Waza Logone floodplain
in Cameroon. IUCN, Gland, Switserland.
Loth, P. (ed.) (2004). The return of the water. Restoring the Waza Logone floodplain in Cameroon.
IUCN, Gland, Switserland.
Louette, M. (1981). The birds of Cameroon, an annotated check-list. Verhandel. Kon. Acad.
Wetensch. Lett. Schone Kunst. Belg., Kl. Wetensch. 43 (163): 1-295.
MacKinnon, J., K. MacKinnon, G. Child and J. Thorsell (1986). Managing Protected Areas in
the Tropics. IUCN, Gland, Switzerland.
Maley, J. (1981). Etudes palynologiques dans le bassin du Tchad et Paleoclimatologie de l’Afrique nord tropicale de 30 000 ans à l’epoque actuelle. These Sc, Travaux et Documents de
l’ORSTOM, Paris 129: 586. Consulted from Pp. 45-52 in Coetzee, J.A. and E.M. van Zinderen Bakker (eds.) Palaeoecology of Africa and the surrounding Islands. Vol. 13. Balkema,
Rotterdam, The Netherlands. Also consulted through Maley, J. 1993. Chronologie calendaire des principales fluctuations du lac Tchad au cours du dernier millenaire: le role
des données historiques et de la tradition orale. Pp. 161-163 in Barretau, D. and C. von
Graffenrie (eds.) Seminaire MegaTchad, ORSTOM, Paris, France.
’t Mannetje, L. (1987). Measuring quantity of grassland vegetation. Pp. 63-95 in ’t Mannetje,
L. (ed.). Measurement of Grassland Vegetation and Animal Production. CAB International,
Wallingford, UK.
Marchand, M. (1987). The productivity of African floodplains. Int. J. of Env. Studies 29: 201-211.
Margolius, R. and N. Salafsky (1998). Measures of Success: Designing, Managing and Monitoring
Conservation and Development Projects. Island Press, Washington DC, USA.
Marion, L., P. Ulenaers, P. and J. van Vessem (2000). Herons in Europe. Pp. 1-32 in Kushlan,
J.A. and H. Hafner (eds.) Heron Conservation. Academic Press, London, UK.
Martin, A., P. Bonnet, D. Bourzat and R. Lancelot (1996). Importance of livestock production
and its economic contribution to the countries of the Lake Chad Basin Commission. Pp. 7996 in CIRAD/CTA Livestock Atlas of the Lake Chad Basin. CTA, Ede, The Netherlands.
Mayaka, T.B. (2002). Wildlife Co-Management in the Bénoué National Park Complex, Cameroon: A Bumpy Road to Institutional Development. World Development 30: 2001-2016.
McCarthy, T.S., W.N. Ellery and A. Bloem (1998). Some observations on the geomorphological impact of hippopotamus (Hippotamus amphibius L.) in the Okavango Delta, Botswana.
Afr. J. Ecol. 36: 44-56.
McCully (2001). Silenced rivers. The Ecology and Politics of Large dams. Zed Books, London, UK
and New York, USA.
McShane, T.O. and S.A. Newby. (2004). Expecting the Unattainable: The assumptions behind
ICDPs. Pp. 49-74 in McShane, T.O. and M.P. Wells, Getting Biodiversity Project to Work.
Columbia University Press, New York, USA.
References
311
Mefit Babtie (1983). Development Studies in the Jonglei Canal Area. Technical Assistance Contract
for Range Ecology Survey, Livestock Investigations and Water Supply. Final Report. Vol. 2,
Background and Vol. 3. Vegetation Studies. Glasgow, Rome and Khartoum, Sudan.
Metcalfe, S. (1994). The Zimbabwe Communal Area Management Programme for Indigenous Resources (Campfire). Pp. 161-192 in Western, D., R.M. Wright and S. Strum (eds.)
Natural Connections. Island Press, Washington D.C., USA.
Middleton, B. (1999). Wetland Restoration. Flood pulsing and Disturbance Dynamics. Wiley,
New York, USA.
MINEF (1994). Cameroon Ministry of the Environment and Forestry. Law No. 94-01 of 20
January 1994 to lay down forestry, wildlife and fisheries regulations. Pp. 5-49 in A compendium of official instruments on forest and wildlife management in Cameroon. Ministry of
Environment and Forestry, Yaoundé, Cameroon.
MINEF (1995). Cameroon Ministry of Environment and Forestry, Decree No. 95-466-PM of
18 December 1995 to lay down the conditions for the implementation of wildlife regulations. Pp. 119-151 in A Compendium of Official Instruments on Forest and Wildlife Management in Cameroon. Ministry of Environment and Forestry, Yaoundé, Cameroon.
MINEF-IUCN (1997). Plan Directeur d’Aménagement du Parc National de Waza et ses alentours.
Unpublished report, Yaoundé, Maroua, Cameroon.
Mitchell, B. (2002). Resource and Environmental Management. Prentice Hall, Harlow.
Mohammadou, E. (1988). Les lamidats du Diamare et du Mayo-Louti au XIX siecle (Nord Cameroun) (Vol. XXII). Institute for the study of Languages and Cultures of Asia and Africa
(ILCAA). Tokyo, Japan.
Moritz, M. (2003). Commoditization and the pursuit of piety: the transformation of an African pastoral system. Dissertation, University of California at Los Angeles, Los Angeles, USA.
Moritz, M., P. Scholte and S. Kari (2002). The demise of the nomadic contract: arrangements
and rangelands under pressure in the Far North of Cameroon. Nomadic Peoples 6: 127-146.
Morton, A.J. and E.A. Obot (1984). The Control of Echinochloa stagnina (Retz.) P. Beauv. by
harvesting for dry season livestock fodder in Lake Kainji, Nigeria-A modelling approach.
J. Appl. Ecol. 21: 687-694.
Moss, G. (1991). The Trainers Handbook. Moss Associates, New Zealand.
Mullié, W.C., J. Brouwer, S.F. Codjo and R. Decae (1999). Small isolated wetlands in the
Central Sahel: a resource shared between people and waterbirds. Pp. 30-38 in Beintema,
A. and van J. Vessem (eds.) Strategies for Conserving Migratory Waterbirds. Wetlands. Int.
Publ. 55, Wageningen, The Netherlands. www.wetlands.org/pubs&/pub_online/StratMigratWBs.pdf.
Mundy, P.J. (1972). Waza, Cameroun-access and avifauna. Bull. Niger. Orn. Soc. 9: 16-18.
Myers, N. (1997). The rich diversity of biodiversity Issues. Pp. 125-138 in Reaka-Kudla, M.L.,
D.E. Wilson and E.O. Wilson (eds.) Biodiversity II. Understanding and Protecting Our Biological Resources. Joseph Henri Press, Washington DC, USA.
Nachtigal, G. (1987). Sahara and Sudan. Volume III. The Chad Basin and Baguirmi (translated by Fischer & Fischer). Hurst & Company, London, UK.
Nepal, S.K. and K.E. Weber (1994). A Buffer Zone Concept for Biodiversity Conservation: Viability of the Concept in Nepal’s Royal Chitwan National Park. Environmental Conservation
21: 333-341.
Neumann, R. (1997). Primitive Ideas: Protected Area Buffer Zones and the Politics of Land in
Africa. Development and Change 28: 559-582.
Nguinguiri, J.C. (2001). Guide pour la Formation en Gestion Participative des Ressources
Naturelles. UICN-BRAC, Yaoundé, Cameroun.
312
Floodplain rehabilitation and the future of Conservation & Development
Niamir, M. (1990). Community forestry: herders’ decision-making in natural resource management
in arid and semi-arid Africa. FAO, Rome, Italy.
Niamir-Fuller, M. (1999a). Towards a synthesis of guidelines for legitimizing transhumance.
Pp. 266-290 in Niamir-Fuller, M. (ed.) Managing mobility in African rangelands: the legitimization of transhumance. IT Publications, London, UK.
Niamir-Fuller, M. (1999b). Managing mobility in African rangelands: the legitimization of transhumance. London. IT Publications, London, UK.
Niamir-Fuller, M., and M. Turner (1999). A review of recent literature on pastoralism and
transhumance in Africa. Pp. 18-46 in Niamir-Fuller, M. (ed.) Managing mobility in African
rangelands: the legitimization of transhumance. IT Publications, London, UK.
Nicolson, S.E. (1986). Climate, drought and famine in Africa. Pp. 107-129 in Hansen, A. and
D.E. McMillan (eds.) 1986. Food in sub-Saharan Africa. Lynne Reiner Bouolder, Colo, USA.
Nienhuis, P.H., R.S.E.W. Leeuwen and A.M.J. Ragas (1998). Discussion and State of the art:
new concepts emerging. Pp. 353-367 in Nienhuis, P.H., R.S.E.W. Leeuwen and A.M.J.
Ragas (eds.) New Concepts for sustainable Management of River Basins. Backhuys Publishers,
Leiden, The Netherlands.
Njifortii, H.L., T. Schrader and F. Toornstra (1991). LEARN: A Methodological Challenge for
Rapid Environmental Assessment. Landscape & Urban Planning 20: 173-181.
Ngog Nje, J. (1983). Structure et dynamique de la population des girafes au Parc National de
Waza, Cameroun. Rev. Ecol (Terre Vie) 37: 3-20.
Ngog Nje, J. (1984). Régime alimentaire de la girafe au Parc National de Waza, Cameroun.
Mammalia 48: 173-183.
Ngog Nje, J. (1995a). Central and West Africa. education and training requirements. Pp. 124126 in Robinson, R. (ed.) Africa Heritage 2000: The future of Protected Areas in Africa.
Proceedings of the IUCN Commission on National Parks and Protected Areas African Regional Working Session. National Parks Board, Pretoria, South Africa.
Ngog Nje, J. (1995b). La formation des spécialistes de la faune sauvage et de gestion des aires protégées en Afrique. Unpublished report, Garoua Wildlife College, Garoua, Cameroon.
Njiforti, H.L. (1997). The Biology and Management of Wild Helmeted Guineafowl (Numida
meleagris galeata Pallas) in the Waza Region of North Cameroon. PhD thesis Wageningen
University, Wageningen, The Netherlands.
Njoya, I.S. (2001). L’Ecole de Faune de Garoua (hier-aujourd’hui-demain). Garoua Wildlife College, Garoua, Cameroon.
Noble, I.R. and R.O. Slatyer (1980). The use of vital attributes to predict successional changes
in plant communities subject to recurrent disturbance. Vegetatio 43: 5-21.
North, M.E.W. (1963). Breeding of the Black-headed Heron at Nairobi, Kenya, 1958-62. J. E.
Afr. Nat. Hist. Soc. 24: 33-63.
Noss, A.J. (1997). Challenges to nature conservation with community development in central
African forests. Oryx 31: 180-188.
Noupa, P. (1995). Etude des interactions écologiques entre herbivores sauvages et domestiques dans
l’écosystème de la plaine de Waza (Cameroun). Mémoire maitrise en sciences naturelles
appliquées. Université de Liege, Liege, Belgium.
OAG Münster (1991). Report of the Ornithological Expedition to Northern Cameroon. Biologische
Station Münster, Germany.
Oates, J.F. (1999). Myth and reality in the rainforest: how conservation strategies are failing in West
Africa. University of California Press, Berkeley, USA.
Ortega-Rubio, A. and C. Arguelles-Méndez (1999). Management Plans for Natural Protected
Areas in Mexico: La Sierra de la Laguna Case Study. Int. J. Sustain. Dev. World Ecol. 6: 6875.
References
313
Ostrom, E. (1990). Governing the Commons. The Evolution of Institutions for Collective Action.
Cambridge University Press, Cambridge, UK.
Paillisson, J.-M., S. Reeber and L. Marion (2002). Bird assemblages as bio-indicators of water
regime management and hunting disturbance in natural wet grassland. Biological Conservation 106: 115-127.
Parker, I. and S. Bleazard (2001). An impossible dream. Some of Kenya’s last colonial wardens
recall the game department in the closing years of the British empire. Librario, Kinloss, Moray,
UK.
Parr, J. (1998). Discussion Paper for Protected Area Managers Writing a Management Plan.
Tigerpaper 25: 9-13.
Parsons, J. (1977). The effect of predation by Fish Eagles on the breeding success of various
Ciconiiformes nesting near Kisumu, Kenya. J. Nat. Hist. 11: 337-353.
Perrings, C. (1998). Resilience in the dynamics of economic-environment systems. Environmental and Resource Economics 11: 503-520
Pettet, A. (1976). The avifauna of Waza National Park, Cameroon, in December. Bull. Niger.
Orn. Soc. 12: 18-24.
Pfadenhauer, J. and F. Klotzli (1996). Restoration experiments in middle European wet terrestrial ecosystems: an overview. Vegetatio 126: 101-115.
Pimm, S.L. (1984). The complexity and stability of ecosystems. Nature 307: 321-326.
