Initiatives in local management of dryland forest areas of the Sudano-Sahelian zone

Material Information

Initiatives in local management of dryland forest areas of the Sudano-Sahelian zone final technical report : DFID R6510 Forestry Research Programme
Alternate Title:
DFID R6510 Forestry Research Programme : final technical report
Kerkhof, Paul ( Author, Primary )
SOS Sahel International UK ( contributor )
SOS Sahel International UK. Forestry Research Project ( contributor )
Place of Publication:
SOS Sahel International UK
Publication Date:


Subjects / Keywords:
Great Britain. Department for International Development ( LCNAF )
United Kingdom. Department for International Development
SOS Sahel International UK. Forestry Research Project
أفريقيا -- الساحل
Afrique -- Sahel
Sahel ( LCSH )
SOS Sahel International UK
SOS Sahel (Organization : London, England) ( LCNAF )
Arid regions ( LCNAF )
Spatial Coverage:
Africa -- Sahel
16.024646 x 13.321854


General Note:
VIAF (name authority) : SOS Sahel (Organization : London, England) : URI
General Note:
VIAF (name authority) : Kerkhof, Paul : URI
General Note:
VIAF (name authority) : Great Britain. Department for International Development : URI

Record Information

Source Institution:
SOAS University of London
Holding Location:
Special Collections
Rights Management:
This item is licensed with the Creative Commons Attribution, Non-Commercial License. This license lets others remix, tweak, and build upon this work non-commercially, as long as they credit the author and license their new creations under the identical terms


This item has the following downloads:

Full Text


DFID R6510 Forestry Research Programme



This publication is an output from a research project funded by the United Kingdom
Department for International Development (DFID) for the benefit of developing
countries. The views expressed are not necessarily those of DFID. R6510 Forestry
Research Programme.


Forest resources in the Sahel are used by some 60 million people. These resources
have rapidly diminished in the past era of central control. Policy changes over the
last 10 years have been in favour of local governance of forest resources although
the implementation leaves something to be desired. Effective engagement of local
people is expected to result in more sustainable and equitable forest management.
The research project aims at understanding the constraints to scaling up, and ways
to overcome these constraints.

Few reviews of dryland forest management exist. In a review of the early 1990s,
local community control of the African dryland forests was _ still exceptional
(Shepherd, G.,1992). Much has changed in the meantime, as the examples of an
institutional study by CILSS show for the west African Sahel (CILSS, 1997). But
many examples are based on the wood energy sector, which has many pitfalls for
local governance (Foley, G.,1998). Review of more integrated projects shows that
the legal and institutional framework of such projects tends to collapse after project
completion (e.g. Arzika, M.S., 1996).


The project addresses the purpose level objectives “Development and promotion of
techniques for sustainable management of forest resources through local
governance systems”. This was done through action research in three SOS Sahel

projects (in Niger, Mali and Sudan), case studies in other areas, and national level
studies, over the period 1996-1999.


SOS Sahel has been working in the Sahel since 1985, first on tree-planting activities,
and later on the management of forests and natural resources with local people. The
research was conceived as a way of doing two things:

Firstly, it was felt that, after a long period of implementation, particularly in the
Sudan, it was timely to try to draw together lessons learned. Project staff involved in
day-to-day implementation lack the time and resources to conduct such an analysis
in their own area and country, and thus a research project with an researcher was
deemed the best way forward. For SOS Sahel programme staff, an analysis of its
own experience, and of that of other projects in the same country was_ an incentive
for support to the research project.

Secondly the comparative approach essayed here, pulling together not only the
diverse experience of each of the participating countries, but also the contrasting

institutional context in each, was of particular interest to the SOS Sahel
management team and board. Comparative understanding of the Francophone and
Anglophone context was of special concern.


The research emphasis was on action research in three sites through fifteen visits
over a three year period. Local stakeholders and extension staff were actively
involved in monitoring and research, with a focus on ecological, socio-economic and
institutional aspects. Study of national level institutions and legal frameworks was
undertaken with the assistance of four national consultants of the three countries. A
project maturity workshop, in Bamako Mali in 1999, put together and discussed 20
studies of 30 participating researchers. The project produced 16 internal reports and
18 publications.

Action research in the three projects involved repeated fieldwork by the researcher
along with project staff and villagers, the latter being involved on a more permanent
basis. The very nature of action research accompanying and drawing lessons from
project implementation necessitated ecological, economic and sociological or
institutional enquiry. In total, 15 visits were made to the three countries. Action
research also required adjustment of the research agenda to local needs and
opportunities, i.e. those of the projects and those of the local communities.

Other projects were also studied and a wide range of documents, especially grey
literature, was collected and analysed, some of which required visits to other
Sahelian countries. Relevant national and regional institutions were analysed in the
three countries with an emphasis on the forest service. Researchers from 6 Sahelian
countries gave their inputs at the synthesis workshop which aimed to draw
conclusions pertinent to the Sahel zone.

The expertise in the research project included extension staff who did most of the
village level research, market surveys, and wild foods data collection, etc. The
research process at this level was largely determined by local capacity, which was
therefore a test of the applicability of research results. For the higher level
institutional studies, 4 national consultants were engaged. National lawyers
specialised in environmental management were engaged to analyse the legal

The implementation of the research project differs from that initially planned in
several ways. One of the four countries selected for research, Eritrea, could not be
included for reasons of security. On the other hand, the scope of the research has
been broadened by the national level research, and by the employment of national
researchers. (The programme of activities in the project document was limited to
field level research. )


The full report of the research follows upon this summary technical report and upon

the ‘policy implications’ executive summary. The chief outputs from the research
include the following.

1. Thirteen case studies of local governance and of projects supporting local
governance have been prepared and reported in various ways (see dissemination
outputs below). The details are in the following table:


project case studies 7 studies in 4 countries

cases of local governance at village |3 studies covering 16 villages in 3
level based on fieldwork by researcher | countries

or national consultant
other cases of such governance based | 3 cases in 3 countries
on grey literature and staff interviews

2. The research has drawn conclusions on appropriate forest management
structures for local governance in the Sahel, with emphasis on the following


forest appropriation and guarding

Kerkhof, P., in prep.; and Kerkhof,

rule making


zoning and codification of management
unions of local institutions

monitoring Aziz, RA. and = Kerkhof, P,1998:
Kerkhof, P.,1996a; Kerkhof, P.,1997b;

Kerkhof, P.,1999g

3. Participatory resource monitoring methods have been developed by the
research project in two countries. Due to the nature of the Sahelian woodland, the
emphasis was given to ecological monitoring instead of wood volume monitoring.
Simplified forest management tools have been developed through the research,
such as panoramic photography which is used as a participatory environmental
monitoring tool. This has contributed to the project toolbox, but it should be noted
that such technical innovations are of secondary importance to successful local


Two methods are different from usual practice in the Sahel:


Panoramic photography by local people, which includes the preparation,
documentation and storage of the panoramas by local communities (Aziz, R.A. and
Kerkhof, p.,1998; Kerkhof, p.,1998; Kerkhof P.,1999b).

Ecological and forestry management research included simplified woodland
inventories with methodological development in tune with local capacity. Forest
resource inventory was of locally relevant forest products rather than cubic metres of
wood (Aziz, R.A. and Kerkhof, p.,1998 and Kerkhof, P.1999b). Some methodological
development was pursued in economic monitoring of marketed and subsistence
forest exploitation, such as wild food production by women and exploitation of minor
forest products by artisans (Kerkhof et al.,1998; Kerkhof, p. and Siddig, F.,1998:
Kerkhof, p.,1999g). This economic research included PRAs, market surveys,
producer observation and household surveys.

4. Criteria for effective local governance were identified as a combination of local
and national criteria. Local criteria those which are broadly under the control of local
communities, the most important one being the representativeness of the local
management institution vis-a-vis user groups whose livelihoods depend on the
resource. However, several important criteria are dependent on the national legal
and _ institutional frameworks which vary from country to country, with certain
crosscutting characteristics for the Francophone Sahel. (Appropriate national
institutional and legal frameworks are reported in Ali, A.M.,1999; Anon,1999; Bacha,
A.K.,1999; Madougou, D.,1999a; Mazgoub, T.,1999, Kerkhof, P.,1999g.)

5. Assessment of local forest management donor strategy prevailing in the
Sahel, and especially of the domestic energy strategy of the World Bank. This
strategy has been analysed through field study, review of published and unpublished
literature and through interviews in Mali and Niger. Concern about the constraints of
this strategy and possible negative impacts are being disseminated (Kerkhof, p.,
Siddiq, f. and Damango, B.,in prep.; Kerkhof, p.,in prep.).

6. Policy dialogue and legal innovations in individual countries. As a follow-up
of the Bamako workshop, country delegations notably from Sudan and Burkina Faso
have prepared for policy level meetings with an emphasis on law reform. (see Anon,

7. Most of the dissemination outputs are in both English and French. They

scientific and technical papers; two versions of the final report, one of them to be
very widely disseminated; a Sahelian workshop report; training manuals; and a policy

A list of dissemination outputs to date can be found at the end of this section.

8. Conclusions have been drawn on appropriate forest management structures for
local governance with emphasis on simple and efficient regulation of forest
exploitation. These conclusions demonstrate that institutional and distribution issues,
rather than technical forestry aspects, are the key to successful local forest
management. Some of the institutional issues can be addressed at local level, but
others depend on national level policy and law reform. The research has contributed
to discussions on policy and law reform in several countries.



The research has contributed to the DFID policy of poverty alleviation through a
better understanding and acceptance of more efficient and more equitable forest
management in the Sahel. A more profound characterisation of local resource
management capacity, including that of women has been developed; insights into
the resource-management relationship between sedentary and mobile pastoral
Sahelians have evolved; and greater understanding of poor rural Sahelians by
relatively well off urban policy-makers has been embarked upon.

Key indicators are the initiatives taken by those who influence policy, as witnessed
by the written contribution of influential African participants at the Bamako workshop.
The research findings are presently circulating in national level committees,
government and non-governmental, in various countries. The country specific follow-
up which African delegations have prepared is expected to impact on target
institutions. Indications are that central forest services will cede more responsibilities
to local government institutions.

The main promotion pathways for the research results are

(1) follow-up of the Bamako workshop through national groups

(2) wide distribution of the final technical report in short and longer versions, and

(3) regional networking.

SOS Sahel has a continuing presence in most of the countries concerned which
facilitates support to national initiatives. The results of the research have also been
promoted through a series of publications (see the list at the end of this section).

Various forms of follow-up are independent of the research project but some action
to promote the research findings is desirable. Promotion of networking between
African researchers engaged in local governance of natural resources requires
continued external assistance at this stage. English/French translation and exchange
between West and Northeast Africa are important, particularly in matters of
institutional and legal reform. Proposals for training and capacity building workshops
(probably one Anglophone and one Francophone) are being worked on.


KERKHOF, P. (in prep) Local management of Sahelian forests. Macmillan, UK.
KERKHOF, P., (in prep) From state to local management of the Sahelian forest.

FOLEY G. (Ed.) London: SOS Sahel UK.


AZIZ, R.A. and KERKHOF, P. (1998) Forest monitoring by villagers. Social Forestry
& Environment, April 1999, no.4,pp.5-6.

KERKHOF, P. (1997a) SOS Sahel (GB) research project: “Local management of
dryland forest’. European Tropical Forestry Research Network (ETFRN) News, July
1997, no.20,pp.13-14.

KERKHOF, P. (1999a) Local forest management in the Sahel. European Tropical
Forestry Research Network (ETFRN) News, summer 1999, no.28,pp.6-7.


ALI,A.M. (1999) Institutional analysis of forest management in the Sudan. Workshop
on local forest management in the Sahel, Bamako, 13-16 September, 1999, 26 pp.
ANON (1999) Local management of Sahelian forests. Proceedings of the Bamako
workshop, 13-16 September 1999. SOS Sahel UK, 1 Tolpuddle St. London N1OXT,
UK. 17p. English and French.

BACHA,A.K. (1999) Etude des institutions nationales de gestion des forets. Rapport
de synthese. Workshop on local forest management in the Sahel, Bamako, 13-16
September, 1999, 21 pp. French.

DAMANGO,B., OUSSEINI, S. and KERKHOF,P. (1998) Analyse institutionnelle des
Alamodious, Bankass. Workshop on local forest management in the Sahel, Bamako,
13-16 September, 1999, 13 pp. French.

KERKHOF,P. (1998a) La photographie panoramique. Seminaire sur la foret seche
de l'Afrique de l'Ouest, Ouagadougou, 16-20 november 1998. French. 5pp.
KERKHOF,P. (1999b) Theme 1: local institutions. Keynote address to Workshop on
local forest management in the Sahel, Bamako, 13-16 September, 1999, 5 pp.
KERKHOF,P.,SIDDIQ,F. and DAMANGO.B. (in prep) La sylviculture Sahelienne au
carrefour de la gestion centralisee et de la gestion locale. Actes du seminaire sur la
foret seche de l’Afrique de l'Ouest, Ouagadougou, 16-20 november 1998. French.
MADOUGOU,D. (1999a) Etude sociologique des agents forestiers. Cas du Niger.
Workshop on local forest management in the Sahel, Bamako, 13-16 September,
1999, 1/pp. French.

MADOUGOU,D. (1999b) Etude socio-economique du village de Tientiergou apres
six ans de gestion locale de foret villageoise, Arrondissement de Say, Niger.
Workshop on local forest management in the Sahel, Bamako, 13-16 September,
1999, 24 pp. English and French.

MAZGOUB, T.M. (1999) Law and local forest management in the Sudan. Workshop
on local forest management in the Sahel, Bamako, 13-16 September, 1999, 29 pp.
English and French.



KERKHOF,P. (1997b) Panoramic photography. 7pp. [one day training workshop for
extension staff in El Ain, Sudan, 15 March 1997: plus distribution to Sudanese
organisations]. English.

KERKHOF,P. (1996a) La photographie panoramique. 7pp. [one day training course
in Bankass,Mali, 2 December 1996 and in Takieta, Niger, 20 November 1996].


KERKHOF,P. (1996b) Mali field report. London: SOS Sahel UK. 9 pp. French
KERKHOF,P. (1996c) Niger field report. London: SOS Sahel UK. 9 pp. French
KERKHOF,P. (1997c) Sudan field report. London: SOS Sahel UK. 25 pp. English
KERKHOF,P. (1997d) Mali field report. London: SOS Sahel UK. 19 pp. French
KERKHOF,P. (1997e) Niger field report. London: SOS Sahel UK. 16 pp. French
KERKHOF,P. (1997f) Sudan field report. London: SOS Sahel UK. 38 pp. English
KERKHOF,P. (1997g) Mali field report. London: SOS Sahel UK. 40 pp. French
KERKHOF,P. (1998b) Mali field report. London: SOS Sahel UK. 28 pp. French
KERKHOF,P. (1998c) Sudan field report. London: SOS Sahel UK. 33 pp. English
KERKHOF,P. (1998d) West Africa field report. London: SOS Sahel UK. 29 pp.

KERKHOF,P. and SIDDIQ,F. (1998) Socio-economic analysis of some woodlands in
Kordofan. London: SOS Sahel UK. 34pp.

KERKHOF,P., DAMANGO,B. and GUEGUERE, R. (1998) Analyse socio-
economique de la foret de Tyi. London: SOS Sahel UK. 38pp. French.

KERKHOF,P. (1999c) Sudan field report. London: SOS Sahel UK. 40 pp. English
KERKHOF,P. (1999d) Mali field report. London: SOS Sahel UK. 22 pp. French.
KERKHOF,P. (1999e) Sudan field report. London: SOS Sahel UK. 13 pp. English.
KERKHOF,P. (1999f) Local forest management in the Sahel. policy brief. London:
SOS Sahel UK. 1 p.

KERKHOF, P. (1999g) Local forest management in the Sahel. Final Report.
SHEPHERD,G. (Ed.). London: SOS Sahel, nd.

ON fi fn stl sill yin,



It has been difficult to generalise from the material presented in this report, since the
Sahelian climate operates in long and erratic cycles, and most of the case studies

are from only the last 5 years. Nevertheless the SOS Sahel research project has
been able to contribute to the understanding of Sahelian natural forest management

in the following ways.

In the early 1990s, it was thought that forest resources under communal
management were regressing and that on farm trees increasingly constituted the
main remaining source of tree products for local people. Present findings confirm
that forest resources under local control include agroforestry systems, patches of
privately owned natural forest, related resources such as game, and community
forest. In fact community control over forest resources is strongly increasing rather
than regressing.

A newly observed factor is forest appropriation: forest resources on public land are
increasingly (re)appropriated by local groups, the appropriation is accompanied by
rule making and enforcement, and by the application of local law.

The present evidence shows that local legitimacy of the managing institution is the
key condition, along with the informal support of the local administration. However,
more formal ownership is probably a condition for successful forest management in
the longer term.

The right fit between the size and nature of the management institution and that of
the forest is a key issue.

Where overarching local institutions exist, they fulfil a useful institutional role. They
are an important mediator between village institutions in case of conflict and they
have an important union and lobby role vis-a-vis the state.

The economics of local management of Sahelian forest has rarely been researched.
SOS Sahel research suggests that while average revenues directly attributable to
the forest are very modest outside the peri-urban impact zone, intra-community
differences are large and certain poor social groups depend highly on such

In 1990 it was felt that the Sahelian climate was drying and natural vegetation
degrading. Recruitment was poor and woodlands in recession. However, rains over
the last 5 years of the decade have been particularly good, and the vegetation has
responded. Management intensity has probably increased, in part as a result.

The question now is not whether local management is an alternative to central
management, but how it can be best done.


Recommendations in the report are presented in the form of conditions for

successful management. A set of ten criteria is proposed.

Criterion 1: the basic criterion for management of Sahelian woodlands is the
livelinood dependency of the user groups attached to a particular resource. Where-
ever such dependency exists, either economic or socio-cultural, the woodland is
subject to management.

Criterion 2: Management transaction costs should be very low, given the low
revenue generation capacity of most Sahelian woodlands.

Criterion 3: Management should aim at long term change, and be measured against
environmental standards instead of wood volume.

Criterion 4: Local forest management should be based on indigenous knowledge
systems. External technical knowledge may be supportive but should not be imposed
on local management. A systems approach to R and D is required.

Criterion 5: The different stakeholder groups amongst the settled communities have
to negotiate their rights and responsibilities before sustainable management is

Criterion 6: Pastoral interests should be represented in the resource negotiation
process. These interests are located at strategic rather than operational levels.

Criterion 7: Local governance of forest resources is a multi-sectoral engagement,
which needs support from multi-sectoral institutions. Local government is the
appropriate institution at local level, and the environmental and planning agency is
the appropriate institution at the national level.

Criterion 8: Short-term institutional arrangements may facilitate local forest
management in individual projects, but in the long term, unambiguous legal support
should be secured.

Criterion 9: As long as modern law is not functional, local law guarantees
enforcement of management rules. Projects should encourage the development and
application of local law.

Criterion 10: Good governance must be achieved in local management institutions,
as much as in local and national government.




INTRODUCTION ...........cccsssscsscsssscsscesvscescesnscesessscesessescesessescesessescesessnscesessescsessescesesseacosessessescssessescssesseecsseeseees 1
LOCAL FOREST MANAGEMENT IN THE SAHEL PART T: THE CONTEXT............cscsssssssssevscseeseoees 4
2. EVOLUTION OF SAHELIAN FOREST MANAGEMENT. .........sscsscsssscsscsevscsscsenscsscssscsscssescsscssescssceseees 4
2.1 THE EMERGENCE OF THE FORESTRY SECTOR wu... ccccecsesessesseeseeseeecseesseecseeaseeeaeeaseeeneeaseseneeaeeeeneeaseas 4
2.2 1960-1990: CENTRALISATION AND TREE PLANTING Wie ccecceseeeeneeseeecseeseeecneeseeeeneeaseesneeaseeeneeasees 5
2.3 19908: DECENTRALISATION, DEMOCRACY, SECTORAL INTEGRATION uo... ccc ccceeeeenseteeeeneeneees 6
3. WOODLAND ECOLOGY AND MANAGEMEEN T............scscsssscsscsssscsscsesscsscssescsscssnscsscssescsscssescsscssesssseeseees 8
3.1 ENVIRONMENTAL CHARACTERISTICS oo. ccccccccecssessesessenseessenevsecseneesecseavseceenavseceenavsessenavseseenevaeeeenees 8
3.2 WOODLAND ECOLOGY 00... ccceceeeescnsesecseseseceesaesecseavsessenavsecseavsecseavsecseavsecsenavseceenavaeseenavseseenasaeseenaes 9
3.3 INSTABILITY OF SAHELIAN WOODLANDS 00. cccccecetececneeteseeeceesececnevsenscnsesecsensvsessensvsessenteneetens 10
3.4 WOODLAND MANAGEMENT ooo. cccccccceecscnsesecscseseceenevsecscnsvsececevsecscevsecscnessecsenseseesenssiensensvsessenteneetens 14
3.4.1 ASSESSMENT OF PRODUCTIVITY AND QUOTA voicccccccccccsccstetesesecnseeecnseescnseseeesesaessecesescussescnsesaeaes 14
3.4.2 FOREST MANAGEMENT PLAN vuececccccccceccssecscesecnsesecneeescseescuseseessecesecsseescaesseesecsesscesescasercnaesaeaes 16
3.4.3 FOREST REHABILITATION... .ccccccccccecsccccssescsseeecesecnseescuesescsseescusessessecesscacescsessessecsessscaesescaeeeseaesaeeaes 17
3.4.4 FOREST EXPLOITATION .uoccccccccceccceseesceeescneseceseceesecsesescsscescesessessecesscuseescnsesseesecsessecaeseecaserenaesaeeaes 18
3.5 CONCLUSIONS ooo. cccccccceeeescnsesecscnsesevsecscnevsecscevsecsesvsecsensvsecscsvsecsensvsecscnsvsecscevsecsenseseesentsseesensvsestenseneetens 20
4, ECONOMICS OF LOCAL WOODLAND MANAGEMENT..........csssscsssscsscsssscsscesnscessesscosessnscesensnsonses 21
AL INTRODUCTION oo cccccccccscnscnsesecsensesecsenacsecseavsecseavsecseavsecscavsecseavsecsenassecaeaavsecseavsesaenavieseenasaeseeneed 21
4.2 METHODOLOGY o.oo cceeccecnseteseeseseesecsenacsecsenavsecseevsecsenavsecseavsecsenavsecseavsecaeassessenessessenaviesaenasaeeeenee 21
4.3 FIREWOOD MARKET ECONOMICS 00... ccccceeccccssesessensesecseseesecscnaesessenavsecsenavsecacavsavsessenavsessenavseseeasaeeees 24
4.4 BROAD VALUES OF SAHELIAN WOODLANDS... ccccceeseesenseneeseeecaesecscaesecseevsesaeavaeseenaeaeeeenaed 28
4.4.1 THE KELKA FOREST ..ccccccccccccccececessscnseescnsesecesecseesecsesescaeescesecsessecessecnasescaeseeesecsesecesescnaseeceaetaeaes 29
4.4.2 BANKASS oc cecccccccccecetcseescsseescnecseesecessecsesscsececesecsessecesscasescusessessecaesecseescnaeseeesecsessecseseecasercnaesaeeaes 30
4.4.3 KORDOFAN, SUDAN wicceccccccccccsecccesssscnseescnsesscesecsessecsesescaeescusecsessscaesecsseescaesseesecsessecesescaeeecsaesaeaes 34
4.4.4 LIVESTOCK ECONOMICS. .ccccccccccccesseceseescsececesecneesecnesescaeescuseseessecsesscacescaeseeesecsesseceseecaseeceaetaeaes 36
4.5 COST OF FOREST MANAGEMENT 0.0... ccccccecccceseeeesenseseceesavsecsenacsececnavsecscnavseceenavseceeavsesaenavaesaenasaeseenees 38
4.6 CONCLUSIONS 00... ccccececccsssecsensesecsecseneesecsenavsecsesavsecsenavsecsenavsecsenavsecaenavsecsenavsecsenavseceeavseseenavaessenasaeseeneed 41
5. NATIONAL INSTITUTIONS AND LEGAL FRAMEWORK .............cccsscssssssscsssessscsscessscsessescsessscesonses 43
S.LINTRODUCTION 0.0. cic cccccesecsecsenseecsenevsecscevsecsensvsecsensvsecscsvsecscnsssecscevsecsensvsesscnsesessenssiesscnsvsessenseneetens 43
5.2 THE FOREST SERVICE ou... cccccscccccesesecsenseecscnsvsecscnevsecsenevsecscsvsececnsvsecscnevsessenevsessenssseesensssessensvsessenseneetens 44
5.3 LOCAL GOVERNMENT 00... cccccccccesecsenseececevsecscsevsecsenevsecscevsecscnsvsecscnevsecsesvsecsensssecsenssseesensvsensenseneetens 48
5.4 MULTISECTORAL AGENCIES... cccccceccsecscnsesecsensesecsenevsecscsvsececnsvsecscnsvsecseevsecsensesessenseseesensssestenseneetens 31
5.5 LAW AND LOCAL MANAGEMENT ooo. ccccccececcnsetecsensesececnsvsececnevsecscsvsecscnsvsecsensesessensvseesensvsessenteneetens 52
D.5.1. LEGAL FRAMEWORK ve cecccccccccccesecesesscnseescnsesecesecseeecsesescuaeescusesaessecsesecasescuaesseesecaessecaesecaseeesaesaeeaes J2
5.5.2. FISCAL REFORM wiccccccccccccsscssescssessessecnsssscnecescnseescesecsessecsesescaeesceaecseeecsesecaeeescaeeesesesescaseecsaesaeeaes 56
5.5.3. GOOD GOVERNANCE Lu. cecccceccccceseceeseceseescnsececeecaeeecsesescaeescescuacescaecseesscsessecaesscnaesesesesaesaseeceaesaeaes 56
5.6 CONCLUSIONS ooo. ccccceceeeccnsesecscnsesevsecscnsvsecscevsecscnevsecsesvsecscsvsecscsvsecscsvsecsenevsecsensesessenssseeseneveessenseneetens 37
INSTITUTIONS ...........cccssscssssssscsssesssssssessscsssssnscosonssscosensnscsensnscsensescsensesosessesoseesescesensescesensescesesseecesensesoeseeses 58
6. LOCAL COMMUNITY PERFORMANCE IN FOREST MANAGEMENT..........cscssssssscsscsssscsscesesoeses 58
6.1L INTRODUCTION occ ccccccecccnscnsesecsesesecseacsecseavsecscnavsecsenavsecscaavseceenavsecsenavsecaeassecseevseseenavaesaeasaeseenees 38

