Floodplain rehabilitation and the future of conservation & development : adaptive management of success in Waza- Logone, Cameroon

(1)

Floodplain rehabilitation and the future of conservation &

development : adaptive management of success in

Waza-Logone, Cameroon

Scholte, P.T.

Citation

Scholte, P. T. (2005, November 23). Floodplain rehabilitation and the future of conservation & development : adaptive management of success in Waza-Logone, Cameroon. Institute of Environmental Sciences (CML), Leiden. Retrieved from https://hdl.handle.net/1887/4290

Version: Not Applicable (or Unknown)

License: Licence agreement concerning inclusion of doctoralthesis in the Institutional Repository of the University of Leiden

Downloaded from: https://hdl.handle.net/1887/4290

Note: To cite this publication please use the final published version (if

(2)

FLOODPLAIN REHABILITATION AND THE FUTURE OF CONSERVATION& DEVELOPMENT

(3)

Tropical Resource Management Papers, No. 67 (2005); ISBN 90-6754-953-3

(4)

Floodplain Rehabilitation and the Future

of Conservation & Development

Adaptive management of succes in

Waza-Logone, Cameroon

(5)

(6)

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 im-proved management without substantial inputs was out of scope under the reign-ing economic conditions. In Waza-Logone, an area still endowed with spectacular wildlife, we could start this unique experiment, improving the livelihoods of fish-ermen and pastoralists, rehabilitating wildlife habitat and, ‘on top’, triggering an improved management system. This study follows the quest of preparing, imple-menting, 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, 1993-1997) and at the Ecole pour la Formation des Spécialistes de la Faune (Garoua, 1998-2000). Temporarily back in the Netherlands, various organisations request-ed 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 anoth-er challenge when teaching an MSc course on rangeland management at the Uni-versity of Ndjamena. I was captured by the exchange of experiences in professional education. In 1998 I was asked to initiate Garoua’s community conservation cur-riculum. Supervising students during their field research has been amongst my most rewarding experiences.

(7)

The process of writing 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 Waza-Logone and Garoua work, became small projects with their own objectives and con-clusions. 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 large-ly descriptive and the ones that did not sufficientlarge-ly contribute to the connecting (sci-entific) thread of this study, see annexed list of background publications.

The first half of this study ‘Impact of reflooding’ is predominantly based on eco-logical 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 person-nel 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.

(8)

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 flood-ing season and not necessarily the calendar year when the consequences of the refloodflood-ing 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 + flood-ing) 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 descrip-tions not specifically related to ecological monitoring (Chapters 2, 8-11), years are indicat-ed 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).

(9)

(10)

PART I

Introduction

1 General Introduction and Study Outline

2 The Ecological History of Waza-Logone: Constructing a reference

(13)

(14)

1

(15)

(16)

1.1 Introduction to African floodplains and their rehabilitation

‘From a European’s point of view Nuerland has no favourable qualities, unless its severity be counted as such, for its endless marshes and wide savannah plains have an austere, monotonous charm. It is throughout hard on man and beast, being for most of the year either parched or a swamp. But Nuer think that they live in the finest country on earth and, it must be admitted, for herdsmen their country has many admirable features’

Evans-Pritchard (1940) introducing the ecology of the Sudd (South Sudan), the largest stretch of seasonally flooded grassland in Africa.

African floodplains

The fertility of floodplains is legendary and some of humankind’s main civilisa-tions, 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 seasonal-ly 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 sta-tus (WCMP 2004).