Pirot, J.Y., P.J. Meynell and D. Elder (2000). Ecosystem Management: Lessons from around the
world. A guide for Development and Conservation Practitioners. IUCN, Gland, Switserland.
Pitkin, B. (1995). Protected Area Conservation Strategy (PARCS): Training Needs and Opportunities Among Protected Area Managers in Eastern, Central and Southern Africa. Biodiversity
Support Program, Washington, USA.
Poché, R.M. (1974). Notes on the roan antelope (Hippotragus equinus (Desmarest)) in West
Africa. J. Appl. Ecol. 11: 963-968.
Polet, G and J.R. Thompson (1996). Maintaining the floods – hydrological and institutional
aspects of managing the Komadugu-Yobe River basin and its floodplain wetlands. Pp. 73100 in Acreman, M.C. and G.E. Hollis (eds.) Water Management and Wetlands in Sub-saharan Africa. IUCN, Gland, Switserland.
Polet, G. (2000). Waterfowl and flood extent in the Hadejia-Nguru wetlands of north-east Nigeria. Bird Cons. Int. 10: 203-209.
Pollack, M.E. and J.F. Horty (1999). There is More to Life Than Making Plans: Plan Management In Dynamic, Multiagent Environments. Artificial Intelligence magazine 71-83.
Pretty, J.N., I. Guijt, J. Thompson and I. Scoones (1995). Participatory Learning and Action, A
Trainer’s Guide. IIED Participatory Methodology Series, IIED London, UK.
Prins, H.H.T. and H.P.V. de Jeugd. (1993). Herbivore population crashes and woodland structure in East Africa. J. of Ecol. 81: 305-314.
Pulliam, H. R. and B.J. Danielson (1991). Sources, Sinks, and Habitat Selection: A Landscape
Perspective on Population Dynamics. American Naturalist 137 (suppl.): S50-S66.
Quammen, D. (2003). Saving Africa’s Eden, 204 National Geographic, 50-72 (September 2003).
Rauchenberger, D. (1999). Johannes Leo der Afrikaner. Seiner Beschreibung des Raumes zwischen
Nil und Niger nach dem Urtext. Orientalia Biblica et Christiana 13. Harrasswitz Verlag.
Wiesbaden, Germany.
Rees, W.A. (1978a). The ecology of the Kafue lechwe: the food supply. J. Appl. Ecol. 15: 177-191.
Rees, W.A. (1978b). The ecology of the Kafue lechwe: soils, water levels and vegetation. J. Appl.
Ecol. 15: 163-176.
Rettie, D.F. (1995). Our National Park System. University of Illinois, Urbana, USA.
Rice, W. R. (1989). Analyzing tables of statistical tests. Evolution 43: 223-225.
314
Floodplain rehabilitation and the future of Conservation & Development
Rietkerk, M, P. Ketner, L. Stroosnijder and H.H.T. Prins (1996). Sahelian rangeland development: a catastrophe? Journal of Range Management 49: 512 - 519.
Rietkerk, M. (1998). Catastrophic vegetation dynamics and soil degradation in semi-arid grazing
systems. Ph.D. thesis, Wageningen Agricultural University, Wageningen, The Netherlands.
Rietkerk, M., M. Boerlijst, F. van Langevelde, R. HilleRisLambers, J. van de Koppel, L. Kumar,
H.H.T. Prins and A.M. de Roos (2002). Self-organization of vegetation in arid ecosystems.
The American Naturalist 160: 524-530.
Riney, T. (1963). A rapid field technique and its application in describing conservation status
and trends in semi-pastoral areas. African Soils 8: 159-257.
Robin (1954). Le livre des sanctuaires de la nature. Bibliothèque Géographique. Payot, Paris.
Robins, E. (1970). The Ebony Ark. Black Africa’s Battle to save its Wildlife. Taplinger, New York,
USA.
Rogers, A. and P. Taylor (1998). Participatory Curriculum Development in Agricultural Education. A training guide. FAO, Rome, Italy.
Roshier, D.A., A.I. Robertson, R.T. Kingsford and D.G. Green (2001). Continental-scale interactions with temporary resources may explain the paradox of large populations of desert
waterbirds in Australia. Landscape Ecology 16: 547-556.
Rothwell, W.J. and H.J. Sredl (1992). Professional Human Resources Development: roles and competencies. 2nd Edition. Vol II. HRD Press, Amherst, USA.
Roux, F. and G. Jarry (1984). Importance, Composition et Distribution des Populations d’Anatidés
Hivernant en Afrique de l’Ouest. Muséum National d’Histoire Naturelle, Paris, France.
Roux, F. and G. Jarry (1986). Denombrements d’Anatidés Hivernant dans le Bassin Tchadien,
Janvier 1986. Muséum National d’Histoire Naturelle, Paris, France.
Roux, F. and G. Jarry (1987). Denombrements d’Anatidés Hivernant dans le Bassin Tchadien,
Janvier 1987. Muséum National d’Histoire Naturelle, Paris, France.
Saberwal, V.K. and A. Kothari (1996). The Human Dimension in Conservation Biology in
Developing Countries. Conservation Biology 10: 1328-1331.
Salafsky, N. (1994). Ecological Limits and Opportunities for Community-based Conservation.
Pp. 448-471 in Western, D., R.M. Wright and S. Strum (eds.) Natural Connections. Island
Press, Washington DC, USA.
Salafsky, N., R. Margoluis and K. Redford (2001). Adaptive Management: A Tool for Conservation Practitioners. Biodiversity Support Program, Washington DC, USA. (www.fosonline.org/Resources.cfm).
Saleh, A. (1994). La pêche dans le parc national de Kalamaloué: une activité à stimuler, à tolerer,
ou à interdire? Rapport de stage optionnel. Université de Dschang, Dschang, Cameroon.
Saleh, A. (1995). Impact de la reinondation de la plaine du Logone sur la peche: Cas particulier des
villages riverains du Parc National de Waza. Rapport de stage d’insertion professionelle.
Université de Dschang, Dschang, Cameroon.
Sayer, J.A. (1982). The pattern of decline of the Korrigum Damaliscus lunatus in West Africa.
Biol. Conserv. 23: 95-110.
Sayer, J. (1991). Rainforest Buffer Zones. Guidelines for Protected Area Managers. IUCN, Gland,
Switzerland.
Scheffer, M. , S.H. Hosper, M.L. Meijer and B. Moss (1993). Alternative equilibria in shallow
lakes. Trends Ecol. Evol. 8: 275-279.
Scheffer, M., S. Carpenter, J.A. Foley, C.Folke and B. Walker (2001). Catastrophic shifts in
ecosystems. Nature 413: 591-596.
Schneider, J.L. (1994). Le Tchad depuis 25000 ans. Geologie, Archeologie Hydrogeologie. Masson, Paris.
Scholes, R.J. and R. Biggs (2005). A biodiversity intactness index. Nature 434: 45-49.
References
315
Scholte, P. (1996). Conservation status of cranes in north Cameroon and western Chad. Pp.
153-156 in Beilfuss, R.D., W.R. Tarboton, N.N. Gichuki (eds.) Proceedings of the 1993 African Crane and Wetland Training Workshop. www.savingcranes.org/data/program/pp153156.pdf
Scholte, P. (1998). Waza National Park in Cameroon. Pp. 46 in Sherbenin, A. and M. Freudenberger: Migration to Protected Areas and Buffer zones and Protected Areas: Can we
stem the tide? Parks 8: 32-53.
Scholte, P. (2000a). Inventaire de la biodiversité. Ressources en espèces sauvages. Pp. 50-53
in GEPIS. Vers une gestion durable des plaines d’inondation sahéliennes. IUCN, Gland, Switserland and Cambridge, UK.
Scholte, P. (2000b). Towards consensual park management planning in Africa. Oryx 34: 8789.
Scholte, P. (2003a). Immigration: a potential time bomb under the integration of Conservation and Development. Ambio 32: 58-64. (www.iucn.org/themes/cem/cem/region/cameroon.htm)
Scholte, P. (2003b). Curriculum development at the African Regional Wildlife colleges, with
special reference to the Ecole de Faune (Cameroon). Environmental Conservation 30: 249258. (www.eldis.org/fulltext/CurriculumDev.EnvCons-Scholte.pdf).
Scholte, P. (2005). At the interface of legislation and wildlife management: a decade of experiences in consensual protected area management planning in Cameroon. Pp. (in press)
in Wandesforde-Smith (ed.) The Future of Conservation in Africa: Law, Biodiversity, Livelihoods and Development. Cambridge University Press, Cambridge UK.
Scholte, P. E. Pamo, S. Kari, S. Kersten and P. Kirda (1996a). Floodplain rehabilitation in N.
Cameroon: expected impact on vegetation, pastoralists and wildlife. Pp. 492-493 in West,
N. (ed.) Proceedings Fifth International Rangeland Congress. Society of Range Management,
Colorado, USA
Scholte, P., S. Kari and M. Moritz (1996b). The involvement of nomadic and transhumant pastoralists in the rehabilitation and management of the Logone floodplain, North Cameroon.
Issues paper 66, Drylands programme. IIED, London, UK.
Scholte, P., A. Saleh, K. Bobo and B. Boukar (1998). Who gets the fish? Assessing competition between piscivorous Birds and the fishery on the Logone floodplain of Waza National
Park, North Cameroon. Pp. 378 in Farina, A., J. Kennedy and V. Bossu (eds.) Proceedings
International Ecology Congress. INTECOL, Florence, Italy.
Scholte, P., S. de Kort and M. van Weerd (1999a). The Birds of the Waza-Logone Area, Far
North Province, Cameroon. Malimbus 21: 16-50.
Scholte, S. Adam, S. Kari and J.H. Mbouche (1999b). Walking a tightrope: Using PRA in a conflict situation around Waza National Park, Cameroon. PLA Notes 35:7-12. (www.iied.org/
sarl/pla_notes/pla_backissues/documents/plan_03502.PDF)
Scholte, P., S. de Kort and M. van Weerd (2000a). Floodplain rehabilitation in Far North
Cameroon: expected impact on bird life. Ostrich 71: 112-117.
Scholte, P., P. Kirda, S. Adam and B. Kadiri (2000b). Floodplain rehabilitation in North Cameroon: impact on vegetation dynamics. Applied Vegetation Science 3: 33-42.
Scholte, P., S. Kari, M. Moritz and H.H.T. Prins (2005). Pastoralist Responses to Floodplain
Rehabilitation in North Cameroon. Human Ecology (in press).
Scholte, P. and J. Brouwer 2005. The Relevance of Key Resource Areas for Large-Scale Movements of Livestock: are Sahelian floodplains an example? Pp. (in press) in Prins, H.H.T.
and F. van Langevelde, Large Herbivore Dynamics in Fragmented Landscapes.
Scholte, P. and I. Hashim (2005). Red-fronted Gazelle. Pp. (in press) in Kingdon (ed.) Mammals of Africa. Academic Press, London, UK.
316
Floodplain rehabilitation and the future of Conservation & Development
Schouten, M.G.C. (2001). Nieuwe Heelmeesters: Wetenschap en praktijk van natuurherstel. Inaugural address Wageningen University, Wageningen, The Netherlands.
Schrader, T. (1986). Les yaérés au Nord du Cameroun: pâturages de saison sèche (?) Aspects socioécologiques du développement pastoral dans la plaine inondable du Logone. Leiden University,
Leiden, The Netherlands.
Schricke, V., P. Triplet and P. Yésou (2001). Contributions Françaises à la connaissance des
oiseaux d’eau Palearctiques hivernant dans le Delta du Senegal. Alauda 69: 135-148.
Sethna (2004). What’s Hysteresis. www.lassp.cornell.edu/sethna/hysteresis/WhatIsHysteresis.html. Accessed 27 March 2005.
Shambaught, J., J. Oglethorpe and R. Ham (2001). The Trampled Grass. Mitigating the impacts
of armed conflicts on the environment. Biodiversity Support Program, Washington DC, USA.
Shepherd, J. (2004). The Ecosystem Approach. Five Steps to Implementation. IUCN, Gland, Switserland and Cambridge UK.
Sherbenin, A. de and M. Freudenberger (1998). Migration to protected areas and buffer zones:
can we stem the tide? Parks 8: 38-53.
Sighomnou, D., (2000). La plaine du Yaere dans le Nord Cameroun: une experience de restauration des inundations. Seminaire GIRN-ZIT, 20-23 Juin 2000, Bamako, Mali.
Sinclair, A.R.E. and P. Arcese (1995). Serengeti in the Context of Worldwide Conservation
Efforts. Pp. 31-46 in Sinclair, A.R.E. and P. Arcese (eds.) Serengeti II. Dynamics, Management and Conservation of an Ecosystem. University of Chicago Press, Chicago, USA.