6.2. FOREST PROTECTION 0c eccccecssecscnsesecsesesecsesevsecscavsecseavsecsenavseceenavsecscavseceeavsessenavseseenavseseenaeaeeeeneed 59

6.3. RULE ENFORCEMENT |... ccc cccccceseesscnseecsesesecseavsecsenevsecsenavsecsenavsecacaavsesecnavseceeavseceenavsesaeavaesaenasaeeeenaed 62
6.4. TRANSPARENCY... cccccccseeecscnsesecsensesecscavsecsenavsecsenavsecsenavsecsenavsecsenavsececnavsecaeaasseseeavsessenavaeseenasaeeeeneed 65
6.5. MANAGEMENT PLANNING o.oo ccceccccessseesensesecseseseceesavseceesavsecsenavsecscnavsececassecsesenavseseenavseseenaeaeseeneed 67
6.5.1 FORMULATION OF MANAGEMENT PLANS .uiececcccccccccccseccetetnsesecneeescuscescnseseeesecesecnesesenaeseeesecaeeaeengs 67
6.5.2. IMPLEMENTATION OF MANAGEMENT PLANS. u..cccccccceccsseteeseceseescuseescnseseceecesaecnesescaesseseceseengs 69
6.6. COMMUNICATION osc cccceceeccscssetecsensesecscnsvsecscavsecscnavsecsenavsecsenavsecaeavsecscavsecacnavsecseevseseeavsesaenasaeseenaed 70
6.7. FOREST RESOURCE MONITORING... cccccccesssecseeseesseeesecseneesececnavsecsenavseceeavsecsenevseseeavaeseenavaeseeneed 71
7, LEGITIMACY OF LOCAL INSTITUTIONS ..........ssscsssssssssssssssssssessssssscssssossssssssssessssossssssssssssssssesssoosees 73
TA INTRODUCTION 00 ccccccccececnscesesecsensesecsenecsecseseesecscavsecsenavsecsenavseesenavseceenavsecaeavseseenasseseenavsesseasaeeeeneed 73
7.2, ETHNICITY AND REPRESENTATIVITY o.oo. cccccccccceseessenseesseeesecsenaesecseavseceenavsecsenavsesseavaeseenaeaeeeenae 73
7.2.1 CASES FROM EL AIN, SUDAN. .cccccccccccccescstscscetesnsesecnseescusesscnsesaeesecesecesescuaeeseesesaesseceseecnaseeseaeeaaeaes 74
7.2.2 CASES FROM WEST AFRICA wiccecccccccccccescsteccetecnsesecsesescuseescssesseesecessscesescaeeesesessesecesescnaseeseaeeaaaes 79
7.3. SOCIO-ECONOMIC DIFFERENTIATION 000. ccccccccssseesenseeecseneeseeseneesecseavseceenavseceenevsessenavseseenaveeeeneed 82
7.3.1 WOMEN IN FOREST MANAGEMENT INSTITUTIONS 0. .ccccccccccsessceesscuseescuseeeceseseesecuesescnaeeecnaetaeeaes 8&3
7.3.2. PASTORAL INTERESTS oo.ccccccccccceccceseescuseescusececesesecesecsessscsaeescnsecsessecsesecnesescaesssesecsessecesescnaeeeseaetaaaes 84
7.3.3 INTERNAL AND EXTERNAL ORGANISATION .ui.cccccccccesccsssscsteseeesecnsesecueeescusesecesessesecessscnaseeceaetaeaes 86
TA. CONCLUSIONS o.oo. eccccccccsssecsensesecsecsensesecscnavsecsesavsecsenavsecsenavsecseaavsecsenavsecsenavsecsenasseceeavseseenaviessenasaeseeneed 89
8. SUSTAINABILITY OF LOCAL FOREST MANAGEMENT. ..........cccsssscssssssscsssessscssesscesessescsessnscesoeses 90
8.1. ELEMENTS OF PROJECT DEPENDENCY 0... ccccccecccseesetecscnsesecececsecscevsecscessecscnsesessensssensensvsessenteneetens 90
8.1.1 FOREST REHABILITATION. .cccccccccccsecccssescssececesecnsesecneesscuseescusessessecesescssescnsesseesecaesecaesescasercnaesataes 91
8.1.2 OTHER PROJECT INVESTMENTS ...cccccccccscccsssccesecssesecueeescuseescuseseeeseceseecnseescaesseesecsessecesescnaseecsaesaesaes 92
8.1.3 ILLEGITIMATE POLICTES...cccccccccccsececesetesesecsseescusesecesecessecsecescsacescusecsessecaessecsesescaeeeeesecsesaseeceaesaeeaes 93
8.2. ELEMENTS OF INDEPENDENT FOREST MANAGEMENT ....0o oc ccccccceeecceseteceeneeecscsesesscnsssensenseneetens 94
8.3. CONCLUSIONS ooo. eccccecccccsseeecscnsessvsecsesvsecscsvsecsensvsecsensvsecscsvsecscnsssecsensvsecsensvsecsesvsecsensssenscnsvsessenseneetens 97
9. CONSTRAINTS TO PUBLIC RESPONSIBILITY.............scsssscssssssscssesssscsssesssceseesscoseesscsessscosessnscesonses 97
9.1. LEGAL VACUUM OF LOCAL MANAGEMENT 00. occccccccceeecseneetececnsesecscevsecscnsesessensssesscnsvsensenseneetens 97
9.3. POWER OF STATE AGENCIES VIS-A-VIS LOCAL INSTITUTIONS occ cceeeeceteesereeteesenseneeeens 104
DBD. MALL ecccccccccc cece scesecnsessesecssecseesecsessscseceecusecssssecsesecsesescnaeeeessecsessecescesecseesecesesesecsesecesesenaeeeseaesaees 105
9.3.2. SUDAN viicccccccccceccscesscesecssesecsesescsseescesecsessecessscsecesceseceeesecsesecsesesceseeecssecsesecneseecnesesecessecesesenaeeeseaeeaees 106
9.4, CONCLUSIONS ooo. ccecescsesseeseeseesseecsseseecseeseecseesseecssessescaeesseseneeaeescneeseesenseseesensessesenesseeseneeseeseseneetens 108
10. OPPORTUNITIES FOR THE NEXT DECADE ...........scscsssscsscssvscsscseescsscsenscsscssescsscssescseescsscssnscsseeseees 109
10.1. BUILDING ON EXISTING CAPACITIES 00... ccccccccccecseecscnseeecsensesecscsvsecscnsvsecscsvsessensssessensssessessentenees 109
10.2 PROJECT TOOLS TO ENHANCE LOCAL FOREST MANAGEMENT... cccccceccesseteesenseteteeneeeneens 112
10.2.1 INTRODUCTION .cccccccccccccesesecseescsseeecnsesseesecsessscsseescssecseeecesesceasescsaesscasescaeceeesecsesescnesescaeseeeaesas 112
10.2.1. SIMPLIFIED FOREST INVENTORY .ccccccccccsccssecsseseceseescuseescusececesecessceeescseseeesecessscnesescnaeseeeaeeas 113
10.2.3. LOCAL PRODUCT INVENTORY wccecccccccccccescssecesesecneeescuseescuseceeesecesescnscesescasesenseseseseceseeeneseneeas 114
10.2.4. NATURAL REGENERATION ASSESSMENT .o.ccccccecccceccssessseseseeceeesecneeescnecescnseseeesecesscnesescnaeeeseaeeas 116
10.2.5. PANORAMIC PHOTOGRAPHY \.oiccccccccccecescceeescnseeecneceeesecsesescaeescusecaeesecesscasesesesseesecaeseeeeseneeas 117
10.2.6. MAPPING .iccccccccccccccccccccssecssesecneeescsseescnsesseesecsesescseescesecaessecesescaeescesecaesecesescasesecsesscesescnaeeeseneeas 121
10.2.7. INVOLVING PASTORALISTS AND MOBILE USERS. .occccccccccccscssscceesscuseescnsesecesecnessscnesescnseeeeeaeeas 126
LL. INTRODUCTION ooo cccccceceeesecsenseecseevsecscnevsecscnevsecsensvsesscnevsecscnevsecsesssecsesssessensesessentssestenseneetens 127
11.2 CRITERTA OF WOODLAND MANAGEABILITY ooo. ccccceccccececsensetececnsesececnsvsessensssessensvsessenseneetens 128
REFERENCE S ..........scccsssssossssessssssesssssscesssssesssssscsssssssssssssssssesscsnssessssssosssssssessssssessssessossssssosssssooseesesoosees 141
ANNEXES. .......cccsssossssssssossessscosoesssoossnsnscoseesnscsonsescsonsscosessescosensesensssensesoosensesoesensesonsensescesensescesensesoesensesceseeses 131
ANNEX I. PROJECT PROFILESG............:cssscssssssscsssessscsssessscsseesscosenssscsessscsensescsessnscesensescesessescoseesnsceseeses 131



AND PRODUCER OBSERVATION. 0......:sscsssesssesssesssesssesssssssesssesesssssssusesesesiecssessssssiesssessresssesssesssesssessseessesets 164


ANNEX II.3. EXCHANGE RATES SE£/UKE 2.1995-1.1999 oooccccccccssssssssssssessseesseesseesseesseesesseesssesssesssessseessesees 166

ANNEX II.4. COST MODELS OF LOCAL FOREST MANAGEMENT ........cccsccsscsssessseessessseesseessessseessesssenees 167

List of Tables

Table 1: Three different local forest management systems

Table 2: Cost of establishment for a 2,000 ha village forest

Table 3: Cost of maintenance

Table 4: Overall project cost per unit of local management

Table 5: Military hierarchy in Nigerien forest service

Table 6: Redistribution of fines collected by foresters in the case of Niger

Table 7: Hierarchy of central, local and traditional forms of government in three

Table 8: The total project cost in US$ per year per village, market or other
relevant social aspects as reported by the project.

Table 9: The process of village registration

Table 10: Interventions of forest agents and response from local communities,
individuals and Alamadiou in August 1998.

Table 11: Opportunities and constraints of local product inventory in woodlands
of El Ain and Bankass

Table 12: Opportunities and constraints of panoramic photography

Table 13: Cost comparison for different mapping techniques, roughly in order
from low cost to high cost

Table 14: Firewood markets established in Niger and Mali

List of boxes

Box 1:
Box 2:
Box 3:
Box 4:
Box 5:

Some examples of rent extraction in Niger

Immigrant groups assert forest ownership

Conflicting leadership in Newa village forest management
Appropriation and management of a forest in Samori
Sergue organises itself against excessive state powers




El Ain





Walde Kelka


300 kgs firewood
1 cu.m

A Traditional forest management institution in Mali
“Cercle” or district in the 5th Region of Mali where the SOS
Sahel UK project PPEB and later PAGE operates

Cellule Combustibles Ligneux, Mali

Department for International Development

A “Cercle” of district in the 5th region of Mali where NEF Mali is

A location in North Kordofan, Sudan, where the Forest
Management Project (NFMP) is operating

Project title, Danish/World Bank funded project

Food and Agriculture Organisation

A unit of land equivalent to one acre, approx. 0.4ha
Geographic information system

Geostationary positioning system

Location in Douentza where NEF Mali is operating

Region in Sudan where SOS Sahel UK is operating

Unit of land in Sudan

Near East Foundation

Natural Forest Management Project

Projet Agroforestier Ouest Zinder

Projet pour la protection de l'environnement a Bankass
Participatory Rural Appraisal

Projet d’Utilisation des Sols et des Forets

Location of the PAFOZ project

Village forest in Niger where the rural firewood market system is

Turkana Rural Development Project, Kenya

Supra-village level institution in Kelka forest

= 1 stere
= 670 kgs


In the 1960s and 1970s, an era of central control in Sahelian States, the orthodox
view was that local people were not capable of properly managing their forest
resources. The forest management tasks were considered be too complex for local
people, and their objectives were seen to be short term and local capacity
insufficient for modern forest management. However the failure of central control
forced governments and their donors to search for alternatives, experiments in
village woodlot programmes and joint forest management were a departure from the
orthodoxy and yielded valuable experience.

At the same time, forestry strategies shifted from tree planting to management of the
natural vegetation. In the harsh Sahelian environment it usually makes more sense
to manage the remaining forest rather than restore it where it no longer exists. Shifts
in paradigm were made towards the end of 1980s and led to a new generation of
Sahelian forest management strategies, one which aims at understanding and
developing the much more complex options for natural forest management by local

The initial questions posed in the SOS Sahel study were centred around the nature
of local institutions dealing with the forest, and the nature of the management tasks
performed by them. An important aim was to adapt forest management tools and
make joint forest management more participatory. This soon led to asking more
basic questions about such forest management.

It is evident that conventional forest definitions, concerned with crown cover
percentage, are hardly useful for Sahelian forest management. It is also clear that
many of the forests have always been used and managed by local communities
without the use of conventional management tools. The colonial powers first, and
then the state have weakened but rarely erased “local management’. A dichotomy
appears to exist between, on the one hand, what the state and technical assistance
agencies call forest management, and on the other hand, the way rural Sahelians
measure, communicate, control and harvest their forest.

More appropriate definitions are proposed in the context of this study. The following
definition has been prepared by the workshop on local forest management in the
Sahel, held in Bamako, 13-16 September 1999 (SOS Sahel, 1999):

“The Sahelian forest is any land in the Sahel with a forest statute, customary or
modern, with some woody biomass’ (reference). This definition accepts that a woody
vegetation cover percentage is not useful. Many of the Sahelian village forests are
only slightly covered by trees and shrubs. Yet they are forests under the prevailing
statute, and they may be a key natural resource for some local user groups.

Local management was defined at the workshop as “negotiated management by

stakeholders whose livelinood depends on the forest’, rather than by proximity to the

forest, so that non-local users with critical dependence on the resource are not

excluded. Negotiation is an essential element of the definition to ensure rational
management. The rational of the forester or the firewood cutter, may not be rational
to herder or artisan. The negotiation of their interests is the basis for rational local

In classic views on forestry, the state, supported by development agencies, insists
on its view of the nature of the forest in a given ecozone. If it is thought that the
forest is degrading, is invaded by less desirable species or is turning into a desert,
conventional wisdom insists on prescriptive measures. These would include a crown
cover percentage, a wood volume, an annual wood harvest through a formal
management plan, and nationwide protection of certain tree species.

A thesis of this study is that the objectives set by the forest service on behalf of the
state are strongly sectoral and are not legitimate. Another part of the thesis concerns
so-called modern forest management, which usually includes geographic
partitioning, biometrics, remote sensing, mathematical modelling and a standard
management plan. The thesis is that:

(a) harvest quotas in the Sahel are not useful since the ecology and economy are far
too diverse and variable to be usefully captured in a site specific model.

(b) conventional forest management tools are not useful since they are not
understood by the forest managers.

(c) such tools are usually too expensive for Sahelian forest management, where the
revenues per unit area are extremely modest.

In short, it appears as if there is a real dichotomy between forest management as
conceptualised by the state, and that defined by local institutions. Decentralisation is
likely to provide new opportunities to local management. Local institutions are
increasingly enabled to get on with the sort of forest management they were used to.
However, the physical and human environment have greatly changed since the
precolonial time, when many of the local institutions were established. Will local
forest management lead to socio-economic improvement, equitable development,
civil peace and sustainable landuse ? The experience to date is limited but it does
help to outline the opportunities and constraints of the next decade.

The present report is based on a programme of action-research, surveys, networking
and communication over three years in the Sahel, with a geographical emphasis on
Mali, Niger and Sudan. Three SOS Sahel field projects were the basis for field
research. The experience from a much larger number of projects in 6 countries has
been incorporated, and networking culminated in the Bamako workshop involving 30
researchers and practitioners from 8 countries.

The report is organised as follows. Part one describes how the forestry sector has
evolved, and it gives the ecological, economic and institutional setting. The Sahelian
woodlands are an integrated natural resource of which the ecology and economy are
explored in chapters 2 and 3, with variability as the common denominator.

The climatic characteristic of irregular wet and dry cycles is a major determinant of
the Sahelian woodland ecology. Socio-economic variability is both spatial and
temporal. Large differences in economic structure exist between neighbouring
villages, let alone within the Sahel as a five thousand kilometre wide region. The
interseasonal and interannual variation of economic activity is equally important. The
combined ecological and socio-economic variability is also the key problem of any
model of Sahelian forests indicating how much they can sustainably yield.

The institutional, policy and legal environment is presented in chapter 5. This is the
environment which ultimately determines to what extent local management can be
scaled up. It is a turbulent environment of orthodoxy and innovation, sectoral and
integrated approaches, dirigiste and democratic tendencies.

The core of this report is the second part, which is concerned with the extent to
which local communities are actually managing their woodlands. Chapter 6 explains
the praxis of local management as the alternative to state defined management. If
given the opportunity, local communities do take charge of their natural resources.
What are the nuts and bolts of local forest management, how are cost and benefits
distributed, how are conflicts resolved ? A crucial characteristic of local institutions is
their legitimacy, which is described in chapter 7. Evidently, there are constraints to
local management, some of which may be insurmountable, others which may be
solvable. These are discussed in chapters 8 and 9. Finally, the opportunities of local
management, with and without project support, are presented.

The third section of the report contains the conclusions and recommendations. It
also summarises the understanding gained over the last three years. The most
important case studies have been enumerated in the annex.

In order the enhance the readability of the report, names of places have been
simplified and are explained in the glossary. The major credit for this report should
go to the many Sahelian researchers and practitioners, villagers and project staff,
whose work is the foundation of this study. Responsibility for the opinions expressed
rests with the author and with SOS Sahel International UK.

Paul Kerkhof, 1999.



Precolonial practice ought to be the starting point for a description of change in forest
management, but descriptions are few and sketchy. Comprehensive natural
resource management systems existed in some areas such as the Niger delta under
the Peuhl empire of Macina (Diakite,N.). Even so, written sources of information are
rare. Codified forest management was only brought about by the introduction of
colonial practice.

Forestry introduced by the French and British colonial powers at the beginning of the
20th century was shaped by several factors. It was based on the European forestry
tradition, moulded in the 19th century with German forestry as the cradle of the
sustainable forestry concept. German foresters and their concepts have had an
important influence on the forestry institutions in neighbouring countries as they were
often invited to help shape these institutions (Shepherd, al.1998). European
forestry of the 19th century was concerned with both natural forest management
through single tree selection systems and with plantation forestry for degraded
areas. The emphasis of "modern" or "scientific forestry was increasingly on
clearcutting and planting of forest blocks at the expense of more natural and
selective forms of forest management (Kuechli,C.1997).

Another influence on colonial forestry practice in the Sahel was a sense of urgency
to protect forests against the people who had always lived in or near them. In 1900,
a Forestry Code was instituted in the French Sahel which eroded local access and
management rights. The forest service retained the monopoly over decisions
concerning who could harvest in which forest and under which conditions
(Ribot, J.1995). Wood cutting permits were the key instrument in a process which
increasingly allowed outsiders to exploit local forests. Other instruments included
transport permits, taxation, price setting, and land expropriation for gazettment. In
the process, rural people essentially lost legal control over commercial exploitation of
the forests they were accustomed to manage.

The nature of the emerging forest service can to some extent be explained by the
European model. A sociological study of the German forester explains the
instruments of exclusivity and monopolisation over the forest resource applied by
foresters at the time (Heeg,W. 1974). Language, uniforms and mysticism were
among the instruments which helped maintain exclusivity of foresters over the forest
resource. In the Sahelian context, the militaristic nature of the forest service was
particularly pronounced, at least in the French colonies. The first Malian foresters


were soldiers who returned from the first world war and who were employed to
protect the colonial forests. A military hierarchy has characterised forestry services in
the francophone Sahel from the colonial time until today.

A difference between the French and the English colonial administration in the Sahel
is the broad assumption of state ownership of land and natural resources in the
French Sahel as opposed to reliance on indirect rule in the English colonies of the
west African region (Mortimore,M.1997). In the Sudan, local forest management
systems continued to exist to a large extent until independence, when respective
governments through the forest service assumed formal control at the expense of
customary authority.

Colonial forestry was mostly concerned with the steady supply of fuel to trains and
river steamers. Firewood plantations along rivers and railways became an
expression of modern or scientific forestry, often managed on the basis of forest
blocks. This has influenced post-independence forestry strategies which focused on
urban firewood production in forest blocks, and at a later stage, village woodlots.


The evolution of Sahelian forest management can, unsurprisingly, be seen as a
product of changing political and economic conditions. In the newly independent
states, one party state authority was introduced and reinforced at the expense of
customary authority and traditional resource management. This greatly boosted the
role and size of the forest service given that forest resources are a major natural
resource in the Sahel.

Efforts were made to further gazette forest land but with a modest impact due to the
operational and budgetary constraints of a lengthy reservation process. The main
impact of forest resource centralisation was on public land. The greatly increased
number of foresters and forest service stations on public land over the period 1960-
1990, along with legislative changes, had a strong centralising effect on forest
management and weakened local management institutions. Forestry concepts were
reduced to those accepted in the forest service. In the meantime, the Sahelian forest
ecosystem was degrading, and this was seen to be as a result of climatic instability
and also of the activities of hard pressed but irresponsible rural people.

The State applied two forest management strategies on public land. Natural forest
management was controlled through the permit and tax system and through policing.
Secondly, tree plantations were promoted as a response to the perceived
environmental crisis. In the 1970's, the strategic emphasis was on large scale peri-
urban plantations, mostly for urban energy (Foley,G. et al, 1997).

The general failure of this strategy led to the emergence of the village woodlot in the
1980's. Village woodlots were much smaller, they were situated on public land, which
projects conveniently but ambiguously called "village land", and villagers were
expected to be the beneficiaries and to meet an important part of the cost. In


technical terms, village woodlots were the same plantation model of species,
spacing, planting, weeding and protection but at a small scale. By the end of the
1980's, the village woodlot strategy was considered a failure both in terms of
technique and of community participation (Kerkhof,P.1990).

The important role of natural forests and natural regeneration was increasingly
recognised in the 1980's along with the failure of the plantation model. The Nigerien
FLUP project and Kenya's TRDP were pioneers in the development of natural forest
management strategies. The techniques developed in FLUP relied on physical soil
measures, mulching, tree planting and reseeding combined with protection against
livestock. By the end, the cost appeared to be too high and sustainable production
was lower than anticipated but FLUP was the first major attempt at joint forest
management in the Sahel (Christofferson,K.A.,et al.1993).

In Kenya, TRDP developed the idea that local institutions should be responsible for
forest management, and it prepared a legal framework for this model
(Barrow,E.C.G.,1987). From that time, the idea of legal recognition of local
institutions which carry the major responsibility for local management has spread to
a larger number of countries, albeit at project rather than at policy level. Also the idea
of joint forest management, i.e. joining the state and local communities for the
management of forest resources, has spread. The joint forest management idea
applies in the first place to state forest reserves and, to some extent, to public land.

In the 1980's, international organisations and donors attempted to reform national
forest services and forest policy, aiming to redress the balance from a the bankrupt
policing strategy toward extension. The Swiss Development Cooperation invested
considerably in the Malian forest service training and infrastructure, and so did the
Dutch and Swiss donors in Burkina Faso. The World Bank (IDA) supported change
in the Sudanese forest service and elsewhere.

The outcome of their commitment and that of other donors appears to be mixed. A
greater interest within the forest service in participatory approaches has been
observed (e.g. Kabore,V. et al.,1985, for Burkina Faso). On the other hand, many
structural problems have not been addressed and donor support may have had an
perverse impact through the legitimacy and resources it provided to the forest
service as a centralised management institution. Broad political and economic reform
in the Sahelian societies were required for substantial change in forest management.


The 1990's were marked by a number of related macro-economic and political
changes which had a bearing on natural resources management in the Sahel.
Structural economic adjustment was imposed in all economic sectors throughout the
region. For the first time, recruitment of foresters halted and operational funds were
squeezed. It was increasingly accepted that central control of forest resources was
no longer a viable option. The option of local governance although not willingly
embraced by administrators, was increasingly seen as the only realistic option.


At the same time as the forest resource declined the proportion of forest service
income derived from permits declined steadily in favour of income from fines,
increasingly imposed an arbitrarily ‘blanket’ manner. This oppressive behaviour put
the foresters as the agents of the state in closest contact with the poorest people, in
the front line when political change occurred.