(17)

16 Part I – Introduction

No and name Country Area at peak flood (km2) Floodplain type

1 Senegal Delta Senegal 8 000 Coastal delta 2 Senegal Valley Senegal 5 000 Fringing floodplain 3 Inner Niger Delta Mali 30 000 Internal delta 4 Niger fringing plains Nigeria 5 000 Fringing floodplain 5 Niger Delta Nigeria 36 000 Coastal delta 6 Volta River Ghana 8 500 Fringing floodplain 7 Benoué River Nigeria 3 100 Fringing floodplain 8 Benoué River Cameroon 1 000 Fringing floodplain 9 Logone Floodplain Cameroon 5 000 Fringing floodplain 10 Chari and Logone Chad 63 000 Fringing floodplain

11 Congo River Congo ? Fringing floodplain

12 Barotse plain Zambia 10 750 Fringing floodplain 13 Kafue Flats Zambia 4 300 Fringing floodplain 14 Okavango Botswana 17 000 Internal delta 15 Shire River Malawi 1 000 Fringing floodplain 16 Kifakula Depression Congo 1 500 Fringing floodplain 17 Kamulondo Congo 12 000 Fringing floodplain

18 Sudd Sudan 92 000 Fringing floodplain

19 Tana Delta Kenya 1000 Coastal delta

(18)

Since about 5000 BC, when the earliest systematic colonisation of the Nile, Meso-potaminan 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 re-source base on which local communities and wildlife depend, with devastating impacts (McCully 2001).

Few of the large African seasonally flooded grasslands are effectively protected, al-though 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 assess-ments 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 hydrolog-ical interventions in African floodplains, showing that their costs, including envi-ronmental 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 orig-inal functions.

(19)

Unfortunately the knowledge about African wetlands remains poor (Denny 2004) and the few studies on the impact of processes, such as water releases from reser-voirs, 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 meas-ured 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 con-servation approach, studied by predominantly natural sciences. Larger-scale reha-bilitation efforts almost always associate people, and require interactions of natu-ral and human sciences. The subsequent increasing complexity makes it particu-larly 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, nature rehabilitation cannot be isolated from the rehabilitation and management of the landscape...nor can the landscape be considered in iso-lation from various socio-economic functions. The integration of these layers of interest forms one of the biggest and most urgent challenges of spatial planning in the Netherlands’ (Schouten 2001).

and

‘...experiences in Europe with the rehabilitation of large rivers are rare, relative to smaller streams, due to the large numbers of interwoven societal demands on the river, the connected economic costs and the complexity of the physical and biological systems involved. Proposals and concepts for large river restoration are much more abundant than demonstrations in the field. The empirical large-scale testing of the models, connected to the political willingness of catchment-scale restoration, are forming the real bottlenecks’ (Nienhuis et al. 1998).

The Everglades (USA) hosts the best-known large-scale (sub) tropical wetland

rehabilitation program, aiming at restoring parts of the 5000 km2degraded area

(Kiker et al. 2001). Compared to African seasonally flooded grasslands, the Ever-glades’ 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 approxi-mately 10 billion USD. In contrast, the budget of the Waza-Logone floodplain

reha-bilitation totals 20 million SFr, i.e. 1% of the recovery cost per km2of the

Ever-glades (Kiker et al. 2001; IUCN 1999). Despite the multitude of technical studies,

(20)

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 possi-ble 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 restora-tion 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 con-servation (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

‘Mourir pour les éléphants’ (Verschuren11970), ‘Les guardiens de la vie sauvage’

(Dupuy2& Dupont 1984 and ‘An impossible dream’ (Parker & Bleazard32001).

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 pro-tected against Africans’. The new ‘leitmotif’ was ‘conservation with, instead of against people’ and ‘win-win scenarios’ (Inamdar et al. 1999). This new conserva-tion paradigm, so-called Community Conservaconserva-tion, was considered a means of re-conciling Conservation and Development by ensuring that the interests and knowl-edge 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 re-sources. Alternatively, compensations should be provided for the opportunity

(non-1 – General introduction and study outline 19

1 Former director of wildlife of DRC (former Zaire). 2 Former director of wildlife of Senegal.

(21)

exploitation) costs of protected area resources by the initiation of small-scale devel-opment projects. For some, community conservation was also considered a prag-matic 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 Con-servation concept in Africa was based on experiences of which the CAMPFIRE

ini-tiative on wildlife utilization in Southern Africa is one of the best known (Child4