Sinibaldi,I., K. Schmidt, P. Scholte, I. van Duren, F. Corsi, J. Brouwer and H. Prins (2004).
Dependence of large mammals and other flagship species on water and water management in
Africa. A literature review. Scientific report commissioned by WNF, Zeist, The Netherlands.
Sinsin, B., A.C. Tehou, I. Daouda, and A. Saidou (2002). Abundance and species richness of
larger mammals in Pendjari National Park in Benin. Mammalia 66: 369-380.
Skead, C.J. (1964). Black-headed Herons, Ardea melanocephala (Vigors and Children), nesting on a krans. Ostrich 35: 236.
Sklar, F.H., H.C. Fritz, Y. Wu, R. van Zee and C. McVoy (2001). The design of ecological landscape models for Everglades restoration. Ecological economics 37: 379-401.
Snelson, D. (1993). CEC Protected Areas and Biodiversity Strategy: Case study of Training and the
College of African Wildlife Management. Unpublished report, IUCN/WWF Regional Office,
Eastern Africa, Nairobi, Kenya.
Snelson, D. and A. Lanjouw (1997). Institutionalizing in-service training in protected area authorities in Africa. Protected Areas Conservation Strategy (PARCS) Project. African Wildlife
Foundation & Wildlife Conservation Society, Washington, USA.
Spinage, C. (1998). Social change and conservation misrepresentation in Africa. Oryx 32: 265276.
SPSS (1993). SPSS for Windows, release 6.01. SPSS Inc..
SPSS (1999). SPSS for Windows, release 9.0. SPSS Inc..
Steiner, A. (2004). News ReleaseWorld Congress Bangkok www.iucn.org/congress/documents/
press/2004-11-25-cem.pdf (accessed 22 March 205).
Stenning, D.J. (1957). Transhumance, migratory drift, migration; patterns of pastoral Fulani
nomadism. Journal of the Royal Anthropological Institute 87: 57-73.
Stone, R. (1997). What’s Your Role? Training for Organizational Impact. A guide for training officers in protected area management. African Biodiversity Series, no. 5. Biodiversity Support
Program, Washington DC, USA.
Szép, T. (1995). Relationship between west African rainfall and the survival of central European Sand Martins Riparia riparia. Ibis 137: 162-168.
References
317
Tambekou, A., F. Djaoro and B. Boukar (1997). Rapport de synthèse des mouvements migratoires
dans les villages périphériques du parc national de Waza. Internal report Waza-Logone project, Maroua, Cameroon.
Tapsell, S.M. (1995). River restoration: what are we restoring to? A case study of the Ravenbourne River, Kondon. Landscape Research 20: 98-111.
Tarboton (1977). The status of communal herons, ibis and cormorants on the Witwatersrand.
S. Afr. J. Wildl. Res 7: 19-25.
Taylor, J. (1972). The Black-headed Heron. Animals 13: 688-689.
Taylor, P. (1997). Participatory curriculum development for agricultural education and training: experiences from Viet Nam and South Africa. Training for Agricultural and Rural Development 1997-98: 4-15.
Tchamba, M. (1996). Elephants and their Interactions with People and Vegetation in the WazaLogone Region, Cameroon. PhD thesis, Utrecht University, The Netherlands.
Tchamba, M.N. and P. Elkan (1995). Status and Trends of some large mammals and ostriches in Waza National Park, Cameroon. Afr. J. Ecol. 33: 366-376.
Terborgh, J., C. van Schaik, L. Davenport and M. Rao (eds.) (2002). Making Parks work. Strategies for Preserving Tropical Nature. Island Press, Washington DC, USA.
Thibault, M. and S. Blaney (2001). Sustainable Human resources in a Protected Area in Southwestern Gabon. Conservation Biology 15: 591-595.
Thiollay, J.M. (1978). Les plaines du Nord Cameroun. Centre d’hivernage de rapaces paléarctiques. Alauda 46: 314-326.
Thomas, L. and J. Middleton (2003). Guidelines for Management Planning of Protected Areas.
IUCN, Gland, Switserland.
Thompson, K. (1985). Emergent plants of permanent and seasonally-flooded wetlands. Pp. 43107 in Denny, P. (ed.) The ecology and management of African wetland vegetation. Junk Publishers, Dordrecht, The Netherlands.
Thorhaug. A. (1980). Recovery patterns of restored major plant communities in the United
States: High to low altitude, desert to marine. Pp. 113-124 in Cairns, J. (ed.) The recovery
process in damaged ecosystems. Ann Arbor Science, Ann Arbor, USA.
Thouless, C. R., M. Allen, C. Coetzee, H. Dublin, H. Mahamat, Mohamadou, A.D. Njoh, H.
Peters, P. Scholte and M.N. Tchamba (1995). Management of conflict between humans and
the migratory Waza elephants. Consultants’ report. Waza-Logone Project, Maroua, Cameroon.
Tiffen, M., M. Mortimore and F. Gichuki. (1994). More people, Less erosion. Environmental
Recovery in Kenya. John Wiley & Sons, Chichester, UK.
Tobias, C., and C. Vanpraet (1981). Notes d’écologie Soudano-Sahelienne. Quelques relations
sols-végétation dans le parc national de Waza (Nord Cameroun). Revue Science et Technologie 1: 51-81.
Triplet, P. and P. Yésou (2000). Controlling the flood in the Senegal Delta: do waterfowl populations adapt to their new environment. Ostrich 71: 106-111.
Trolliet, B. and O. Girard (2001). Numbers of Ruff Philomachus pugnax wintering in West
Africa. Wader Study Group Bull 96: 74-78.
Tuntivanich, P. (1989). Curriculum development in veterinary science: a case-study from
Thailand. Pp. 238-251 in van den Bor, W., J.C.M. Shute and G.A.B. Moore (eds.) SouthNorth Partnership in Strengthening Higher Education in Agriculture. Pudoc, Wageningen,
The Netherlands.
Turner, M.D. (1999). The role of social networks, indefinite boundaries and politicalbargaining in maintaining the ecological and economic resiliency of the transhumance systems
of Sudano-Sahelian West Africa. Pp. 97-123 in Niamir-Fuller, M. (ed.) Managing mobility
in African rangelands. IT Publications, London, UK.
318
Floodplain rehabilitation and the future of Conservation & Development
ULG Consultants limited (1998). Training Needs Assessment. Unpublished report, SADC
Wildlife Management Training Programme, Harare, Zimbabwe.
UNDP (1998). Hwange National Park Management Plan 1999-2002, UNDP, Harare, Zimbabwe.
Unesco-MAB (2002). Periodic Review for Biosphere Reserves. www.Unesco.org/mab.
Vabie, M.B. (2000). Problèmes liés à l’utilisation des Méthodes Participatives: Enseignements
tirés de l’application sur le terrain des PRA/RRA dans certains pays de la Sous-Région de
l’Afrique Centrale. Arbres, Forets et Communautés rurales 15 & 16: 49-55.
Vaillant, A. (1956). Les pâturages naturels du Nord-Cameroun. Unpublished report.
Vanclay, J.K. (2001). The Effectiveness of Parks. Science 293: 1007.
Van de Ban, A.W. and H.S. Hawkins (1988). Agricultural Extension. Longman, Essex, UK.
Van den Bergh, J.C.J.M., A. Barendregt and A.J. Gilbert (2004). Spatial Ecology-Economic Analysis for Wetland Management. Modelling and Scenario Evaluation of Land-Use. Cambridge
University Press, Cambridge, UK.
Van den Bor, W. (1989). Higher Agricultural education in development. Pp. 3-20 in van den
Bor, W., J.C.M. Shute and G.A.B Moore (eds.) South-North Partnership in Strengthening
Higher Education in Agriculture. Pudoc Wageningen, The Netherlands.
Van den Bor, W., J.M. Bryden and C.T. Jongmans (1995). Rethinking higher agricultural education in a time of globalization and rural restructuring. European Journal of Agricultural
Education and Extension 2: 29-40.
Van der Kamp, J., M. Diallo and B. Fofana (2002a). Dynamique des populations d’oiseaux
d’eau. Pp. 87-141 in Wymenga, E., B. Kone, J. van der Kamp, L. Zwarts (eds.) Delta Intérieur du Niger. Ecologie et gestion durable des resources naturelles. Wetlands International/
RIZA/Alterra/Altenburg & Wymenga, Veenwouden, The Netherlands.
Van der Kamp, J., M. Diallo, B. Fofona, E. Wymenga, E. (2002b). Colonies nicheuses d’oiseaux
d’eau. Pp. 163-186 in Wymenga, E., B. Kone, J. van der Kamp, L. Zwarts (eds.) Delta Intérieur du Niger. Ecologie et gestion durable des resources naturelles. Wetlands International/
RIZA/Alterra/Altenburg & Wymenga, Veenwouden, The Netherlands.
Van der Kamp, J., B. Fofana and E. Wymenga (2005). Ecological values of the Inner Niger
Delta. Pp. 155-177 in Zwarts, L., P. van Beukering, B. Kone and E. Wymenga (eds.) The Niger,
a lifeline. Effective watermanagement in the Upper Niger Basin. Wetlands International/
RIZA/Altenburg & Wymenga, Veenwouden, The Netherlands.
Van der Valk, A.G. (1981). Succession in wetlands: a Gleasonion approach. Ecology 62: 688696.
Van der Valk, A.G. (1985). Vegetation dynamics of prairie glacial marshes. Pp. 293-312 in
White, J. (ed.) The Population Structure of Vegetation. Junk, Dordrecht, The Netherlands.
Van der Valk, A.G. (1992). Establishment, colonization and persistence. Pp. 60-102 in GlennLewin, D.C., R.K. Peet and T.T. Veblen (eds.) Plant succession, theory and prediction. Chapman & Hall, London, UK.
Van der Weide, M. (2002). Blauwe Reiger. Ardea cinerea. Pp. 82-83 in SOVON. Atlas van de
Nederlands broedvogels 1998-2000. Vogelonderzoek Nederland 2002, Nederlandse Fauna 5.
Natopma; Natuurhistorisch Museum Naturalis, KNNV uitgeverij & European Invertebrate
Survey-Nederland, Leiden, The Netherlands.
Van Meel, van R.M. (1997). How to augment effectiveness and flexibility by curriculum development in agricultural higher education. A case for a study programme on biological diversity. Pp. 97-108 in Charrier, A. and R. van Haarlem (ed.) Biodiversity, an issue in Higher Education. Proceedings of the third European Conference on Higher Education for Agriculture. Wageningen University, Wageningen, The Netherlands.
References
319
Vanpraet, C.L. (1977). L’écologie et l’amenagement du Parc National de Waza. PNUD/FAO,
Rome, Italy.
Venter, J. (1995). South African overview. Pp. 128-129 in Robinson, R. (ed.) Africa Heritage
2000: The Future of Protected Areas in Africa. Proceedings of the IUCN Commission on National Parks and Protected Areas African Regional Working Session. National Parks Board,
Pretoria, South Africa.
Verschuren, J. (1970). Sterven voor de olifanten (Mourir pour les Elephants). Het Wereldvenster,
WNF, Zeist, The Netherlands.
Walker, B., S. Carpenter, J. Anderies, N. Abel, G.S. Cumming, M. Janssen, L. Lebel, J. Norberg, G.D. Peterson, and R. Pritchard (2002). Resilience management in social-ecological
systems: a working hypothesis for a participatory approach. Conservation Ecology 6(1): 14.
[online] URL: http://www.consecol.org/vol6/iss1/art14/
Walters, C.J. and C.S. Holling (1990). Large-scale management experiments and learning by
doing. Ecology 71: 2060-2068.
Wanzie, C.S. (1988). Sociability of Buffon’s kob (Kobus kob kob Erxleben) in Waza National
Park, Cameroon. Mammalia 52: 21-33.
WCPA (2004). World Database on Protected Areas. www.sea.unep; wcmc.org/wdbpa/index.htm.
Weber, B., and A. Vedder (2001). In the Kingdom of Gorillas. Simon and Schuster, New York,
USA.
Wellcomme, R.L. (1979). Fisheries Ecology of Floodplain Rivers. Longman, London, UK.
Wells, M., K. Brandon and L. Hannah (1992). People and Parks. Linking Protected Area Management with Local Communities. World Bank/WWF/ USAID, Washington DC, USA.
Wesseling, J.W., C.A. Drijver, E. Naah, A.Namba, A. Zuiderwijk, D. Ngantou, P. Scholte and
A. Tchappi (1994). Waza Logone flood restoration study. Delft Hydraulics, Delft, The
Netherlands.
Wesseling, H., E. Naah, C.A. Drijver and D. Ngantou (1996). Rehabilitation of the Logone floodplain, Cameroon, through hydrological management. Pp. 185-198 in Acreman, M. and
G.E. Hollis (eds.) Water management and Wetlands in Sub-Saharan Africa. IUCN, Gland,
Switserland.