Decentralisation of government is probably the key political change in favour of local
forest management. The political revolution in Mali of 1991 is the most illustrative
example of change from a centralised to a decentralised system of government
within this decade. In other Sahelian countries, the rate of change is less
pronounced but is nevertheless clearly identifiable. The Code Rural in Niger has
defined an alternative legal model of resource management which it is now
implementing, despite the operational constraints to widespread application. The
rural firewood market model developed in Niger is an experiment in local resource
exploitation albeit with very limited local governance.

In Sudan, The Forest Act of 1989 put in place a system of local forest ownership but
the forest registration process is so centralised that its application has proved to be
almost impossible. The recently established system of federal governance is much
more promising. Locally elected Rural Councils, roughly equivalent to the Malian
Communes, are expected to be empowered to support the village forest registration
process. In Burkina Faso and Senegal, the decentralisation process is slow but is
likely to have a steady impact on the ability of local communities to manage their
forest resources.

The notion of sectorally integrated land resource management has gained ground in
the 1990's. In many countries, "environment" has for a long time been identified with
"forestry". In all Sahelian countries, the forest service is in actual fact a sectoral
institution, with few professionals other than foresters. More genuinely
environmental, and thus multisectoral institutions have been established in the
1990's throughout the region in the form of National Environmental Planning
Secretariats. Although these secretariats are in principle under the President's or
Prime Minister's Office, their de facto authority as a cross ministerial organisation is
extremely limited. Mali is the only country in the region which has attempted to
integrate sectoral Ministries into one Rural Development Ministry, so far with mixed

Structural adjustment has had an adverse impact on the efficacy of environmental
planning and policy making (Bass, al,1998; Speirs,M. and Secher
Marcussen,H,1998). The inflation of institutions and policies has gone together with
a reduced capacity for implementation. For local forest management, structural
adjustment is thus a double edged sword: it has helped to enforce forest service
withdrawal from management of village forest resources, but it has also inhibited the
development of alternative, sectorally integrated support mechanisms.

An underlying current in local resource management is increased democracy and
accountability. There is uncertainty about the direction and outcome of this process
in many countries. Local forest management may enhance the scope for democracy
and accountability, but it needs to be supported by a much broader process of
societal change.


The Sahel is usually defined as the zone adjacent to and South of the Sahara desert,
excluding the horn of Africa. The rainfall is the major abiotic determinant but there is
no universally accepted definition of the Sahel. French colonial maps divide the
Sahel into the northern Sahel, with an average rainfall of 100-300mm/year and the
southern Sahel with an average annual rainfall of 300-500mm/year. In this report, the
Sahel is defined as having 100-GO0mm average annual rainfall and it includes east
Sudan'. Nevertheless, the distinction is not clearcut and some cases outside this
zone are included in the report.

A frequently used term in literature is “drylands” or “arid and semi-arid lands”. In the
Sahelian woodland literature regions such as the Casamance, with a rainfall of some
1,000mm, are included because they are part of Senegal (Jackson,J.K.,1983).
Various international drylands forestry conferences recently held in west Africa
include areas of 750-1,500mm (Bruns,S.,,1995). The Sahel in a political sense
may include countries such as the Gambia, but the ecology of such areas is very
distinct from what this report defines as the Sahel.

Another essential abiotic characteristic of the Sahel is the mosaic of run-off and run-
on. What is all too often simplified in development projects and policies as erosion, is
in actual fact a dynamic, spatial pattern of redistribution. This is important in that a
rainfall event in one area may have a quite different effect in a neighbouring area,
which can be conceptualised as a mosaic of run-off, transfer and run-on (Stafford
Smith,M. and Pickupp,G.,1993). A multitude of productive natural resources in the
Sahel, including many woodlands and forests, would not have existed without such

Water is important as both the medium and the substance of redistribution. Soil
nutrients are redistributed across the landscape by water, wind, fire and grazing. The
subsurface redistribution of both water and nutrients, which reappears at the surface
as streams and marshes, or is tapped by deep rooted woody vegetation, allowing
woodlands to grow in zones where rainfall does not support such vegetation, is still
poorly understood.

1 The Sudanese Agricultural Research Corporation has recently changed the policy, inherited from colonial
times, which conceives Darfur and Kordofan ecologically as a part of the east African zone. These zones are
now seen to be part of the west African agro-ecological zone (Hasabelrasoul,F.M.,1999.



The ecological, economic and legal definitions vary considerably, so that, for
example, what is called “village forest” as defined under statutory terms in Sudan
may not be a forest at all in ecological terms. But also within ecology, a large number
of classifications exist. A widely accepted classification in East Africa defines forest,
woodland, bushland, shrubland and grassland, plus various combinations.
Woodland, for instance, has trees of up to 20m in height and a canopy cover of not
more than 20 percent (Pratt,D.J. and Gwynne,M.D.,1977).

A different classification, prepared for UNESCO, defines forest, woodland, bushland,
thicket, shrubland and grassland, as well as various combinations and edaphic
formations. Forest is defined as a continuous stand of trees at least 10m tall, with
crowns interlocking. Woodlands have trees of at least 8m and a canopy cover of
40% or more, bushland has trees of 3 to 7m and wooded vegetation in shrubland is
up to 2m high (White,F.). The case studies in this report cover all of these vegetation
types. Bushland, shrubland and grassland are the most commonly found categories,
and woodland is the economically most valuable category found in most project
cases. In this report, the terms forest and woodland are used interchangeably.

In many case studies, a large part of land which is statutorily defined as village forest
is dominated by bushland and grass savanna with a dynamic composition of woody
and herbaceous cover. This constantly changes over time and space between
treeless grassland, denuded areas and wooded thickets. Intermediate wooded
grasslands are seen by some as the most desired state in terms of maintenance of
productive capacity and biodiversity, but disciplinary orientation may determine
perception. What the forester conceives as reforestation may be seen as bush
encroachment by the range manager.

Five fundamental determinants exist in the savannah ecology: plant available
moisture, plant available nutrients, fire, herbivory and anthropogenic influence. Plant
available moisture is by far the most significant factor. Under drought conditions,
annuals disappear rapidly while hardy shrubs may survive for years. Upon improved
rainfall conditions, annuals may recover in a few weeks, but a forest may require

How the herbaceous-ligneous balance changes depends to an important extent on
the composition of these five determinants. Fire is advantageous to species which
recover quickly such as annuals. When areas are fully protected from fire and
herbivory, woody vegetation may dominate in many ecotopes on the long term
(Moss,J.M.S.,1996). Under a no grazing regime herbaceous vegetation will
increasingly dominate turning these woodlands into treeless savannahs. This has
been confirmed in various livestock exclosure trials (e.g. Peltier,R.,1994).

While fire may be more destructive to trees than to grasses in the long term, grazing
and browsing tend to reduce the fuel load for fires, and may prevent the
development of a thick herbaceous cover which inhibits tree regeneration. A


moderate fire regime may stimulate the productivity of both herbaceous and woody
vegetation. A research project in Burkina Faso found that protection against fire has
reduced production of annuals by 40% and perannials by 25% (Bruns,S. et al.,1993).
Research in Niger found that herbivory hardly had any impact on initial Combretum
coppice survival (Peltier,R.,1994).

A large number of experiments may be quoted, but universal relationships cannot be
established. Simple causal relationships may be established in a site specific context
but they cannot be generalised. One reason is, that a given determinant may have
varying impacts. In the case of fire, for instance, this could be the frequency, timing
and temperature of fire occurrence. Secondly, the various determinants are
interactive. Thirdly, a short term impact may be different from a long term change,
which is poorly understood (Stafford Smith,M. and Pickupp,G.,1993). Slow
processes may operate at the speed of a humans lifetime which makes it difficult to
observe and analyse, whereas most experiments rarely continue for more than 5 or
10 years.

Grass production has frequently been researched in the context of pastoralism and
range management. Herbaceous matter degrades and recovers much faster than
trees and shrubs under varying rainfall conditions, which allows a more rapid
assessment. The tree forage biomass, important for small livestock and camels, may
be much greater than that of the herbaceous vegetation in some Sahelian forests. In
one Nigerien forest, 2,000t. dry matter of tree forage was measured against only
130t. dry matter of the herbaceous vegetation (Ousseinie,G.,1997).


The ecological equilibrium theory assumes that an orderly and directional process
exists by which plant communities are succeeded by others until a persistent,
characteristic plant community dominates becoming the climax community for that
particular site. In many Sahelian sites this is often seen to be forest or woodland. If
this vegetation were to be disturbed, as happens normally under influence of farming
and pastoralism, it would overtime and if left undisturbed ultimately return to the
forest climax. Restoring the forest climax is also an ecological justification in forestry
for rehabilitation and management of woodlands.

A number of researchers feel that in arid environments with a high drought
frequency, such as the Sahel, a disequilibrium exists in which abiotic disturbances
are dominating. The idea is that the ecosystem is disequilibrial, event driven, and is
characterised by non-linear dynamics. Climatic instability is an important element of
a disequilibrium ecosystem (Moss, J.M.S.,1996; Behnke,, 1993;
Coppock,D.L.,1993). Five cases are described here to illustrate the particular
dynamics of Sahelian woodlands.

1. The most comprehensive natural forest research in the Sahel has been done on
the retracted vegetation patterns (such as “tiger bush”) which constitute a common


vegetation type in south east Niger and which are found in parts of the Sahel
receiving approx. 300-500mm/year.

Retracted vegetation such as the tiger bush is a woodland on terrain with a very
gentle slope (0.5-2%). Soils are lateritic with high run-off on patches of denuded soils
which alternate with vegetation strips and patches which absorb the run-off. The
shape and size of vegetated patches vary as a function of rainfall, soil and slope
(Ambouta,J.M.K., 1997). The profile of these vegetated areas consists of
herbaceous plants upstream, which catch some of the run-off. Litter and branches
are the most important micro-environment for natural regeneration, they catch and
stabilise soil and seeds and protect seedlings from browsers. Subsequently, a
mixture of woody and herbaceous vegetation exists and termites are very active. The
increased soil porosity and soil organic content result in complete infiltration of run-
off, and allow growth of Combretum nigricans, Gardenia sokotensis and other
species which are normally found in the Sudanian rather than Sahelian

The grass vegetation on the upstream side of the vegetated strips and patches tends
to colonise bare soil, under conditions of reasonable grazing pressure and rainfall.
The woody vegetation follows suit, and so does the edaphic zone of run-off
infiltration. The trees on the downstream section of the vegetation strip receive
increasingly less run-off, and they slowly die back, beginning with species normally
found in the Sudanian zone. The entire vegetation strip, usually 10-40m wide, is
slowly moving upslope. During important drought events such as 1982-84, part of the
tiger bush died off. 10 years or so after the drought, part of the run-off is no longer
absorbed by the tiger bush and it has not regenerated. Tiger bush in Sudan has
largely disappeared since the drought (Wickens,G.E. and Collier,F.W.,1971).

Highly retracted and less retracted vegetation may co-exist on a given locality such
as a plateau, with similar rainfall and soil conditions. Research in Niger suggests that
the above ground biomass and productivity of highly retracted vegetation is higher
than of relatively dispersed vegetation under similar conditions. The biomass of
highly retracted vegetation in the case study is 17.5t/na and of dispersed vegetation
it is 15t/na (Ambouta,J.M.K.,1997). In terms of firewood volume (wood over 4cm
diameter), highly retracted vegetation has a stock of 14cu.m versus 8cu.m for the
dispersed vegetation. Vegetation retraction offers protection against drought and,
once established, such vegetation will only very slowly disperse under improved
rainfall conditions.

2. A study of the forest of Bahn in northern Burkina Faso compared vegetation
patterns between 1955 and 1984, when the average rainfall was 510mm/yr. The
rainfall over this period was lower than in the previous 46 years, when it was
710mm/yr. In 1955, only 27% of the sample area (1,500ha) was covered with trees,
bare soil covered 6% and the remainder (72%) was grassland. By 1984, trees
covered 40%, bare soil 55% and grassland only 5%.

In 1955, the trees were mostly dispersed in the grass savannah, and the largest
contiguous patch of woodland covered 1/7ha By 1984, the distribution of the tree


vegetation had completely changed. Many of the small patches of woodland that
existed previously, had either died off or had been absorbed by a much larger patch
of woodland, which was confirmed by the large number of dead trees which were
found outside the woodlands of 1984. In 1984, the largest contiguous patch of
woodland covered 500ha, which is 73% of all tree cover in the sample area

The researcher attributes the changes observed in woodland cover and structure to
a combination of reduced rainfall, and the increased run-off from outside and inside
the sample area. Outside the sample area, clearing for agriculture has been a source
of run-on, and inside the sample area bare soils have contributed to run-on in certain
woodlands. Despite a reduction of 28% in average annual rainfall, run-on has
increased considerably, and forest cover increased from 27% to 40%. Many small
patches of woodland died but in the form of contiguous woodland in run-on sites
forest productivity has much increased. The dynamics of this forest resembles that of
the tiger bush but at the scale of several thousand hectares, unlike the woodland
ecology described for south east Niger.

3. Acacia mellifera vegetation in North East Africa may have a cycle of regeneration
and die-back over large areas. Acacia mellifera forms dense thickets which prevent
herbaceous production and protect the thicket against fire. Natural regeneration does
not occur and eventually the bush dies off and is replaced by grassland
(Harrison,M.N. and Jackson,J.K.,1958). An Ethiopian study found that high livestock
pressure in savannah lead to gradual encroachment by Acacia mellifera. The bush
savannah is being replaced by dense thickets, forcing the Borana pastoralists to
migrate and abandon the area, but in the long term, the thickets are dying off due to
lack of recruitment. A herbaceous vegetation recolonises the area and pastoralists
are likely to return. The whole cycle is expected to last 60-100 years

Woodlands in North Kordofan, Sudan, with an average rainfall of about 300mm, are
dominated by Acacia mellifera which is the main source of construction wood and
firewood. In 1990-1993 researchers were concerned about the lack of A.mellifera
recruitment and an important tree planting scheme was initiated (Seif el Din,A.G.,1992). During these years, the average annual rainfall was 241mm, which is
well below average. Over the years 1994-1998 the average annual rainfall rose by
58% to 380mm and A.mellifera recruitment increased steeply. This rendered earlier
woodland restoration strategies, such as tree planting in microcatchments, quite
meaningless (Kerkhof,P and Siddig,F.,1998). By 1999, 58,000 trees has been
planted in microcatchment schemes over a 9 year period, while about 3 million
Acacia mellifera had regenerated naturally since 1994.

4. In the 1980's, arid lands forestry projects such as TRDP in Kenya were concerned
about the lack of natural regeneration in Acacia tortillis forests (Kerkhof,P.,1990).
Riverine forests of aged A.tortillis were lacking recruitment and it appeared that in
the continued absence of regeneration, possibly as a result of browsing extensive
forests would die back. Ecological research elsewhere, however, found that A.tortillis


may regenerate abundantly but not under its own canopy (Moss,J.M.S.,1996). A
spatially irregular pattern of A.tortillis forest can be anticipated in low rainfall areas
instead of the model of homogeneous forest formation and rejuvenation commonly
found in forests of high rainfall areas.

The various cases quoted originate from different geographic areas but they have
various aspects in common. Low and variable rainfall has a big impact on the
woodland ecology, which is subject to important long term changes. The spatial
patterns of the woodlands vary over time and may oscillate between “denuded soils”
and “dense woodlands’. Ecological characteristics such as recruitment and die-back
evolve accordingly. These dynamics may be conceived as a steady state in some
cases, but they appear to be more or less irreversible in others. The Nigerien tiger
bush or Acacia mellifera in Borana, for instance, may be seen as stable vegetation
over larger areas even though it dies off locally. However, in the case of Bahn or the
tiger bush in Niger and Sudan seriously affected by drought, vegetation changes
appear to be non-linear.

The most important conclusion for this study is that tree stock and productivity for a
given site may not evolve according to the principle, by which sustainable wood
production can be achieved through management, determined by site specific
factors such as rainfall and soils. In spite of management activities such as tree
planting and fire control, woodlands may deteriorate and die off or alternatively,
woodlands may expand in the absence of such measures. Woodland changes and
forestry interventions in Sahelian woodlands should be evaluated against the
particular ecological


Little is Known about more indirect functions of woodlands such as the provision of
water and nutrients to agricultural production systems through redistributive
mechanisms, such as the run-off / run-on mosaics mentioned before. Livestock
corraling is another form of redistribution. Livestock owned by villagers produce
manure which is distributed by farmers on their fields. Livestock of nomadic
pastoralists is kept in temporary corrals which, once abandoned, may be cultivated
by neighbouring villagers, ¢.g. in Kordofan.

The relationship between woodlands and dry season water availability is also poorly
understood. One idea is that forest and woodland attract rain and produce water,
hence the need for reforestation and forest protection. Agro-hydrlogical studies
elsewhere in Africa have demonstrated that generally the contrary is true
(Loerup,J.K. and Hansen,E.,1997). Although such studies have apparently not been
conducted under Sahelian conditions, the same inverse relation between water
availability and forest cover may exist.



Sahelian woodland management can be classified in two distinct categories:
interventions by foresters through the forest service and through projects, and
management practice by local people. Although forest service management
objectives vary from place to place, they tend to concentrate on sustainable,
optimum wood production. A range of management activities are undertaken to
achieve this, which may include:

- forest inventory and planning, which is the basis for any of the subsequent activities
- division into compartments

- fireline establishment and maintenance

- tree planting or enrichment planting

- direct seeding

- cutting trees of little use and retention of valuable species

- guarding against grazing, browsing and cutting

- clearfelling, simple coppice, coppice with standards or other forms of exploitation.


Forest inventory is undertaken so as to determine stock, productivity and annual
offtake. The key issue is what should be measured and how it can be measured with
sufficient precision and at reasonable cost. Initially, only trees qualified as sawn
timber or usable for other purposes in the colonial economy were measured. The
definition of what should be measured has varied constantly ever since. Forest
inventories in the beginning of the 1980’s under FAO and UNEP, for instance, used
the 10cm diameter threshold as a minimum for wood volume estimates (FAO, 1991).
This seriously underestimates firewood volume and more recent inventories
establish a 4cm threshold. Other studies again propose wood weight as the reliable
estimate for wood energy, which is complicated by variable moisture content
(Nygaard,R.,1998). Such estimates may be air dry or oven dry. The continuously
changed definitions in forest inventories makes comparison between them very

The annual yield or productivity is still more difficult to assess. In the case of single
age stands, typical in European forestry where most of the techniques have been
developed, it may be the average annual increment measured over the rotation, or it
may be the actual increment for a given year. In Europe, reliable data exist for key
species such as spruce and douglas fir. But in natural European forests with many
different species, age groups and exploitation systems, reliable production tables are
hardly available. If reliable estimates are rarely available for European mixed forests,
they cannot realistically be expected for Sahelian forests.

The earliest estimates are from Clement, who estimated Sahelian forest productivity
at 0.5 cu.m/ha/yr. However, very few of the sites observed by Clement fall within the
Sahelian zone (Jackson,J.K.,1983). According to Jackson, who presented the first


major review of Sahelian forest management, forests with a rainfall of less than
400mm can be considered to be virtually unproductive, and should be preserved on
environmental grounds with only dead wood being collected. This does not tally with
SOS Sahel's experience from north Kordofan, where village woodlands in the 250-
400mm rainfall zone are actively managed, exploited and play a role in the local
economy (Kerkhof,P. and Siddig,F.,1998).

In the 1980’s and 1990's, a number of Sahelian forestry projects have estimated
standing stock and forest productivity. Two examples are provided here to
demonstrate the difficulties of arriving at reliable estimates. The first example is
Takieta forest in Niger. In 1987, the FLUP project did a forest inventory and found a
standing stock of 0.2t/ha (0.3 stere/na)(Anon, 1988). In 1991, 1992 and 1995 minor
inventories have been undertaken by various projects using different methods
(Kalla,B.,1992; Carr,S.; Habou,A.). In 1996, a more important study, using different
methods again, arrived at an average stock of 4.6 stere/na (Mounkaila,M.,1997), but
the results are not considered useful by project staff, since volume tables from a very
different ecological zone were used (Kerkhof,P.,1997). In any case, the difference in
stock estimates of 0.3 stere in 1987 and 4.6 stere in 1996 is inexplicable.

In a second Nigerien example, the Gusselbodi forest, inventories were conducted
using more or less comparable techniques over a 10 year period. In the early 1980's,
the FLUP project estimated productivity at 0.5 stere/ha/yr which was revised to 1.0-
1.5 stere/na/yr later in the 1980’s (Ichaou,A. and d’Herbes,J.M.,1997). Two
inventories after project termination arrived at a productivity of respectively
0.25stere/ha/yr and 0.34stere/ha/yr (Hopkins,C.,1992). One problem faced by
forestry researchers is that woodland management is not fully controlled, and sample
areas may be over or underexploited for some period of time. Complete protection
against any exploitation over a long period of time is rarely feasible and even
monitoring of forest exploitation is hardly possible.

In spite of inherent difficulties, stock and productivity estimates continue to be made
by projects who feel there is no way around them. For one category of projects,
particularly those of NGO’s, the standard forest inventory is an obligation even
though the results are hardly used or considered useful?. Some project staff feel that
the inventory was necessary to justify the project strategy of involving local people
vis-a-vis the forest service, since “it takes the pressure off’, even though the results
may not be used in any way.

In the rural firewood market projects, forest inventories are held because sustainable
yield, in the form of annual permissible quota, is seen as a cornerstone of the project
strategy. A local forest management strategy without a quota is considered
impossible. Energie-ll, for instance, found an average productivity of 0.66 stere/ha/yr
for the Tientiergou forest in Niger, which served as the basis for determination of an

* In the French Forestry literature as recently as 1998, many European mixed forest management
concepts are described as ‘chaotic’ and as being without scientific basis. | argue that productivity
estimates under a multitude of Sahelian management regimes cannot reliably be produced in the
absense of clear silvicultural concepts.


annual quota (Peltier,R.,1994). However a project survey 6 years after establishment
of the market system in one of the two surveyed forests, found that local people were
concerned about forest depletion, despite indications that they observed the quota
system (Giraud,S.,1998).

A Malian firewood productivity study of 1998 concluded that none of the surveys and
experiments in the country allowed reliable productivity estimates (Sylla,L.M. and
Nouvellet,Y., 1998). The rural firewood market system requires productivity
estimates, whether reliable or not, and the researchers propose a mathematic model
in which productivity is determined by rainfall and by crown cover. Yet modelling in
absence of substantial empirical data is generally found to be unreliable.

An alternative to the forestry principle of sustainable yield, such as a firewood quota,
may have to be developed. This may be conceptualised in the form of minimum
environmental standards to which local management is to adhere (Ribot,J.C.,1995).
Such standards should have long term, area wide indicators which "see through"
short term, site specific changes. It should also avoid an overemphasis on
(sub)sectoral concerns through the focus on “environment” as a multi-sectoral


The formal forest management plans resulting from inventories, are more or less
based on standard procedures in forestry. The rural firewood market system
provides guidelines for the production of these plans, which should include :

-general environmental and socio-economic background information

-maps showing boundaries of the forest and of forest blocks

-forest exploitation plan, with annual quota

-possibly an investment plan with tree planting and other rehabilitation measures.

In actual practice, few Forest Reserves have a formal management plan. In the case
of North Kordofan State, none of the 160 Reserves have a written management plan,
other than one plan produced by an expatriate adviser. In the Zinder Department in
Niger, for instance, most of the Forest Reserves no longer exist de facto, and there
are no management plans. In the case of rural firewood markets supported and
supervised by a project, plans exist but they are much simpler than what has been
prescribed (e.g. Energie Il). In a number of projects, natural forest management
plans have been written by external consultants for villages (e.g. NEF Mali,1997;
CARE International/Mali, 1998) but application is unlikely to take place. The following
management activities have been undertaken to varying degrees.



1. Forest boundary demarcation and protection. This is the single most important
management activity for both government and locally owned natural forests. In the
case of State Forest Reserves, deployment of guards to protect the forest against
tree cutting, grazing and possibly trespassing is the traditional task of the forest
service, which can apply highly punitive sanctions. Nevertheless, many State Forest
reserves are poorly guarded due to lack of funds, and others have disappeared

In the case of locally owned forests, protection is the key management activity which
is executed even if no other form of forest management takes place. Protection
tends to be much more diversified and very modest sanctions are applied, but under
good institutional arrangements, such protection may be much more effective than
that by the forest service. In State Forest Reserves and in some of the village forests
demarcation has been done, but in may cases markers have been destroyed.

2. Division of the forest into compartments, whereby each compartment may
represent an annual cut. This is done in all state forests where management is
prescribed under the Rural Firewood Market system. Instead of a large number of
blocks, as recommended by FLUP, the Rural Firewood Market system recommends
a smaller number, e.g. 6 blocks per village forest, in order to simplify management
and reduce the cost of boundary delimitation.

Energie-Il found that dividing the forest into blocks is useful in order to

- facilitate control e.g. by the relevant village institution since it will be Known which
part of the forest can be exploited in a given year.

- ensure an even distribution of exploitation

- allow easier ecological monitoring

However, these arguments can be turned around, and it can equally well be argued
that in an undivided forest estate exploitation is more evenly distributed. Ecological
monitoring may be easier if locally identified vantage points are used instead of
externally imposed blocks. In many village forests, monitoring of the whole forest is
necessary to avoid illegal exploitation by outsiders, at least for a part of the dry

Most locally managed forests which are not under the rural firewood market system,
do not have a system of forest blocks. SOS Sahel found that different parts of the
village forest may have different names such as "The Forest Situated Near The
Village Limit" or "The Forest Situated Around The Ancient Village". Different parts of
the forest are useful to varying degrees for a host of uses: livestock, domestic
energy, artisanal and construction wood, fruit, and medicine. The formal partitioning
into forest blocks which represent annual cuts is perhaps only useful in the case of
single purpose forest management systems.


