The Senegal River:

Flood management and

the future of the valley

Adrian Adams
Adrian Adams has lived for twenty years in Senegal and works with a farmers
organisation in the Senegal River Valley. In her latest book (with Jaabe So)A
Claim to Land by the River: a Household in Senegal 1720-1994 she depicts
the 20 year struggle of a farmers organisation to defend its vision of people-
centred development against the State development corporation in charge of
irrigation projects in the Senegal River Valley (see Haramata N32, page 23).
For further information, please contact the author at the following address: BP
11, Kounghani via Bakel, Senegal. Fax: +221 983 52 56. E.mail:
[email protected].
 CONTENTS

INTRODUCTION 1
THE 1970S: THE FUTURE ACCORDING TO OMVS 1
IRRIGATED CROPPING: AN IMPASSE 3
ARTIFICIAL FLOODING: A BROKEN PROMISE 6
IS THERE NOW AN AGRICULTURAL POLICY FOR THE RIVER? 9
THE CAYOR CANAL AND FOSSIL VALLEYS PROJECTS 12
THE MANANTALI ENERGY PROJECT 15
FLOOD SUPPORT: WHAT IS POSSIBLE? 20
REFERENCES 26

MAPS

Map of the Fossil Valleys: the Cayor Canal project 13

Manantali hydropower project: Supply lines 16
INTRODUCTION

The River Senegal rises in the Fouta Djallon and flows northward through
increasingly arid land; when it finally turns west towards the ocean, it borders
on desert. In these areas of low rainfall, the rivers annual flood is necessary to
life. Towards the end of the rainy season, it overflows its banks and floods the
broad alluvial plain of the middle valley, where crops are grown in the dry
season after the waters have receded. The valleys agricultural production
systems traditionally followed the seasonal rhythm of the river: rainfed
cropping and pasturing on the jeeri uplands, followed by flood-recession
farming and grazing on the waalo lowlands. Over the period 1946-1971, it is
estimated that on average 312,000 hectares were flooded every year on both
banks of the river, and 108,000 hectares cultivated; on the Senegalese side of
the river, 65,000 hectares were cultivated. (OMVS-IRD, 1999).

Projects to develop the potential of the river, based since colonial times on
irrigated rice-growing, have never taken this age-old production system into
account. Since the 1960s, rainfall and river flooding have declined
considerably, and in some years have been non-existent. The drought simplified
things for would-be developers by allowing them to behave as if the valleys
traditional agricultural and pastoral production systems were a thing of the past,
and the future belonged to irrigation alone. When Senegal ratified the OMVS
programme, this clean sweep policy became irrevocable. The planned dams
would not stop the rains, but they would make it possible to cut down
considerably on annual flooding.

THE 1970S: THE FUTURE ACCORDING TO OMVS

The Organisation pour la Mise en Valeur du Fleuve Sngal (OMVS), the

members of which are Senegal, Mali and Mauritania, was established in 1972.
It replaced the Organisation des Etats Riverains du Sngal (OERS), founded
in 1968, which was broken up after the withdrawal of its fourth member,
Guinea. Its appointed task was to promote and co-ordinate the development of
the river basin on the territory of the three member states.

In the year of its founding, OMVS stated its aims as being: to provide a secure
and steadily improving livelihood for the inhabitants of the river basin and
neighbouring areas; to safeguard as far as possible the ecological balance of the
river basin; to make the economies of the three member states less vulnerable
to climatic conditions and external factors; and to accelerate the economic
development of the member countries by the intensive promotion of regional
co-operation.

In 1973 it announced its programme, based on the construction of two dams:

one upstream, at Manantali in Mali, which would retain the waters of the
Bafing; and one at the mouth of the river, which would stop salt water entering
the Delta and lower valley. The programme included three components. The
first was irrigation. The second was navigation: the river was to be made
navigable between Saint-Louis and Kayes throughout the year. The third was
energy, with the construction of a hydro-electric power station at the foot of the
Manantali dam.

How would irrigation contribute to achieving the declared objectives? The

twelve-volume Programme intgr de dveloppement du bassin du Sngal
stated that the aim was to bring under irrigation between 300,000 and 400,000
hectares of land, which would yield two crops a year. Between 1975 and 1983,
the area under cultivation would increase from 6,356 hectares in the rainy
season and 3,250 hectares in the dry season to 72,841 and 58,776 hectares
respectively. In other words, the area under cultivation would grow
fourteenfold in nine years: an average annual increase of 13,500 hectares. The
main crops would be rice and wheat, while the areas devoted to sorghum and
maize would gradually diminish. For twenty years, the rivers annual flood
would be artificially maintained, to allow for traditional flood-recession
farming, but the level of flooding would be reduced from year to year; by the
end of the twenty-year period, it was expected that the entire population of the
valley would be engaged in irrigated farming (PNUD-OMVS, 1974).

It was reckoned at the outset that the programme as a whole would extend over
40 years, with an overall budget of 800 billion CFA francs, 280 billion of
which would be spent on agriculture. At the end of the 1970s, the cost of the
Diama dam was estimated at 34 billion CFA, that of the Manantali dam at 102
billion. By the early 1980s, funding had been obtained for the first phase and
work could begin .

Many people warned of the predictable consequences of the OMVS

programme:. agronomists, economists, sociologists and journalists, most of
them foreigners. Their warnings went unheard; and the people whose interests
they were defending the inhabitants of the valley were not even consulted.

Box 1. Critics of the dams

Just under a year ago, France asked for alternative solutions to be considered. A study was
carried out, but only a few insiders got to see it... and quickly hid it away. No wonder: for
it showed not only that the OMVS programme was not essential to food security in the
region, but that in many respects it was the worst possible solution... (...)

More and more voices are being raised to demand that the project be re-examined before
embarking on such a gigantic undertaking , and that the full range of conceivable solutions
be considered . This would make it possible, within a reasonable period of time, to devise a
programme ensuring the harmonious and productive economic and social development of a
region which, after a long period of neglect, is now in danger of being handed over to
sorcerers apprentices (Bessis, 1981).

Who will benefit from these big dams? OMVS tells us that a key objective is to guarantee
and improve farmers incomes. Once again, the farmers are being used as a pretext.
Supposedly, it is all being done for them. But if we look more closely, we find that the dams
have already benefited:
research consultancies, which have made billions of francs in fees and are hoping to
make a lot more;
the bureaucrats of the OMVS and their counterparts in each of the three countries.

Once the decision is taken, and construction work begins, the major beneficiaries will be the
big public-works contractors, along with the firms that supply them with all kinds of
equipment... The planned operations will be a source of considerable profit for all these
vested interests. When the dams are completed, they will leave the farmers and the countries
concerned to try and make the best of it all; but they will wash their hands of the failures
and difficulties which are bound to occur (Dumont, 1981).