Western, D. (2003). Conservation science in Africa and the Role of International Collaboration. Conservation Biology 17: 11-19.
Westoby, M. (1980). Elements of a theory of vegetation dynamics in arid rangelands. Israel
Journal of Botany 28: 169-194.
Westoby, M., B. Walker and I. Noy-Meir (1989). Opportunistic management of rangelands not
at equilibrium. Journal of Rangeland Management 42: 266-74.
Wetten, J.C.J. van and P. Spierenburg (eds.) (1998). Waders and waterfowl in the floodplains of
the Logone, Cameroon, January, 1993. WIWO report nr. 67. Leiden, The Netherlands.
Wetlands International (2002). Waterbird Population Estimates-Third Edition. Wetlands International Global Series No. 12, Wageningen, The Netherlands.
Wild, R.G. and J. Mutebi (1997). Bwindi Impenetrable Forest, Uganda: Conservation through
Collaborative Management. Nature and Resources 33: 33-51.
Wilkie, D.S. and J.F. Carpenter (1999a). Can nature tourism help finance protected areas in
the Congo Basin? Oryx 33: 332-338.
Wilkie, D.S. and J.F. Carpenter (1999b). The potential role of safari hunting as a source of revenue for protected areas in the Congo Basin. Oryx 33: 339-345.
Wilkie, D., E. Shaw, F. Rotberg, G. Morelli and P. Auzel (2000). Roads, Development and Conservation in the Congo Basin. Conservation Biology 14: 1614-1622.
Wilson, R.T. (1982). Environmental changes in western Darfur, Sudan, over half a century
and their effects on selected bird species. Malimbus 4: 15-26.
320
Floodplain rehabilitation and the future of Conservation & Development
Wit, P. (1975). Preliminary notes on the vegetation of Waza National Park. FAO report, Rome,
Italy.
World Heritage Centre 2002. Periodic Report 3. Africa (2002). www.unesco.org/whc.
Zeltner, J.C. (1988). Les pays du Tchad dans la tourmente 1880-1903. Racines du Présent. L’Harmattan, Paris, France.
Zeltner, J.C. (1997). Les pays du Tchad et la montée des périls 1795-1850. Racines du Présent.
L’Harmattan, Paris, France.
Zon, A.P.M. van der (1992). Graminées du Cameroun. Wageningen Agricultural University
Papers 92-1. Wageningen, The Netherlands.
Zwilling, G.U. (1940). Unvergessenes Kamerun. Zehn Jahre Wanderungen und Jagden 1928-1938.
Verlag von Paul Baren, Berlin, Germany.
321
Summary
Introduction
The fertility of floodplains is legendary. Fluctuations in water level create a seasonal cycle of flood and drought, allowing a high primary production, abundant
wildlife and high human population densities. Since the 1950s, demand for irrigation water and electricity has increasingly disturbed the natural flooding regimes
of rivers in Africa. In 1979, the Waza-Logone floodplain in semi-arid Cameroon
followed this fate by the construction upstream of the Maga dam and its embank2
ment, greatly reducing the flooding intensity in an area of 1500 km , including
Waza National Park. As a consequence, annual grasses invaded productive perennial grasslands, reducing the carrying capacity of the area for livestock, fisheries
and wildlife. The cultivation of irrigated rice, the main purpose of the Maga dam,
was largely a failure, as testified by the use of less than half of the irrigation scheme’s
capacity and continuing import of rice. In 1994, after intensive local consultations, the Waza-Logone Integrated Conservation and Development Programme
re-opened a watercourse previously blocked by the embankment along the Logone
2
river. The reinstatement of the natural flooding regime in an area of 600 km
aimed to restore both wildlife and human use of the desiccated floodplain. This
study questions if, and how, reflooding may lead to the restoration of the Waza-Logone
floodplain to its pre-dam structure and conservation and development functions. It does
so by looking at the ecological history of the floodplain over the past two centuries
(Ch. 2) and analysing the impact of the reflooding on the vegetation (Ch. 3 and 4),
the waterbirds (Ch. 5), the antelopes (Ch. 6) and pastoral use (Ch. 7). Based on
these analyses the integration of development and conservation and conflicts
between the two are discussed (Ch. 8), as well as the usefulness and limitations of
conflict mitigating management planning (Ch. 9). This is then related to present
and possible future training of protected area managers in Africa (Ch. 10), and
how these trainees view their own training in disciplines that have been lacking
so far (Ch. 11). A synthesis and final conclusions are presented in Chapter 12.
The ecological history of the Waza-Logone floodplain
In Chapter 2 I argue that the reference image of the pre-Maga dam period when the
‘floodplain was still intact’ was biased by the period from which information was
consulted. Changes in hydrology, vegetation, land use, and especially wildlife have
occurred regularly throughout the last two centuries, but in several cases the system’s resilience has led to their recovery. The intensity and speed of changes in
322
Floodplain rehabilitation and the future of Conservation & Development
the post Maga dam period, triggered by lower than average rainfall and man-induced
drought, were, however, unprecedented in recent times. When studying the impact
of reflooding, there is a need to distinguish between the different causes of these
changes.
The impact of reflooding in Waza-Logone
In Chapters 3 and 4 the main question is: Does reflooding lead to a 100% perennial
grass cover with the same floristic composition and production that existed prior to the
Maga dam construction? And what are the mechanisms through which these floristic
and production changes take place?
In Chapter 3, results are presented from vegetation studies from 1984 onwards
along a transect covering always annually flooded, reflooded since 1994 as well as
desiccated parts of the floodplain. Fragmentary information suggests that in the
early 1980s in extended parts of the area perennial grasses died. This can be largely attributed to the dam construction, although there were indications that the condition of especially Vetiveria nigritana was already poor due to the mid-1970s
drought. From 1993 to 1999, the floristic composition has been monitored in a
grid in the centre of the reflooded zone. Following the reflooding, the cover of
perennial grasses, most notably Echinochloa pyramidalis and Oryza longistaminata,
increased again from 41 to 75% in the reflooded zone. The cover of annual species,
most notably Sorghum arundinaceum, a dominant annual grass only since the mid1980s, decreased in the reflooded zone from 58% to 23%. If the observed conversion rate of annual into perennial grassland is extrapolated, a recovery towards a
‘100%’ perennial state was most likely reached after the 2003 flooding season.
This was confirmed by qualitative observations in the transect in 2002. In terms
of species composition, this rehabilitation was only partial as of the previously
dominant perennial floodplain grasses, Vetiveria nigritana did not show any sign
of re-establishment despite its good condition elsewhere in the floodplain. The initial hypotheses of a ‘full’ recovery, in five years, to the pre-Maga state from the
1970s was therefore rejected. The layout of our study, with plots both in and outside the reflooded area, showed the large impact of reflooding compared to a low
impact of merely average rainfall. The gradual recovery of Oryza longistaminata
and Echinochloa pyramidalis grasses can be explained by lateral rhizomatous growth.
The tussock grass Vetiveria nigritana depends for its multiplication on seed dispersal, a less secure strategy that may depend on entirely different environmental
conditions. This holds in particular for the direction of flooding water, and therefore the sediments it carries, which has remained different from the pre-dam situation. With future reflooding one should take into account a period of at least ten
years for the recovery of rhizomatous grasses. If rainfall is lower than average, this
period may be much longer. Full recovery that includes again a predominant pres-
Summary
323
ence of Vetiveria nigritana may, apart from favourable climatic conditions, also be
dependent on the restoration of soil fertility.
Flood depth has been frequently used to explain the distribution of plant species,
but its relation with vegetation production has remained ambiguous. In Chapter
4, we studied the relation between flooding and above-ground biomass at the end
of the flooding season, as well as the dry season, to assess the impact of reflooding on vegetation production. Above-ground biomass of a combination of all species
and of the individual perennial grasses Oryza longistaminata and Echinochloa pyramidalis showed a positive linear relationship with maximum flood depth up to
one meter. These relations became stronger during the two years following the
water raising, showing the lag in reaction time to floodplain rehabilitation. Aboveground biomass data from other major floodplains in the three main African geographic regions showed a similar relation with maximum flood depth up to one
meter. Dry season regrowth, important because of its high nutrient quality in a
period of forage scarcity, was not directly related with maximum flood depth. A
comparison with other African floodplains hints at an exponential relation of
regrowth production if maximum flood depth exceeds one meter. Below this maximum flood depth, the timing of burning has an overwhelming influence on the
(low) regrowth. Although the mechanism of the relationship between maximum
flood depth and above-ground biomass is not yet fully understood, presented data
allow an assessment of the impact of the 20 cm water level rise due to the reflooding. This way it may be estimated that wet season above-ground biomass has
increased by approximately 10% in the first reflooding year increasing to approximately 37% in later years. For future reflooding I postulate the same two-year
time-lag and a vegetation production increase following the presented maximum
flood depth – above-ground biomass equation. In addition, dry season regrowth
production is expected to increase only after several years, when sufficient rhizome biomass has been accumulated.
Does reflooding lead to a recovery of waterbirds?
In Chapter 5, I attribute the drop in numbers of the area’s flagship species, the resident Black-crowned Crane, from 10 000 in 1973 to 2 000-2 500 individuals in 1992
to the Maga dam construction. The impact of the reflooding was monitored by
January total counts from 1992-2000. Total waterbird numbers in the dry season
increased from 60 000 to 105 000, whereas the number of species surpassing the
international 1% criterion doubled from 6 to 12. The increase in Anatidae corresponds to their recovery over most of West Africa following the droughts in the
1980s. The increase in especially Ciconiiformes in Waza-Logone was not paralleled
by similar trends in other main West African floodplains, suggesting that floodplain
rehabilitation has played an important role. Amongst the water birds that showed a
trend less than the general two-fold increase, were the large predominantly piscivorous Marabou and Yellow-billed Storks. This is attributed to the repeated destruction
324
Floodplain rehabilitation and the future
of their colonies by (fisher) men, a recent development in Waza-Logone that elsewhere in West Africa has already been under way for years. In contrast, a Black-headed Heron colony increased from 750 to 2500 nests between 1993-2003. The exceptional colony size, a multiple of the next largest known colony, suggests that besides
improved habitat due to reflooding, protection also played a vital role. It can be concluded that the increase in waterbirds in Waza-Logone is due to a combination of
factors: improved rainfall (especially Anatidae); floodplain rehabilitation (especially
omnivorous Ciconiiformes); and protection measures (selected Afrotropical Ciconiiformes). Hypotheses for the recovery of waterbirds by future reflooding should
take into account their long-term dynamics based on regional changes (e.g. rainfall)
and site-specific changes especially human protection and prosecution.
Does reflooding lead to the recovery of antelopes to their pre-dam numbers?
Antelopes, the subject of Chapter 6, are amongst the most prominent wildlife in
Waza National Park that has witnessed changes in rainfall, hydrology and human
encroachment during the last decades. To assess their impacts, we reviewed 26 surveys, comprising total, transect and localised counts, both aerial and terrestrial held
from 1962 to 2001. Numbers of Kob antelopes, a key floodplain species, crashed
from 20000 to 5000 in the period 1979-1983 following the Maga dam construction
that also coincided with a period of rinderpest. Numbers further declined to 2000 at
the end of the 1985 drought. The Kob population numbers increased between the
late 1980s and the early-mid 1990s from 2000 to 5000 individuals. The initial
recovery in Kob numbers following the reflooding did not continue beyond 1997.
Estimated Korrigum (‘Topi’) and Roan numbers dropped already in the early 1970s
and slightly recovered to respectively 2000 and 1000 in the 1990s. The disappearance of Waterbuck, soon after the construction of the Maga dam, has become a symbol of the area’s degradation. Two Waterbuck individuals were observed once again
in 1998 and 2003. The diversity of counting methods notwithstanding, the drop in
Kob numbers and the disappearance of Waterbuck can be attributed to the construction of the upstream Maga dam in 1979 and subsequent low rainfall. Yet Kob
population structure data showed that its reproduction capacity was not severely hit.
Kob recovered in the late 1980 and early-mid 1990s, but recovery did not continue
beyond that, despite increased flooding. The postulated explanation is increased
competition with livestock due to the rapidly increasing cattle grazing intensity (see
below). Direct persecution, made possible by the dramatic drop in number of park
guards, has probably played an important role as well. Floodplain rehabilitation
hypotheses should therefore not only be based on water management but include
feedback mechanisms of conservation effectiveness as well.
Does reflooding lead to full recovery of pastoral use of the floodplain?