In the case of village forests studied in Kordofan, a number of villages have
partitioned their forest into two or more units, with the following classification system:
- Natural forest/planted forest. The latter usually covers a very minor part of the
forest in most villages, but it has the highest level of protection.

- Rich/poor forest cover. The former tends to be used for construction wood, while
areas of poor forest cover are protected against green tree cutting.

No form of delimitation has been done in any of the forests which have been divided
into blocks or categories, since all villagers know where the different blocks are

Some projects deal with state forests for which local management is being
developed. They often find it useful to partition the forest into areas adjacent to the
villages, or a group of villages. In Tientiergou, Energie-Il recommended that the size
of a forest managed by a village should be limited to 4 km7. This implies that in many
cases, the forest is situated at 4-8 kms from the village centre, given that agricultural
land is situated close to the village. In Takieta, SOS Sahel found that groups of
villages want to manage a piece of state forest (Takieta Forest Reserve). The size of
forest considered manageable by villagers appears to be not much different from
what was found in Tientiergou.

3. Enrichment planting, with or without soil and water conservation measures, has
been widely practised and still figures in many project plans, together with direct
seeding. Tree planting may not be recommended since cost are always
disproportionate to returns. Direct seeding has been applied by some projects in
combination with tillage or organic litter. It is recommended by Energie-ll and is used
by various Sudanese projects, but the same reservations have to be made with
respect to economic returns.

4. Firebreak establishment, early burning and fire fighting is practised to varying
degrees, and is stimulated by various forestry projects. This management practice is
not often found in forests under local management, but local people do fight fires in
their own forest. Firebreaks were established in most of the Kordofan village forests
studied but they are not maintained, although firefighting on village land, not just the
forest is done. In many other forests studied, such as Segue, Kelka and Samori in
Mali, firebreaks do not exist and many of the forests are probably too large for
effective firebreaks.


1. Clearfelling. This has been tested by a number of projects and research stations.
The ecological effects are very important, since coppicing is poor and the woody
vegetation tends to disappear entirely. Fields abandoned several decades ago can
easily be traced on aerial photos since colonisation by woody vegetation is mostly
limited to few species such as Guiera senegalensis.


2. Simple coppice, or coppice with standards, with a minimum girth limit for each
species. The FLUP project recommended a height of cut of 30 cm, whereas the
Energie-lIl silvicultural recommendation is "as close to the base of the tree as
possible", although it recognises that for practical (economic) reasons trees should
be cut at e.g. 30 cm height.

SOS Sahel found that under local management, construction wood is generally cut
selectively, leaving many stems of a multi-stemmed tree untouched, and thereby
preserving much of the productive capacity. Furthermore, villagers apply a more or
less random selection of trees. In some cases, informal groups of woodcutters
decide to work together and create forest extraction roads and paths in denser
patches of forest. This applies to most forms of small size construction wood, which
appears to be produced under the most sustainable form possible. However, in
some categories of forest exploitation the whole tree is cut and dies, at least in the
lower rainfall areas (e.g. 300mm). This is usually done for certain forms of artisanal
wood and for large poles.

Energie-ll reports that the woodcutters of Tientiergou prefer to cut at the beginning of
the dry season when the trees are still full of sap and hence easy to cut. They found
that cutting in the middle of the dry season is often fatal for trees, and cutting at the
end of the dry season allows trees to regenerate throughout the rainy season and
grow beyond the reach of browsers. It is likely that shoots require a great amount of
water in the first year and in the absence of sufficient water the tree dies. For each
tree cut, Energie-ll found an average of 10 shoots growing in the year following the
cut (Peltier,R.,1994).

SOS Sahel, however, found that commercial tree cutting is commonly done in the
middle of the dry season, for plain economic and probably insurmountable reasons:
at the beginning of the dry season much labour is required for the agricultural
harvest. This was found for all research sites and may well be generally valid for the

3. Many other forms of forest exploitation exist but are poorly researched. Forestry
research and development projects have primarily been concerned with wood
production, and especially firewood production. Livestock production is generally
acknowledged as an important aspect of woodland management but it has been
given little attention. In many projects of the 1980’s, livestock herding woodlands was
discouraged on the grounds of damage to young trees and shrubs. Woodland
management projects at the time, encouraged tree planting as a form of forest
rehabilitation, and such plantations are quite vulnerable to livestock pressure.

Research on the interaction between livestock and Sahelian tree vegetation
exploited by local people suggests that under most conditions, livestock does not
seriously affect tree productivity. In a Nigerien experiment, a few highly palatable
species such as Boscia angustifoloa and Pterocarpus erinaceus were seriously
affected by livestock (Achard,F.,1997). In measurements of regeneration of the
dominant tree species in Kordofan woodlands, significant differences were found in


recruitment of key tree species under varying fire regimes, but livestock had not
negatively affected tree regeneration (Kerkhof,P.,1997). Recent woodland
management projects no longer recommend livestock exclusion. The question
whether, intensive firewood production regimes such as the rural firewood markets
negatively affect livestock production, has hardly been researched. However some
researchers have voiced concern about the impact of firewood market programmes
on livestock production (Paris,P.,1997; Madougou,D.,1998).

Other forms of exploitation include collection of fruit, leaves and flowers, production
of pharmaceutical products, oils, gums, and dyes. Apart from important commodity
crops such as Acacia senegal, very little is known about their management. In SOS
Sahel study sites, woodland management rules prohibit green fruit collection in all
cases. Grazing and browsing is prohibited or controlled in a few forests, away from
tree planting sites. But so far very little information exists about productivity under
different management regimes.


1. Sahelian woodland ecology is affected by strong climatic, hydraulic and biological
fluctuations which should be understood before recommending management
systems. Due to the site specific nature of these fluctuations, local management is a
rational response if it is accepted that local managers have a good knowledge of
these fluctuations. Standard forestry planning mechanisms developed in Europe are
less suitable to the Sahelian conditions.

The sustainable yield concept, which is at the heart of European forestry, is
problematic. It aims at a quantified assessment of annual wood yield as the basis for
management, as it is applied in the firewood quota system. It is at odds with the
complex natural, multi-purpose forests in the Sahel. If standards are necessary,
alternative concepts may have to be developed such as "minimum environmental

2. Many of the prescriptions for woodland management in project documents such
as tree planting, direct seeding, compartmentalisation and grazing control have a
poor ecological or economic basis. In a situation where there has been little success
in researching and developing sound technical packages, more attention should be
given to local knowledge systems. Whereas agricultural research and extension
agencies have appreciated indigenous farmer knowledge for some time, this is not
the case for indigenous forestry knowledge in the Sahel.

3. From the few available case studies it appears that woodland management by
local groups is more sophisticated than the general literature suggests.




The economic value of the Sahelian woodlands and the economic impact of different
management regimes are not often analysed. Woodland management projects have
primarily been run by foresters, who have pursued innovations in forest inventory,
rehabilitation and exploitation. Development agencies have been interested in
institutional innovations, sometimes with extensive anthropological or institutional
research. But although projects have an interest in economic impact of woodland
management, few if any economists have been engaged except for the study of
firewood markets. In most projects, neither the economic data nor the methods and
expertise to obtain them are available.

The stakes for Sahelian forestry could be high. Lessons of the past demonstrate that
successive strategies, from peri-urban plantations through to village woodlots and
natural forest rehabilitation, were technically feasible but had a very poor economic
efficiency (e.g. Foley,G. et al,1997; Kerkhof,P., 1990). It is evident that contemporary
forestry strategies also depend on economic conditions which ought to be monitored.

Present local forest management strategies are dominated by the rural firewood
markets approach in the francophone Sahel and by village forest gazettement in
Sudan. Firewood markets are not only a dominant current strategy but also a
particular case where market economics is highly developed. This category of
projects has a strong monitoring system for one subsector, urban firewood, but little
interest in other woodland values.

Woodland values other than firewood need to be analysed since they may be more
important. This has been done in three case studies in this report in some detail:
Bankass and Kelka in Mali, and Kordofan in Sudan. The present chapter also
includes a review of economic evaluation methods applied in the Sahelian
woodlands. Natural resource economics is poorly developed and the discussion
about methods is as pertinent as the data generated.

Finally, an assessment of the cost of management is made based on project
expenditure data. A cost benefit analysis is not possible due to insufficient data, but
the cost data do serve as an indication of the economic efficiency of the various local
woodland management strategies.


One reason for the lack of attention to woodland economics is perhaps the past
performance of economic tools in development projects (Davies, J. and Richards,M.,
1998). In the 1970's and 1980's cost benefit analysis was the predominant economic
tool used in forestry project appraisal. The outcome was usually a favourable internal


rate of return for the project appraised. But poor development impact in many
projects has discredited not only cost benefit analysis but probably also the role of
economics in forestry development.

Cost benefit analysis has also been used in evaluation, such as the Majjia Valley
windbreaks in Niger (Dennison,S.E., 1996; Rorison,K.M. and Dennison,S.E.,1987;
Dennison,S.E.,1988). Again, the outcome of the economic model was very positive
whereas the long term outcome of the windbreaks is disappointing.

Several explanations can be put forward for the poor performance of cost benefit
analysis in forestry projects. The main reason is probably the gross simplification of
the models used and the overoptimistic assumptions applied. Physical data such as
forest productivity were rarely available and data from elsewhere, such as research
stations, was used instead. In most cases, the interactions between forestry,
livestock and agriculture were underestimated. Market information was often patchy,
simplifying both the range of products and the interseasonal and _ interannual
variations. Cost elements such as taxation, fines, bribes and confiscation may not
have been included, and the subsistence economy was often poorly appraised.

Questionnaire survey, usually aimed at households, is another widely used tool in
economic analysis, which may be used as an input to a cost benefit analysis. This
tool has the potential benefit of capturing internousehold variation, but it has severa
limitations. Many questionnaires deal with a few variables out of the many which
influence the forestry sector. Comprehensive questionnaires, some of them
consisting of 700 questions, have also been used in Sahelian forestry, but the
reliability of the data is not known (e.g. Niger: Giraud,S., 1998). Reliable data on
issues such as natural resource exploitation, herd size and even grain harvest may
be hard to obtain, either because they are related to wealth or because they involve
illegal activity. Finally, many large household surveys are poorly understood by loca
people and project staff, and have thus little impact on the development process
(Chambers, R., 1997).

In the 1990's, two methodological innovations have taken place in the economics of
natural resources: participatory economic analysis, and environmental economics
(Winpenny,J.T.,1993; Davies,J. and Richards,M., 1998). Davies and Richards
distinguish three different traditions, while recognising important overlap:

1. Economic and financial cost-benefit analysis
2. Environmental economic analysis
3. Participatory economic analysis

Participatory economic analysis, has put local producers and consumers at the heart
of the analysis. Instead of external analysts implementing large household surveys,
local people have a central role in economic assessment with the assistance of a
facilitator. One of the strengths of participatory analysis is the incorporation of non-
market values, which a cost-benefit analysis tends to put in a footnote or leaves out
altogether. The tools are derived from participatory rural appraisal, such as resource


mapping, transects, sociogrammes, group discussions and ranking exercises. The
economic appraisal is thus a thematic PRA, one which focuses on the economics of
a local forest resource. IIED's Hidden Harvest programme has been a major
promoter of participatory resource economics, with some case studies in African
drylands (e.g. Pretty,J.N. and Scoones, |.,1989; Bishop,J. and Scoones,|.,1994;
Campbell,,1995; IIED, 1995; Adaya,A.L.,,1997).

On the other hand, participatory economic analysis is characterised by a low level of
quantification. The perception and relative values of local people are emphasised
which inhibits quantification, and local measurement units make it hard to apply the
information on a wider scale. PRA supporters maintain that it is better to be
approximately right than to be precisely wrong. Triangulation, the PRA expression of
data corroboration, is seen as the main tool to deal with bias. However, it seems that
this is often not applied in practice, which raises the question how “approximately
right” the outcome is. Furthermore, the way PRA is usually applied is as a one-off
survey, in which case it fails to capture important variables. Internousehold variation
is poorly understood since PRA tends to focus on groups.

Environmental economic analysis has been developed in order to value the many
non-marketed cost and benefits of the environment (Winpenny,J.T., 1993). Many of
the methods are duly poorly applicable to developing countries, but during the 1990's
some have been adopted by forestry economics in Africa and elsewhere. (Emerton,
L., 1996; Brown, K. and Emerton, L., 1997). In some of the case studies, local units
of measure have replaced money and symbols have been used instead of written
text, which overlaps with PRA principles. The final outcome of the assessment is in
monetary terms, and economists are divided about the usefulness of the method.

The most useful contribution of environmental economics to forestry in developing
countries is probably conceptual rather than practical. Forestry economic
assessment in the Sahel has often been dominated by a few marketed forest
products, and by urban woodfuel in particular. Environmental economics has
conceptualised a much broader framework of values, marketed and non-marketed,
direct use and indirect use, use and non-use values. This may help to put marketed
firewood and construction wood in a more realistic context of woodland values.
Firewood may be the most visible product for townspeople, but for local actors,
grazing, bequest to the next generation, and options for future land development
may be more important.

Sahelian forestry economics literature is modest and is mostly based on household
and market surveys and producer interviews, usually dealing with the woodfuel
cycle. Some studies have attempted to assess the livestock production value of
woodlands, and a few have attempted to quantitatively assess a wide range of



The Nigerien Energie-lIl project (1989-1996) has built on the firewood market model
developed in the 1980's by FLUP, and the Guesselbodi forest project in particular.
FLUP developed the concept of natural forest restoration and controlled exploitation
by local communities. It concentrated on ecological and technical innovations, with
management based on a contract between the forest service and local woodcutters

The main contributions of Energie-Il have been in scaling up of the FLUP project
centred approach and achieving and a_ significant reduction of woodland
management cost. It has also developed an improved fiscal policy, which taxes
firewood from uncontrolled woodland exploitation more than firewood sold under
agreement with the forest service, through the rural firewood market (Foley,G. et al,
1997). The market model was adopted on a much larger scale than in previous
projects to both public and gazetted forest lands. On the regional scale, the Energie-
Il model is being adopted in Mali, Senegal, Burkina Faso and in other countries.

Three different rural firewood market models are analysed here. In all three cases,
the issue is one of firewood commercialisation by local institutions regulated one way
or another through the forest service. In Energie-ll (Niger), the emphasis is on public
land, in Kita (Mali) the project is mainly concerned with gazetted forest. In the Kelka
(Mali), firewood production is exclusively done on public land. All three projects are
pursuing the establishment of rural firewood markets for which the legal context has
been created in the two countries, but they have different strategies and are in
different stages of development.

Over the period 1994-97 the Kita project has established 48 rural firewood markets
which, in 1996, have commercialised 10,710 stere of firewood, with a gross revenue
of FCFA 17 million. The revenue accumulated over the project life time represent a
benefit of USD 2 per inhabitant, against a project cost of USD 30 per inhabitant (10).
This financial cost benefit ratio does not include indirect benefits such as capacity

The project has set up a system which allows local associations to benefit from the
productive gazetted forests in this region, which has an average rainfall of 600-
1,100mm per year (11). Kita is considered the first rural firewood market system
functioning in Mali, but a part of the functions, including tax refunds, have been
executed by the project instead of the state. The Phase Ill evaluation notes that
highly profitable groundnut production is now the main threat to the forest, due to
illegal forest clearing. It follows that firewood marketing opportunities do not stop

local people from converting forest into agricultural land if they anticipate higher
benefits from such alternatives.

Fifteen villages in the Kelka forest in Mali have been supported by the Near East
Foundation (NEF) in order to appropriate the forest resources on village land, and in
particular the commercial exploitation of firewood. An extensive household survey


carried out by NEF Mali in the 15 villages reveals the importance and dynamics of
firewood markets (NEF Mali, 1997). The Kelka forest covers about 100,000 ha and is
situated along the tarmac road north of the regional town of Mopti, which offers
exceptional opportunities to urban firewood commerce. Over 40% of Mopti's
domestic firewood consumption in 1998 came from the Kelka forest (CCL, 1999).

The firewood subsector is the most important revenue earner in the Kelka, but
analysis of the 5 year dynamics shows that the volume is constantly decreasing,
from 35,700 steres in 1992 down to 15,200 steres in 1996, representing a decline of
18-20% per year (NEF Mali, 1997). Firewood revenues fell from FCFA 64 million in
1992 down to FCFA 39 million in 1996, without allowing for the devaluation and
inflation of the FCFA over this period. In the Kelka, although local forest
management and related institutions have existed for a longer time, the rural
firewood market approach is recent and is motivated by the need in Niger to legally
strengthen the local institutions and to obtain the partial tax refunds typical of the
market system.

Energie-Il has led to the creation of 60 rural firewood markets in the Niamey basin
which cover about 125,000 ha of forest and sold 75,000 stere of urban firewood in
1995. The gross income before taxes and other charges was FCFA 129 million in
that year. The system is currently under pressure as donor funds have not been
available since 1997, but the legal and tax innovations instituted with support of
Energie-ll have remained in place (Foley,G. et. al, 1997).

In all three projects, a main result of the rural firewood market system is that outside
traders can no longer legally exploit local forest resources. Local exploitation is
legalised under a contract between the forest service and local cooperatives, and the
tax system allows for fuelwood tax refunds to local management institutions. The
permit system enables individual wood cutters to purchase permits locally, so that
costs are reduced. An important flow of revenues has been secured in villages by
local stakeholders who have few alternative sources of income and tend to be of the
poorer groups.

Management costs in the Kelka forest appear to be lower than in the other cases.
This may be explained by the effective functioning of local institutions which existed
prior to the rural firewood market system, and have been strengthened with project
assistance over a long period of time. Local management in the Kelka is much more
comprehensive than in the other two cases since local institutions assume
responsibilities for guarding, enforcement, fining and intervillage cooperation vis-a-
vis state agencies. None of these responsibilities are reported for cooperatives in
Energie-Il and Kita. Monitoring of rural firewood markets in Niger shows that illegal
exploitation by outsiders is seen as a serious problem by local people, who consider
that the forest service is the only competent institution to deal with this

An interesting aspect of the Kelka and Energie-ll cases is that local forest
management has been functioning for at least five years and various economic


indicators have been measured. The Kita experience is somewhat less relevant for
the purpose of comparison, since it is situated in the Sudanian and Sudano-Sahelian
zone, it has only been functioning since 1996, and forests are not locally owned.
Table 1 makes a summarised comparison of the three projects.

The rural firewood market system has constraints which are not apparent from the
table. Firstly, the tax differences created between firewood from a rural firewood
market and uncontrolled exploitation elsewhere is not sufficient to compensate for
the lower efficiency of the rural markets. In the case of Niger in 1993, the last year
for which tax collection data are available, only 30% of the taxes due on firewood
were collected, a figure which has probably dropped further since 1993.

In Mali, the tax collection rate was only 15-20% in 1997 (Konandji,H., pers.comm.,
1998). This implies that the average taxes paid for uncontrolled exploitation may
have dropped to a level which fiscally punishes rather than subsidises rural firewood

Secondly, the returns from firewood exploitation remain low compared to
opportunities for market oriented agriculture wherever such opportunities present
themselves. In the case of Kita, important forest areas are being cleared for the
lucrative peanut market despite the rural firewood markets. In Bankass, a study on
forest clearing for agriculture demonstrates the economics of such conversion
(Trench,P. et al., 1997). About 6,100 ha of forest, the size of a few village forests,
has been cleared over the period 1992 to 1996 in the Sourou valley mainly for the
cultivation of rice. The 1996 rice harvest was estimated at 5,276 tonnes with a
market value of FCFA 615 million (USD 1.23 million), which is three times more than
the gross annual revenue of the three firewood market case studies taken together,
at no donor or government cost.

TABLE 1. Three different local forest management systems. n.a.= not available

Average annual rainfall 400-700mm 360mm 600-1 ,100mm
Population concerned na. 10,000 inhabitants 30,000 inhabitants

Number of associations or
firewood markets

60 (1995) (1)

15 (1992-96) (2)

48 (1996)

Forest under management

125,000 ha (3)

106,000 ha

various gazetted forests,

area n.a.

Wood marketed per year

75,413 stere (1995)

24,617 stere (avge)

10,710 stere (1996)

Gross income before taxes | FCFA 129 million FCFA 52.5 million FCFA 17 million
& local charges
Net annual income of wood | FCFA 84 million FCFA 47 million FCFA 12 million


Taxes, charges, levies, etc.
paid as % of gross revenue

34% excl. hired labour

11% incl.hired labour (4)

29% incl. repayment donkey
carts (5)

Part of taxes refunded to
village development

66% plus some funds for
village woodlands


none but changes being
prepared to this effect

in near future

Management cost (6) 6% apparantly negligable 1.6%

Revenue per wood cutter, | n.a. FCFA 43,969 FCFA 13,918 plus donkey
per year carts for some (7)
Annual revenue from local | n.a. FCFA 8.6 million FCFA 3.5 million

firewood commerce

Cost of project/year USD 2.5 million na. USD 0.6 million

1. Limited to the Niamey basin, which represents 90% of the total.

2. In the case of the Kelka, this is identical to the number of villages, but not necessarily so in the other projects.

3. Figure derived from available reports.

4. The available figures do not allow to establish distinguish between taxes, levies, etc. and hired labour cost. Furthermore, village specific rates
exist for levies (included in the data).

5. For those wood cutters who obtained a donkey cart on credit and thus have to repay. Details unknown. Hired labour is probably not included in
the figure.

6. This cost is included under line taxes, levies, etc. It is a revenue for the local market manager.

7. For those who entered the credit scheme. Repayment details not known.

Rural firewood markets are a practical option if as in the case of Kelka a significant
forest resource is available at a strategic geographic location with a good transport
system. Clearly, the majority Sahelian forests do not meet these criteria. Even for the
Niamey firewood basin the ambitious Energie-ll plan was limited to 250,000ha of
woodlands over a 10 year period, which is still only 18% of all woodlands in the

The quota system may not be considered functional in silvicultural terms. In Kankani
village, the model rural firewood market quoted in the World Bank technical report,
the quota is unrelated to local forest resource availability and productivity. It is simply
taken from the average for the given agro-ecological zone (Direction de I'Hydraulique
et d'Environnement). Monitoring in Tientiergou, Niger, and in the Kelka, Mali, also
indicates that a quota system may not guarantee sustainability (Giraud,S., 1998;
Madougou,D., 1998, NEF Mali, 1997b).

Finally, the cost of market establishment and maintenance may be too high if it relies
on support by external agencies such as the forest service. Establishment costs are
estimated by Energie-Il as varying from FCFA 1,270/ha to FCFA 8,440/ha depending
on the type and size of market (Foley,G. 1997). Maintenance of the system depends
on a complicated market and fiscal administration which may not be sustainable
without external funding (Foley,G.,pers.comm.). The forest service is essential in
guarding and enforcement, which further increases the cost. Many foresters continue
to insist on active forest restoration such as tree planting, thus adding another cost
element. In Mali, rural firewood market establishment is contracted to NGO's and
private operators, which is expected to reduce cost, but control and maintenance
remains with the forest service.

In the early 1990's it was observed that Guesselbodi provided a model, but that the
approach adopted, radical though it had been, had not gone far enough. The costs
were far too high, fuelwood produced could not compete on the market without a
substantial subsidy, and ecological sustainability was at risk (Christopherson,K.A.,et
al,1993). The lessons drawn in the early 1990's for Guesselbodi seem to be valid at
the end of the decade for the rural firewood market: the social cost is high, the price
of firewood sold in the markets may not be competitive, the delegation of control


from the state to local institutions has not gone far enough and ecological
sustainability may be at risk. In addition the rural firewood market system captures
only one part of woodland economic value and neglects the broad range of non-
firewood products and services.


In three recent studies (two carried out under this research project and one reviewed
for it), a range of economic woodland values have been assessed. The
methodological constraints in economic valuation are important because of
interseasonal and interannual variations, the relevance of indirect uses and non-
marketed uses, option and non-use values, and the sensitivities involved in forest
exploitation. In the Kelka, one of the three cases, a questionnaire survey of 500
households has been used in order to evaluate the forestry, livestock, agriculture
and artisanal sectors in 14 villages.

The two SOS Sahel case studies presented here have used complementary
methods over a period of two years:

- Thematic PRA which includes resource mapping, product ranking, wealth ranking,
group discussion and producer interviews.

- Household and market survey.

- Producer observation as a variant of participant observation, and monitoring of
production by producers themselves.

The PRA approach created a positive rapport between local people and the
researchers, and enabled villagers to fully understand research objectives. In data
collection terms, PRA has been important to obtain a broad overview of the different
stakeholders, products, markets, prices and priorities. It also allowed an appreciation
of non-market values in relation to marketed products and services. Wealth ranking
provided first insights in socio-economic differentiation between and within villages.
Producers were identified to participate in research at a later stage.

The household surveys were very limited in sample size and in scope. Limited
sample size allowed extension staff to take charge of data analysis and thus be able
to do such surveys on a regular basis. The scope of the surveys was limited to
issues which are neither controversial, (forms of tree cutting), nor sensitive, (wealth
in cereals and livestock). In some villages, surveys have been repeated by local staff
in the second year in order to capture some of the interannual variation. Market
surveys were held seasonally to assess price fluctuations, volume and number of
traders as well as their origin.

Forest utilisation was explored through monitoring of key producers and markets
over two seasons. Producers were interviewed, and for some of the products, the
production cycle was monitored. In other cases monitoring was undertaken by the
producers themselves.