The development of the Senegal River is a vital task for the peoples of the River and the
countries through which it flows. However, if it were to be carried out with the aims and
methods envisaged at present, it would be better for them that it not be undertaken at all
(Adams, 1977).

IRRIGATED CROPPING: AN IMPASSE

There were trials of irrigated rice-growing in the Delta as early as the 1950s.
But it was not until 1973 that SAED, the state corporation responsible for
developing agriculture on the left bank of the river, extended its activities from
the Delta to the Valley itself. At Nianga, in Podor province, it constructed
dikes around 10,000 hectares of flood-plain land and, in 1975 built the first
large-scale irrigation scheme, 650 hectares. In 1975, it extended its activities as
far as Matam, where an agent of SATEC (a French consultancy), sent to the
area in order to improve yields of sorghum production, created three small
irrigated rice-growing schemes, 25 hectares in all, with 150 drought-stricken
farmers; and Bakel, where irrigated rice-growing was introduced by a French
agricultural technician who had come to help a local farmer improve
subsistencefarming in the area .

Small village-level irrigation schemes (primtres irrigus villageois, or PIV)

developed rapidly in the Valley: from 20 hectares in 1974, to 7,335 hectares
(i.e. 29% of the total area brought under irrigation) in 1983, and 12,978
hectares by 1986. Should this be regarded as a success? That is how it was
seen at the time. But it proved to be a dubious success, as it was based on
exceptional technical, economic and social conditions. The PIV were developed
during a time of drought, which devastated rainfed crops, severely reduced the
areas available for flood-recession farming, and killed off much of the areas
livestock and fish. Thus for local farmers, irrigated agriculture was the only
possible activity at the time. Irrigation equipment was supplied free of charge,
and SAED provided inputs and services in the form of seasonal advances at
subsidised prices (over 50% in some cases), to be repaid after the harvest.
Irrigation networks were built on the light soils of the river bank, which were
not highly valued in the traditional system of agriculture and were thus
generally available for irrigation. .

If the drought and the subsidies had continued, it would have been possible to
go on believing that the government and local farmers were pursuing the same
objectives. But the changes which occurred in the 1980s (improved rainfall,
withdrawal of subsidies and tougher conditions for access to credit, locating of
irrigation schemes on flood-recession land surrounded by dikes to prevent
flooding) brought latent contradictions out into the open . Structural adjustment
programmes entailed new policies, such as promoting private-sector activity
demanding that farmers take on responsibility for their own affairs, and
reducing levels of State intervention, which heightened inequalities between
poor farmers hard hit by the withdrawal of subsidies, and those with the means
and connections required to take advantage of the new circumstances.
Thenceforth, irrigation began to create problems for social cohesion and access
to land, especially in the Valley.
At the end of the 1980s, more than two-thirds of irrigation schemes were
located in the Delta, defined for our purposes as the area downstream from
Dagana: large schemes built by SAED, many of which had been rehabilitated
and transferred to groups of producers, and schemes belonging to the rapidly
growing private sector. In this area, where there had been little farming before
the advent of irrigation, irrigated farming was often undertaken by new
farmers with other means of subsistence; it was highly mechanised, and most
of the harvest was marketed. The main difficulties declining yields, pollution
and salinisation of poorly drained lands were due to the anarchic spread of
poorly-constructed private irrigation schemes,, without any proper planning of
water use. There were also serious difficulties with the repayment of loans.
Delta land requires a level of investment which few of its present users,
whether peasant farmers or or private developers , seem able to achieve.

At the same time, less than a third of irrigation schemes mainly rough-and-
ready first-generation PIV were located in the Valley (upstream from
Dagana). In 1989-1990, only 121 of the 215 irrigated PIV in the dpartement
of Matam were wholly or even partly under cultivation. All the first-generation
PIVs were run-down and no longer in use. Some of the second-generation
schemes, established on flood-recession farmland, had yielded poor results. In
the dpartement of Bakel, a study of 823 hectares of PIV revealed that 40% of
the total area was not being farmed in 1988-1989.

In the Valley, as opposed to the Delta, little of the harvest is marketed, and
there are few non-agricultural resources apart from emigration. In a subsistence
society, when production costs rise, farmers will decide to use their resources
of money and labour elsewhere because those resources are limited and
essential to their survival. But what is in itself a rational choice in difficult
circumstances, could have the unfortunate effect of excluding peasant farmers
(defined as persons whose only resource is agriculture) from irrigated farming.
And who would benefit? In the Valley, as opposed to the Delta, applications
for land come mainly from local residents: traditional leaders, traders, men
working abroad, who register of their familys land in their own names, thus
using their traditional rights as a launching-pad for commercial agriculture,
usually on fairly small irrigation schemes.

By the end of the 1980s it was clear that if poor farming families were to
continue to depend on irrigated farming alone for subsistence, the only way
they could make ends meet, apart from emigration, which would deprive
them of the necessary manpower would be by working as agricultural
labourers, which implies the existence of a flourishing private sector. The
danger is therefore that in the absence of modes of agricultural production other
than irrigation (in particular flood-recession farming), independent peasant
farmers will cease to exist as a class, as a result of pauperisation and
proletarianisation; whether directly, when those lacking the means to irrigate
are deprived of access to water, and therefore to land; or indirectly, through
dependence on jobs created by alienating a growing proportion of irrigable
land, which will in turn jeopardise young mens chances of gaining access to
irrigated farming .

Diama was completed in 1986, Manantali in 1988; the two dams were
officially inaugurated in 1992. As we have seen, they were built during a
period of growing difficulty for irrigated agriculture, with mounting debts,
ecological degradation of the Delta, and irrigation schemes being abandoned in
the Valley. A few years later, it became clear that the production goals put
forward to justify the OMVS programme were far from being attained. Not
only had the social objectives (a secure and improved livelihood for the
inhabitants of the River basin) not been achieved; in at least one respect, they
had been deliberately set aside. For during the first few years after the dams
came into service, when there was a clear need for artificial flooding to provide
the inhabitants of the Valley with a degree of food security, OMVS did not
keep its promise.