In Chapter 7, responses of mobile pastoralists to the floodplain rehabilitation program were assessed through interviews with leaders of over 100 pastoral camps,
held at the end of each grazing season from 1993 to 1999. We registered changes in
325
Summary
2
number of camps and herds, and the time spent in the 600 km of the Logone floodplain that was reflooded in 1994. The first year, few pastoralists anticipated the reflooding or its impact, and the increase in grazing intensity was caused by a prolonged stay of pastoralists who already used the area for transit. The following three
years showed a sharp increase in the number of camps, which stabilised from 1997
onwards. The number of cattle herds doubled, two-thirds of which was due to newcomers and one-third to ‘natural’ increases. The average time cattle herds were present in the reflooded area increased by 60%. Overall, grazing intensity, expressed as
cattle density, increased linearly following the gradually recovering perennial grasslands to three times its value in 1993. It is expected that this increase has levelled off
-2
at a maximum of about 100 cattle km around 2003. Almost all observed changes
were attributed to the reflooding, the only factor that differed between the studied
floodplain and surrounding areas from where pastoralists moved away. No ‘territorial blocking’ nor ‘chaotic overshoot’ scenario has taken place with the reflooding.
The Ideal Free Preemptive Distribution scenario, assuming that any increase in perennial vegetation is consumed by cattle with the first individuals using the territories
pre-empting them, best explained pastoralist responses to floodplain rehabilitation.
Pre-emptive use might be reflected in the customary rights that pastoralists have to
campsites, which they have occupied for a long time. One of implications of the Ideal
Free Pre-emptive Distribution is that with reflooding, apart from a short lag in pastoralist reaction time, all additionally produced forage is consumed, leaving few
incentives for pastoralists to refrain from grazing inside Waza National Park. I postulate this Ideal Free Preemptive Distribution scenario for future reflooding. Unfortunately, this scenario predicts continued grazing of livestock inside Waza National Park, unless strict measures are taken.
Enhancing Conservation & Development integration by management
planning and training
If rehabilitation is ecologically successful, how may then the ecosystem’s functions that
underlie the balance between Conservation and Development be assured?
In Chapter 8 I discuss the risks of Integrated Conservation and Development Projects (ICDPs) in particular the overshoot of their successes. ICDPs aim to stimulate conservation without the previous negative experiences for local people, but
pay little attention to their long-term impact such as immigration. The rehabilitation of the Logone floodplain, the core activity of the Waza-Logone ICDP, aimed
at restoring the area’s services for both wildlife and human populations. The
reflooding has led to a 34% increase of sedentary fishermen and a multiple number of temporary fishermen. Whereas livestock intensity tripled, Kob antelopes have
not increased, indicating a reduction in their competitiveness. The virtual disappearance of wildlife in nearby Kalamaloué NP, due to advanced human encroach-
326
Floodplain rehabilitation and the future of Conservation & Development
ment, is therefore a bleak perspective for Waza NP. Examples from the Central
African Republic, Galapagos, Nigeria and Zimbabwe also showed that in openaccess systems, improvement in living standards (development) may stimulate
immigration, jeopardising the stability necessary in protected areas (conservation).
Most ICDPs lack demographic monitoring, masking its possible immigration risk.
To counter the immigration risk in Waza, a management policy was formulated
based on local stakeholder categorisation and subsequent privileges. The implementation of the management plan that offered development activities only in villages outside of the national park resulted in the voluntarily displacement of a village out of Waza NP. Whereas decades of oppression from park authorities towards this village only caused frustration, the new more balanced policy of the
‘carrot and stick’ yielded a tangible result. It is further recommended that ICDPs
should be involved in regional land-use planning to discourage development activities that stimulate immigration.
In Chapter 9, management planning is discussed as means to get a ‘grip on the
whole situation’. Management plans, such as the one formulated for Waza NP, are
generally considered to be indispensable for the daily management of protected
areas. Historically, their focus was on relatively straightforward technical or ecological issues. Since the 1990s, management plans have increasingly been based on
the consensual interpretation of pressing management issues, often related to the
legal context of protected areas. The most important product of the modern management plan is a consensus building process based on negotiation among stakeholders. This chapter asks whether the high expectations for management planning as an interface between protected area management and wildlife law and policy are realistic. The analysis first sketches protected area management policy in
Cameroon, particularly the limitations of the legal context in which it occurs. This
is followed by an account of the formulation process for the Waza NP management
plan, a process that has led to a consensus on sensitive key issues among local communities, local and national authorities, and international environmental NGOs.
This particular exercise in management planning was a learning process for all
concerned. The analysis questioned, however, whether the resulting management
plan was sufficiently realistic to lead to its successful implementation. The Waza
case was not unusual, as shown by a comparison with management planning experiences in other protected areas in Cameroon and other African countries.
Protected area managers generally have little input in protected area planning and
community conservation. Their lack of capacity in these disciplines was not unique
to Waza-Logone, but also identified as a major constraint for Integrated Conservation Development Programs elsewhere in Africa. In the chapters 10 and 11 I present the experiences at the Garoua Wildlife College, Cameroon, in its endeavours
to reinforce the capacities of protected area personnel in Waza NP and more generally in West-Central Africa.
Summary
327
Regional wildlife colleges in Cameroon, Tanzania and, recently, South Africa have
been responsible for the training of 4000 protected area managers in Africa. Training need assessments called for major curriculum reforms, which were developed
and implemented in the late 1990s. Chapter 10 analyses the factors that influenced this curriculum reform in the colleges’ endeavour to respond to new developments in African wildlife management. Since 1979, the curriculum of Garoua
Wildlife College has changed only gradually, whereas work placement subjects,
selected by students and their employers, have quickly responded to developments
in wildlife management, such as by an increase in the number of people-oriented
subjects. In the new curriculum, Garoua’s mid-career students appreciated biology and inventory disciplines for their relevance, as well as courses in other disciplines tailored to conservation practice. The curriculum reforms implemented at
Garoua depended on the presence of interested lecturers with an appropriate background, often obtained by additional training. The curricula of the regional wildlife
colleges at Garoua, Mweka (Tanzania) and SAWC (South Africa) showed important differences as a result of regional characteristics and differing visions of the
wildlife management profession. All three colleges have struggled to establish a
feed-back system for continuous curriculum review. Increasing the exchanges
between the colleges could further develop the curricula. While pursuing necessary changes in curriculum and institutions, care should be taken not to reduce
the colleges’ sustainability.
Training need assessments for African protected area managers have revealed the
need for people-oriented training. In the late 1990s, the Garoua Wildlife College,
Cameroon developed long (diploma and certificate) and refresher courses in community conservation for mid-career managers and guards from Waza NP and
other West and Central African protected areas. Through lectures, case studies
and Participatory Rural Appraisal exercises, the courses emphasised the development of skills for tuning principles of people participation to the conservation
objectives of protected areas. Chapter 11 reviews the trainees’ evaluations of these
courses to appreciate their relevance and support their further development. Diploma students judged the course as highly relevant because of the acquired analytical skills, whereas certificate students considered them of medium relevance
only. The reaction on short refresher courses varied as a function of the use of
cases from either the trainees’ professional experience or from the fieldwork location. The reactions of trainees to this learning opportunity suggest that protected
area personnel are not ‘attitude-limited’ as often suggested. Their constraints to
develop a more people-oriented work style lie largely in the areas of knowledge
and skills. These findings motivate increased efforts to implement training of protected area personnel in community conservation, preferably early in their careers.
328
Floodplain rehabilitation and the future of Conservation & Development
Synthesis
In the concluding Chapter 12, the outcome of the floodplain rehabilitation is reviewed by comparing the reflooding responses that, because of their different
response times, caused human supported resources (cattle and Black-headed Herons) to become dominant. I analyse the contributions of management planning
and capacity building to correct these undesired effects of the otherwise successful reflooding. The implementation of the management plan and training program has been successful on a local level, as testified by the improved relationship
between park authorities and local communities. Poor human resource management in the Cameroonian wildlife sector has overshadowed several results however.
I subsequently review different concepts of ecosystem change and management
that help to understand and communicate the observed changes of floodplain
resources. Panarchy, a holistic empirical concept, allows observed ecological and
socio-economic developments to be linked in a temporal and spatial hierarchy. This
hierarchy stresses the time lags between resources and the instalment of reflooding, as well as with capacity building, management and developments on which
should be anticipated. Only by considering the separate ecosystem elements,
causal explanations of the observed developments can be highlighted. Reflooding
induced the rehabilitation of the floodplain, not its restoration, as the 2000-2003
‘end result’ differed from the pre-dam situation with regard to the areas’ hydrology, vegetation composition, livestock-antelope ratio, etc.. In addition, the recovery
process of several floodplain resources (tussock grasses, antelopes) has followed a
different path and speed than the degradation following the dam construction.
This hysteresis has led to the land use state outcome that differed from the pre-dam
situation in its conservation value. I subsequently discuss the concept of resilience
that illustrates the remarkable capacity of the Waza-Logone ecosystem to overcome disturbances. As this recovery not always leads to the identical reference situation, I introduce the concept of evolving states to explain that observed (vegetation) states are not necessarily static, but may undergo cyclic succession.
At the end of this thesis I analyse the pursued overarching adaptive management
approach. I therefore revisit the monitoring observations and insights to reformulate the floodplain rehabilitation expectations. I call special attention to the role
of a ‘vision’, including not only targets for rehabilitation but also understanding
the required enabling environment. Empirical concepts of change and exchanges
with other experiences played an important role in the development of this vision.
One may conclude that adaptive management, based on long term monitoring of
separate ecosystem components and a regular review of targets within an overall
vision, forms a useful framework for the rehabilitation and management of tropical (floodplain) environments.
329
Samenvatting
Inleiding
De vruchtbaarheid van overstromingsvlakten is vermaard. Fluctuaties in waterniveaus creëren een cyclus van overstromingen en daaropvolgende droogten, met
als gevolg een hoge primaire productie, talrijk wild en vaak een hoge bevolkingsdichtheid. De vraag naar irrigatiewater en elektriciteit heeft sinds de jaren vijftig
echter in toenemende mate de natuurlijke loop van rivieren in sub-Sahara Afrika
veranderd. Dat geldt ook voor de Waza-Logone vloedvlakte in het droge NoordKameroen. De Maga-dam en een dijk werden aangelegd, waardoor de jaarlijkse
2
overstromingen in een gebied van 1500 km , inclusief het Waza National Park,
afnamen. Eenjarige grassen verdrongen de meerjarigen, waardoor de draagkracht
van het gebied voor vee, visserij en wild sterk afnam. Het cultiveren van geïrrigeerde rijst, de belangrijkste reden voor de bouw van de dam, werd een mislukking. Minder dan de helft van de irrigatiecapaciteit wordt gebruikt en er wordt nog
steeds rijst geïmporteerd. Na intensief lokaal overleg opende in 1994 het geïntegreerde natuurbeschermings- en ontwikkelingsproject Waza-Logone een zijrivier
die eerder door de dijk van de Logone rivier was afgesneden. Een gebied van zo’n
2
600 km werd hierdoor opnieuw bevloeid met als doel zowel de natuur als het
menselijk gebruik ervan te herstellen. Deze studie onderzoekt of en hoe herbevloeiing
kan leiden tot het herstel van de structuur en de natuurbeschermings- en ontwikkelingsfuncties van de Waza-Logone overstromingsvlakte van voor de dam.
De ecologische geschiedenis van de Waza-Logone overstromingsvlakte
In hoofdstuk 2 beredeneer ik dat het referentiebeeld van de periode voor de dam,
toen de vloedvlakte nog intact leek te zijn, beïnvloed was door de geraadpleegde
informatie. Veranderingen in hydrologie, vegetatie, landgebruik en vooral wild, hadden vaak al veel eerder plaatsgevonden. De veerkracht van het gebied heeft meestal geleid tot herstel. De intensiteit van veranderingen in de periode na de dam, nog
verergerd door geringe regenval, waren echter ongekend. Het is daarom belangrijk onderscheid te maken tussen de herkomst van de geobserveerde veranderingen.
De invloed van herbevloeiing in Waza-Logone
De leidende vraag in de hoofdstukken 3 en 4 is: Leidt herbevloeiing tot 100 procent
meerjarig grasland met dezelfde floristische samenstelling en productie van voor de Maga-
330
Floodplain rehabilitation and the future of Conservation & Development
dam? En wat zijn de mechanismen die leiden tot deze veranderingen? In hoofdstuk 3 presenteren we de resultaten van vegetatiestudies vanaf 1984 langs een
transect dat loopt van de jaarlijks overstroomde vloedvlakte, door de herbevloeide
naar de sinds de Maga dam uitgedroogde delen van de vloedvlakte.
Fragmentarische informatie suggereert dat begin jaren tachtig op veel plaatsen in
het gebied meerjarige grassen stierven. Dit was grotendeels het gevolg van de
komst van de dam, hoewel er indicaties zijn dat vooral Vetiveria nigritana al in
slechte conditie was door de droogte van midden jaren zeventig.