Producer observation is a useful step beyond oral data collection. It not only verifies
information collected during discussions and interviews, but it also explores issues
which are not easily discussed or which have been omitted. The technique is only
possible if there is a positive, trusting relationship. It is a time consuming method so
that a local researcher, may have to be engaged.

Certain productive activities such as wild fruit collection are problematic to research
through recall or observation. Monitoring by local producers with assistance of a
literate villager may be the only option. It has, for instance, not been possible to
measure the amount of fruit collected and consumed before arrival in the village.

One-off surveys are usually characterised by a variety of frustrations and
weaknesses, and longitudinal monitoring is preferred. The need for longitudinal
monitoring means that local people such as extension workers and literate villagers
have to be in charge.

Net revenues and net returns to labour have been calculated for the main revenue
earning activities and total economic value is discussed. The household surveys
have provided insights into the relative importance of the different sectors of the
village economy. The concept of stumpage value may be defined as the residual
value after subtraction of all exploitation cost and the reasonable profit of the
entrepreneur from the sales price (Bertrand,A., 1993; Davies,J. and Richards,M.,
1998). This has been used by some researchers, but may not be useful in the
context of Sahelian forestry. Stumpage value assumes that the standing tree is a
part of the market economy, which is rarely the case in the Sahel, where it should be
seen as a part of the subsistence economy.


The Kelka forest in the district of Douentza is customarily owned by 14 villages,
supported by NEF Mali. In the Kelka, forest exploitation is the major revenue earner
(48% of total), followed by livestock (33%), agriculture (10%), migration (6%) and
others (3%), (figures based on a 5 year average 1992-96). In the forestry sector,
firewood production accounts for 81% of this revenue, firewood trade for 15%,
baobab leaves 3%, construction wood 1%. The average revenue of FCFA 58 million
from forestry per year (1992-96) is very important, but it is rapidly decreasing at an
annual rate of 19% over the 5 year period (NEF Mali, 1997).

However, looking at revenues alone greatly underestimates the size of the
agricultural sector. The large majority of agricultural products are used for
subsistence purposes, and do not appear in the revenue assessment. Using the
local market price for cereals, the average market value of cereals alone is FCFA
260 million per year, or 4.5 times that of the forestry sector. This puts the relative
economic importance of the agriculture and forestry sectors in a different context.
The size of the livestock sector is probably underestimated in that nomadic


pastoralists use the Kelka forest, but are not incorporated in the survey, except for
the Peuhl livestock "resident" in some of the Kelka villages.

Forestry products are also partly used for subsistence. However, only 4.2% of
firewood produced is for subsistence, the remainder is for sales. Subsistence
production plays a more important role for other forest products, but such products
only make a small contribution to the forestry sector of the Kelka.

The interaction between the agriculture and livestock sectors follows the pattern of
reduced sales from livestock at times of good agricultural production, and vice versa.
1994 was a year of outstanding cereal production, almost double that of other years.
In the same year, cattle sales dropped by 48%, and small livestock sales by 59% of
the average sales over the four years which had relatively poor agricultural
production. Forestry revenues do not appear to be visibly influenced by the
agricultural production for a given year. The steady decrease of firewood revenues is
explained by exhaustion of dead wood stocks which originate from previous

The study does not mention non-market and non-use values. Some of the indirect
uses such as grazing and browsing have been mentioned, but it is not clear to what
extent livestock values depend on the Kelka forest. It is not clear how reliable
livestock ownership data is, nor how reliably respondents have been able to recall
production and sales information over the five year period. The methodological
description of the study is minimal and it is not clear whether any data corroboration
has been carried out. Nevertheless, it is clear that the firewood subsector is
economically very important though rapidly decreasing. In 1998, the Kelka supplied
41% of the firewood consumption of the regional town of Mopti, and the figure is
probably much higher for the period 1992-1996 (CCL,1998).


In the district of Bankass, not far from the Kelka, woodlands on the Bandiagara
escarpment are mostly owned by Dogon villages which have been founded in the
Seno plains in a process of linear migration (Gallais,J.1975). SOS Sahel UK through
the Bankass Environment Project (BEP) has provided assistance to the villages
since 1993 and economic monitoring of forest exploitation has been done in five
villages over the period 1997-99.

A thematic PRA done in 1997/98 provided the general overview of the key
stakeholders present in the area:

- the agricultural village of Tyi, which customarily owns about 1,900ha of the scarp
forest. Villagers do not depend on it for revenue, except for fruit collection by the
women, since they have important alternative income from dry season gardening.

- The agricultural village of Kobo, which owns a neighbouring part of the forest and
heavily depends on the scarp forest for dry season revenue.

- The Peuhl herder settlements of Wini and Jile, wno depend on the scarp forest
resources for wet season livestock production.


- Artisans in the villages of Kobo of Dounde, who need the scarp forest for supplies
of raw material.

The PRA also showed that various environmental and spiritual values were equally
important as some of the direct uses such as hay and firewood. Dry season
gardening, for instance, is perceived as highly dependent on the steady streamflow
from small canyon forests which are well protected against exploitation. In both Kobo
and Tyi, spiritual sites are used by people from distant locations as far as Bamako
and lvory Coast, who seek benediction.

The study did not attempt to put money values on any of these non-marketed
categories. Spiritual values are linked to the fetish which is clouded in secrecy and
thus hard to investigate. The Travel Cost method may appear useful for visitors to
sacred sites, but the kind of monitoring system needed for this method cannot be
reliably established in this particular cultural context. Market production has been
monitored for firewood, various kinds of construction wood, furniture and other
artisanal products and for wild fruits. Livestock production has been studied but the
outcome is of limited use due to wealth related sensitivities and other complications.

Agriculture is the most important economic activity in subsistence terms for the
farming communities of Tyi and Kobo. The household survey data on cereal
production is unlikely to be reliable since the head of household, who is in charge of
the main millet granary, is rarely willing to provide accurate information, particularly
when there is a surplus. A conservative assumption is that the Seno plains are
generally self sufficient in cereals. Based on standard cereal consumption rates and
market prices of cereals, it is evident that despite modest revenues agricultural
production is the main economic activity.

In Tyi village, dry season gardening is an important source of revenue for most
households, with an estimated average net income of 59,000FCFA/producer for the
1999 season. In the other villages, dry season gardening is negligable since water is
not easily available. In Kobo, forest exploitation and processing are important for
62% of households, with more or less similar proportions for the different wealth
classes. The estimated net annual revenues for the most important revenue earning
forest exploitation and processing activities are for the 1999 season:

- construction wood for producers with carts 56,000FCFA

- construction wood for producers without carts 14,000FCFA

- bed and stool makers 85,000FCFA (see annex 1 for details)

Except for some of the bed makers and a few artisans, forest exploitation is limited to
3-5 months of the year during the dry season after the harvest. Many of the
households who are engaged in forest exploitation claim this activity allows them to
pay their taxes and purchase cereals to make up for the deficit. Wealth ranking in
Kobo shows that forest exploitation is quite evenly distributed over the socio-
economic categories. About half of the households in the category "poor" are not
engaged in forest exploitation nor do they have significant alternative sources of
income. Such households are characterised by lack of labour, i.e. households of 3-4


members (incl. children), as opposed to 7-8 members for families engaged in forest
exploitation. It appears that the very poor are not able to exploit the forest resource
for income generation, but the poor with some available labour depend highly on
forest exploitation in order to avoid extreme poverty. Extreme poverty is locally
defined as not being able to pay taxes and having to beg food from the neighbours.

The poor category in Kobo also has an above average share of construction wood
producers without carts. Interviews and producer observation enabled estimation of
net returns to labour for the three major wood cutting activities for the 1999 season:

- construction wood production without carts: 77FCFA/hr

- construction wood production with carts: 288FCFA/hr

- bed production: 167FCFA/hr

The first category has the highest frequency but also the poorest return to labour,
and half of the producers included in the sample were considering pulling out. All
sampled producers without carts are in the category "poor" of the wealth ranking.

Commercial firewood production is rarely practised. All the construction wood
producers in the sample stated that the benefits are too low. This is corroborated by
the market survey which shows that the volume of firewood sold in the weekly
markets is minimal and the number of firewood sellers is much smaller than e.g.
sellers of wild foods. An interview combined with producer observation with a woman
who was a professional producer suggests that net returns to labour are 40FCFA/hr,
a rate which is satisfying to her but which would apparently drive the men out of

In the PRA fruit production was ranked by the women in agricultural communities
among the most important uses of the forest. In Tyi village, the household survey
suggests that richer households are more effective producers than poor households,
given that they produce 5-10 times more than their poor neighbours. Monitoring of all
fruit brought home daily by four households for the 1998 wet season, corroborated
with market surveys, allowed an estimation of the market value of the fruit. Women
brought home on average 3,700FCFA/hnousehold(see annex 2). Although the
amount is modest compared to the income from gardening or from construction
wood, it is high if compared to income from firewood, and to other sources of
women’s income.

Livestock has a buffer function, it is sold in years of poor agricultural productivity in
order to purchase food. In years of good crop production, crops are sold and
livestock is purchased. For the two Peuhl communities in the area, livestock rearing
is the principal economic activity, but household and producer surveys have not
been held in these communities and another tool has been used instead. Much of
the Dogon cattle is managed by the Peuhl “attached” to the Dogon community
concerned. Seasonal movement includes dry season grazing in the Niger delta.

In Tyi, livestock production has been evaluated using veterinary service statistics,
which are more reliable than the administrative statistics which are the basis of
taxation. Livestock is valued using the local market prices and an average sales rate,


which is a simplification given that livestock may have other values, at least for the
Peuhl. Milk sale, for example, has not been included in the evaluation. The volume of
livestock sales is based on a standard rate which the Livestock Department in Mali
has established at 10% of the herd size. Another rate is applied to herd size increase
in most years, with losses occurring during drought years. When applying the local
market rates, the 1997 livestock production value in Tyi, which does not distinguish
between Peuhl and Dogon livestock, is thus estimated at 4 million FCFA. Livestock
depends on the scarp forest resource for about 3 months of the year so that the
contribution of the scarp forest to livestock production in this village may be
estimated at a quarter of this value (3 months out of 12), i.e. 1 million FCFA. This is
an estimate of the sales and not of the residual value, since the cost of production
have not been subtracted.

The herders do not pay for access to the scarp forest as a grass and browse
resource. They do pay for access to dry season browse in the form of F.albida
branches on cropland, which is negotiated between herders and the local forest
management institution rather than with individual farmers.

When combining the various sources of information, forest resource exploitation for
income generation may be ranked approximately as follows for a local forest
management unit of 10 villages (Alamodiou) in this area:


. Construction wood

Fruit and edible flowers and leaves


. Other artisanal products (mortars, tool handles, etc)




Ranking based on revenues does not take the subsistence economy into
consideration. Firewood is accorded a much higher priority in PRA ranking than in
the analysis of revenues. An important part of the fruit is not sold but consumed, a
part of the construction wood is collected instead of bought, and local medicine is
collected but not sold at the weekly market. It has not been possible to estimate the
volume of the major subsistence products and if it had been, valuation would have
posed another constraint. Neither has any attempt been made to value the important
environmental services and spiritual uses of the forest.

The limitations on the study preclude an overall resource valuation in market terms.
The household survey simplified socio-economic relations which are much more
complicated the concept of household. All the same, some conclusions can be
drawn from socio-economic monitoring over a 2-year period in these villages:

1. The economic benefits of the forest resource are spread over a wide range of

sectors such as livestock, food, construction wood, artisanal production and
domestic energy. No single economic activity dominates at the scale of the local


forest management institution, although specialisation certainly occurs at the level of
individual households and villages.

2. A high degree of specialisation in forest resource use may exist between
communities, characterised by extensive trade relations between them. It is probably
in the economic and social interest of individual villages not to monopolise forest
resource use on their customary territory, but instead to support an “economic
community” and leave resource regulation to the supra-village level management
institution (Alamodiou).

3. Socio-economic differences within the village have an impact on forest resource
exploitation. The better off families with more available labour and capital are more
effective and efficient at exploiting the forest resource. The forest is particularly
important for households which have few alternative income earning opportunities.
The poorest households hardly capture the benefits of forest exploitation, due to lack
of available labour.


Much of the El Ain area in North Kordofan is situated in the northern Sahel, with an
average annual rainfall of less than 300mm. Woodland management in the area has
been supported by SOS Sahel (UK) since 1989. Due to the arid conditions, both
agriculture and livestock are marginal income earning activities. This area has a very
poor economic status. Food-for-work is a common intervention and migration is
widely practised in which entire villages may leave during the dry season. In 1998
and 1999 small household surveys were held in three villages (n=56) complemented
by a PRA and market and producer surveys.

The results show that the woodlands provide a very modest contribution to overall
household revenues, 11% in the most positive case. In all three villages, households
classified as "rich" (average 12% of village households) spent more on woodland
products than they earned from them. Rich households buy woodland products from
the poor, who constitute the majority of the village population. The poorest people,
due to lack of labour, do not have the capacity to exploit the woodlands to a
significant extent. The figures confirm those of another survey held in North
Kordofan, which suggest that natural resources make up for 8% of household
revenues for the year of research. All of these surveys were held in years of fair to
good rainfall, and it is most likely that in years of poor rains and crop failure, the
share of the natural resources sector in household revenues rises steeply. The
natural resources revenues exclude livestock production, which is seen as a
separate economic activity, although it is clearly dependant on natural resources.
They also exclude most of the artisanal activities, which depend partly on natural

The variation between individual households is high and therefore the average does
not reflect individual interests. A small number of households depend almost entirely


on income from the sales of construction wood, charcoal, fruit, etc. although
subsistence agricultural activities normally interrupt these professional activities. A
number of artisans who produce rope and furniture also depend almost entirely on
these activities for income generation.

During the reasonable good rains of 1997-8 rich households had a surplus
agricultural production, poor households had a deficit, which is the normal situation
for this socio-economic group. The rich households have not sold much of their
livestock, except for those with unusual cash constraints. In years of poor crop
production, livestock and natural resources are the most important economic
activities for those who do not migrate during the dry season.

The volume of marketed woodland products is distributed as follows for the 1997/98

1. Construction wood 43%
2. Charcoal 38%
3. Fruit 11%
4. Firewood 2%
5. Other (incl. gum) 7%

In the subsistence economy different priorities exist which have been studied using
PRA tools. Firewood, perfume, grass, furniture, construction wood and tools are
amongst the most important woodland related uses. They are hard to quantify but in
the case of construction wood in one village, estimates based on the number of
buildings and their useful lifetime, suggests that subsistence production is in the
order of 3-6 times more important than the marketed production. The value of
subsistence plus marketed construction wood can be estimated at USD 37/
household/year. In this estimate, subsistence construction wood has been valued at
the same rate as marketed construction wood.

The cost of production is negligible in most of these categories. However, a form of
taxation is imposed by the forest service in the order of 5-10% of the value of wood
products offered at the weekly market. The official forest product taxation system is
not functional in rural areas. As transport costs incurred to obtain the permit at a
distant government office is usually higher than the anticipated gross revenue,
people take the risk of woodland exploitation without a permit. The forest service in
need of financial resources raids the markets from time to time to collect taxes.

A comparison of forest product utilisation rates and standing stock has been made
for some villages. While productivity studies have not been done, an estimated
sustainable productivity rate of 7.5% of standing stock suggests that most village
woodland reserves cannot meet demand. Either the non registered woodlands will
continue to play a major role in provision of wood products, or villagers will have to
purchase products from elsewhere, and/or change consumption patterns. These
different scenarios have been observed among the village forests in the project area.
In villages with extensive woodlands outside the registered village forest, they have a


high level of protection. In villages with modest reserves outside the registered
forest, people fear that these reserves will soon run out. In villages without any
significant natural resources outside their registered forest, traditional consumption
patterns have already changed; for example they now construct banco houses
(made from materials purchased from distant places) instead of houses made from
local materials.

The Kelka, Bankass and North Kordofan case studies demonstrate the variations
found in the Sahel. The Kelka is strongly influenced by the urban market
opportunities and the rural firewood market, which although very different from the
Nigerien concept, is a relevant system. This is not so in the other two cases, where
firewood is of minor importance in the market economy. These two cases also
demonstrate the variability which can be found between villages and between
households in terms of household revenues derived from the forest. In Bankass, an
economic community of a large group of villages makes sense given the intervillage
specialisation, in Kordofan villages operate as autonomous units with intravillage

It has not been possible to measure the effect of rainfall and agricultural productivity
on the forestry sector, since 1997 and 1998 were both years of good rainfall. It is
assumed that in years of poor rainfall, the relative importance of the forestry sector is
high in comparison to other sectors.


Valuation of the livestock production function of the forest is difficult in the case of
animal rearing by the settled population and still more difficult in the case of nomadic
and transhumant pastoralists. The main conceptual problem is how to measure the
residual value of the forest resource as a factor of livestock production. Practically
this is constrained by getting access to mobile pasotralists and local sensitivity to
sharing information on livestock which is related to taxation and wealth attributes of

There are examples of commoditisation of the grazing resource, especially in the
Niger delta, where range ownership and livestock production are very distinct
(Diakite,C.N.1993). Herders pay for access as a function of the number and type of
animals and number of days. But the residual value concept seems to be of little use
to the three case studies and probably to most situations in the Sahel, where access
to range is more or less free, although dry season water is often paid for and
numerous other obstacles exist which raise the cost of livestock production.

In the Kelka and Kordofan case studies, livestock estimates value locally owned
livestock and not the forest resource as a livestock production factor. In the case of
Bankass, livestock is kept in the forest during the rainy season and is fattening
during that time. It is possible, to provide a value of the forest as a factor of livestock
production equivalent to the amount of time livestock spend in that particular forest.


However, this does not attempt to value the forest as a factor in livestock production
under different forest management regimes, e.g. state control until 1993, and
management by local institutions since 1993.

In the case of Baban Raffi forest near Maradi, Niger, an estimation of livestock value,
based on a 5% annual offtake, is FCFA 85 million. This is about 3 times as much as
revenues from the 22 rural firewood markets of this forest at the time(Paris, P.,1994).
No attempt is made to estimate the impact of different management regimes in
Baban Raffi, but it is observed that clearing for agriculture and wood cutting are the
two major threats to sustainable livestock production. A study in the Tientiergou
forest in Niger, which has a highly profitable rural firewood market, found that even
the firewood producers see livestock production as their first economic concern

The only available assessment of economic impact under different forest
management regimes is from the Day forest in Djibouti. In the project area and in
control areas where the project doesn’t work, the productivity of different landuse
zones inside and outside the forest has been assessed. The effect on nomadic
livestock production has been estimated through various economic techniques
(Nessim, J.A.; Richards,M.1994). Major methodological constraints appear to be:

1. Project interventions include deferred grazing, establishment of fodder plots, etc.
Assumptions are made about increased forage availability over a 30 year period, but
this may be unrealistic in the light of the problematic experience of such interventions
with regards to institutional sustainability. The study itself notes that generating
timely physical estimates of environmental change is a major analytical weakness.

2. Since market values of forage do not exist the assumption is made that the price
of sorghum can be used to construct a market with a price based on sorghum
importation and transport cost, with a 30% increase to correct for lack of digestible
protein in sorghum. It is argued that pastoralists would have no other alternative but
to buy imported sorghum in the case of continued environmental degradation. Which
appears highly artificial as an assumption.

Even if it were possible to value livestock production under different forest
management regimes, additional complications arise. Forests have a nomadic
livestock emergency function during serious droughts and they can be a strategic
resource when trekking between dry season and wet season grazing areas. The
amount of fodder of such forest may be modest in terms of average annual
requirements, but may have a key function in the sustainability of the overall pastoral
livestock system.

An increasingly important aspect of livestock economics is the cost of fines and
damages paid to farmers and to the administration, in the case of livestock intrusion
into agricultural fields. Various authors assert that the fines imposed on herders have
become so high that farmer's income from fines may be more important than the
value of the crops, particularly if situated on livestock corridors. Other forest uses


such as wood cutting, pruning and pollarding may be surcharged through
confiscation, taxes and fines. Unlike agriculture, livestock production and modest
levels of tree cutting are sustainable forestry practices. It is ironic therefore that
relatively sustainable forest exploitation is taxed whereas clearing for agriculture, is
rewarded by fiscal policy and practice and as a result of land use and tenure
regimes. The cost imposed on sustainable forest utilisation may stimulate woodland
conversion into cropland, even though the combined value of forest and livestock
products and services may well be superior to that of agriculture, at least in the
northern Sahel.

The conclusion for livestock valuation of the Sahelian forest under different
management regimes is that there are no widely accepted methods available.
Livestock production as a function of the forest resource is often undervalued or
simply neglected in many projects. The few studies available on livestock values in
the forest environment suggest that livestock is important, and frequently more
important than any other single forest product.


None of the study projects or case studies have themselves systematically assessed
the cost of local forest management. Project decisions tend to be made by budget
considerations (is there a budget line 7) and by the need to achieve the project
outputs in the logical framework. Under the project approach cost implications of a
policy intervention to society, to local communities and to individuals are not often
analysed. Local forest management may have very different cost implications at
these levels, such as:

-individuals: fines and confiscation due to changed management
-communities: meetings, surveys, boundary demarcation, guarding, court case

-government/donor: forest inventories, extension services, policing.

Table 2. Cost of establishment for a 2,000ha village forest

Rural firewood market establishment, Niger 1.1 average 5,000FCFA/ha: 10million FCFA

Village forest registration, Sudan 2.1 7,5 million S£/village forest plus indirect cost (n.a.):

Fireline establishment, Sudan 3.1 1FFWday/ha: not practised in other cases: 1,500US$ for
a 5*4km shape forest

Rehabilitation: Asa rule, cost far too high for large scale implementation

- nursery

- transport

- planting

- fencing or guarding, and possibly replanting

Sources: 1.1 Foley,G. et al.1997. 2.1 Kerkhof,P.1997 and Kerkhof,P.1999 3.1 Kerkhof,P.1998, based on a village forest of 4km*1km(400ha) and
standard food-for-work rates and conversion to SE based on dry season market rates.

The various models of local forest management are so different that cost comparison
between the models is of limited use. Such comparisons nevertheless are
nevertheless indicative for the potential of scaling up. Available figures are presented


in tables 2, 3 and 4. Total project cost per forest or per community yields another
indication, but it does not separate out the non-forest management components
(mesures d'accompagnement), research and policy development, etc.

Table 3. Cost of maintenance

1. Rural firewood market:

1.taxation + refund system 1.Unknown. Taxation is poorly functioning in Niger and Mali
2.local market manager 2.6% of sales

3.forest service policing, 3.personell, operational cost

4.quota revision, resource monitoring 4.only with donor funding

2. Full local management:

1.communication, meetings 1.15,000FCFA/village/yr (24US$)

2.guarding and enforcement 2.25,000S£/village/yr or 10,000FCFA/village/yr (avge

3. Standard forest inventory 22 mio.FCFA/forest of 83,000ha estimated equivalent at 2.5

mio.FCFA/village forest of 4,500ha e.g. frequency once per
5 years. 358US$/village forest/year

4. Participatory monitoring systems 1.65,000FCFA/village forest of 1,000ha

1.Local product inventory 2.83, 000F CF A/village forest of 1,000ha

2.Panoramic photography in either case, e.g. once per 5 years

5. Policing by forest service Cost to local communities 100,000FCFA/village (161US$)

plus individual fines or 5-10% of market sales in Sudan

SOURCES: 1.mostly based on Foley et al,1997 2.1 estimate mainly based on written accounts of Assaga Anda
Alamodiou,Bankass. Cost are lower for small management units. 2.2 Assumes that cost are equal to average income from fines and/or local
taxes, figures based on written and oral accounts.

Table 4. Overall project cost per unit of local management





KITA, MALI 52,000 USD na.


These cost data do not clearly illustrate the human resources requirements.
Relatively complex data collection or data management systems such as GIS and
GPS, used in Energie-ll, require trained human resources. Two different cost models
may be construed:

1) highest possible cost model: directed firewood market with quota based on regular
standard forest inventories, combined with a high level of policing and monitoring by
the forest service, and with investments such as tree planting.


2) lowest possible cost model: local institution covering 10 villages making, applying
and enforcing rules without formal quota, management plans and taxation systems,
and without investments and codified monitoring.

The highest cost model is highly realistic in terms of present policies, institutional and
legal framework. Any project proposal based on this model would be highly
acceptable to the administration and to the forest service in particular. However, it
carries a prohibitive cost in financial and human resources terms. The lowest cost
model is financially highly feasible but can only be realised in the presence of a
development agency which has sufficient weight to create an artificial policy
environment in which local management can thrive, and is thus costly in an indirect

A cost benefit-analysis may be attempted on the basis of available data on the
revenue flow from woodlands under local management, combined with cost data
available from different management regimes. Socio-economic studies from two
different areas provide estimates of revenues from the forest resource (see annex 2):

Kelka, Mali: 4.900,000 FCFA/village/year
Bankass, Mali: 1.500,000 FCFA/village/year

The estimates do not include subsistence products, environmental services and non-
marketed values such as option values and bequest values which have surfaced in
the PRA. Two cost models may be formulated which are based on actual practice:

1. A high cost model based on the cost of forest management under the firewood
market model with quota system, monitoring and policing by the forest service, but
without forest rehabilitation measures. For a 2,000ha village forest the cost estimate
is 10 million FCFA. The estimates for maintenance vary greatly with a minimum of
about 400,000FCFA/yr to over 2 mio.FCFA/yr if donor cost (table 4) are

If this were applied in the Kelka, the cost would be 8%-40% of the total revenues, in
the case of Bankass cost would be equal to total revenues, and in El Ain the cost of
the management system would largely exceed revenues from the forest. The cost to
villagers’ time is not included but it is relatively modest since many of the
management functions are the responsibility of the forest service.