ARTIFICIAL FLOODING: A BROKEN PROMISE

In 1986, the year before the Manantali dam came into commission, there was a
good flood. In 1987 there was no flood at all, as the reservoir behind the dam
was being filled. In 1988, there was an artificial flood of modest but
satisfactory proportions: harvests would have been good but for an invasion of
locusts. But in 1989 OMVS caused major losses by allowing two floods to take
place. An initial natural flood, from the rivers two unregulated tributaries,
receded fairly rapidly, and farmers had begun to sow their flood-recession
crops when a second artificial flood, unleashed for technical reasons,
drowned the seedlings in low-lying areas. Many farmers, lacking seed and
labour, were not able to sow another crop. In 1990, despite drought and the
almost total failure of rainfed crops, OMVS decided not to release any water at
all, but to retain all the water to test the reservoirs storage capacity.
In 1991, with tests of the reservoirs storage capacity completed, an artificial
flood might have been expected. In the Middle Valley, 1991 had been a year of
drought; it was an excellent opportunity to show how the dam could benefit
local communities. On September 1 and 2, approximately 1000 m3/s was
released to join the natural flood of the unregulated tributaries. Two days later,
the river reached a flow rate of approximately 2500 m3/s at Bakel, the
minimum required to produce flooding. If the release rate of 1000 m3/s had
been maintained for about a week, there would have been a good flood and
good harvests on the waalo. But all OMVS wanted to do was keep the water
level in the reservoir at around 206 metres above sea level; once this had been
achieved, they reduced then stopped the release of water. The water level in the
reservoir was then kept at 207 and subsequently 208 metres, close to its
maximum capacity. Even so, so much water was flowing into the reservoir
that it was necessary to make periodic releases of 1000 and even 1500 m3/s.
When the flow of the river reached 2500 m3/s, part of the flood-plain was
inundated and farmers sowed their fields when the water receded. A few
weeks later, as in 1989, a second release from the Manantali dam drowned a
large part of their seedlings.

In September 1992, OMVS released enough water to cause a flow of between

1900 and 2400 m3 at Bakel. It was enough to flood the lower-lying depressions
but not enough to ensure good waalo harvests over the whole of the flood-plain
area. The rains had been poor along much of the Valley and the releases of
water had been, as the OMVS consultants freely confessed, below the
minimum needed to provide a useful flood (Horowitz, 1993).

This failure on the part of OMVS to fulfil its commitments is all the more
striking when we consider that during this period, there was a significant
change in the Senegalese governments position on the issue. OMVS position
had always been that artificial flooding should be a temporary measure, lasting
only ten years. But as early as 1984, the then Minister of Planning, Cheikh
Hamidou Kane, suggested a more flexible position, whereby the annual flood
would be maintained until such time as the farming population of the Valley
had gained access to enough irrigated land to satisfy their basic needs.
Moreover, he said, It may prove necessary to maintain artificial flooding if its
suppression is likely to cause the degradation of natural eco-systems and the
destruction of existing agro-pastoral systems of production (Rpublique du
Sngal, 1984).
In 1987, a team from the Institute for Development Anthropology (IDA) in the
US, embarked on a research programme in Senegal known as the Senegal
River Basin Monitoring Activity (SRBMA). This showed that, area for area,
flood-recession farming yielded better results than irrigated agriculture for an
equal investment of labour and money, while minimising risks. The work of
this team, of a high scientific quality, successfully defended the idea that
permanent controlled flood releases from Manantali, raising the level of the
river to that attained in times of natural flooding, was justified in that it would
increase levels of production, income and employment, while protecting the
environment. They also claimed, contrary to OMVS consultants, that there was
no incompatibility between controlled flooding and the production of electricity.
Their work became a basic reference for any discussion of the future of
agriculture in the Valley (IDA, 1991).

Box 2. IDA: an alternative use for the dams

The solution to these problems does not lie in giving up irrigation but in seeing it in clearer
perspective as one and not necessarily the most important element in a complex
production system. The overall productivity of this system and the manpower invested in it
can be improved by appropriate management of the Manantali dam. But if people are to
make the necessary investment in sustainable agriculture, they must be guaranteed on-going
access to the basic factors of production, in particular land (both irrigated and flood-
recession) and water (p. 23).

We recommend that the dam be managed in such a way as to support a diversified

production system, in which the guiding principle is the concern to achieve equitable
economic growth without damage to the environment. This would mean the dam managers
releasing from the Manantali reservoir the volume of water needed to maximise production
in the flood-plain area, while safeguarding the environment and taking into account the
needs of irrigation and the production of hydroelectric energy (p.39).

The time has come to adopt a new approach to the development of river basins, for a
number of reasons. Firstly, there is growing evidence of the ecological, economic and socio-
political costs of river basin development strategies which greatly reduce annual flooding.
Secondly, more and more stress is being put on development which favours the majority of
rural people on low incomes. Thirdly, African governments are giving greater attention to
the decentralisation of responsibility for decision-making and management. Fourthly, there is
a growing awareness... that success in managing the environment depends on the
participation and advancement of the rural poor. Development strategies which impoverish
these groups and degrade some of Africas most productive eco-systems as has happened
with the development of many river basins in Africa are really a thing of the past (p.318).
Source: Horowitz, Salem-Murdock et al, 1990
When the results of IDAs work were presented at a seminar in Dakar in
November 1990, they were favourably received by the Senegalese government,
which, with its Left-Bank Master Plan (Plan Directeur de la Rive Gauche),
was about to opt for maintaining a yearly artificial flood. However, the then
OMVS High Commissioner declared that this research was an affront to the
authority of OMVS, the only body authorised to decide how the water in the
Manantali reservoir should be used. Moreover, the IDAs hydrology expert
was advised that it was dangerous to raise questions about artificial flooding,
as farmers might begin to think they were entitled to it.

Introduced in this way, the issue of artificial flooding could have led to an in-
depth debate on agricultural policy for the Valley, and the need to choose
between two objectives: making the dams profitable as quickly as possible, and
ensuring the survival of family-based farming. Fairly conducted, such a debate
might have reached the conclusion that opting for the first objective would
ensure that the second could never be achieved; whereas the reverse was not
true. However, despite a few gestures in that direction, such a debate never
took place. Although the question remains alive, its outcome seems more
doubtful than ever; even though it is now widely acknowledged that of all the
consequences of building a dam, changes in the rivers downstream flow are
the most harmful of all for the natural and human environment.

IS THERE NOW AN AGRICULTURAL POLICY FOR THE RIVER?

Thus, by the early 1990s. the development of the Senegal River was in a state
of crisis, having achieved neither its economic nor its social objectives. The
Master Plan for the Integrated Development of the Left Bank (Plan Directeur
de Dveloppement Intgr de la Rive Gauche / PDRG), drawn up in 1990 and
adopted by the Senegalese government in 1994, seemed to recognise this fact.
Its introduction presented a highly critical account of the past twenty years, and
stated that towards the end of the 1980s, so many setbacks and fears for the
future caused something of a change of direction.

The text continued: In the end, there was a rethink of the development
strategy chosen, the idea of maximising irrigated areas being replaced by one of
integrated and harmonious development, seeking to achieve the best possible
compromise between social imperatives (self-sufficiency in food for the local
people), economic imperatives (a return on investment) and ecological
imperatives (restoration and protection of the environment). And it concluded:
The Master Plan arose from this context, and aims to define a development
strategy for the left bank over the next 25 years (Rpublique du Sngal,
1994). In at least one area that of the annual flood this strategy was indeed
new.