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 Com-munity Conservation and pleaded for a return to the ‘fortress Conservation’ ap-proach (Oates 1999; Spinage 1998). This so-called ‘pro-park lobby’ is in an unlike-ly 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. Com-munity Conservation has its limits when wildlife resources or socio-economic cir-cumstances do not allow their sustainable use in sufficient quantities compared to the opportunity costs of the protected areas (Adams & Hulme 2001). The poten-tial 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. Ex-periences 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;

(22)

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 Conser-vation-Development approaches that should also take into account the risks of development. At present Conservation-Development approaches through Inte-grated 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 Com-munity Conservation, the role of protected area managers has been further erod-ed into a caretaker with an agenda (i.e. management plan) fillerod-ed in at the nation’s capital. This situation contrasts the early days of independence when African Park

wardens were amongst the best-educated personnel in their country.5Early

grad-uates 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’.

1 – General introduction and study outline 21

(23)

1.3 Reflooding in Waza-Logone

6

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 re-flooding aimed to rehabilitate vegetation, birds, wildlife, pastoral grazing and fish-eries 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 monitor-ing of the impact of the refloodmonitor-ing 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 con-struction 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

rainfall,7the 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 collabo-ration 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 pro-duction. 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 suc-cessful and in the early 1990s half of the irrigated rice scheme area was left un-used. 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 Conserva-tion Union (IUCN), the Institute of Environmental Sciences (Leiden University), the Netherlands Development Organisation (SNV) and WWF-The Netherlands.

(24)

1 – General introduction and study outline 23

(25)

The overall goal of the project was to assist the government of Cameroon in pur-suing 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.

Amongst the specific objectives8were:

• 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 op-tions, 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 lit-tle 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

document, to prepare and implement a 1000 km2reflooding in one step, looked

increasingly grotesque.

Although not planned and budgeted for, subsequent discussions held at the Waza-Logone 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 crop-land, 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 communi-ties and authoricommuni-ties the pilot floodplain reflooding was initiated by breaching the

embankment, reflooding from September 1994 onwards 180 km2of desiccated

floodplain and raising water levels in an additional area of approximately 600 km2.

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.

(26)

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 rela-tively 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 pro-grammed with local communities, sedentary as well as nomadic, protected area authorities, traditional and administrative authorities and the Ministry of Environ-ment and Forestry. Experiences with the impleEnviron-mentation of this manageEnviron-ment 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 as-sumed flooding – productivity relationships. Monitored changes in waterbird and antelope numbers questioned a simple recovery mechanism and highlighted in-teractions of wildlife with competing land uses, in particular pastoralism that showed a steady increase in grazing intensity. With an impact zone of several hun-dred 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 avail-able 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 plan-ning and traiplan-ning. 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.

(27)

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 dia-gram shows the link between available information and the reference image for floodplain rehabilitation. Presented information also stresses the ecological signif-icance of Waza-Logone, even in its degraded post-dam period, justifying the reha-bilitation 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

Waza-Logone floodplain to its pre-dam structure and conservation and development functions?

In other words, How resilient is this floodplain environment after the dam

construc-tion? 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 fol-lowing 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 peri-od 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 overexploit-ed again. In this study the question of ecosystem restoration is answeroverexploit-ed 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 mecha-nisms with which these changes take place?

(28)

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, char-acterised by the perennial grasses Echinochloa pyramidalis, Vetiveria nigritana,

Hyparrhenia rufa and Oryza longistaminata (Wesseling et al. 1994). This optimism

was based on observations in a 40 km2area in the central, relatively well flooded

floodplain where the water level was raised by the construction of a small dam (Drij-ver & 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; Leen-dertse 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 range-land 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 sud-den 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

(29)

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 occa-sional, qualitative observations. The only quantitative information on the occur-rence 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 Black-crowned 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 rela-tively favourable rainfall that characterised the period since the reflooding, may also have influenced the observed waterbird numbers. This also holds for protec-tion, 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 promi-nently 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 popula-tions 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 es-timated 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

(30)

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 graz-ing intensity than resources available, a territorial scenario with a lower grazgraz-ing 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 Management 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 ana-lyse if comparable risks exist in other Integrated Conservation-Development Projects (ICDPs) and how these are dealt with. The immigration risks motivated the devel-opment 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 whole: Management planning