2. A low cost model has no establishment cost of any significance. It involves
meetings, communication, guarding and enforcement cost in the order of
25,000FCFA/village/year for a large management unit, and perhaps double that for a
small management unit. If local product inventory and panoramic photography were
to be added, 31,000FCFA/village/year should be included. The total management
cost range is 25,000FCFA - 81,000FCFA /village/year. For the three case studies
this is in the order 1-10% of the total revenues. Cost calculations are limited to
financial cost and do not include the villagers' time. This model is only possible on a
large scale in a policy environment which favours local governance.



1. The majority of economic studies in Sahelian local forest management deal with
specific sectors or subsectors, in particular firewood. There are few studies which
present a quantitative evaluation of the broad economic interests of local people in
the forests they manage. Studies which analyse the interests of the different socio-
economic strata are rare, and none of the available studies have valued the forest
economics of both farming and pastoral communities.

2. Forest exploitation is hard work and is generally avoided by those who have
alternative, sources of income. Exploitation of locally managed woodlands benefits
the poor more than other economic groups. The poorest group, however, may not
have the means to exploit the woodlands for the generation of significant revenues.
Forest exploitation is often considerably more efficient if tools and equipment are
available such as donkey carts, and if wood cutters are well organised. Those with
access to some capital and labour exploit the forest more efficiently.

3. Economic activities tend to be opportunistic, especially in the northern Sahel.
Opportunism is recognised as a characteristic of nomadic pastoralism, but it is
equally characteristic of farming communities, in their own way. Crop production is
highly variable and it is often inadequate for the majority of farming families. A range
of complementary economic opportunities are seized such as temporary or long term
migration, food for work, herding and forest exploitation or processing. This is not to
say that all commercial forest exploitation is opportunistic. A small group of
professionals have an established pattern of forest exploitation and processing.

Income from forest exploitation may fluctuate considerably as a function of crop
production and other economic opportunities. The implication for forest management
is that such management should respond to opportunism. The imposition of an
annual yield or quota based on silvicultural considerations does not make sense
since it does not respond to fluctuating economic needs. Firewood projects should
not be surprised if they find that local wood cutters do not meet the annual quota, or
surpass It.

If society requires gaurentees of forest sustainability, it makes more sense to
establish long term ecological indicators such as minimum environmental standards,
rather than short term silvicultural indicators.

4. In some cases, woodland exploitation offers important urban market opportunities.
In these cases, a reasonable level of management cost seems to be acceptable. In

the large majority of woodlands, however, revenues are very low and local
management cost should be kept at an absolute minimum to avoid driving many
poor people who use the forest into extreme poverty.

Many forest management packages are costly in financial and in human resource
terms. They are particularly costly in comparison to the benefits of forest


management under most conditions. Various cost minimisation opportunities can be

- Leave delineation of woodland boundaries to villagers and customary authorities
and/or representatives of local government.

- Abandon the "scientific quota system and leave quotas to local institutions. Also
leave forest inventories and other forms of monitoring to villagers. The administration
may want to use remote sensing information, GIS and other tools to monitor long
term changes, the cost of which should not be charged to local forest management.

- Drop any pressure on local management institutions to make investments in forest
restoration (amenagement). Whilst this is an old recommendation many projects still
make such investments or plan to do so.

- Encourage local institutions to undertake guarding and enforcement. Cases which
cannot be resolved locally will be dealt with by the administration and the justice

The quality of local forest management should not be measured by the principle of
sustainable maximum wood yield. Different communities make different choices
based on dynamic values and needs, economic or otherwise. Some villages
establish exclusive user rights over a part of their woodlands, whereas others share
their forest resources with neighbours. Some impose strict forest conservation
measures, others may prefer gradual conversion into an integrated landuse system
of forestry, livestock and agriculture, a mixed landuse system which often makes
most sense.


The methodological constraints in Sahelian forest valuation are numerous, and are
due in particular to the complexity of livelihood systems and cultural diversity.
Economic comparison is constrained by variable definition, and sometimes studies
do not explain how measurements have been made. In some studies, livestock,
forestry and artisanal production are seen as different, in others they are seen as
integrated since they concern the same natural resource. The Bankass and Tyi
studies suggest that the keys to sound methodology are:

1. Good relations between researchers (or project staff) and local stakeholders. For
example, the relationship between the Sudanese project and the forest service, has
constrained relations between researchers and local stakeholders, which
enthusiastic PRA and a lot of patience did not significantly change. In Bankass, the
clear institutional distinction between project and the forest service enabled the
necessary positive rapport for producer observation.

2. A regular presence instead of one-off surveys. This is only practical if local staff
are capable and motivated to monitor socio-economic change, and methods are
adapted to local capacity. For example, household surveys which can be analysed
with a pocket calculator instead of SPSS. The usual statistical analysis of variance


cannot be applied, but the results may be more reliable than those conventionally

3. Corroboration through a range of methods which may include PRA, wealth
ranking, household and market surveys, producer surveys and observation of the
production cycle. This confirms conclusions drawn by other researchers, who did a
detailed comparison of PRA and household survey in natural resource valuation
(Davies, J., et al., 1999).



In sociology a distinction is made between an organisation and an institution. An
organisation is a distinct body with a legal identity such as a line Ministry, set up to
achieve certain objectives, while an institution is more than an organisation: it
includes structured and persisting patterns of norms, behaviour and relationships,
put into practice by organisations. It includes formal rules underpinned by law as well
as informal rules which are usually supported by ideology (Bass,S. et al., 1998). This
report uses the term institution for both institutions using formal laws and those using
informal rules.

The national level institutions which bear an impact on local forest management are
defined by law, by policy as a principle of action, and by a mission which can be
expressed in norms, behaviour and relationships. Despite the ethnic, linguistic,
geographic and climatic diversity in the Sahel, dominant trends can be determined.
Institutions and law appear to be relatively durable phenomena in forestry as
opposed to policy, of which numerous versions are adopted and discarded.

In the 1990s numerous forestry policy initiatives were taken in order to integrate new
objectives, in particular to improve participation and integrated resource
management in the direction of joint forest management. The forest service has had
to cope with many new policies. This led to policy inflation contrasting with a reduced
capacity for implementation consequent on structural adjustment. Rather than
merely cutting back on resources Institutional reform of the forest service is
considered necessary, while learning from experience in other sectors.

A contrary argument holds that the objectives of local forest management are so
different from those of the forest service, that the two should not be joined. The
forest service has the capacity to protect and manage national forest reserves, which
should be considered as a _ structurally different task from natural resource
management by local communities. The implication is that capacity building in
institutions other than the forest service is required.


Nevertheless, the forest service is presently the major institution at national level. It
is backed up by a legal framework and it is responsible for much of the policy making
in the sector. However new multisectoral institutions emerged in the 1990s and are
assuming some responsibility, these include environmental agencies at national level
and elected government at local level.


While organisations in the public or in the private sector may not have the persistent
patterns of norms, behaviour and relationships characteristic of institutions,
sociological studies in various countries suggest that such characteristics do exist in
the forest service. Foresters have traditionally been trained to become the guardians
of the forest, with a specific jargon, set of codes, symbols and culture.

The very beginning of forestry may be traced back to medieval Europe, where kings
and other notables created the job of managing the forest estates (e.g.
Gerner,H.,1998). There was no forestry education but, as far as written records go,
these people became the first professional foresters of a kind. Their main interest
was protection of the forest against the commoners, an aim which they pursued by
vigorous means.

The more contemporary form of professional forestry is seen to originate from
Germany from where it spread to neighbouring countries in the 19th century, and
subsequently to Africa (Shepherd, G., et al.,1998). A comprehensive sociological
study of the German forester gives an insight of the strong socialising character of
the profession and its education system at the time (Heeg,B., 1972). While these
characteristics have undoubtedly been modified in the various countries, the forestry
services in the Sahel have maintained some of it.

In the francophone Sahel, the colonial government established the forest service as
a quasi-military institution, in the traditions of the French forest service, Office
National de Forets (ONF). In the ONF, foresters still carry firearms and are subject to
military hierarchy, and so do the Sahelian counterparts. The current organisation or
the Nigerien forest service is presented as an example in Table 5. In Sudan and
other countries with a British colonial history, the military character and espirit de
corps is less pronounced although foresters may also carry firearms.

Table 5. Military hierarchy in the Nigerien forest service (Madougou,D., 1998)

Category |Class_ |Corps military rank
1 Ingénieurs Commander Lieutenant- |Colonel or
A colonel General
2 Ingénieurs Captain Commander
1 Conseiller under-Lieutenant | Lieutenant


2 Contréleur Deputy-chef Aspirant Underlieu-
C 1 Agent Sargent Major Adjudant
D Préposé Sargent Sargent-chef
E 1 Garde Brigadeer Brigadeer-

Regulation of commercial forest exploitation was legally attributed to the forest
service. Although protection of trees and forests against local people, i.e. the policing
function, was the basic preoccupation. A major legal instrument was gazettement,
through which local rights were abrogated in order to establish state forest reserves.
Another instrument was the forest exploitation permit system, which was applied
outside the gazetted forest. One peculiarity of the forest policing system is the
distribution of a part of the fines to the foresters, a characteristic which dates from
the medieval European system. The system of distribution in Niger is presented in
table 6 as an example.

The dominant political ideology in the newly independent states of the Sahel was
marxist or socialist. The land and much of the forest resource was essentially
appropriated by the state and by the forest service as the competent authority in the
forest estate. The authority of the forest service was thus increased to include all
forest exploitation, be it for commercial purposes or for subsistence, though some
access to forest byproducts remained in place for local users.

Another change in the forest service was brought about by the major droughts and
subsequent flow of development assistance. This introduced policy innovations such
as support to tree planting and management outside the forests, renewable energy,
desertification control and popular participation. It lead to some reform in the forest
service and to a change in attitude, at least among the more progressive foresters. It
also extended the authority of the forest service into the kitchen, where Malian
foresters, for instance, used to fine villagers for not having improved stoves.
Furthermore, additional financial resources helped to execute the policing functions
more effectively. Deforestation and Firewood gap hypotheses prevailed in the 1970s
and 1980s (Foley,G., 1999). This put a further emphasis on certain forest values and
subsectors at the expense of others.

Table 6. Redistribution of fines collected by foresters, in the case of Niger
(Madougou,D., 1999).

Beneficiary Percentage of fine|Repartition of sum paid
paid out out

Agent and possibly assistant 10% 40%

Forestry staff in local district 4% 16%

National directors 03% 12%

Central directors and deputies 02% 08%

Chiefs of central services and their | 02% 08%



Other forestry agents in| 03% 12%

Forestry staff attached to Minister | 01% 04%
(Conseillers, Secrétaire Général,
....) and elsewhere (DEP, DAAF)

Total 25% 100%

The Forest Codes tended to define broad legal terms while the elaboration of their
application through rules and regulations (Voie Reglementaire) was to a large extent
left to professional foresters. This led to excessive control. Clauses on collective guilt
remained in many forestry codes, making accessories to forestry crimes equally
culpable with the perpetrator. This clause has been widely abused leading to annual
fines for whole villages, a practice which was still reported in 1999. Recourse in
disputes between local people and Forestry Agents was in many cases only to the
Forestry Director (Ribot,J., 1995).

In the 1990's the trend of a widening domain and increased authority of the forest
service has been reversed, in part aS a consequence of macro economic policies
such as structural adjustment and through decentralisation and increased
democracy. Also, as the impact of the forest service on rural people has been
recognised as excessively negative in some countries their authorities have been
curtailed by government.

Considerable differences have emerged between the Sahelian countries. In Sudan,
the forest service (FNC) is authorised to regulate all forest exploitation, apart from by
products, through a permit system including production on village and private
woodlands. “Forests” includes all trees and shrubs. The 1989 Forest Act supports
the establishment of village woodlands with land titles accorded to the community,
however exploitation is controlled by the FNC through a management system, and
through a permit and tax system for commercial exploitation.

The FNC is self-financing at the State level. North Kordofan State, for instance, has
few productive forests and much of the income is derived from taxing and fining
villagers and traders. The tax on forest products is variable but villagers often avoid
obtaining permits as they find the transaction cost higher than anticipated revenues
from forest exploitation so that they avoid obtaining the permit. In order to maintain
the flow of revenues, the forest service regularly confiscates forest products at
markets and along roads.

Sudanese forestry policy is to turn 25% of the national territory into a forest reserve,
up from 15% in the 1980s. Presently only a minor fraction is gazetted. Given that
almost all forests in the gazettement process are state forests rather than community
forests, the policy effectively implies expropriation of most traditional community
forest resources, used by farming communities and pastoralists. About 96% of the
professional cadre are foresters and only 1-2% have social and environmental
science qualifications (Adil M.A., 1999). It is obvious that the FNC is driven by the
ideology of forest control and protection as well as financial extraction introduced by


the colonial forest service. But under the present dynamics of decentralisation from
the centre to the States this approach is under great pressure (Tarig,M., 1999).

In Mali, the forest service was amongst the most repressive in the region until the
1991 revolution, but has been reformed since. In 1995 new laws on forest
exploitation were adopted which have reduced the authority of the forest service:

- trees in land lying fallow for up to ten years remain the property of the owner and
can be managed without permit (except for protected species). This used to be 5

- forests under the authority of the Rural Communes are subjected to a management
plan by the competent agency which is not necessarily the (national) forest service.
The text is ambiguous at this point, but it may allow a Commune to establish its own
technical service.

- A rural woodfuel market system along the lines of the Niger model will provide
improved opportunities for local forest management in gazetted and non-gazetted

Furthermore, the forest service is no longer a sectoral Ministry but has been
reorganised with other technical services into a Rural Development Ministry. The
forest control functions of this Ministry were integrated with matters such as seed
quality control, animal diseases, etc. The personnel of the various sectoral Ministries
was amalgamated and reorganised, which probably had a considerable impact on
the mission and ideology which underlies the forest service. But in 1998, the Malian
government reversed this institutional innovation and the forest service was re-
established as the Department of Nature Conservation, responsible for policing in all
forest areas (Bacha,A., 1999).

In Niger, the forest service has experimented since the 1980's with forms of local
forest management through donor assisted programmes which led to the
development of the rural woodfuel market system, often seen as a breakthrough in
local forest management. Furthermore, the Nigerien model of natural resource
tenure development is one of continuous experimentation in which the forest service
has allowed considerable freedom to pilot projects.

Nevertheless, the forest service has retained tight control over forest legislation.
Recent attempts at legislative innovations in the light of the Code Rural, supported
by GTZ, have not been successful (Anon, 1995). The committee charged with
legislative review consisted entirely of forest service cadre, so the only non-forester
invited, a lawyer, did not turn up. The review concluded with a recommendation to
uphold a maximum of control functions. Important stumbling blocks to local forest
management, such as high fees on valuable forest products such as ronier
(Borassus aethiopium) are maintained. In the field forestry agents maintain an
extensive system of rent extraction and continue to instil fear in local communities,
see box 4.


Box 1. Some examples of rent extraction in Niger

1. The PRSAA agricultural training project incorporated pruning and pollarding in the
training and extension scheme. Farmers found that, once they applied their skills,
foresters were the next to visit them and fine them for pruning and pollarding.

2. The Energie-ll project experimented with selected farmers on how to best prune
tress. The project researchers coordinated their fieldwork with staff of the local forest
office. To the surprise of the researchers, the same foresters went around to the
farmers the day after the research visit, in order to fine them for pruning trees.
(Madougou, D., 1999)

In Burkina Faso and Senegal, the forest service has also experimented in
participatory approaches, but local groups are usually treated as syndicates
implementing a forestry service plan and are rarely given management
responsibilities (Kane,O. and Winter,M., 1997; Painter, T. and Sanou,S.,1997).

The current mission and ideology of the forest service and of individual foresters in
the Sahel is variable. Many of the younger foresters tend to be interested in training
and extension rather than policing, but in the hierarchic organisation they have little
power and little opportunity to put ideas into practice. Many of the more progressive
foresters tend to leave the service and work for NGO’s.

The forest service will undoubtedly maintain authority over gazetted forests, either
through traditional policing and management, or through joint forest management
schemes in which local user groups play a role. The role of the forest service in
integrated resource management by local groups on public land is likely to remain
limited. The structure and the mission of the forest service, as well as the sociology
of the traditional forester go against the grain of integrated, local governance of
natural resources. The Malian model of alternative technical services at the
Commune level, although untested as yet, should be an interesting experiment. Also,
alternative institutions of a multi-sectoral nature are being established, such as
environmental agencies at national and at regional level, and various forms of local


International policy is increasingly important in local affairs. In the 1990's a number
of international policies have impacted on the natural resources sector, in particular:

— the principles of good governance;

— the transformation from centrally planned economies to market economies,
accompanied by privatisation and structural adjustment;

— the globalisation of economies;
— decentralisation (Bass,S. et al., 1998).


Privatisation through land titling is one element of change. Some researchers argue
that Jeffersonian ideals underlie recent policy trends in land privatisation. The ideas
of a link between a regime of private property of a solid, roughly egalitarian class of
peasants and democracy is gaining momentum in West Africa (Elbow, K. et al.,
1996). Another idea is that privatisation is the only way out of the "tragedy of the
commons", and land titling may resolve endless resource conflicts.

While privatisation and market based strategies have gained strength in the
agricultural commodity sector, the existence of market failures in the natural
resources sector is characteristic, with land privatisation policies falling short in the
case of shared natural resources. On the contrary policies aimed at decentralisation
of government provide a significant impetus to local forest management.

The OECD/Club du Sahel characterise change in the West African Sahel essentially
as: (a) transformation of the Sahelian rural economy with urban markets and roads
as catalysts for economic change (b) decentralisation of government and
administration, supported by legal adjustment and (c) insecurity and conflict (OECD,
et al). Although decentralisation of government is a regional characteristic, its
content and speed varies greatly from country to country. Various francophone
Sahelian states have established the Commune (or Commonauté Rurale), and

Sudan the Rural Councils as key institutions for elected local government (see table

Senegal is the first country to experiment with elected local government. Since 1972
locally elected councils have owned their natural resources though under a high level
of central government control. In Senegal the 1993 Forest Code aimed at local
participation but leaves the elaboration of forest management plans to the forestry
service. Management plans are required for all forests, including those belonging to
local government and private individuals. The forest service monitors implementation
and may revoke management plans, even though there is no specific appeal
procedure. Forest product taxes collected are not under local control, although the
Minister may allocate 25% of such taxes to local government (Ribot,J.C., 1995).

In Niger, some Communes have been established since the legal context was
created in 1964 but only 21 of the required 200 communes programmed at the time
had actually been established by 1994. The communes consist of about 50 villages,
a size which restricts representativity of local government.

In Mali, decentralisation has been high on the agenda since 1991. Rural Councils
were elected in 1999 and they are now major interlocutors for rural development
assistance. These Rural Communes will have a considerable degree of autonomy in
local affairs, although precise roles and responsibilities have not been finalised. It is
a possibility that the Rural Commune may establish its own local forest service
independent of the national forest service.

Constraints to representativity concerns the size of the Commune. In Niger, about 50
villages make up a Commune, in Mali and Senegal this is 10-20 villages. The


appropriate unit of local forest management may evidently not be that of the
Commune. A further criticism of decentralisation concerns landuses which cross
Commune boundaries, such as nomadic pastoralism (Bacha,A.K. and
Tessougue,M.,1996; Mortimore, M., 1997). It is obvious that a choice must be made
between smaller units of local government, which are probably more representative
of the electorate, and larger units which are may respond better to mobile landusers.

Sudan has elected Rural Councils which to date have little power over forests and
natural resources. In the current process of decentralisation the power transfer from
the centre to the States is being negotiated. The forest service wants to uphold
central authority over natural resources, but the government is likely to transfer
authority over 60% of the resources to the States, which are expected to further
decentralise to the Rural Councils. The Rural Councils are large, comprising a
hundred or so villages, but unlike in West Africa the Sudanese resource tenure
legislation allows resource appropriation by individual villages (Government of
Sudan, 1989; Tarig,M., 1999).

Table 7 provides an overview of the hierarchy of government and administration in
the three countries. This does not reflect the wide variations within each country, nor
does it reflect the role of the judiciary, land commissions, Sultans, religious leaders,

Table 7: Hierarchy of central, local and traditional forms of government in three

National government National government National level
= political/administrative = political,administrative = political\administrative
State. authority: Governor Region. Gouverneur Departement.Prefet
Province: Commissioner Cercle. Commandant Arrondissement.Sous-Prefet
Rural Council, elected government | Arrondissement (Chef) Chef de Canton/Groupement
approx 100 vill. traditional authority
Nazir, traditional authority Commune Rurale (No elected local government in rural
elected government approx. 10 | areas)
Umaa, traditional authority Chef de poste: administration
Sheikh, traditional authority Chef de Village/Campement | Chef de Village/Campement
and Village Council (elected) (state/traditional) and Chef de Terre | traditional authority

Elected local government with important powers over natural resources is thus a
very recent phenomenon in the Sahel. The trend is to transfer power from the forest
service to local government. The resistance put up by the forest service is
understandable and can probably be overcome. There is another, more important
concern, that decentralisation of government without empowering village level
authority carries risks. It is feared that in order to fund running costs local
government will sell natural resources. This fear is expressed both in central


government and in the villages. It is evident that Malian villages well endowed with
forest resources fear that the Commune may sell the forests on village land. It is felt
that the Sudanese States and Rural Councils may do the same thing. Whereas
villages are likely to use their resources sustainably, since they will be faced with the
consequences of overexploitation. Elected officials in charge of ten to several
hundred villages may be less concerned with the long term consequences.


One of the criticisms of the forest services is the mono-disciplinarity of institutions
which are formally in charge of a multisectoral resource. By 1999 no significant
sociological, economic and environmental expertise is present in any Sahelian
service. But environmental and other multi-sectoral agencies have been established
throughout the Sahel which have assumed some of the responsibilities previously
taken by the forest service.

In West Africa, national environmental action plans have been prepared and in both
Sudan and west Africa, environmental committees and secretariats exist. In Niger,
the "Principes Directeurs d'une Politique pour un Développement Rural of 1992
provides the general framework for all interventions in rural development, and the
PNEDD (Plan National d'Environnement pour un Développement Durable) was
published in 1995. A secretariat (CNEDD, Committee National pour l'Environnement
et un Développement Durable) has organisational responsibilities and UNEP
coordinates amongst donors. In Burkina Faso environmental planning is organised
through PANE (Plan d’Action National d'Environnement).

In Mali the PNLCD (Programme National de Lutte Contre la Desertification) of 1987
provides the general principles while the Schema Directeur du Secteur Rural works
out sectoral plans under the principles of the PNLCD. The main environmental policy
instrument is the PNAE (Plan National d'Action Environnementale) which is
coordinated by a permanent secretariat and supervised by an_ interministerial
committee chaired by the Prime Minister. A similar set-up exists in Sudan, where the
Higher Council of the Environment and Natural Resources has been established
under responsibility of the Prime Minister with a permanent secretariat for day to day
matters in the Ministry of the Environment and Tourism.

While the various countries each have their own environmental policy and planning
system and follow their own rhythm, broad issues are identical. One conclusion is
that of policy inflation (Bass, S., 1998; Speirs,M. and Secher Marcussen,H., 1998).
An increasing number of policies promulgated, along with a reduced implementation
capacity of the institutions expected to guide and implement those policies.
Secondly, environmental plans tend to rely on a rhetoric which lacks realism. This is
in part due to the overwhelming impact of structural adjustment and other macro-
policies which have an adverse effect on environmental conservation, and partly due
to dependence on models of environmental conservation which are increasingly
seen as outdated (Ribot,J.C., 1995; Leach,, 1997). Finally, the planning


process is still highly centralised, although stated intentions are to open up the
process to all stakeholders.

Both the Malian and the Sudanese environmental agencies have been set up for
interministerial coordination under the Prime Minister. In both cases, the
interministerial meeting has only once taken place over the last 5 years and the
agencies have recently been located in the Ministry of the Environment, thus losing
the scope for intersectoral coordination. In the Malian administration, sectoral
Ministries were reorganised into one Ministry of Rural Development in 1995. This
was reversed in 1998, when most of the forestry tasks and most of the foresters
were relocated in a different Ministry, and thus regained their disciplinary nature
(Bacha,A., 1999). It has become apparent that the direction of institutional reform
with reference to local forest management in the Sahel is variable.

All the same, multisectoral agencies now exist and take charge of matters which
were previously the exclusive domain of the forest service. These agencies cover a
wide range of disciplines including foresters who do not use the military insignia and
similar symbolism. The agencies have a very modest staffing but receive a relatively
large amount of external support, whereas donor support to the forest service is now
very modest.



Four levels of law may be identified with respect to local forest management:
constitutional and international law, forest law, other laws pertinent to local forest
management, and local laws and bye-laws. The Sahelian countries have their
individual legal framework, but a number of issues apply throughout the region.

In the area of environment and rural development, international conventions have
been signed by many Sahelian states, such as the Biodiversity Convention and the
Convention against Desertification, which imply a certain concession of central state
sovereignty. Signing up to Agenda 21, for instance, engages African states to apply
principles such as "flexible planning which allows optimal participation of local people
and collectivities."

Similarly, CILSS and UNEP have mandates in the formulation of national and
regional planning. International conventions are general by their very nature, but are
superior to national laws in the legal hierarchy, just under the constitution (Sani,A.M.,
1996). Both international conventions and the constitution support local forest
management in a general sense, but there may be a wide gap between international
conventions and laws of application.