The PDRG stated, it is true, that irrigated farming was the best way of
achieving social objectives (self-sufficiency in food and job creation), a view
which seemed to bear little relation to the prevailing situation at the time the
Plan was drafted. And, after political arbitration on the part of the Senegalese
authorities, in collaboration with the Word Bank, it opted for scenario A, out
of a list of five possible scenarios (see box 3). This scenario envisaged
irrigating as much land as possible without jeopardising other uses of the water
(environment, flood-recession farming , hydro-electric power).

The Plan did nevertheless contain an explicit commitment, however cautious,

to maintaining artificial flooding on a permanent basis. And, for the first time,
it defined a strategy which gave a clear albeit limited place to maintaining
flood-recession farming as an enduring component of the Valleys agriculture.
In fact, scenario A guaranteed 33,000 hectares of flood-recession farmland and
63,000 hectares of grazing land. The two initial phases of the PDRG were also
said to be compatible with scenario B1, which would guarantee 50,000 hectares
of flood-recession farming. It is worth remembering that, in the 1946-1971
period, the average area flooded was estimated at 312,000 hectares on both
banks and 65,000 hectares on the Senegalese bank. The maximum area of
flood-recession farming in the years 1970-1979 was 62,200 hectares; the
minimum area 10,700 hectares (OMVS-IRD, 1999).

But what has become of this Master Plan? It seems to have disappeared without
trace. For OMVS, it is as if it never existed. In 1994, after a good flood, the
farmers had begun sowing what promised to be an extensive area. Then a
second flood, caused by emptying the Manantali reservoir, killed off their
seedlings and inundated the cultivable areas for several months; that years
flood-recession farming had to be abandoned. The exceptional flood of 1995
was just as fortuitous, due to a decision not to retain water so that the dam
could be inspected. An IRD aerial survey and photographs showed that, in the
dpartement of Podor, the area brought under flood-recession farming in 1995-
96 was double that estimated by OMVS in 1970-71. This shows how important
this form of agriculture is for the local people, and how quickly they can adapt
to changed circumstances.(Le Roy, 1997). But the following year, the flood
failed to materialise.
Box 3. The five PDRG scenarios

Z The high-productivity scenario, aiming in the long term to develop the largest
possible irrigated area (154,500 ha of food crops) and as much hydro-electric
power as possible, but with no guarantees for flood recession farming or the
environment if the natural flood should fail to materialise.
A major increase in the irrigated area (to 88,000 ha), with artificial flooding to
A ensure over 33,000 ha for flood-recession farming (land submerged for more
than 2 weeks) and approximately 63,000 ha for grazing and forests (land
submerged for less than 2 weeks).

B1 A moderate increase in the irrigated area to a total of 53,000 ha, with artificial
flooding to ensure the inundation of over 107,000 ha, of which 50,000 ha for
flood-recession farming .

B2 Base-line scenario, aiming at long-term preservation of the area so far brought

under irrigation. This scenario would include artificial flooding to inundate
122,000 ha, of which 57,000 ha for flood-recession farming .

C Maximum artificial flooding, guaranteeing 67,500 ha of flood-recession farming

and 61,000 ha of grazing and forests. Where irrigation is concerned, it would
represent a retreat, with only 14,500 ha allowed; it would also cut back on the
production of hydro-electric power.

Source: Rpublique du Sngal, 1994

According to SAEDs own statistics, the total area covered by irrigation

schemes on the left bank in 1995 was 71,751 hectares; and the area actually
cultivated, both seasons included, was 29,792 hectares. (Some sources claim
that irrigated farming has recently revived, but this seems to be due to a debt
moratorium, and thus represents only a temporary improvement , not a lasting
solution). However, SAED is pursuing an ambitious construction and
rehabilitation programme for irrigation schemes, costing a total of 178 billion
francs CFA. Part of this programme has already been carried out, and there is
funding available for a further part. If SAED were operating within the
framework of the PDRG, which advocates devoting 88,000 hectares to
irrigated agriculture, should it not at least have suspended work on new
schemes?

In 1995, the World Bank approved the Agricultural Sector Adjustment

Programme (Programme dAjustement Secteur Agricole - PASA) presented by
the Senegalese government, and this was used as a basis for the Agricultural
Sector Investment Programme (Programme dInvestissement du Secteur
Agricole - PISA). Just like the SAED and OMVS programmes in the past, the
PASA sets both economic and social objectives: while aiming at an average
annual growth rate of 4%, it aims to improve food security, rural incomes, and
natural resource management. But the strategies it proposes deregulation,
reducing the level of state intervention, land-tenure reform are simply a
return to the New Agricultural Policy of the 1980s. The same is true of the
agricultural projects and programmes formulated as part of PISA, clearly
based on the idea that food security will be ensured by massive investment in
irrigated rice-growing, and that growth of 4% can be achieved by a large
increase in the production of paddy (which is meant to almost quadruple by the
year 2000). Similarly, it is still cash crops that are expected to create jobs in
rural areas. And above all, as far as the Valley is concerned, there is no
mention of the PDRG, even though it was adopted by the government only a
year earlier. It was like a return to ten years before.

In fact, the governments actions gave no indication that it had in any way
reconsidered the agricultural policy it had conducted in the Valley since
independence a policy based exclusively on irrigation. On the contrary, the
Valleys future prospects continued to be described in terms identical to those
current in the 1970s. For instance, the Minister of State for Agriculture stated
in 1997: The only solution for us is irrigation (Sud, 17/11/97). It would
seem that the adoption of PDRG, which at least allowed room for doubt on
this issue, was an aberration.

Thus the government refused to acknowledge, despite ample evidence, that its
agricultural policy for the Valley was in difficulty, let alone revise it to take
account of reality. A total impasse was reached. Is this why the 1990s
witnessed the promotion of other schemes for the Valley the Cayor Canal,
the Fossil Valleys, the Manantali Energy Project which relegated agricultural
development to the background ?

THE CAYOR CANAL AND FOSSIL VALLEYS PROJECTS

The Cayor Canal project was intended to divert water from the Senegal River
to the Dakar region via an open canal connecting Lake Guiers with the Cap
Vert peninsula. It was supposed to satisfy the regions drinking water needs and
at the same time provide 8,500 hectares of irrigated land in the Louga, This,
Diourbel and Dakar regions. The cost of the first tranche of engineering work
was estimated at 76 billion francs CFA 1, with an additional 37 billion CFA
francs for constructing irrigation schemes. This project, which was the subject
of a full technical and financial study, now seems to have been shelved because
of reservations on the part of the funding agencies. If one compares it with the
Manantali Energy Project, with its 1,400 km of power lines and overall cost of
223 billion francs CFA (for which funding seems to have been secured), the
main difference appears to be that the Cayor Canal, which was to have been
built by the engineering corps of the Senegalese army, did not hold out the
prospect of lucrative contracts for foreign companies.