Participatory studies showed that exploitations from inside Waza National Park, despite their illegality, have continued under the presence of, till recently, fair num-bers 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

(31)

from inside Waza National Park. The plan also allowed the creation of a manage-ment committee in which, for the first time, park authorities and local communi-ties discussed problems related with the park management. I discuss the formu-lation process of this management plan and review its outcome based on an inde-pendent 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. internation-al 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 train-ing 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

plan-ning. These pilot courses were developed at the Garoua Wildlife College during the Waza management plan formulation and implementation. Presented experi-ences are also relevant for other protected areas in West and Central Africa.

PART IV – Floodplain Rehabilitation and the Future of Conservation-Development: 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 contribu-tions 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 tar-gets as programmed in Waza-Logone (IUCN 1999) and possibly in other (African) floodplains, are discussed in section 12.6. I conclude with reviewing the overar-ching adaptive management approach that is introduced below.

(32)

1.5 The quest for adaptive management

Adaptive management, can be loosely defined as the learning by doing, relies on an accumulation of credible evidences to support a decision that demands action (Walters & Holling 1990).

In addition to the discipline specific discussions introduced in the preceding para-graphs, I would like to present this book as test of an integrated approach of trop-ical 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 ecosys-tems. 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 Hol-ling (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, mon-itoring 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 manage-ment. For some floodplain resources, i.e. waterbirds, lack of any quantitative ref-erence information allowed us initially only to pursue a trial and error approach as outlined in the previous paragraph. For other resources, most notably perenni-al vegetation and floodplain antelopes, we started with a single hypothesis ‘requir-ing 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 man-agement and alternative hypotheses of floodplain rehabilitation were formulated, amongst others related with favourable rainfall conditions and unfavourable con-ditions related to human pressure.

This thesis follows a thread of adaptive management: From [I] the conduct of

in-ventories 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 solu-tions are planned through management planning and reviewing (Chapter 9) and

capacity building is undertaken for future Conservation & Development (Chapters

(33)

10, 11). [V] In the synthesis the contributions of the undertaken approach and the initially postulated hypotheses are evaluated and reformulated.

(34)

Floodplain Rehabilitation and the Future

of ‘Conservation & Development’

A PHOTOGRAPHIC OUTLINE

(35)

PART I – Introduction

The ecological background of the floodplain

The Adamaoua mountains, situated 500 km south of Waza-Logone, 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.

(36)

The ecological background of the floodplain 35 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’.

(37)

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

(38)

The ecological background of the floodplain 37 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.

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.

(39)

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.

(40)

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.

(41)

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.

(42)

The floodplain people and the resources they use 41 Young men herd the cattle of the agropastoralists.

(43)

The ecological impact of the Maga dam

Deserted quarter of Mahé, in the north-east corner of Waza NP. After the Maga dam construction approximately one third of the sedentary and mobile population of the floodplain left.

Dead Kob antelope. Following the Maga dam construction the Kob pop-ulation 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.

(44)

The ecological impact of the Maga dam 43 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.

(45)

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 Came-roon, 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.

(46)

The lead-up to the reflooding and the first water coming through 45 Breaching of the embankment in 1994.

(47)

Car stuck near Andirni

The Waza-Logone proj-ect 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 fur-thest monitoring sites.

Tracks become naviga-ble for canoes during the flooding season.

(48)

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 flood-ing and vegetation dynamics were intensively monitored. In 1994, prior to the reflood-ing, the grid was largely cov-ered by Sorghum arundina-ceum, a stout annual grass that invaded the floodplain during the absence of flood-ing 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.

(49)

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 reflood-ing. In 2003, almost the entire reflooded area was covered with perennial grasses again.

In the foreground (bi-) annual vegeta-tion with its large open spaces, which allowed a recolonisation by the perenni-al floodplain grasses through laterperenni-al rhi-zomatous growth (from the back-ground).