Tussen 1993 en 1999 werd de floristische samenstelling gevolgd in een grid in het
centrum van de invloedzone van de herbevloeiing. Na de herbevloeiing is de bedekking door meerjarige grassen, vooral Echinochloa pyramidalis en Oryza longistaminata, toegenomen van 41 naar 75 procent. De bedekking van eenjarige grassen, met name Sorghum arundinaceum die dominant was sinds midden jaren tachtig, nam af van 58 naar 23 procent. Als de geobserveerde conversiesnelheid van
eenjarig naar meerjarig grasland wordt geëxtrapoleerd, is een 100 procent meerjarig graslandsituatie waarschijnlijk in 2003 bereikt. Observaties in het transect
uit 2002 bevestigen dit beeld.
Wat soortensamenstelling betreft heeft de natuur zich slechts gedeeltelijk hersteld
omdat Vetiveria nigritana nog ontbreekt, ondanks de goede conditie van deze soort
elders in de overstromingsvlakte. De aanvankelijke hypothese dat binnen vijf jaar
de situatie volledig hersteld zou zijn, moest daarmee worden verworpen. De opzet
van de studie, met proefvlakken in en buiten het herbevloeiingsgebied, toont de
sterke invloed van herbevloeiing aan en de relatief lage invloed van regenval in de
betreffende periode. Het geleidelijke herstel van Oryza longistaminata en Echinochloa pyramidalis kan worden verklaard uit laterale groei van rhizomen. Het polgras Vetiveria nigritana daarentegen is voor vermeerdering afhankelijk van zaadverspreiding; een onzekere strategie die waarschijnlijk van volstrekt andere milieuomstandigheden afhangt. Toekomstige herbevloeiing moet rekening houden met
een periode van tenminste tien jaar voor het herstel van rhizoomgrassen. Een volledig herstel, inclusief Vetiveria nigritana is, naast regenval, mogelijk afhankelijk
van het herstel van bodemvruchtbaarheid.
Vloeddiepte wordt regelmatig gebruikt om de verspreiding van plantensoorten te
verklaren. De relatie met vegetatieproductie is echter onduidelijk. In hoofdstuk 4
bestudeer ik de relatie tussen overstroming en bovengrondse biomassa aan het
einde van het vloedseizoen en tijdens het droge seizoen, om zo de invloed van herbevloeiing op de vegetatieproductie te bepalen. Bovengrondse biomassa van alle
soorten samen en die van de individuele soorten Oryza longistaminata en Echinochloa pyramidalis, laten een positieve lineaire correlatie zien met maximale vloeddiepte tot ten minste een meter. Deze correlaties werden sterker in de twee jaar
na de herbevloeiing. Bovengrondse biomassagegevens van andere overstromings-
Samenvatting
331
vlakten in de drie belangrijkste geografische zones in Afrika laten een soortgelijke
relatie zien. Hergroei in het droge seizoen, dat vanwege het hoge nutriëntengehalte essentieel is voor de voedselvoorziening van vee en wild, bleek niet gerelateerd aan maximale vloeddiepte. Een vergelijking met andere overstromingsvlakten suggereert een exponentiële relatie van hergroei met vloeddiepte indien die
meer dan een meter bedraagt. Hoewel het causale verband tussen bovengrondse
biomassa en maximale vloeddiepte niet bekend is, kunnen we dankzij de gepresenteerde gegevens de invloed van grotere waterdiepte door herbevloeiing op vegetatieproductie bepalen. De schatting is dat de hoeveelheid bovengrondse biomassa in het eerste overstromingsjaar met 10 procent en in latere jaren met 37 procent is toegenomen. Voor toekomstige herbevloeiing veronderstel ik eenzelfde vertraging van twee jaar en een toename van de vegetatieproductie, waarvoor de berekende relatie tussen bovengrondse biomassa en maximale vloeddiepte geldt. Hergroei neemt waarschijnlijk pas jaren later toe wanneer voldoende rhizoom biomassa geaccumuleerd is.
Leidt herbevloeiing tot herstel van het aantal watervogels? In hoofdstuk 5 beredeneer
ik dat de afname van het aantal Zwarte Kroonkraanvogels, Waza-Logone’s paradepaardje, van tienduizend in 1973 tot tweeduizend exemplaren in 1992, toe te schrijven is aan de komst van de Maga-dam. De invloed van de herbevloeiing hebben
we gevolgd aan de hand van totaaltellingen in de maanden januari van 1992 tot
2000. Het aantal watervogels, geteld in het droge seizoen, nam toe van zestigduizend tot 105 duizend. Het aantal soorten dat het 1-procentscriterium overstijgt,
verdubbelde van zes naar twaalf. De toename van Anatidae (eend-achtigen) komt
overeen met het populatieherstel na de droogten in de jaren tachtig elders in WestAfrika. De toename van Ciconiiformes (ooievaars en reigers) in Waza-Logone werd
niet geëvenaard door de andere West-Afrikaanse overstromingsvlakten. Dit kan
hoogstwaarschijnlijk worden toegeschreven aan de herbevloeiing. Watervogels die
minder dan gemiddeld toenamen, waren de voornamelijk visetende Maraboe en
Nimmerzat-ooievaars. Dit teleurstellende resultaat kan toegeschreven worden aan
de regelmatige vernietiging van kolonies van deze soorten. Dit is een recent fenomeen in Waza-Logone dat elders in West Afrika al langer plaatsvindt. De bevindingen rond een Zwartkopreigerkolonie die tussen 1993 en 2003 toenam van 750
naar 2500 nesten, contrasteren hiermee. De uitzonderlijke grootte van de kolonie
suggereert dat naast een verbeterd habitat door herbevloeiing ook bescherming
een rol speelt.
Geconcludeerd kan worden dat de toename van watervogels in Waza-Logone het
gevolg is van een combinatie van factoren: toenemende regenval (dat geldt vooral
voor de Anatidae), vloedvlakteherstel (met name omnivore Ciconiiformes) en beschermingsmaatregelen (specifieke Afrotropische Ciconiiformes). Hypotheses voor
het verdere herstel van de watervogelpopulaties door toekomstige herbevloeiing
dienen rekening te houden met lange-termijn dynamica, gebaseerd op regionale
332
Floodplain rehabilitation and the future of Conservation & Development
veranderingen (regenval) en gebiedsspecifieke veranderingen, met name menselijke vernietiging en bescherming.
Leidt herbevloeiing tot het herstel van antilopenpopulaties tot de pre-dam aantallen?
Antilopen, onderwerp van hoofdstuk 6, behoren tot het meest opvallende wild in
Waza National Park. Veranderingen in neerslag, hydrologie en menselijke druk
hebben grote gevolgen gehad voor hun aantallen. Om deze gevolgen te bepalen,
onderzochten we 26 totaal-, transect- en gelokaliseerde tellingen, uitgevoerd tussen
1962 en 2001, zowel vanuit de lucht als vanaf de grond. Het aantal Kob-antilopen,
een belangrijke overstromingsvlaktesoort, daalde dramatisch van twintigduizend
tot vijfduizend in de periode tussen 1979-1983: de periode na de dam en een
bijkomende runderpestepidemie. Aantallen daalden verder naar tweeduizend aan
het einde van de droogte in 1985. De Kob-populatie steeg tussen eind jaren tachtig
en eind jaren negentig van tweeduizend naar vijfduizend. Het geschatte aantal
Korrigum (‘Topi’) daalde begin jaren zeventig al en herstelde zich enigszins tot
tweeduizend en duizend exemplaren in de jaren negentig. Het verdwijnen van de
Waterbok na de komst van de Maga-dam, werd een symbool van het verval van het
gebied. In 1998 en 2003 werden er echter weer twee gesignaleerd.
Ondanks de diversiteit in telmethodes kan de afname van Kob en Waterbok toegeschreven worden aan de Maga-dam en daarop volgende periode met weinig neerslag. De populatiestructuur gegevens van de Kob laten zien dat de reproductiecapaciteit niet verminderd was. Aantallen van de Kob herstelden zich tot de midden 90er jaren, maar stokte vervolgens, ondanks de toegenomen overstroming ten
gevolge van de herbevloeiing. Een mogelijke verklaring hiervoor is de toenemende
concurrentie met het aanwezige vee (zie hieronder). Directe vernietiging heeft
waarschijnlijk ook een rol gespeeld. Het sterk afgenomen aantal parkwachters
heeft hier ongetwijfeld invloed op gehad. Verwachtingen met betrekking tot resultaten van vloedvlakteherstel moeten daarom niet alleen gebaseerd zijn op waterbeheer, maar ook rekening houden met de effectiviteit van de natuurbescherming.
Leidt herbevloeiing tot het volledige herstel van veehouderij in de overstromingsvlakte? In
hoofdstuk 7 onderzoeken we de reacties van nomadische veehouders op het herstelprogramma. Dat gebeurde door middel van interviews in meer dan honderd
veehouderkampen aan het eind van de droge seizoenen van 1993 tot 1999. We
registreerde veranderingen in de hoeveelheid kampen en kuddes en in de tijd die
2
ze doorbrachten in de herbevloeide 600 km van de Logone overstromingsvlakte.
In 1994, het eerste jaar van de herbevloeiing, anticipeerden maar een paar veehouders op de veranderingen. De begrazingsintensiteit nam toe doordat veehouders
er langer bleven. In de drie daarop volgende jaren nam het aantal kampen sterk toe,
maar dat stabiliseerde vanaf 1997. In totaal verdubbelde het aantal rundveekuddes, waarvan tweederde bestond uit nieuwkomers, en eenderde dankzij natuurlijke toename. De begrazingsintensiteit, uitgedrukt in runderdichtheid verdrie-
Samenvatting
333
voudigde en nam lineair toe en volgde daarmee de geleidelijk herstellende meerjarige graslanden. De verwachting is dat de toename in 2003 is gestabiliseerd bij
een maximum van 100 runderen per vierkante kilometer.
Praktisch alle geobserveerde veranderingen kunnen toegeschreven worden aan de
herbevloeiing: de enige factor waarin de bestudeerde overstromingsvlakte en de
omringende gebieden van elkaar verschilden. Met de herbevloeiing heeft geen territoriale afscherming noch chaotische overexploitatie plaatsgevonden. Het Ideal
Free Pre-emptive Distribution-scenario, dat aanneemt dat iedere toename van meerjarige vegetatie wordt geconsumeerd, verklaarde als beste de reacties van veehouders op het herstel van de overstromingsvlakte. Preventief gebruik, waarbij de
eerste individuen die de territoria gebruiken ze ook leeg maken, is mogelijk te verklaren uit de gebruiksrechten. Een van de gevolgen van Ideal Free Pre-emptive Distribution is dat met herbevloeiing al het extra geproduceerde veevoer wordt geconsumeerd, waardoor veehouders niet gestimuleerd worden buiten Waza National
Park te blijven. Ik postuleer ditzelfde Ideal Free Distribution-scenario ook voor toekomstige herbevloeiing. Helaas betekent dit ook dat er ook in de toekomst vee aanwezigheid blijft in Waza National Park, tenzij hier strikte maatregelen tegen worden genomen.
Versterking van de integratie van natuurbescherming en lokale
ontwikkeling door middel van beheersplanning en training
Als het herstel ecologisch succesvol is, hoe kunnen dan de functies van het ecosysteem die
aan de basis liggen van de balans tussen natuurbescherming en lokale ontwikkeling worden gegarandeerd? In hoofdstuk 8 analyseer ik de risico’s van geïntegreerde natuurbeschermings- en ontwikkelingsprojecten (ICDP’s), met name het gevaar dat ze
door hun succes hun doel voorbijschieten. ICDP’s beogen natuurbescherming te
stimuleren zonder nadelige gevolgen voor een lokale bevolking. Ze besteden echter
maar weinig aandacht aan de gevolgen voor de lange termijn, zoals immigratie.
Het herstel van de Logone overstromingsvlakte, de kernactiviteit van het Waza-Logone ICDP, stond in het teken van een verbetering voor zowel natuur als mens.
De herbevloeiing heeft geleid tot een toename van 34 procent van het aantal sedentaire vissers en een veelvoud daarvan aan seizoensvissers. Terwijl de veehouderijintensiteit verdrievoudigde, nam het aantal Kob antilopen niet toe. Als gevolg van
toenemende menselijke druk verdween in het naburig gelegen Kalamaloué National Park nagenoeg al het wild: een afschrikwekkend voorbeeld voor Waza National Park. Voorbeelden uit de Centraal Afrikaanse Republiek, Galapagos, Nigeria en
Zimbabwe laten zien dat in dergelijke ‘open-acess’-systemen een verhoogde levensstandaard immigratie kan stimuleren die de noodzakelijke stabiliteit ondermijnt. Het ontbreekt de meeste ICDP’s aan demografische gegevens, waardoor dit
334
Floodplain rehabilitation and the future of Conservation & Development
immigratierisico verborgen blijft. Om het immigratierisico in Waza-Logone te
verlagen is een beheersbeleid opgezet dat is gebaseerd op een categorisatie van
lokale belanghebbenden en hun privileges (zie hieronder). De uitvoering van het
beheersplan stimuleerde ontwikkelingsactiviteiten in de dorpen buiten het park
en leidde tot het vrijwillige verplaatsen van een dorp uit Waza National Park.