Map of the Fossil Valleys and Cayor Canal projects

<Sorry, map not available>

Conceived under the auspices of the Ministry of Hydraulics after the PDRG
had been drawn up, the Fossil Valleys Revitalisation Programme (Programme
de Revitalisation des Valles Fossiles / PRVF) is a plan to restore water on a
permanent basis to 3,000 km of former watercourses: the valleys of the Ferlo,
the Saloum, the Sine, the Baobolon, the Car Car and the Sandougou. The
water would be drawn from the Senegal River at two places, one situated
upstream from Matam, the other at Keur Momar Sarr. The latter site has
already been used, in 1994, to restore water to approximately 150 km of the
lower Ferlo valley, where pilot farms have been set up. The aim of PRVF is
to make better use of the water made available by the construction of the
OMVS dams, by diverting to central areas of Senegal the excess water that
would otherwise be discharged into the sea during the annual flood. The idea is
that this would give a great boost to Senegals economy, in particular
agriculture (with 75,000 hectares to be brought under irrigation ), livestock
farming, inland fisheries and forestry. The estimated cost is 30 billion francs
CFA.

It is only possible to draw water from the Senegal River at the time of the
annual flood (15 August15 October), but the flood is insufficient to guarantee
the quantity of water needed to irrigate these 75,000 hectares without
jeopardising other needs. The fact is that the water regarded as being
discharged into the sea serves to maintain, for a minimum of 2 weeks, the
water level required to submerge low-lying flood-recession farmlands. The
projects promoters nevertheless claim that the volumes of water they plan to
draw off are still well below the quota allotted to Senegal by OMVS. They say

1 Equivalent to US$ 122 millions.

that this quota amounts to 6 billion cubic metres of water annually, while the
main uses being envisaged (PDRG, Cayor Canal, Fossil Valleys) would only
account for 5.4 billion cubic metres. However, this calculation disregards the
water needs of the Manantali Energy Project.

It would seem that the decision to postpone the PRVF was due to protests from
the Mauritanian government. This may be only a tactical retreat. The press
reported remarks made by high-placed government officials during an
electoral period, it is true suggesting that the project has not been shelved. As
far as the lower Ferlo valley is concerned, talks are still going on with potential
investors, and at the end of 1998 the Ministry for Water Development
announced that a study to measure the impact of the project in terms of water
use had been finalised and would be submitted to the heads of state of the
OMVS countries. But it seems unlikely that this project will survive the
installation of turbines at Manantali. Conceivably, in the mind of the decision-
makers, writing off flood-recession farming in the Valley is one thing, but
giving up the prospect of extra electricity for Dakar, Nouakchott and Bamako
is quite another.

THE MANANTALI ENERGY PROJECT

The energy component of the OMVS programme, planned since 1977,

languished for many years due to lack of funding. Then, at the end of June
1997, the World Bank approved a loan of US$ 38 million to help fund the
installation and operation of hydro-electric turbines at Manantali dam. The
Diama and Manantali dams together cost almost 200 billion CFA. The Energy
Project will cost a total of 223 billion CFA, 18.7 billion of which is to be paid
by the governments of the three OMVS member countries. Of the remaining
205 billion, 136 billion will be raised by loans and 67.3 billion in the form of
donations. The main contributors, apart from the World Bank, are France,
Germany and the European Union.

The project involves the installation at Manantali of a hydro-electric power

station with a capacity of 200 megawatts (MW), capable of supplying
approximately 800 gigawatt-hours (GWh) of electricity per annum in a year of
average river flow. In the original version of the project, this electricity was
destined for the development of the mining industry in Mali. According to the
current plan, Manantalis electricity is mainly intended to supply the cities of
Dakar, Nouakchott, and Bamako. It will be transmitted by a network of more
than 1,400 kilometres of landlines, which alone will cost 114 billion francs
CFA: 326 km for the eastern line (Manantali-Kita-Bamako) and 821 km for the
western line (Manantali-Kayes-Matam-Dagana-Sakal) - with a branch from
Matam to Kadi (87 km) and another from Dagana to Rosso and Nouakchott
(226 km) joining the existing Sakal-Tobne line which supplies Dakar.

According to the World Bank, the aims of the project are as follows: to reduce
the cost of electricity in each of three countries; to help service the debt
incurred in building the Manantali dam; to improve the efficiency and
reliability of the three countries electricity networks; to promote private-sector
participation in running the project and other future projects in the Valley; to
set up an effective organisation to build and operate the project installations and
to limit the negative impact of the project and the Manantali dam on the
environment and health; and to support the traditional agricultural sector
downstream of the dam by rational management of the Manantali reservoir
(World Bank, 1997).

Manantali hydropower project: Supply lines

<Sorry, map not available>

As far as reducing the cost of electricity and improving the efficiency of the
networks is concerned, the energy from Manantali (52% of which will go to
Mali, 33% to Sngal and 15% to Mauritania) will reduce the price of
electricity for the Senegalese consumer by only 8%, and will supply only 15%
of SENELECs needs (present capacity 330 MW). It will merely postpone for
three years the need for further investment. Moreover, the electricity
generation projections used in calculating the profitability of the project are
based on hydrological data for the period 1950-1994. If they had been based on
data for the years 1974-94 a period of poor river flow which corresponds
more closely to present-day realities the anticipated amount of energy
generated would drop from 804 to 547 GWh; and the energy savings
anticipated by the governments of the OMVS countries would fall from 22% to
17%. It would have been better to make a thorough examination of the
alternatives natural gas from Diamnadio, the hydro-electric resources of the
Gambia River, reducing the price of fuel by doing away with state subsidies to
the Socit Africaine de Raffinage before embarking on a project the costs of
which (both direct and, as we shall see, indirect) are very high, and the
profitability less certain than one is led to believe.
As for helping to service the debt incurred for the construction of the dams,
roughly 65% of the funding was in the form of loans, and in 1996 the amount
required to service the debt was around 12 billion CFA. The plan is that the
Socit de Gestion de lEnergie de Manantali (SOGEM Manantali energy
management company), to which OMVS has had to delegate its responsibilities
in this field, will pay 43.8% of the interest on the debt out of its revenues, and
the remainder will be paid for by revenues from irrigation (and navigation, if it
ever gets off the ground). However, the loans to fund the Energy Project will
considerably increase the debt: from 47.4 billion CFA in 1997 to 184.6 billion
in 2001.