(50)

Impact of the reflooding on vegetation production

Impact of the reflooding on vegetation production 49

This study clarified the link between maximum flood depth and vegetation production. With a 20cm rise in maximum waterlevel through reflood-ing, above-ground biomass increased with approxi-mately 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.

(51)

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.

(52)

Impact of the reflooding on waterbirds 51 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 breed-ing 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.

(53)

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 bene-fit 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.

(54)

Impact of the reflooding on antelopes

Impact of the reflooding on antelopes 53

Park guard counting animals at Gobe waterhole in Waza NP. At the end of the dry season, only a hand-ful 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.

(55)

The responses of pastoralists to the reflooding

Discussions with agropastoralists from Fadaré. Responses of mobile pastoralists to the floodplain rehabil-itation 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.

(56)

The responses of pastoralists to the reflooding 55 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.

(57)

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 com-pletely 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.

(58)

The risks of a Conservation & Development project that is too successful 57 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.

(59)

The difference that management planning can make

Practising a new fire policy in Waza NP. Traditionally, management plans focus on technical and ecologi-cal issues, which were also included in the Waza Management Plan.

Elephant poached inside Waza NP. Also with increased local community involvement in conservation, anti-poaching remains an important activi-ty because of the continuing presence of well-armed commercial poachers. Removing mud out of the

(60)

The difference that management planning can make 59 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 com-munities are consulted on park management issues.

(61)

View of Halé village

Building the new village of Halé, outside the national park. Whereas decades of oppression towards Baram, the village in Waza NP, only caused frustra-tion, the new more balanced policy of the ‘carrot and stick’ resulted in the voluntary move-ment of more than half of its population to Halé.

Water pump that was installed at Halé with the support of the Waza-Logone project.

(62)

The importance of developing capacities of protected area managers

The importance of developing capacities of protected area managers 61

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 con-straint for Integrated Conservation & Development Programmes elsewhere in Africa.

(63)

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.

(64)

The importance of developing capacities of protected area managers 63 Garoua students practising Participatory Rural Appraisal, a field exercise in the newly developed Com-munity 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 imple-ment training for protected area personnel in community conservation.

(65)

(66)

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’

(67)

(68)

2.1 Historical periods and main ecological characteristics

Rainfall and flooding are the dominant factors, which trigger changes in vegeta-tion composivegeta-tion and producvegeta-tion (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 re-silience 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, hydrol-ogy and the impact of human interventions and, above all, the quantity and quality of available sources of information. Information sources prior to the 1930s encom-passed 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 lev-els 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 & Levt-zion 1981; Rauchenberger 1999). One often has the impression that these second-ary 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 millen-nium, 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

Floodplain rehabilitation and the future of conservation & development : adaptive management of success in Waza- Logone, Cameroon

Floodplain rehabilitation and the future of conservation &

development : adaptive management of success in

Waza-Logone, Cameroon

Floodplain Rehabilitation and the Future

of Conservation & Development

Adaptive management of succes in

Waza-Logone, Cameroon

Preface

Table of Contents

PART I

Introduction

1

1.1

Introduction to African floodplains and their rehabilitation

1.2

Introducton to ‘Conservation & Development’ and capacity

building needs

1.3

Reflooding in Waza-Logone

1.4

Research questions and study outline

1.5

The quest for adaptive management

Floodplain Rehabilitation and the Future

of ‘Conservation & Development’

A PHOTOGRAPHIC OUTLINE

PART I – Introduction

The floodplain people and the resources they use

The ecological impact of the Maga dam

The lead-up to the reflooding and the first water coming through

PART II – The Impact of Reflooding in Waza-Logone

Impact of the reflooding on vegetation dynamics

Impact of the reflooding on vegetation production

Impact of the reflooding on waterbirds

Impact of the reflooding on antelopes

The responses of pastoralists to the reflooding

PART III – Enhancing Conservation – Development Integration by

Management Planning and Training

The risks of a Conservation & Development project that is too successful

The difference that management planning can make

The importance of developing capacities of protected area managers

2

The Ecological History of Waza-Logone:

Constructing a reference image for

floodplain rehabilitation

2.1

Historical periods and main ecological characteristics