In hoofdstuk 9 bediscussieer ik beheersplanning als middel om greep op de situatie te krijgen. Beheersplannen worden gezien als onmisbaar voor het dagelijkse
beheer van beschermde gebieden. Van oudsher ligt hun focus op technische of
ecologische onderwerpen. Sinds de jaren negentig zijn ze vooral gericht op consensus, gebaseerd op onderhandelingen met een groot aantal belanghebbenden.
In dit hoofdstuk vraag ik me af of de hoge verwachtingen van beheersplanning als
schakel tussen het beheer van natuurgebieden enerzijds en natuurbeschermingswetten en beleid anderzijds, realistisch zijn. Allereerst analyseer ik het natuurbeschermingsbeleid in Kameroen, in het bijzonder de juridische beperkingen ervan.
Daarna beschrijf ik de totstandkoming van het beheersplan voor het Waza National Park. Dat heeft geleid tot een consensus over onderwerpen die gevoelig lagen
bij lokale gemeenschappen, lokale en nationale autoriteiten en internationale milieu-NGO’s. Deze beheersplanning was voor alle betrokkenen een leerproces. De
analyse roept echter de vraag op of het managementplan realistisch genoeg was
om uitgevoerd te worden. Hierin verschilt de Waza-case overigens niet van beheersplannen van andere natuurgebieden in Kameroen en elders in Afrika.
Natuurbeschermingspersoneel wordt over het algemeen nauwelijks betrokken bij
planning en community conservation. Het ontbreken van vaardigheden in deze disciplines is niet uniek voor Waza-Logone, het is ook een probleem bij ICPD’s elders in Afrika. In de hoofdstukken 10 en 11 presenteer ik de ervaringen van het
Garoua Wildlife College bij het verbeteren van de vaardigheden van het natuurbeschermingspersoneel in Waza National Park en elders in West-Centraal Afrika.
Regionale opleidingsinstituten in Kameroen, Tanzania en recentelijk Zuid-Afrika,
hebben meer dan vierduizend Afrikaanse managers in de natuurbescherming opgeleid. Onderzoeken naar trainingsbehoeften toonden eerder aan dat er grote veranderingen nodig waren in de curricula. In de loop van de jaren negentig zijn die
dan ook aangepast. Hoofdstuk 10 analyseert welke factoren van invloed zijn geweest
op de pogingen van de opleidingsinstituten een antwoord te vinden op de nieuwe
uitdagingen in natuurbescherming.
Sinds 1979 kende het curriculum van het Garoua College slechts geleidelijke veranderingen. Dit stond in schril contrast met de snelle ontwikkelingen die de studenten in hun stages tegenkwamen. Dat gold in het bijzonder voor mensgerelateerde onderwerpen. Midcareer-studenten waardeerden in het nieuwe curriculum
de relevantie van vooral biologie en survey technieken. Dat gold ook voor andere
Samenvatting
335
disciplines die sterk op de praktijk van de natuurbescherming waren gericht. Het
doorvoeren van hervormingen in het curriculum was in Garoua afhankelijk van
de interesse van trainers met de juiste achtergrond. De curricula van de regionale
natuurbeschermingsopleidingsinstituten in Garoua (Kameroen), Mweka (Tanzania) en SAWC (Zuid Afrika) zijn onderling erg verschillend. Dit is het gevolg van
regionale accenten en uiteenlopende visies op het natuurbeschermingsvak. Alle
drie de instituten worstelen met het instellen van een feedbacksysteem voor een
continue ontwikkeling van de curricula. Uitwisseling tussen de drie colleges kan
een stimulans zijn voor verdere ontwikkeling ervan. Tegelijkertijd moet echter
ook de continuïteit van de opleidingsinstituten niet uit het oog worden verloren.
Onderzoeken onder Afrikaans natuurbeschermingspersoneel brachten de behoefte aan training in mensgerelateerde onderwerpen aan het licht. Eind jaren negentig heeft het Garoua College daarom cursussen ontwikkeld in community conservation voor haar mid-careerstudenten, alsmede voor beheerders en wachters van Waza
National Park en andere beschermde gebieden in West-Centraal-Afrika. Door middel van hoorcolleges, casestudies en PRA-oefeningen, werden vaardigheden ontwikkeld om het betrekken van de lokale bevolking af te stemmen op de doelstellingen van beschermde gebieden.
In hoofdstuk 11 bespreken we de evaluaties van cursisten op de ontwikkelde cursussen. ‘Diploma-studenten’ (HBO-niveau) beoordeelden de cursussen als zeer
relevant vanwege de opgedane analytische vaardigheden, terwijl ‘certificate-studenten’ (MBO-niveau) ze slechts als gemiddeld relevant beschouwden. De reacties
op de korte opfriscursussen varieerden, afhankelijk van de mate waarin praktische
cases werden gebruikt waarmee de cursisten vertrouwd waren. De reactie van de
cursisten geeft aan dat de betrokkenheid van natuurbeschermingspersoneel groter is dan vaak wordt verondersteld en dat het ontbreken van een mensvriendelijker werkstijl veroorzaakt wordt door een gebrek aan vaardigheden en kennis. Deze
bevindingen laten zien dat pogingen om natuurbeschermingspersoneel meer te
betrekken bij community conservation, liefst ook vanaf het begin van hun carrières, zinvol zijn.
Synthese
In hoofdstuk 12 evalueer ik het herstel van de overstromingsvlakte door de reacties van de afzonderlijke ecosysteemelementen te vergelijken. Door de verschillende reactietijden heeft dit geleid tot een dominantie van door mensen ondersteunde hulpbronnen zoals vee en Zwartkopreigers. Ik analyseer vervolgens de bijdragen van beheersplanning en training om deze ongewenste effecten van de overigens succesvolle herbevloeiïng te corrigeren. De uitvoering van het beheersplan
en het trainingsprogramma waren op lokale schaal succesvol, zoals de verbeterde
336
Floodplain rehabilitation and the future of Conservation & Development
verhoudingen lieten zien. Slecht personeelsbeleid in de Kameroense natuurbeschermingssector heeft echter vele goede resultaten overschaduwd.
Ik bespreek vervolgens verschillende aspecten van veranderingen van ecosystemen en management die de geobserveerde veranderingen kunnen verklaren. Panarchy, een holistisch empirisch concept, maakt het mogelijk de geobserveerde
ecologische ontwikkelingen en interventies te verbinden in een hiërarchie van
ruimte en tijd. Deze hiërarchie laat een vertraagde reactie zien, niet alleen tussen
de hulpbronnen en de herbevloeiing, maar ook tussen beheer, capaciteitsopbouw
en ontwikkelingen als immigratie. Alleen door de afzonderlijke ecosysteemelementen te bestuderen kunnen causale verbanden worden belicht. De herbevloeiing heeft een herstel van de overstromingsvlakte bewerkstelligd en geen restauratie.
Het eindresultaat uit 2000-2003 wijkt af van de pre-dam situatie met betrekking
tot hydrologie, de floristische samenstelling, de vee-antilope verhoudingen, enzovoort. Ook het herstelproces van diverse hulpbronnen van de overstromingsvlakte is anders verlopen dan het degradatieproces na het aanleggen van de dam. Deze
hysterese heeft geleid tot de huidige staat van landgebruik waarbij natuurbescherming een minder belangrijke rol speelt. Ik bediscussieer vervolgens het concept
van veerkracht dat de opmerkelijke capaciteit van het Waza-Logone gebied weerspiegelt om veranderingen te boven te komen. Omdat herstel niet altijd leidt tot
de identieke referentiesituatie, introduceer ik het begrip ‘evolving states’ om uit te
leggen dat de geobserveerde staat van de vegetatie niet noodzakelijkerwijs statisch
is, maar ook cyclische successie kan ondergaan.
Ik besluit deze dissertatie met een analyse van de overkoepelende ‘adaptive management’ benadering. Hiervoor analyseer ik de observaties en verkregen inzichten
om de herstelverwachtingen van de overstromingsvlakte nader te beschouwen. Ik
vraag speciaal aandacht voor de rol van een visie, waarbij niet alleen oog moet zijn
voor de doelstellingen van het herstel, maar ook de rol van ondersteunende diensten daarbij. Empirische begrippen van verandering en uitwisseling met andere
ervaringen speelden een belangrijke rol in de ontwikkeling van deze visie. Ik concludeer dat adaptive management gebaseerd op lange-termijn monitoring van de
afzonderlijke ecosysteemelementen en een regelmatige evaluatie van de doelstellingen binnen een algemene visie, een geschikt kader vormt voor het herstel en
beheer van tropische (overstromingsvlakte) milieus.
337
About the Author
Born in 1962, Paul Scholte grew up in the area around Nijmegen. His passion for
nature found an outlet through the NJN, the Dutch youth organisation for nature
observation. Bird watching in the Ooijpolder and surrounding forests completed
his VWO education at the Canisius College from 1974 to 1980.
In 1987, Paul graduated in Biology at Wageningen University. For a major in animal ecology he studied the link between land use, herpetofaune and birds of prey
in Evros, Greece. During this first stay abroad, spontaneous contacts with local
communities opened his eyes for the social dimension of nature conservation. He
did majors in tropical soil science and vegetation science (mapping) during a oneyear stay in Kenya. A minor was taken in extension science. During the first six years
of his professional career Paul was active in rangeland ecology and management
in Yemen (DGIS) and in Chad where he was project manager (DHV Consultants).
With his colleagues, Paul prepared the first vegetation map of Yemen. In Chad he
supervised studies on rangeland resources and their utilisation and translated
results into policy advises to the Ministry of Livestock.
Employed by Leiden University, Paul headed the sections on national park assistance and ecological research at the Waza-Logone project (Cameroon) from 1993 to
1998. He co-ordinated the formulation of the Waza Management Plan and supervised impact studies of reflooding on vegetation, wildlife and pastoralists. Employed
by DGIS and the Ministry of Environment and Forestry, Paul lectured at the Garoua
Wildlife School (Cameroon) on participatory approaches in wildlife management
and management planning from 1998 to 2000. He also co-ordinated curriculum
reform at this regional college for protected area managers. The present thesis is a
consolidation of the scientific publications he authored on these experiences.
Following his return to the Netherlands in 2000, Paul was independent consultant charged with project formulation, backstopping and evaluation for UNDP and
Worldbank-GEF, IUCN, WWF, BCTF and others in Cameroon, CAR, Chad, Gabon,
Mongolia, Morocco, Senegal and Sudan. Paul was appointed member of the Wadden Sea Council and is member of the Dutch working group Ecology & Development, the Antelope Specialist Group (IUCN/SSC) and the Commission on Ecosystem Management (IUCN/CEM). From 2004 to 2005, Paul was senior biodiversity officer at NC-IUCN, focussing on corporate responsibilities with regard to biodiversity. Late 2005 Paul returned with Stephany, his partner, and children Thyne
and Nika to Yemen as Chief Technical Advisor of the Socotra Conservation and
Development Programme.
338
Background Publications
Research presented in this thesis is based on a series of field studies that elaborated upon methodologies and analysed preliminary experiences. Inventories complemented these background materials.
General
Scholte, P., E. Pamo, S. Kari, S. Kersten and P. Kirda (1996). Floodplain rehabilitation in N. Cameroon: expected impact on vegetation, pastoralists and wildlife. Pp. 492-493 in West, N. (ed.) Proceedings Fifth International Rangeland Congress. Society of Range Management, Colorado, USA.
Scholte, P., W. Mullié, C.Batello, M. Marzot, A. H. Touré and D. Williamson (2004). Wildlife. Pp.
227-257 in Batello, C (ed.) The Future is an Ancient Lake. Traditional knowledge, biodiversity and
genetic resources for food and agriculture in Lake Chad Basin ecosystems. FAO, Rome, Italy.
Birds
Mullié‚ W.C., J. Brouwer and P.Scholte (1995). Numbers, distribution and habitat of wintering
White Storks Ciconia ciconia in the East-Central Sahel in relation to rainfall, food and anthropogenic influences. Pp 219-240 in Biver, O., P. Enggist, C. Marti, T. Salath (eds.) Proceedings
of the International Symposium on the White Stork (Western Population). Basel, Switserland.
Scholte, P. (1996) Conservation status of Cranes in North Cameroon and Western Chad. Pp
153-156 in Beilfuss, R.D., W.R. Tarboton and N.N. Gichuki (eds.) Proceedings of the 1993
African Crane and Wetland Training Workshop. International Crane Foundation, Baraboo,
Wisconsin, USA. www.savingcranes.org/data/program/pp153-156.pdf
Scholte, P., S. de Kort and M. van Weerd (1999). The Birds of the Waza-Logone Area, Far
North Province, Cameroon. Malimbus 21: 16-50.