Private-sector participation in the fruits of the project is a certainty. Quite apart

from all the studies carried out over the last twenty-five years, and the
lucrative contracts for the dams electro-mechanical equipment, civil
engineering, etc. won by European companies, SENELEC (the Senegal
national electricity company) has been privatised (the main acquirer being
Hydro-Qubec), and even the task of producing electricity at Manantali will be
entrusted to a private operator. For, as regards the establishment of an efficient
organisation, the World Bank also stipulated that the newly created SOGEM,
acting through a consortium of companies, should appoint a private operator
to be known as the Socit dExploitation de Manantali (SEM) with a
renewable fifteen-year contract to manage the production of electricity and, by
extension, the day-to-day operation of the dam. SENELEC and the other
national companies will pay SEM, which, after deducting its share of the
proceeds, will pay the rest to SOGEM.

The role of the OMVS High Commission will be limited to monitoring

decisions taken in respect of water management, mitigating adverse effects on
the environment and health, etc. But in this field, unlike that of electricity
production where all organisational and financial matters have been carefully
studied, efficiency seems to count for little: everything has been left very
vague. A plan has been drawn up to minimise and monitor impacts on the
environment (Plan dAttnuation et de Suivi des Impacts sur lEnvironnement -
PASIE), but it explicitly states that it only covers activities directly related to
the Energy Project and which come within the remit of SOGEM, and
therefore does not make OMVS directly responsible. The environmental
monitoring and protection programme and the environmental health programme
have been judged very unsatisfactory when compared to work already done in
this field.
Even the IRD studies on optimising the management of the Manantali reservoir
are relegated to the PASIE framework. Yet these studies are the only evidence
of the World Banks interest in the last of the quoted objectives: supporting the
traditional agricultural sector downstream.

No doubt we should be happy to find, alongside productivity-orientated

objectives (in this case focused on electricity), at least one objective (apart from
health and the environment) presented as being of a social nature or, to use the
Banks terminology, related to the struggle against poverty. We are assured
that the Manantali reservoir will be managed as a multi-use reservoir, and that
artificial flooding will revive the traditional functions of the rivers flood-plain.
But doubt soon creeps in. Can we really believe that, just when Manantali is
about to begin supplying electricity to the three capital cities and generating
income for OMVS, OMVS is going to use the dam to benefit flood-recession
farming and grazing ? A closer look reveals that things are not so simple: the
funding agencies reports contradict themselves and one another.
 Box 4. What should we believe in the reports published by the World Bank...

By addressing some of the key impacts of the existing dam, the project will also improve the
situation of low income rural communities in the river valley... The reservoir management
optimisation program, prepared under the project, will contribute to improve the artificial
flood regime for purposes of traditional agriculture, which is vital for low income rural
households (p.36).

Aa preliminary hydrology study conducted in 1996 by ORSTOM ... indicates that there
would be on average a reduction in flood-recession agriculture from 45,000 to 30,000
hectares ... as compared to the situation without the power plant. (Appendix 11). (World
Bank, 1997)

... and the African Development Bank?

In the long term, the construction of the two dams will make it possible to resolve
population and food problems, through the stabilization of the people and the re-
establishment of the balance of the ecosystem (p.1). The development strategyof the Senegal
river basin aims to establish a balance between man and his environment (p.8)., The
availability of water in the river Senegal all during the year has become possible, thanks to
the Manantali dam. A vast campaign for the occupation of agricultural lands is underway on
both sides of the river.... The significant results of the farming activity in this region which
was formerly unsuitable for agriculture are drawing new populations (p.25). The results
obtained at this stage are impressive: the regulation of the flows of the Bafing in order to
facilitate the double cropping irrigation of the schemes, , and the supply of an annual
artificial flood (p.39). A marked improvement has been observed in the health of the people
living in the valley (p.36). (ADB, 1994)

Although the Energy Project per se will not have any major impact on the environment, it
provides occasion for correcting the negative impact which the Manantali and Diama dams
have had on the basins sensitive ecosystems, traditional flood-recession crops , river fishing
and the populations health. (...) The absence of or the low flood level induced by the
retention of Bafing waters by the dam (nearly 60% of the river flow), seriously disturbed
the basins ecosystems and disorganized its traditional economic activities, as a result of
which the region became the poorest in all three countries. The appearance and increase of
social disparities and malnutrition led to the massive exodus of labour force from the
basin.Moreover , the development of irrigated agriculture and the absence of salt
encroachment on the delta resulted in the proliferation of habitats of endemic diseases, e.g.
bilharzia. There has also been an increase in malaria cases and the appearance of new
resistant strains (p.36). (ADF, 1997)

If the Energy Projects alleged benefits for traditional agriculture are being
emphasised now, might this not be to allay fears until it is too late to protest?
We will return later to the reservoir-management optimisation study funded
by the World Bank and conducted by IRD which has the merit of clarifying
what can be done in terms of flood support. For the time being, let us merely
note that the scenario adopted represents a step backwards compared with the
PDRG: the area of flood-recession farmland is smaller, and nothing is planned
for regenerating flood-recession grazing , groundwater reserves and forestry
resources, nor the areas in which river fish can spawn, which were eliminated
by the dam.

In theory the water management plan, reconciling the three main objectives of
the scheme (supplying water for irrigation, producing electricity and flood
support in aid of flood-recession farming agriculture), was meant to be drawn
up under the supervision of OMVS, with the participation of the local
communities concerned, before the year 2000, and before the recruitment of a
private operator to manage SEM. Indeed, the plan was meant to serve as the
basis for a reservoir management manual to be presented to companies bidding
for the contract to manage SEM, and for a Water Charter (Charte des Eaux) to
be signed by the three countries involved. In fact, however, the reservoir
management manual has already been drafted; it gives priority to electricity
generation, which in times of low water flow is in direct competition with
flood support. And OMVS has launched the procedure for recruiting the
private operator to be responsible for the management and maintenance of the
dam and the Manantali power station. The water management plan no longer
seems to be on the agenda, and preparation of the Water Charter, which was
supposed to be based on a consultation process involving all the local
stakeholders, remains at the planning stage. The Charter is also supposed to
draw on a cost-benefit study funded by the World Bank, which aims to
determine on a socio-economic basis the order of priority to be given different
types of water use. At the time of writing, the cost-benefit study had not yet
begun.

Will the authorities really wait for a water management plan, before initiating
the sequence of events leading to the installation of the turbines? It would
appear that work is under way at the construction site. The contractors for the
eastern electricity line are already on site, and invitations to tender for the
western line have already appeared in the press. Without a water management
plan taking into consideration the needs of all water users, and including precise
and binding instructions to the dam operator, the Valley peasant farmers will
again be the losers, and this time for good. Once management of the dam has
been handed over by SOGEM to a private operator, under an agreement
guaranteeing a certain level of electrical power generation and with clauses
penalising the countries involved in the event of failure to comply, it will be
impossible to change management priorities .
FLOOD SUPPORT: WHAT IS POSSIBLE?