The Forest Acts in the Sahelian countries often comprise all forms of natural
vegetation whereas different rules generally apply to plantations and trees on
cropland. Acts pertaining to the transport of forest products, on the other hand,
generally include woodproducts which originate from cropland and plantations.


Grazing, picking of fruit and food and collection of deadwood in public forests are
usually allowed without a permit. The exploitation of green trees tends to be subject
to authorisation in various forms. As a rule, a permit-and-tax system applies to any
green wood production. In the rural firewood system, the production of green wood is
controlled through the annual quota (CCL, 1997; Foley,G. et al., 1997).

Green wood production in forests under the responsibility of the Malian Rural
Communes should follow a management plan which is subject to control by the
competent authority: probably, the forest service (Republique du Mali, 1995). The
content of a management plan is not described by the Act, but if the management
model of the rural firewood markets is taken as authoritative, it is evidently too
complicated and too expensive for application by local communities (CCL, 1997).

The Malian Rural Communes may have the authority to establish a technical service
other than the forest service. The first Commune elections have only been held in
1999, so that no experience exists to date. If implemented, the model may be a
radical departure from the conventional institutional and legal framework.

In Sudan, exploitation of the gazetted village forests is the responsibility of the
committee elected by the local villagers, but the forest service may “issue directives
or take measures for the protection” of village owned forests (Government of Sudan,
1989). No reference is made to a management plan in the Sudanese Forest Act, so
that a great deal of uncertainty exists. This leaves wide powers to the local forest
service agents.

Decentralisation of governance in Sudan from the centre to the States and Rural
Councils implies that the village forest registration procedure will be decentralised
and that competencies for land registration will be established in the individual States
(Tarig,M., 1999). It follows that monitoring and control will also be the responsibility
of the States rather than of a forest service under a national director, and the States
are expected to decentralise certain responsibilities to the elected Rural Councils. As
in the Malian case, experience will have to be gained with this model of local

Forest Acts are important but there are other relevant acts of a more general nature,
such as land acts and acts on local governance. Wherever such acts conflict with the
forest act they tend to be subordinated by the latter, since the forest act is usually
more specific. But the acts are incomplete in the organisation of natural resource
management. Acts pertaining to land and to natural resource ownership and usufruct
in Sudan may eventually, in the context of decentralisation, become more important
for natural resource management than the Forest Act 1989.

Similarly, acts pertaining to local government in Mali may turn out to be highly
relevant for local forest management as the Communes take on responsibilities. The
Malian Forest Acts 1995-003 and 004 were a precondition for GEF-Dutch funding of
the rural firewood market project and are considered premature in the process of
Commune establishment and local governance in general (Bacha,A., 1999). A more


integrated legal framework for local governance is likely to be adopted sooner or

Legal dispositions on the nature of the forest management committee may or may
not exist in the various forest acts, but such dispositions are generally quite weak.
How can local management committees be best shaped in the existing legislation?
Should they be considered as public or as private institutions ? If they are to be
considered as private institutions, the following options are available in the
francophone countries (Sani,A,M., 1996):

1. Associations, in particular those for Public Good (Niger:L’'ORD.84/6 du fer Mars
1984 with modifications, in Mali: 1959 legislation on Associations) have the
advantages of simplicity and efficiency. Takieta, Kelka and other forest management
committees are registered as associations. As long as pursuing commercial interests
is not a primary objective, local management committees may be conceived as
associations. However, this legal construct poses two major problems:

- Associations for Public Good are created with the objective of supporting education
and assistance; whereas local management committees do not aim at support to
others, but at self supporting activities.

- Associations are based on voluntary adhesion of adults, which is not the intention
of local management committees for natural resources, which are expected to
represent the community. Local committees are usually intended to represent a local
public interest.

2. Cooperatives, or similar. The more sectoral objective of local management
committees for some forms of local forest management such as woodfuel markets
make cooperatives apparently more suitable. However, various constraints exist.
Cooperatives are expected to pursue a specific commercial activity, whereas the
Sahelian forest covers a range of economic sectors. Cooperatives are also based on
voluntary adhesion of adults.

The Burkinabe village hunting associations which are entitled to engage in contracts
with tourist hunters have similar legal restrictions (Ouedraogo,H.,1999). They deal
with a sectoral resource which is controlled by the relevant state agency. Although
the hunting contracts ensure a degree of resource transfer to the local communities,
the arrangement is still far removed from integrated, local resource management.
What is required is much more creative law reform.

In other cases, such as Gestion de Terroir projects, local committees may be defined
by an ad hoc arrangement which has tacit or explicit approval from the
administration, but which is often not supported by the national legal framework.
Projects may develop arrangements through the local administration, such as an
Arrét Prefectoral, a convention with the local forest office, etc., and, it is hoped,
gradually develop legal recognition as a function of their proven efficiency. Almost
invariably such arrangements collapse when the project terminates (Sani,A.M.,


For both Associations and Cooperatives the question is how they can exercise
prerogatives which exceed Private Law. This is only possible if they pursue a public
service function. If an organisation is to be recognised as a public service three
criteria apply:

- With respect to its origin, it is created upon public instigation. In the case of a
multilateral project such as Energie-ll this criterion is met to some degree, but not in
the case of most projects.

- Its mission should be of public interest. The sustainable use of natural resources
could probably meet this criterion.

- Internal and external organisation reflect those of the government service.

This poses limits on the conceptualisation of local management committees as
public institutions. Under current legislation, such committees would be responsible
to government, at the expense of their independence. The experience shows that
local management institutions are effective and efficient if they are relatively
independent from government institutions such as the forest service and the

Another constraint is that the formal organisation of the local management
committee should reflect that of the government service. Where local forest
management institutions exist, they tend to have an internal organisation and an
ideology which widely differs from those found in government. Unless and until good
governance has become standard praxis, local forest management institutions
should not be copies of government, and the legal dichotomy should persist.

The lowest level of legislation is customary and local law. Local law is described as
the emerging, usually unwritten legislation adopted by local communities in the face
of change, whereas customary law is based on tradition (Le Roy, 1985). Since many
local forest management opportunities are recent, they are supported by local rather
than customary law. The state tends to respect local law to an important extent.

Le Roy (1990) suggests that in many African societies conflict resolution is not
predetermined or based on assessing who is right or wrong as is the case in
Western justice systems. Rather, conflict resolution is part of a longer term
management of social relations and often has as its objective to reconcile opposing
groups rather than to sanction or privilege one side over the other. Customary
institutions are more capable than modern institutions in this form of conflict

In the Kordofan case study, most forest management conflicts have been regulated
by local law. A minority of conflicts have been presented to the court, all of which
have been resolved out of court before the opening of a court case (Kerkhof,P. and
Damango,B., 1998). It has been suggested, in order to strengthen local law, to
formalise these rules and incorporate them in Rural Council bye-laws (Tarig,M.,
1999). Various Malian case studies suggest that almost all forest management


conflicts are solved under local law, some of them with involvement of the
administration, but rarely the court (Diallo, Y. and Winter,M., 1997; Kerkhof,P., 1998).
Conventions between the newly established Communes and the local forest
management institutions may formalise and strengthen local law.


Fiscal reform accompanies institutional and legal reform. The issue is one of
reversing resource transfer, which in the past has been from the rural areas to the
centre. Taxes, permits, penalties, and confiscation under the forest act, and extortion
under the pretext of the forest act, have made rural people poorer and have turned
many forest agents into wealthy urban inhabitants. Town based entrepreneurs have
exploited poor forest control to their benefit. In the case of Niger, for instance,
incomes of urban citizens are about 7 times those of rural citizens (Danida, 1999).

The rural firewood market system is intended to reverse this trend by providing sole
user rights to local communities and by differential firewood tax rates in favour of
local producers. But poor application of the tax system may jeopardise the resource

The trend is to use the firewood market system to impose very high taxes on non
firewood products, such as construction material. The forest service, as a national
institution, imposes standards and rates irrespective of regional characteristics. A
species may be declared protected nationwide whereas it makes up more than half
of the trees in a given community forest. A building pole may be taxed at a level
which seems reasonable for capital city standards, but which is exhorbitant for a
rural Sahelian who lives in a very modest hut.

The tax laws do not spell out whether locally sold forest products are subject to
taxes. In the Gaya project, local institutions and project staff managed to obtain an
agreement with the forest service to exclude locally sold palm products from the very
high taxes. But this is unlikely to happen outside a project environment.

If the resource flow from rural Sahelians to urban based traders, forest agents and
forest service is to be arrested, much more significant fiscal reform will be required.
Institutional, legal and fiscal reform should be seen as integrated and should be
much more innovative. It should be adapted to local conditions and favour local


Two qualities of governance by the state are relevant to local forest management.
One is the rule ordered congruence of the state, the second is the lawful behaviour
of state agents and agencies.

If the term state is used for legally empowered and legitimately coercive institutions,
questions are raised about a number of African states. At some stage countries like
Zaire, and Somalia may be considered stateless, but the qualification is not just


about "state" or "stateless" but about the degree to which states are legitimate
(Jessop,B., 1990). Rule ordered congruence between politico-legal institutions in the
state is a measure of state formation. For instance, an ever increasing number of
contradictory rulings from the central government, local government, judiciary and
others are indicative of poor state formation.

Even if congruent laws exist, they may not be applied. Many cases of unlawful
behaviour in forest management by state agents and state agencies have been
recorded. Such behaviour typically includes:

- refusal to pay reasonable compensation, as stipulated in the forest act, to acquire
local usufruct rights in the case of forest expropriation.

- intimidation of local leaders who attempt to defend local rights vis-a-vis commercial
companies or state agencies.

- extraction of moneys and animals for personal benefit of state agents in the name
of the forest act.

Sustainable forest management by local communities can only be secured by fair
governance at higher levels.


The forest service has historically been in charge of policy making and of project
support to community forestry. It has so far been the institution responsible for
supervision of local forest management projects. However, the mission of the forest
service as well as the social psychology of the forester are at odds with the type of
support required for effective local forest management. Institutional alternatives are
emerging in the Sahel, in particular local government at community level, and
environmental and other sectorally integrated agencies at national and regional level.

Institutional, legal and fiscal reform should be conceptualised as an integrated issue.
The process of reform is different in the various countries but generalisations can be
made. In the short term, projects may uphold local forest management but in the
long term, such management should be defined in the framework of public law.
Public responsibilities may be situated in the village, in the Rural Council/Commune
or in a combination of institutions (Abari,M.M. et al.,1997). Local law as a
complement to state law is likely to persist for a long time, because state law is
inappropriate or inefficient. Local law may be formalised in the form of local
government bye-law, but there is little experience to date.

In any case, unlawful behaviour by powerful actors such as commercial companies,
the forest service or corrupt civil servants needs to be curtailed by the practice of fair
governance. Responsible local institutions with legitimate local law cannot
sustainably manage their resources under conditions of poor governance at higher





A local forest management institution is defined here as any community based
organisation which is directly involved in forest management. Such institutions
are characterised by a pattern of norms, behaviour and relationships as opposed
to a government organisation, which is characterised by a legal identity and a set
of formal objectives.

Local institutions presently manage forest resources in their immediate
environment. Part Il of this report analyses how they are performing in the
various management tasks, what constraints they are facing and some of the
opportunities. An analysis of performance raises the question of comparison:
what should their performance be compared with? Historical comparison is not
possible since data on the impact of management are unavailable, and such
comparison is probably not relevant since the setting has completely changed.
Given that the alternative to local management is state management,
performance of local institutions may be compared with that of the state forest
service. The forest service manages gazetted forests and it had, until recently,
almost full responsibility for forests on public land. The comparison is presented
in chapter 6.

Conflict study reveals a great deal about the interaction of the different actors,
those within and those outside the local communities, those who are clearly
visible and those who are not (Lund,C.,1995). During conflicts, historical
arguments are put back on the agenda and actors who seem to have little
importance may turn out to be key players. Recent conflicts in the case studies
have been important events to help understand how local institutions function
internally and externally.

Interviews with the various actors are the main data collection tool. Such data are
subjective but forms of corroboration can usually be obtained. The acceptance
and enforcement of rules in the case of a village forest, for instance, can be
understood through interviews of villagers, inhabitants of neighbouring villages,
an evaluation of forest offences and fines paid up, and by observation of the
forest itself, directly or through occasional photography. Household interviews,
group interviews and discussions with specialist producers yield information
which build up a case description.


Winter (1994) has proposed a characterisation of forest management institutions
in terms of law, of legitimacy, of functionality and of operations. In this
classification, legitimacy is a key aspect which is mostly expressed in terms of
transparency, representativity and accountability. Chapter 7 describes the
process of representation and exclusion of interest groups in forest management
as the key element of legitimacy. Pastoral groups are often a particular case of
exclusion from local forest management. They are often excluded also from the
research and development agenda, so that the available data are scarce.

The functionality and operationality of forest management is to an important
extent determined by financial and human resources. Chapter 8 draws up two
different experiences: forest management praxis inside a project environment,
and outside a project environment. This helps explain the scope for sustainability
and replicability of the forest management model.

Chapter 9 analyses the legal and political constraints to local management. The
outcome of conflicts are usually determined by historical, legal and political
arguments. The judiciary, forest administration and political institutions are
particularly important for local forest management institutions, as obstacles or as
allies. In chapter 10, the immediate future of local forest management is
discussed. Increasing capacities in local communities are emerging opportunities
for local governance, and a range of project tools may help strengthen these
institutions - as long as external support does not choke local initiative.


In the 1970's and 1980's, national forest management strategies aimed at fores
restoration through tree planting and similar measures. The key fores
management strategy of the 1990's is protection of remaining natural forests
rather than rehabilitation of forests which no longer exist. Protection is also the
most essential strategy of local communities which have assumed the contro
over forest resources they had previously lost to the state. Protection agains
uncontrolled exploitation is the basis of any local forest management system. In
all case studies it is found to be the elementary form of forest management.

In the national forest service also, protection was and still is the key managemen
activity. The gazettement of forests is followed by the posting of salaried fores
guards who, in many Sahelian countries, have a military status. An formidable
system of sanctions is in place with penalties of up to 10 years imprisonment for
forest offences. In spite of this, forest protection by the state has been notoriously
ineffective and inefficient.

Very few guards have been deployed in relation to forest area. In many cases,
one or two forest guards have to protect forest reserves of 5,000ha or more, and
sometimes a guard may have to protect several forest reserves at a time. Guards


are notoriously corrupt and the legitimacy of forest expropriation by the state
often poor in the eyes of local people. When signs of illegal exploitation are foun
by a forest agent, the nearest community may be considered guilty and b
penalised. Many forest agents have been associated with repression an
extortion and they are identified with private rent seeking behaviour.
The exploitation rules applied by the state agency are often impracticable. The
permit system of the forest service is impracticable for local people if only
because of distances involved. Permits have often been sold without reference to
rational forest management, since forest reserves did not have management
plans to start with. The forest has often been illegally exploited since local people
or urban based traders managed to circumvent rules and repression.


Forest protection by the state has been unable to maintain the forest resource,
both on public land and on gazetted land. In the Zinder Department of Niger, for
instance, most of the gazetted forests no longer exist de facto even though they
continue to exist de jure. A government survey team noted that most forests and
agroforestry systems outside the forest reserves, on the contrary, were in fair
condition (PUSF,1986). In Kordofan, the forest service has sold off whole forest
reserves in order to raise cash (Adil,A., 1999).

When taking this as the baseline, forest protection by local communities appears
to perform much better in most cases. Guarding may the responsibility of:

- guards who have been employed to protect the whole forest, or a part of the
forest which is insufficiently monitored by the village population. Such guards
may be employed for a specific season, in particular during a part of the dry
season when the risk of illegal exploitation is high. The salary may be paid from
fines or from community revenues derived from sales of forest produce.
Alternatively, guards receive goods confiscated from illegal exploitation, they are
given the right to harvest certain forest products, or they are released from other
community work.

- members of the forest management committee may be in charge of guarding.
The number of people guarding the forest is higher than in the case of salaried
guards, but the committee members may visit the forest on a less regular basis.
In some cases, remuneration is obtained as a percentage of fines or taxes being
paid out.

- all members of the community may be responsible for monitoring of forest
exploitation and inform leaders in case of irregularities. In the case of an arrest,
several male members of the management committee may be recruited for the
occasion. In some circumstances, arrest may be dangerous and a team is
preferred. Another advantage of arrest by a team is that witnesses are present.
In the absence of witnesses, contradictions between the culprit and the guard
may arise which are hard to verify.


In many forests, forest protection is the responsibility of different actors at various
seasons. During the cropping season, most villagers are occupied with cultivation
and women may bring lunch to those working on the land. Also, wild foods may
be abundant and women and children spend some time collecting fruit and
flowers throughout the forest. At this time, the village forests are well monitored
by the whole village population. In the latter part of the dry season, however, few
villagers walk around in the forest. Many men may have migrated for the season,
and there are no wild foods to be collected. It is also the peak time of the
construction season, so that the risk of illegal tree cutting is high. Forest
protection may be the task of specially assigned guards who may at the same
time be engaged in certain forms of forest exploitation.

The assignment of guarding a forest may also be specific to certain sites, instead
of the forest as a whole. Valuable and vulnerable sections of a forest may be
protected whereas others hardly need protection, at least for a given season. In
Abunaanaa village in Kordofan, for instance, hay production is valuable in the last
few months of the dry season. This is the time that urban livestock traders put up
camp close to the forest but just outside village boundaries. Abunaanaa guards
deployed during these months have to protect the peripheral section of the forest.
But the traders will go as far as encroachment during the night to avoid being
caught. The alternative for Abunaanaa is to liaise with the court to force the
traders out of the area altogether.

Whereas protection by the forest service is governed by national standards, local
institutions apply a degree of protection which is highly variable as a function of
local physical, economic and institutional factors. Most rules are very different
between villages and tend to be flexible.

Rules which are found in most forests under local management are:

- no whole tree cutting of important species by non-villagers. Cutting of such
trees by villagers is regulated by licence. Species which are considered "firewood
trees" are not protected. Species which are important for fruit are protected
against any form of cutting.

- protection against charcoal production.

- fruit production is not regulated except for prohibition of green fruit collection.

- protection against clearing for agriculture.

Rules which are found in some of the forests are:

- the entire forest or part of it is protected against any form of tree cutting for
specified period.

- no temporary settlement of nomadic groups for more than some days.

- grazing and hay production regulated in certain areas and seasons.

In forests under the rural firewood market system, firewood production is highly
regulated but legally permitted. The rules are very different from those mentioned
above. Other forms of exploitation are hardly regulated under the firewood


market system.

It is clear that forest protection by local communities is numerically much stronger
than protection by the state. Tne number of guards and informants (people who
report infractions to those responsible) per unit area is incomparably bigger than
in the forest reserves. But what is the impact of forest protection under local
management? Interviews, review of infractions and conflict cases in general, as
well as physical monitoring are some of the means of verification.


In the forest service, infractions are dealt with by punishment of the culprit, or in
his absence, of the nearest community. Alternatively, a proxy is invented which
assures a flow of income. In North Kordofan, for instance, most forest products
are taxable but the tax and permit system is so impracticable that local people
cannot follow it. Instead, the forest service occasionally raids weekly markets and
confiscates forest products. In west Africa, the nearest community may be
punished if evidence of a forest offence is observed and the culprit cannot be
found. The punishment for an infraction is severe and is typically in the order of
10-20% of the annual income of local people. In Sudan, forest infractions are
punishable by prison terms of up to 10 years.

In comparison with the forest service, local communities deal with infractions in
milder ways. In the first place, local communities only sue the culprit if there are
no ambiguities about the identity. Unlike the forest service, forest offences are
not followed up if the culprit cannot be identified. In the most typical case, the
culprit is given a warning and may be required to present his apologies to local

Distinction may be made between infractions by local people and infractions by
strangers. Local villagers are expected to be fully aware of the rules which apply
to the different parts of the village territory, but strangers or visitors can be
excused for not being fully aware. Tamarind may grow in what seems to be
community forest, but which is in actual fact an old, abandoned field where
valuable trees still belong to the person who once cultivated the land. A stranger
collecting the fruits of this tamarind is informed that he has no rights to collect the
fruit in this site without permission of the owner; he will excuse himself and he
may even leave with the fruits he has already collected. But a local villager is
likely to be sued and the harvest will be confiscated.

The way that the culprit reacts is particularly important at this stage. If he is sorry
about his behaviour, or he can reasonably explain why he is not aware of the
rules, and promises not to do such a thing again, he will get away with the forest
offence lightly. This is normal for a first case offence.

A second step in the enforcement of rules is confiscation of produce and possibly

arrest. The culprit is perhaps not behaving respectfully, the offence has been
repeatedly made, or is of a more serious nature. The culprit is brought before
local leaders who will pronounce judgement. If the offence is considered
punishable, any locally appropriate fine may be imposed such as:

- confiscation of the produce, a typical punishment

- ameasure of millet

- a sum of money, which tends to be extremely modest compared to what forest
agents or the court impose

In some instances, fines are quite important. Examples are fines imposed on
nomads who have put up camp in protected zones of the village forest and who
have cut a large number of trees. In El Ain, fines of up to 75,000SE£ have been
reported, and the pastoralist association in the forest of Banh has imposed fines
of up to 15,000FCFA. Illegal firewood production in the Kelka forest, where the
commercial interests in firewood exploitation are very high, is severely punished.
In N’dounkoye village, Kelka, fines of up to 50,000FCFA are imposed.

A third step in the enforcement of rules is necessary if the culprit refuses to
accept a ruling by the local leaders. Recourse is usually possible to higher local
authorities or to a council of local leaders of several neighbouring villages. But at
the end of the day, customary and local law may not be accepted, particularly if
the penalties are high. State institutions may become involved, such as the forest
service and the police and in the end, the court. If the accused is found guilty, the
fine to be paid is often higher than in the village.

The community may ultimately apply the highest possible sanctions which exist.
In Kordofan, this may be complete social exclusion in the case of a village
member, in which all other villagers will refuse to communicate with the person
concerned. Migrants who have settled in the village, such as agro-pastoralists,
may be forced to leave. In Bankass, the fetish may be applied as the ultimate
sanction, which has death as a consequence. Very few examples have been
recorded of severe punishment in recent years as a result of forest offences.

The application of sanctions and taxes may be poverty indexed. In both Malian
and Sudanese cases, poor members of the community may be pardoned for a
forest infraction if it is felt that, due to poverty, they have no options. In the case
of taxes imposed by local management institutions such as the Alamodiou, fund
collection may be poverty adjusted, so that villages known to be very poor do not
pay any contribution at all.

To illustrate the majority of forest offences and ways of dealing them, a number
of short case descriptions of Kordofan is presented here, roughly in order of

1. a group of 3 nomads stayed inside the forest last season. They were told to
leave but they said one of their people was sick. They were allowed to stay for


three days until the person concerned was better, then they left.

2. Abu Hemera and Filia villagers were found cutting trees in a recently
registered forest, and received a warning but kept their produce.

3. A man of neighbouring Burbur village had establisned a cheese factory near
the forest on land belonging to Burbur. He was asked to leave since the large
number of livestock was seen as a threat to a part of the El Goz village forest. He
relocated his factory.
4. Someone of the village itself had goats inside a new plantation in the village
forest. The "official" rule is that a fine of 1,000S£ (0.46US$) should be paid per
animal. But the person apologised, and the fine was not paid.

5. A nomad brought his cattle into a plantation inside the forest and villagers took
his animals which were put in a designated corral in another village. The nomad
had to pay a fee to the owner of the corral but was not fined.

6. Two charcoal makers from another village were found inside the forest. Their
produce was taken and they were fined 5,000SE£ (2.34US$) each, which was
paid up.

7. Someone from another village produced charcoal in the forest and refused to
apologise. He was taken to the police in Banjedid and a trial was going to be
prepared. The matter was resolved by the umda, members of the court and

8. Someone who cut a Baobab and who refused to pay the fine imposed by the
village committee, fought with the Sheikh, was subsequently reported to the
forest service office and was fined S£ 3000 (6.90US$).

9. In november 1998 three people of a neighbouring village, including their
Sheikh, were found cutting trees to produce "mutrags" (building poles). Two of
them were fined 10,000S£ (4.20US$) each, the third 5,000S£ (2.10US$), which
they paid. Their produce was confiscated.

10. A nomad cut trees and was fined 75,000S£ (34.50US$), paid up.

11. Three agropastoralists who were living on El Goz land had cut a large
number of Acacia seyal to feed their livestock inside the forest. They were fined
140,000S£ (73US$). When the time came to pay, they escaped and were never
seen again. In a similar case, a nomadic group established a camp inside the
forest, cutting trees to erect fences. The villagers first went to the project office,
then to the police. However, by the time the police came, the nomads had fled.
They have not come back.

Some researchers distinguish various forms of non-state law. Some forms of
non-state law are embedded in specific normative structures, such as religious
law and customary law (Lund,C.1997). Another category of non-state law may be
called local law and can be seen as a pragmatic adoption in the wake of lega
reform. In this concept, local law in forest management can be seen as a recen
adaptation to changing circumstances which allow local management institutions
to defend their interests and settle a case outside court and outside modern law.
From the case studies it appears that local law is a key element in successfu
local management since modern law and state institutions are usually inefficient,
whereas customary law does not cover the emerging opportunities of loca



State institutions may back up the enforcement of local law through informal
mechanisms. In the case of El Ain, 5 cases of forest offence have been observed
among the 19 village forests over the period 1995-mid 1999 where local
institutions have not been able to enforce penalties. Tne court has been involved
in these cases but four out of five cases have been resolved before it came to a
judicial process, and none has been resolved under the provisions of the Forest
Act (one case is pending).