Scholte, P. (1999). Status of Vultures in the Lake Chad Basin, with special reference to Northern Cameroon and Western Chad. Vulture News 39: 3-19.
Scholte, P., S. de Kort and M. van Weerd (2000). Floodplain rehabilitation in Far North
Cameroon: expected impact on bird life. Ostrich 71: 112-117.
Scholte, P. and R.J. Dowsett (2000). Birds of Waza new to Cameroon: corrigenda and addenda. Malimbus 22: 29-31.
Scholte, P. and P. Robertson (2001). Chad. Pp 177-184 in Fishpool, L.D.C. & Evans M.I. (eds.)
Important Bird Areas in Africa and Associated Islands: Priority Sites for Conservation. Pisces
publications and BirdLife International. Newbury and Cambridge, UK.
Fotso, R., F. Dowsett-Lemaire, R.J. Dowsett, Cameroon Ornithological Club, P. Scholte, M.
Languy and C. Bowden (2001). Cameroon. Pp 133-159 in Fishpool, L.D.C. & M.I. Evans
(eds.) Important Bird Areas in Africa and Associated Islands: Priority Sites for Conservation.
Pisces publications and BirdLife International. Newbury and Cambridge, UK.
Mammals
Scholte, P (2000). Ressources en espèces sauvages. (Inventaire de la biodiversité). Pp 50-53 in GEPIS
(2000) Vers une gestion durable des plaines d’inondation sahéliennes. IUCN, Gland. Switserland.
Background publications
339
Scholte (2001). Notes on the status of antelopes in central and southern Chad. Pp 15-22 in East, R
(ed.) Antelope Survey Update No 8. IUCN/SSC Antelope Specialist Group Report.
Scholte, P. (2005). Dama gazelle. Pp (in press) in Kingdon et al. Mammals of Africa. Academic Press,
London, UK.
Scholte, P. and I. Hashim (2005). Red-fronted Gazelle. Pp (in press) in Kingdon et al. Mammals of
Africa. Academic Press, London, UK.
Pastoralism
Scholte, P, S. Kari and M. Moritz (1996). The involvement of nomadic and transhumant pastoralists in
the rehabilitation and management of the Logone floodplain, North Cameroon. Issues paper 66,
Drylands programme, IIED, London. (also in French).
Kari, S. and P. Scholte (2001). La réhabilitation pastorale de la plaine d’inondation Waza-Logone
(Cameroun): comment consolider sa réussite écologique? Pp 315-316 in Tielkes, E., E. Schlecht
et P. Hiernaux (eds.) Elevage et gestion de parcours au Sahel, implications pour le développement.
Comptes-rendus d’un atelier régional tenu à Niamey, Niger, du 2 au 6.10.2000. Verlag Ulrich
E.Grauer, Beuren, Stuttgart Germany.
Moritz, M., P. Scholte and S. Kari. (2002). The demise of the Nomadic Contract. Nomadic Peoples
6: 124-143
Scholte, P. and J. Brouwer (2005). The Relevance of Key Resource Areas for Large-Scale
Movements of Livestock: are Sahelian floodplains an example? Pp (in press) in Prins, H.H.T.
and F.van Langevelde. Spatial Ecology of Large Herbivores and Pastoralists.
On management
Scholte, P., Saleh, A., Bobo, K and Boukar, B. (1998). Who gets the fish? Assessing competition
between piscivorous Birds and the fishery on the Logone floodplain of Waza National Park,
North Cameroon. Pp. 378 in Farina, A., J. Kennedy and V. Bossu. Proceedings International Ecology Congress, July 1998. Florence, Italy.
Scholte, P., S.Adam, S.Kari and J.H. Mbouche. (1999) Walking a tightrope: Using PRA in a conflict
situation around Waza National Park, Cameroon. PLA Notes 35: 7-12.
(www.iied.org/sarl/pla_notes/pla_backissues/documents/plan_03502.PDF)
Scholte, P. (2000) Migration Policy of the Waza-Logone Project in the Far North of Cameroon:
from recognition to negotiation based on stakeholder identification. Pp. 13-14 in Vabi, M.B.
Linking the conservation of biological diversity to sustainable development. WWF-Cameroon
Programme Office, Yaoundé, Cameroon.
Scholte, P. (2000). Towards consensual park management planning in Africa. Oryx 34: 87-89.
Scholte, P. (2000) Towards collaborative management in Waza National Park: The Role of its
Management Plan. Pp. 41-53 in Bauer, H. and A. Madi (eds.) People, parks and wildlife. Contributions from Cameroon. Proceedings of the Park-People Conference, Maroua February 1998. Centre
for Environmental and Development Studies, Maroua, Cameroon.
Est, van D. and P. Scholte (2001) Linking social and ecological diversity: organisations for adaptive
natural resource management in the Waza-Logone floodplain (North Cameroon). Pp. 55-65 in
Ali, M., P. Loth, H. Bauer and H. de Iongh (eds.) Management of fragile ecosystems in the North
of Cameroon: the need for an adaptive approach. CEDC/CML.
This publication was financially supported by WWF-Netherlands, Scholte
Consulting and Leiden University.
The Waza-Logone project, under the auspices of which a major part of the
fieldwork of this study was undertaken, was carried out by the Government of
Cameroon and IUCN, in co-operation with the Institute of Environmental
Sciences of Leiden University (CML), the Netherlands Development Organisation
(SNV) and WWF-Cameroon, with the financial support of the Dutch Ministry of
Foreign Affairs and WWF-Netherlands. My assignment at the Ecole de Faune
was supported by the Dutch Ministry of Foreign Affairs (DGIS) and the
Cameroonian Ministry of Environment and Forestry.
Acknowledgments
For the financial assistance of the production of this book, I am grateful to the efforts of Esther Blom (WWF-Netherlands). Matthew Parr (NC-IUCN) assisted with
the language editing of the unpublished chapters, Linda Huijsmans with the Dutch
translation of the summary, Mohamed Yesef with the French translation of the abstract. Sjoukje Rienks made a book out of all of this.
I am grateful to the staff of the Waza-Logone project, Michael Allen, Daniel Ngantou, Richard Braund, Maureen Roëll, Jaap Kok and Roger Kouakam as well as
Patrick Dugan and Jean-Yves Pirot of the co-ordinating IUCN-wetland group for
the support to the reported fieldwork. At the Waza-Logone Project I had a wonderful time with my colleagues who greatly shaped the outcome of the study: Saïdou
Kari, Philippe Kirda, Saleh Adam, Bobo Serge Kadiri, Boukar Beladane, Paul Kouamou, Etienne Pamo (Dschang University), the hydrologists Emmanuel Naah, and
Daniel Sighomnou of IRGM and of the park village team Jean Hilaire Mbouche,
Djebba and Madoum. In the floodplain Kaskala was our host and guide. His passing away is a great loss.
In the pastoral camps and villages we were always received with great hospitality
and attention. Yerima Ouamarou, El Hadj Bouba, El Hadj Bello and El Hadj Eli
learnt us the ins and outs of the pastoral life. The Waza National Park villages
Andirni, Mahé, Halé and Zwang were our wet season floodplain homes. Saly of
Halé was one of those immigrants, born in the floodplain, who made a difference.
I would like to memorise the Blama (village chief) of Zwang who lost his life in
the unscrupulous actions of the anti-gang movement in 1998. I am grateful to the
Waza NP wardens Badjoda Daouda, Mahamat Habibou and Saleh Adam for the
trustful collaboration, despite our sometimes different positions and responsibilities towards local communities. The Waza guides, Manga, Moussa Barka and his
son Oumarou, and all the others, not only provided reliable guidance, but were a
continuous source of information and feed-back on our observations. The friendship they offered has left deep marks (and the name of my son, Thyne Moussa).
The Garoua Wildlife College was a stimulating environment for which I am grateful to the various generations of students. The directors Ibrahim Njoya, Jean Ngog
Nje and Andrew Allo Allo, and recently Francis Tarla showed particular interest in
my work. Mayna, Louis Tsague, Dong, Tchadel, Emmanuel Battokok, Rigobert
Azombo, Etienne Hatungimana, Talla, Kwabong and Robert Ndim were inspiring
colleagues. In addition, Battokok and Azombo were the driving force behind the
waterbird counts. I was the last of a generation of Dutch trainers to the Ecole. I am
grateful for the continuing interest of my predecessors Bart van Lavieren, Chris
Geerling, Piet Wit, Ton van der Zon and Floris Deodatus.
Although the Ministry of Environment and Forestry as an institution was sometimes a
source of frustration, the pleasant collaboration with the former directors of
wildlife Denis Koulagna, Ndjoh à Ndiang and Mengang greatly improved our understandings of the Yaoundé bureaucracy. At the CEDC (Centre for Environment and
Development Studies), Madi Ali, Martin Tchamba, Franke Toornstra, Hanson
Njiforti, Hannie Korthof and Waldo were our stimulating discussions partners.
My colleagues at CML, the Institute of Environmental Sciences of Leiden University, tolerated my erratic presence for over a decade. Carel Drijver, who supervised
the 1980s preparatory studies, was the driving force behind the initiation and early
years of the Waza-Logone project. His engaged backstopping gave me a flying
start. Jeroen van Wetten initiated several of the ornithological studies. Edith Roos,
Annelies Oskam, Helias Udo de Haes, Hans de Iongh and Gerard Persoon have
contributed in their various capacities to my work. The continuing feed-back of
Wouter de Groot and that of Hans Bauer, Diny van Est, Ruth Noorduyn, Herwig
Cleuren, Gerard Barendse and Paul Loth enriched my stay in Leiden. Evert Meelis
advised on statistical analysis. I am grateful to Herbert Prins, Pieter Ketner, Frank
Langeveld and their colleagues of the Resource Ecology Group, Wageningen University, for their feed-back and stimulating discussions; Willemien Schouten kept the
communication going. A rewarding part of my work was with students and later
colleagues. Martine Graafland, Marlijn Hoogendoorn, Peter van der Jagt, Saskia van
der Klundert, Bo Oosterhuis and Claudia van der Pot put in a vast effort to produce
the data of the vegetation production studies. The ornithological studies by Selvino
de Kort, Merlijn van Weerd, Ernée Raspe, Suzanne van der Giesen, Corinthe
Zekveld and Hellen Elissen were a background for Chapter 5. Mark Mortiz volunteered in the beginning of the pastoral studies and has continued with us along
the pastoral road ever since.
342
Floodplain rehabilitation and the future of Conservation & Development
I am further grateful to Joost Brouwer and Piet Wit for commenting on many versions of all chapters. Patrick Denny (IHE-UNESCO) reviewed the entire manuscript with formal and constructive informal comments. Approximately 70 reviewers of the various chapters overloaded me with hundreds of hours of work. Ultimately, many of their comments helped to sharpen my analyses and arguments,
resulting in the presented chapters. Many reviewers have remained unknown, but
in addition to above-mentioned people, the following persons have contributed to
the final shape of the chapters. Jan Bakker and Ab Groothuis (Groningen University) and Dick Visser (Nijmegen University) commented on the vegetation studies. Wim Mullié and Jan Wanink stimulated me to open my field notebooks and
publish my ornithological observations. They also commented upon the subsequent
manuscripts, as did Leo Bruinzeel and Eddy Wymenga (AW ecological consultants), Leo Zwarts (RIZA) and Marcel Klaassen (NIOO). Jeroen Bredenbeek and
his WIWO colleagues provided the 1999 and 2000 waterbird counts data, as did
Patrick Triplet (RN de la Baie de Somme) with the Senegal Delta data. Rod East,
Chris Geerling, Bart van Lavieren and Peter van Bree commented on the antelope
chapter. I am also grateful to Jean Ngog, Jean Thal and Louis Tsague (Ecole de
Faune), Daan Bos, Luc Hebou and Hannie Korthof for the use of their unpublished antelope data. Han van Dijk (Africa Study Centre), Leslie Moore and Sean
Overland commented the pastoralists chapter. Geoffrey Wandesforde-Smith (UCDavis) edited the management planning chapter. Nancy Gellman, Deo-Gratias
Gamassa and Freddy Manongi (Mweka), Fanie Greyling and Lynn Pullen (SAWC)
and Rob van Haarlem (Wageningen University) contributed with their information and comments on the curriculum development and training chapters.
Passion for nature made this study fun. I am particular grateful to my old Ooijpolder and Evros mates Jeroen en Wouter Helmer for accompanying me in this passion. With the stimulation and love of my parents I have been able to explore the
directions of my life. Ever since the start of this study, Stephany Kersten has supported me with her love and help. Our children Thyne and Nika show there is
even more in life.
© 2005 Paul Scholte, Amsterdam
Lay out: Sjoukje Rienks, Amsterdam
Photos: Paul Scholte, Amsterdam
Maps: Yde Bouma, Leusden