The importance of the annual flood for the balance of small-scale family
agriculture in the Valley needs no further demonstration. In the first part of this
study, we noted the negative effects of the dams on the annual flood, and the
additional dangers arising from the Manantali Energy Project. What remains to
be discussed is whether it is still possible to save the situation.

It will be recalled that in 1991 the research team from the Institute for
Development Anthropology recommended that water be released each year
from Manantali to support the annual flood, except in the few years when the
level of the reservoir and the flow rate of the river are too low in August and
September. It will also be recalled that the Master Plan for the Integrated
Development of the Left Bank (PDRG), adopted by the Senegalese government
in 1994, opted for a water management scenario (scenario A) which would
allow for both the development of irrigated agriculture (88,000 hectares), and
the delivery of an artificial flood guaranteeing over 33,000 hectares of flood-
recession farming (land submerged for more than 2 weeks) and approximately
63,000 hectares for flood-recession grazing and forests (land submerged for
less than 2 weeks). It also held out the possibility of further benefits, since the
two initial phases would also be compatible with scenario B1 (53,000 hectares
of irrigated agriculture, and artificial flooding guaranteed for over 107,000
hectares, of which 50,000 hectares suitable for flood-recession farming ). Even
today, maintaining artificial flooding is still an official objective. In 1996, the
Institut de Recherche pour le Dveloppement (Development Research Institute /
IRD, formerly ORSTOM) was asked by OMVS to carry out a study on
optimal reservoir management (see box 5).

It will be recalled that the Manantali dam is built on the Bafing, the main
tributary of the Senegal River, which accounts for over half the Senegals flow
as measured at Bakel. The two other major tributaries, the Bakoye and the
Falm, are not regulated. This means that any artificial flood will have two
components: the natural flows of the Bakoye and the Falm, and the additional
volume of water released from the Manantali reservoir. In practice, artificial
flooding is the result of releases from the Manantali dam, calibrated to provide
the desired rate of flow at Bakel by taking into account the contributions of the
unregulated tributaries. The lower the natural flow rates of the Bakoye and the
Falm, the greater the volume of water released from the dam. In the years
since the Manantali dam came into service, the major component of the annual
flood has been the water released from the Manantali reservoir.

Box 5. Terms of reference of the IRD study

The Manantali dam, which can store 11.5 billion cubic metres of water, currently has a
regulating role. It makes it possible to maintain throughout the year the flow rate of over
200 m3/s required for controlled irrigation , and to augment the natural flood supplied by
the rivers non-regulated tributaries. This flood support, the suppression of which had been
envisaged earlier, is vital for safeguarding the environment of the middle valley and
creating the conditions for flood-recession farming, , which is a profitable activity for
farmers.

The reservoirs are currently managed by measuring water flow or water levels, on the
Bafing where it enters the Manantali reservoir and on the Senegal river downstream from
the junction with its unregulated tributaries. These measures, based on data collected as the
season progresses, make it possible to manage the reservoirs during the dry season. But the
absence of rain/water-flow forecasts for the catchment basins of the rivers three tributaries
prevents the dam management from co-ordinating releases from the Manantali dam with the
natural flood from the unregulated tributaries.

The study consists of three distinct but closely related parts:

research on flow rates well upstream from the dam, to improve forecasting of the natural
flood of the unregulated tributaries and the Bafing itself ten days in advance. This
advance warning will make it possible to calibrate flood support from the dam (artificial
flooding), by making it coincide, insofar as possible, with the peak of the natural flood;
or, in the case of very heavy rainfall, to level out the peak of natural flooding. In times of
severe shortage of rain,, it will enable the OMVS High Commission, in conjunction with
the political authorities, to decide how the available water should beused .
research to determine the optimal flood required to maintain flood-recession farming,
replenish groundwater reserves and preserve ecosystems, while minimising losses of
electrical power.
implementing reliable methods for predicting and monitoring flow rates, and for
managing the dam.
Source: IRD-OMVS, 1998.

Initially, computerised simulations of the management of the dam, based on the

available data for the rivers natural flow over the period 1950-1993, were used
to show what the results of the dams operation would have been if it had been
in existence since 1950. The flood hydrograph that emerged, almost identical in
volume to that used for the PDRG, is defined as follows: 0 cubic metres of
water per second at Bakel on day 0; 2,500 m3/s on day 6; 2,500 m3/s on day
11; 0 m3/s on day 55.
Three possible management scenarios were envisaged:

Scenario 1 gives priority to the needs of agriculture (irrigated and flood-

recession ),
Scenario 2 gives priority to electricity generation and the needs of
irrigated agriculture (with no flood support),
Scenario 3 takes into account the three main objectives of the dam
(supplying water for irrigation, generating electricity , and flood support in
favour of flood-recession farming ).

Comparative analysis was based on the amount of electrical energy generated

and the area available for flood-recession farming . As IRDremarks : Under
the conditions of low water flow which have prevailed over the last 25 years or
so in the Senegal River basin, these are the two objectives which compete with
each other. Supplying water for irrigation poses few problems, since it can be
provided on a permanent basis without greatly penalising the generation of
electricity.

On the basis of natural flows for the period 1970-93, which correspond more
closely to present circumstances, the third management option would give the
following results (for the sake of comparison, we also give the results that
would obtain if the dams had not been built).

Multi-use management of the dams River regimen without dams

(Scenario 3)

50,000 hectares of flood-recession farming 50,000 hectares of flood-recession farming

one year in ten, 40,000 hectares every other one year in three
year

Annual average of 30,000 hectares during Annual average of 57,000 hectares

the period envisaged

Two years out of three, some flood- Fewer years in which flood-recession
recession farming is possible farming is possible

Where power generation is concerned, scenario 2 of course gives the best

results, but would not allow for any flood-recession farming at all. The results
of scenario 1 are very similar to those for scenario 3, because when the flow of
the river is weak (as during the period under study), almost all releases of
water would pass through the turbines and could therefore be used for
generating electricity. For scenarios 1 and 3, the same management directives
are envisaged, with two additional directives in the case of scenario 3, relating
to maintaining the level of Manantali reservoir (when the level falls to 195
metres or less, releases are limited in accordance with the priorities decided)
and with electricity generation (release of a flow sufficient to generate 90 MW
if the level of the reservoir is above 182 metres, or the maximum that can be
generated if the level is above the dams overflow threshold).

In 1997, IRD delivered an artificial flood which came close to the reference
hydrograph. According to a rapid survey of SAED agents and people living
near low-lying depressions, this flood was generally judged satisfactory
from the point of view of the area flooded, although in the Matam area it was
considered to have been of too brief duration to flood local depressions
effectively. Though far better than the 1996 flood, which reached only a small
area of land, it was nevertheless not nearly so good as the 1995 flood. From
satellite pictures it was possible to estimate the total area under water. Where
artificial flood support is concerned (the year 1997-1998 being taken as
representative), we estimate the area available for flood-recession farming at
about 70,000 hectares: 45,000 hectares on the left bank and 25,000 hectares on
the right bank (IRD, 1998).

It should however be noted that for the 1997 flood, contrary to the
recommendation in the terms of reference, it was not possible to set up a
system to give ten days advance warning of flows in the upper river basin
(Bakoye and Falm). It was therefore not possible to implement a more
flexible system of flood support, even though, as the IRD report acknowledged,
this would allow for better electricity generation than a fixed-date flood, as
well as a greater frequency of adequate floods than would be the case under
the rivers natural regimen.

Nevertheless, from 1998 on, if the deficiencies of the 1997 flood be remedied,
OMVS will have at its disposal the technical conditions required for delivering
flood support from the Manantali reservoir. The final phase of the IRD study
involves drawing up a dam management manual. The rules to be implemented
will correspond to the objectives set as part of the optimum scenario. All
possible eventualities will be taken into account, to establish an order of
priorities in the event of chronic deficits or conflict between one use and
another.
There remains one vital problem, over and above any technical considerations.
If the flood is to take place in accordance with the stated criteria, it is not just a
matter of achieving the right technical conditions (as we have seen, not all of
them have yet been put in place). There also needs to be the political will to
implement flood support, enshrined in clear and binding instructions to the
operator appointed to manage the Manantali dam. It should still be possible to
achieve this last condition, since the recruitment of an operator is not due to
take place until the member states have finalised the Water Charter. But there
is little reason to believe that the necessary political will exists. In the
conclusion to its summary report, after emphasising that the artificial flood is
a matter of vital environmental, human and economic importance, IRD notes
that: it remains doubtful whether the partner countries have the will to
maintain this artificial flood over the long term (IRD 1998).

Senegal River development schemes have not brought about development. They
have not brought prosperity, except to a few artificial enclaves. The prospect of
the dams becoming profitable seems more remote than ever. In the absence of
massive investment, which seems most unlikely, profitability will not come
from irrigated agriculture, but will depend solely on the Energy Project. Can
one even speak of profitability when, to service a debt which will never be
repaid, a further debt has been contracted; in what is not just a vicious circle,
but a vicious spiral?

Instead of development, there has been destruction. By sweeping aside

production systems which offered a degree of food security, Senegal River
development schemes has made life even more precarious than before for
many of the inhabitants of the Valley. Those who are excluded from irrigated
farming because of the high costs involved; cannot now fall back on the
waalo, because since the dams were built, the annual floods can no longer be
relied on to provide crops and grazing and replenish fish stocks . All the future
seems to hold for them is emigration to the volatile outer fringes of Dakars
urban sprawl.

Things could and should be different. It would still be possible using these
very dams to give some security to small-scale family agriculture, the only
possible basis for prosperity in a country like Senegal, by striving to make the
best possible use of all the Valleys resources. However, nothing, or almost
nothing, in the record of flood management since the founding of OMVS, gives
any hope that those in charge will rethink their position. This raises the
question of whether the real problem is not in fact political, in the broadest
sense of the term. If agriculture in the Valley is in deep trouble, is it not
largely because its peasant farmers, pastoralists and fishermen have had no say
in the decisions affecting their future? The recent Presidential elections have for
the first time brought the opposition candidate to power. It is not known
whether the new President will adopt a different policy towards the Senegal
River Valley. For the moment, one can only hope for the best.
 REFERENCES

Adams, A. (2000) Quel avenir pour la Valle? Dakar

Adams, A. (1997) Le long voyage des gens du Fleuve. Paris.

Adams, A. (1985) La Terre et les gens du Fleuve. Paris.

African Development Bank. (1994) Project completion report, Manantali Dam.

July. ADB.

African Development Fund. (1997) Appraisal report, Manantali energy project.

September. ADF.

Bessis, S. (1981) Faut-il construire les barrages sur le Sngal? Jeune

Afrique, 22 July.

Crousse, B. et al. (1991) La valle du fleuve Sngal: valuations et

perspectives dune dcennie damnagements, Paris, 1991

Dumont, Ren. (1981). Le dfi sngalais. Dakar.

Hollis, G.E. (1990) The Senegal River Basin monitoring activity: Hydrological
issues, parts I and II. Institute for Development Anthropology, Binghamton,
NY.

Horowitz, Michael M. and Salem-Murdock, M. (1993) Development-induced

food insecurity in the middle Senegal Valley. GeoJournal 30.2.

Horowitz, M., Salem-Murdock, M. et al. (1990) Suivi des activits du bassin

du Fleuve Sngal (SRBMA): Rapport final intgr. Institute for Development
Anthropology, Binghamton, NY. Republished as Les barrages de la
controverse: le cas de la Valle du Fleuve Sngal (1994), Paris.

IRD-OMVS. (1998) Etudes de loptimisation de la gestion des amnagements

de lOMVS. Phase I, Summary report. OMVS, Dakar.
IRD-OMVS (1999) Impacts potentiels de la gestion des ouvrages et des eaux
de surface du Fleuve Sngal sur lagriculture de dcrue. Summary report,
provisional draft. OMVS, Dakar.
Le Roy, X. (1997) La place des cultures de dcrue dans les systmes de
production irrigue. ORSTOM, Dakar.

ORSTOM. (1997). Comparaison de trois scnarios de gestion du barrage de

Manantali. ORSTOM, Dakar

PNUD-OMVS. (1974) Programme intgr de dveloppement du bassin du

Sngal. 12 volumes. Norbert Beyrard, Paris.

Rpublique du Sngal, , Programme dInvestissement du Secteur Agricole,

1995-2000. Ministre de lAgriculture, Senegal.

Rpublique du Sngal. (1994) Plan directeur de dveloppement intgr de la

Rive Gauche du Fleuve Sngal (PDRG): document de synthse. Ministre de
lEconomie, des Finances et du Plan, Senegal.

Rpublique du Sngal. (1984). Conseil interministriel sur les perspectives et

stratgies de dveloppement de laprs barrages. Ministre du Plan et de la
Coopration, Senegal.

Rpublique du Sngal. (1995) Programme dajustement sectoriel agricole

(PASA). RdS, Senegal.

World Bank. (1997) Staff appraisal report, regional hydropower development

project. June. World Bank, Washington, DC.