Other Sahelian case studies confirm that enforcement by local institutions works
up to a point, but requires backing by national institutions for some of the forest
offences. The Sudu Baba of the forest of Banh, Burkina Faso, and the village
leaders and sultan of the forests of lake Fitri, Chad, manage forest offences to an
important degree (Ouedraogo,|. and others, 1999; Tidiane,D.A., 1999). In both
cases, enforcement is legitimised by customary authority which has been
weakened by the modern state, but which is still functioning. In the forest of
Banh, this is the Sudu Baba, the assembly of lineage representatives. In Fitri, it is
the traditional village leadership, the area representatives of the sultan and in the
final instance, the sultan himself. But they do require assistance from the
government administration or court for those offenders who do not accept
customary enforcement.

Local law may conflict with modern law if local institutions protect a resource for
which the forest service has issued exploitation permits. Such cases have been
documented for the Kelka and Samori forests in Mali and for Gaya in Niger. Local
institutions have “arrested” the intruders concerned, who are invariably from
outside the community, and may have confiscated the produce. The outcome of
such conflict varies from case to case, but it is evident that forests may be
protected by local communities against external interests, even if such interests
are legally backed up.


Transparency or visibility of forest management expresses the degree to which
stakeholders know the management practice. It is also about an understanding of
the internal and external rules of the forest management institution. To some
extent transparency is a function of accountability, i.e. knowledge of money,
goods and services received and expended.

The forest service has published internal and external rules and ought therefore
have a high level of transparency. But there are large discrepancies between
formal rules of the organisation and practice of forest service agents. A study of
forest service agents' knowledge of forest policy and regulations in Mali, for
instance, shows how poorly agents understand the policies and rules they are
supposed to apply (McClain,R.).


Research in Mali and Niger shows many examples of forest agents illegally
fining villagers. It is not clear whether forest agents and villagers have poor
knowledge of the law, or whether this is rent seeking behaviour against which
villagers have no defence. In many cases receipts are not issued for fines, which
raises questions about accountability.

In Sahelian societies, oral communication is the tradition. The written medium is
of a much more recent nature and although it is increasingly applied, it remains
of modest importance in most of the forest management case studies. The rural
firewood markets, on the contrary, are highly organised due to the quota system
and tax payments and refunds, which require written accounts. In most other
cases of local management, written accounts are absent.

Differences are noticeable between the Sudanese and the west African cases.
Literacy in the Sudanese villages appears to be higher than in the West African
cases. In the Sudanese villages, written accounts are kept by the village
treasurer of fines received, and under the influence of the project, many villages
have written up the rules of forest exploitation. In a number of villages, the
management committee has written up accounts of forest offences and how they
have been dealt with.

In none of the Malian cases other than the rural firewood markets do written rules
exist. Taxes and fines are levied but only rarely are there written accounts. There
are no known examples of management committees keeping written records of
forest offences and of procedures followed. In many committees, the functional
literacy rate is low and, in the case of the Alamodiou, only one out of 5
committees has functional written accounts.

Household level studies of rules, carries out in Malian and Sudanese villages
show the rule interpretation varies considerably. Several dozen different rules
may be mentioned in a single village, some of which are contradictory, and few
households mention the same set of rules. A written prescription of the internal
organisation only exists in the case of the rural firewood markets, whereby rules
are prescribed by the projects. The rules of forest management other than
firewood production are rarely written down in the firewood market system.

The analysis of forest management rules is more complicated than the simple
question of knowledge about rules. Rules mentioned by one source may have a
different meaning for others, depending on the site, season, species and other
variables. The rule "Do not cut green trees", for instance, may be mitigated by
exceptions such as "weedy species", or an exceptional licence provided by the
leadership, or the cutting of minor branches only, or extreme poverty. A range of
exceptions and ambiguities may therefore exist but some rules are clearcut. Fruit
trees such as the tamarind cannot be cut, certain reserves may be out of bounds
for any tree cutters or pastoralists.


Another complication in rules analysis is that inhabitants from neighbouring
communities may simply deny they have knowledge of rules, even if rules are
known. On the whole, transparency is a conspicuous problem in local forest
management which can be explained by the wide range of sites and products
concerned, the poor legal framework and the dynamic nature of forest

he lack of written and published accounts, especially concerning income from
taxes and fines, has led to poor legitimacy in some institutions. In the Alamodiou,
there are several examples of voluntary contributions more or less stopping
because villagers feel that collected funds have been misused or because they
have no knowledge about the use of collected funds. There is a great deal of
overlap between accountability problems and issues of representativity.

In the majority of cases, however, accountability does not appear to be the major
constraint. In many institutions the amount of funds collected is very low and the
few funds collected are to an important extent redistributed to the people who
collected them in the first place, such as the guards. Representation of all user
groups and legal backing are more important concerns.


A key issue in conventional forest management is the production of management
plans. Production of the management plan is often seen to be an important
accomplishment, the completion of a phase of inventory and consultation and the
prelude to more formal and more "scientific" forest management. Such plans may
also be an explicit legal requirement.

Various criteria of community performance in forest management planning can
be defined. One criterion is comparison with state performance, and thus the
degree to which management plans are produced by the forest service for state
forests. Another criterion is the degree to which formal management plans are
produced and understood by the relevant actors. A third criterion is the
application of plans, whether formalised or not.


In two case studies, state managed forests co-exist with community managed
forests within the same administrative territories: the Mirrian Arrondissement in
the Zinder Department of Niger, and North Kordofan State, Sudan. Due to the
Sudanese forest policy, which requires that 25% of the land should be put under
forest reserve, North Kordofan has a large number of gazetted forests, or forests
under the procedure of gazettement. The total is 160 forest reserves, of which
only one reserve, El Ain, has a management plan. As this plan was produced by
an expatriate adviser, none of the reserves in this state have a plan produced by
the forest service itself. In Mirriah, Niger, management plans do not exist, even in


the case of the most important remaining reserve, Takieta, which is being
rehabilitated through project support from various organisations since 1987.

In the case of community forests, different scenarios exist. Community forests
under the rural firewood market system are required to have a forma
management plan as the basis of a contract between the forest service and the
local management committee. The available Nigerien examples, where the
market model has been applied since 1992, show that management plans exis
and that they are generally simplified.

n the market system, the key element of the management plan is delimitation of
he forest with firewood harvest quota for specified blocks. In all known cases,
he plans have been written by project staff while villagers may have been
responsible for forest boundary establishment. Given the complexities involved in
he annual quota system, local communities cannot realistically be expected to
master the planning procedure.

n another category of management plans, project consultants have produced
village management plans after completion of standard forest inventories (NEF
Mali,1996; CARE, 1997). Since the plans have not been produced by project
staff but by visiting consultants, the procedure is another step removed from local
management planning. Some project staff noted that such plans are of no use
since they are not understood by project staff in the first place, let alone by the
ocal forest managers. Others suggest that the plans help in formal recognition of
ocal forest management by the authorities. In neither case does the planning
appears to be of practical use to local management.

n a third category of forest planning, plans do not exist in a codified form. This is
he predominant but least recognised form of forest planning. In this category,
some aspects of forest management may exist in writing but most of it does not.
n Sudan, where land registration is an essential aspect of local management,
he delimitation and registration process is in writing. AS a consequence, the
forest appropriation process in Sudan is more formalised than in any other
Sahelian case of local forest management. In most cases, land is not formally
registered and demarcated, but approximate limits are known to community
members, although conflicting views on precise limits are quite possible.

In almost all cases, partitioning of the forest into zones is quite common. Many
locally managed forests are divided into zones of different ecological and
economic potential, and management rules may apply to specific zones. With the
exception of the rural firewood markets, these zones have not been not mapped
or otherwise codified but they are known to the members of the community. The
different zones may have a geographic name such as "Near The Village Limit" or
be given the name of the predominant species. The villager's understanding of
such stratification is undoubtedly much better, and the investment made is much
lower than the partitioning by surveyors and foresters.


In the case of the rural firewood markets, technicians partition the forest into
blocks which correspond to the annual firewood harvest quota, e.g. on a six year
rotation basis. The corresponding management plan is only concerned with
optimal firewood production. A study of sample villages shows that some
communities do, and others don’t understand the zoning of their forest

Outside the firewood market scheme, management plans originate from the
community owning the resource. Grazing, establishment of cattle camps,
hunting, fruit collection and artisanal wood production, to mention some
examples, may be higher on the planning agenda than firewood production,
depending on local priorities. Rules governing the different management units
have been formulated and the degree of guarding and enforcement is an
indicator of the weight given to the various elements of the plan. But plans have
not been written up by local institutions in any of the known cases.

In Sudan, where literacy is much higher than in the francophone Sahel, some
villages have begun to codify a part of their management plan. The plan typically
includes the land ownership certificate and the map which goes with the
certificate, written rules, and a list members of the management committee. It
may also contain panoramic photos of the forest, taken by villagers of permanent
sample plots. A description of major forest offences may be included. A part of
the codification is project supported and although it is probably sustainable since
it is understood and mostly implemented by the villagers, it may not be continued
and replicated in absence of such support.


Very few data exist on the implementation of forest management plans.
Relatively few formal management plans exist and they are quite recent, the
oldest dating from 1992, the rural firewood markets in Niger. Study of the
implementation of non-codified plans is still harder. Only very tentative
conclusions can be drawn at this point in time.

The Guesselbodi forest in Niger had a firewood market system in place from
1987, but remeasurement of the forest in 1991 and 1992 indicated that the forest
was overexploited and the markets were closed down by the forest service
(Hopkins, 1992). Firewood markets have been functioning in the Tientiergou
forests since 1992, and an evaluation in 1998 indicated that wood cutters in one
village were following the quota system and were respecting the blocks to a large
extent, but they were concerned that stocks were running down (Giraud,S.,1998).
In the second sample village, a large minority of woodcutters were not aware of
the system to follow.

In both villages, the woodcutters were concerned about outsiders exploiting the


village forest, and felt there was not much they could do about illegal exploitation.
In all three cases, uncertainties exist about the implementation of management
plans. Either the plans have not been properly implemented, and as a
consequence sustainability may be at stake. Or implementation was properly
done, but the quota were not well formulated to start with.

In the case of forests managed without codified plans, local control is more
intensive and often respondents are fairly confident about rules being adhered to,
which is confirmed by the imposition of sanctions. In some cases, user groups
voiced concerns about sustainability, in other cases villagers felt that their forest
is now recovering, but there is not much objective evidence.

One study shows how woodlands "inside village forest" have higher natural
regeneration than "outside village forest". Another study shows that fruit trees,
which always have been protected under customary rules, now constitute a
major proportion of all trees. It is not possible to generalise from isolated cases.
Most cases of local management are from the 1990s which are characterised by
relatively good rains. Conclusions about the causal relation between
management plan implementation and forest dynamics are fraught with
difficulties since the impact of the rains may override that of changed
management. Only persistent, long term monitoring is likely to yield relevant
information. Panoramic photos of locally managed forests made by projects and
villagers in three countries date from 1997, which is obviously too early for
meaningful conclusions about vegetation changes.

The conclusion which can be drawn is the state demands more of local
communities in terms of management planning than it performa in its own estate.
Secondly, non-codified forest planning is widely practiced, it responds to cross
sectoral needs and plans originate from the communities managing the forests,
but the planning process is not transparent in the eyes of external stakeholders.
Finally, there is little evidence at this stage about the implementation of
management plans, and about the relationship between management plans and
forest sustainability.


Means of communication by local forest management institutions is a key
performance indicator. By tradition, communication is oral and is channelled
through informal gatherings, village assemblies, markets, festivities and other
occasions. The complexity of communication is very different between the case
studies depending on the size of the institution. In the Sudanese cases, where
forests of a hundred hectares or so are owned and managed by small, individual
villages, communication between members is relatively simple. Pastoralists are
excluded from the management institution, which further simplifies


In the case of the Kelka and the Alamodiou in Mali, a dozen or so villages are
members of one institution which manages an area of tens of thousands of
hectares. Pastoralist groups may be attached to this area and are member,
which further complicates communication. Important and urgent problems may
arise such as a conflict between two villages or intrusion in forest management
by a state agency, and the credibility of local management may suddenly be at
stake. Effective communication may be a key issue in those situations.

Projects tend to assume a part of this responsibility through vehicle transport and
written messages. An institutional evaluation of the Kelka warns of the risk of
poor sustainability when the project organisation assumes responsibility for
communication between villages (Diallo,Y., and Winter,M.,1997). At the same
time, local institutions need to develop more dynamic forms of communication
and not rely on traditional oral communication only. Several cases of
modernisation have been observed as spontaneous practice:

- The use of written notes for convocation of meetings. The weekly markets are
the main mode of distribution.

- The distribution of information notes to neighbouring villages and to pastoral
groups informing them about the rules of forest management.

- The use of the local radio station (against payment) to inform committee
members and the public.

The common denominator is literacy. Even though the leadership may not be
literate, as often happens in the francophone Sahel, a younger, literate villager is
often charged with transcription of the information. Communication may cost
some money, even if it is a minor amount, e.g. 1OOFCFA (0.15USD) in the case
of a short rural radio message.


Changes to the forest resource are of interest to local and to external
stakeholders. Monitoring technologies applied by external agencies have
expanded greatly over the years through the development of improved remote
sensing techniques, geographic positioning systems, geographic information
systems, plotless sampling methods, etc. This has been useful in the rural
firewood projects which have measured the domestic energy resource of peri-
urban areas in great detail.

The typical monitoring tool is the forest inventory which estimates wood volume
based on tree diameter measurements in sample plots. Local people may be
required to assist as casual labour, or to help in general orientation of the
inventory team. Other than that, such inventories are planned and implemented
without any involvement of local institutions. Resource monitoring by local users
as the ultimate forest managers, on the contrary, has not made much progress.
Very few projects have invested in locally manageable resource monitoring



Local users obviously do observe changes in the forest on which they depend, in
other words, local monitoring systems exist, but in a form which is not codified
and which is not objectively verifiable. The advantage of locally observed and
communicated information on forest change is that it is cost free and
comprehensible. “Modern” forms of forest monitoring have the opposite
characteristics: it is expensive and unintelligible for local users, but it is
objectively verifiable.

Yet local institutions may benefit from objectively verifiable monitoring systems
which they master themselves. El Jefil village in Kordofan asserted that their
forest has been well managed and that by 1999 the forest is in better condition
than in 1994, when it formally became their property. The forest service
representative stated that the opposite is true, and that forest management by El
Jefil needs close supervision by the forest service. The existence of objective
monitoring data would have enabled El Jefil to defend the superiority of their
management system vis-a-vis the forest service. Unfortunately, baseline data in
the form of panoramic photos are only a few years old.

The two objectively verifiable monitoring systems developed in Kordofan and in
Bankass are panoramic photography and inventory of local forest products.
Panoramic photos have been taken by villagers from identifiable positions such
as rocks, baobab trees and boundary stones. The relatively open vegetation,
characteristic of the Sahel, has allowed photography of relatively large areas.
Once developed, the photos are kept in the village file, along with the ownership
certificate, the map and other documents. This initiative is unlikely to be taken by
the village, since it requires a camera, but the photos serve as baseline data
which are generally well conserved by the village.

Literature on participatory forest monitoring demonstrates that local communities
do measure their forest resources (Carter,J.,1996). In very few cases, wood
volume is measured and such cases appear to be limited to higher rainfall areas
where local wood cutters sell trees for sawn timber and other high value
products. Also in European forestry, wood volume is only measured if the
products are of high value. In Sahelian village forestry, prices are low and
products are very irregular and difficult to measure. The need for objectively
verifiable forest monitoring should be put in the context of forest economics.




Local institutions are performing in forest management. They guard their forests
in effective ways and they enforce the rules by socially acceptable means. They
plan and monitor forest resource utilisation in their own way, and some may use
the written medium. But how are internal differences in the society represented
and worked out in these institutions?

Development researchers have criticised the concept of community as it has
been applied in many development projects, such as Gestion de Terroir or rural
firewood markets in the francophone Sahel (Anon,1992; Sani,A.M.,1996). In
many projects, internal differences in the villages have been neglected in project
strategies in which village committees have been appointed and are considered
representative without critical review. Constraints such as conflicting interests
between the spiritual land owners and other villagers, or between settled and
transient landusers, have lead to poor policy and project impact.

In this section, the effect of ethnic composition on forest management institutions
and related leadership issues is presented for three cases in three countries.
Specific reference is made to women and pastoralists, stakeholder groups which
are usually underrepresented. The internal and external organisation is described
for a large number of case studies, including the mechanisms of recourse. It also
deals with unions of village institutions.


The basis of resource tenure in the Sahel is spiritual land ownership of the
founding lineage. The ancestor of the founding lineage has appropriated a tract
of land and cleared all or part of it through fire and hoe. Occupation rights were
granted, in popular imagery, by the local spirits, owners of the resource base, to
first arrivals. A pact symbolises this alliance and it is periodically renewed by a
sacrifice offered by the oldest member of the recipient lineage (Anon,1996). Both
lineage and age determine inheritance of leadership.

Others moving onto this tract of land ("immigrants"), temporarily or permanently,
required permission to do so and often tithes or symbolic payments were made.
The immigrants could be removed by the founding lineage under varying
conditions, but generally the longer usufruct rights were held, the more difficult it
became to force out the immigrants. Pastoralists have moved periodically
through with varying claims over local resources. In recent decennia, forcing out
immigrants and pastoralists has become an increasingly important issue since
natural resources have become scarce. But apart from general concepts,
traditional tenure is not at all clear, it is very localised, constantly changing and



In West Africa, the Islamic invasions of the 19th century introduced Islamic
concepts of tenure in which land can be sold or otherwise exchanged (IIED,
1998). The Islamic invasions in west Africa superimposed their tenure systems
on the pre-Islamic systems and clashes have erupted as a result of conflicting
Islamic and pre-Islamic interpretations (Maiga, |. and Diallo, G.S.A.,1995;
Mortimore,M.,1997). In the case of French colonisation, "occupied lands" were
governed according to local custom whereas "unoccupied lands" were claimed
as state property. This principle was not applied in the Sudan and various other
parts of anglophone Africa, which relied to a greater extent on customary
authorities according to the principle of indirect rule (Elbow,K. et al, 1996).

The identification of a chief amongst the founding lineage was widely applied in
west Africa through colonial powers, first through the Peuhl empire of the Dina,
then by European colonialism. In the process of colonisation complex traditional
tenurial institutions were reduced to the chieftaincy, although traditional
institutions continued to exist. At independence, customary institutions were
further weakened. In Mali, a chief appointed by the administration was instituted
to replace colonial or pre-colonial chiefs and spiritual leaders.

In Sudan, the Numeiri Government attempted to eliminate the customary
authorities and replace them with village councils. More recent governments
have recognised customary authority to some extent, and customary authority
presently co-exists with elected village councils. In Niger, ambivalent relations
between the administration and chieftaincy have developed. At various points in
time, national politics either emphasised user rights or lineage ownership rights
(Lund,C., 1995).

The general situation is that traditional institutions have weakened to varying
degrees but continue to exist in many villages as parallel institutions. The state
vested authority in the administrative chief or local committees but the de facto
authority is often shared with traditional authorities, depending on the issues at


Project support to communities wishing to appropriate and manage forest
resources was based on the principle that each village has a leader ("Sheikh")
under whose authority woodland in the village periphery is registered on behalf of
the community. This followed the official procedure of land registration starting
from traditional leadership in hierarchical order: signatures from the Sheikh,
followed by the Umda, and finally the Nazir, before government institutions are


However, over the period 1992-1999, when 19 village forests were registered or
were in the registration procedure, it turned out that in about half of the villages,
the procedure was slowed down or occasionally blocked at the traditional
leadership level. From 1997 to 1999, local leadership was analysed through
repeated interviews. The study shows that the official position in which each
village controls community land through its Sheikh, is no longer tenable.

In all project villages, traditional leaders (Sheikhs) exist who represent the settled
ethnic groups. The origin of these leaders dates back to the settlement of the
ethnic group in the area. A patrilinear system usually determines how leadership
is passed down within the village and how the leader of a new settlement, split off
from the village, is determined. Upon his death, the Sheikh is succeeded by one
of his sons, or in the absence of a son, by a brother. The successor may be his
oldest son, but for various reasons another son may become sheikh, and the
villagers have an important influence on the succession.

Though the Sheikh embodies traditional leadership, other village representatives
tend to have an important voice, depending on their status and personality. The
sheikhs regularly consult representatives of other families in the village.
Discussions with project staff, for instance, are usually held with a group of men
including the Sheikh, who may not have the last word or indeed the most
important word. An influential teacher, for instance, may sometimes act as the
village spokesman in spite of the presence of the Sheikh.

Sheikhs are a traditional form of leadership which follows patrilineal succession,
but villagers may depose their Sheikh in the case of unacceptable behaviour.
The Umda is the next level in the traditional leadership hierarchy, followed by the
Nazir. They have to confirm that such behaviour warrants deposing a Sheikh and
that they will appoint a new leader. Conflicts in the village are in principle solved
within the village. If not, neighbouring Sheikhs will be invited to help solve the
problem; the Umda and Nazir may ultimately be consulted. The Nazir also serves
as a recourse for the sheikh in case of an unacceptable ruling by the Umda.

Three categories of Sheikhs have been distinguished in the project area:

- A Sheikh who represents the lineage of the first cultivator, which owns a large
tract of land. He has considerable powers;

- A Sheikh who represents a group of immigrants and who have relatively little
land and not much power;
- A Sheikh who is actually an assistant of the land owning Sheikh and who
represents a group of people who originate from the lineage of the first cultivator,
but who settled in a neighbouring locality.

[he three different categories are not very distinct but are useful to describe a
continuum between powerful landowning leaders and leaders with few powers.
The actual control over land and natural resources varies from case to case,
including the personality of the Sheikh.


In the view of "Big Sheikh" Eljak:

"A Sheikh is a descendant of the founding family of his territory. My grandfather,
for instance, was recognised as the Sheikh of an area which presently includes
five villages. After my grandfather passed away, an area called Baduga became
the land of one of his uncles. However, this Sheikn never had any sons, so when
he died his elder brother, who was also the Sheikh of Higeina became the Sheikh
of Baduga. He allowed some families to settle in this area and appointed an
assistant Sheikh to represent him. What the project calls "the Sheikh of Baduga"
is in my eyes only an assistant to Sneikh Mohammed Ismael of Higeina. Neither
the assistant Sheikh nor his descendants can ever become a real Sheikh. Only
the son of a Sheikh could possibly become a Sheikh himself"

The formal government position is that each settlement has a Sheikh who is
responsible for community land matters on behalf of the people he represents.
This is the case for community land brought under cultivation by villagers, which
is formally allocated by the Sheikh. Once under cultivation, such land rights are
inherited. But for more important land issues, such as the settlement of
immigrants or of land development companies, or village forest registration, land
appropriation procedures are different. By tradition they are seen to fall under the
authority of the land owning Sheikh who may not live in the village concerned.

This is why, in the eyes of customary authority but not in the eyes of the project,
the Sheikh of Baduga required permission from the Sheikh of Higeina to register
Baduga village forest. Sheikh Eljak, on the contrary, autonomously permitted
registration of a village forest of 200ha and furthermore granted permission to 5
companies for mechanised farming totalling 3,000ha

Apart from the first cultivator descendency, others characteristics may play a
role. Differences exist between immigrant groups of local Arab origin and those
of more distant origin such as the Haoussa and the Nuba. Differences also exist
between groups who are established for a longer time and those who are more
recently established. The personality of a Sheikh also plays an important role.

Box 2. Immigrant groups assert forest ownership

A large number of villages in North Kordofan originate from migrants such as
pilgrims on the move between west Africa and Mecca. Their descendants are
often seen as guests (inhabiting a place for guests, hakura) by the original Arabic
speaking inhabitants, with little access to natural resources. Village forest
appropriation is not an easy task but can be achieved.

Three generations ago the village of Abunaanaa was established by a group of
Haoussa who were given land by the Sheikh of El Mulbas. At the time, the fertile,
riverine forest a few hundred meters from the village was part of the allocated


land. However, in the 1940's landowning leaders have expropriated the riverine
land and sold it to traders from the town of El Obeid. Since that time, Abunaanaa
villagers only have access to this land through crop sharing arrangements or
work on it as paid labourers.

The villagers have access to land for rainfed agriculture but woodland is at some
distance from Abunaanaa and belongs to landowning Sheikhs. When the project
encouraged the village to establish a village forest, it was clear that none of these
Sheikhs would allow Abunaanaa to appropriate a stretch of woodland. The only
option left was to register a piece of treeless land near the village and start from
scratch. In December 1993 the village proposed a site of 34ha but Sheikh Agib of
neighbouring El Mulbas village protested in February 1994, because he had not
been contacted by the project even though he is the competent Sheikh of the
area. Agib is the Sheikh of the Arab tribe Bazaa, who considers the leaders of
migrant villages as his assistant Sheikhs.

Through repeated discussions and attendance of project workshops, Sheikh Agib
was finally convinced that the request from Abunaanaa village is reasonable. The
area was registered and has been rehabilitated by the villagers with project
support and presently has a high degree of protection against all types of
exploitation, except for hay collection towards the end of the dry season. The big
Sheikh of El Mulbas has received hay free of charge for his animals on some
occasions, but Abunaanaa feels secure about its ownership of the forest. By
1999, the area was transformed into a well vegetated bushland with appreciable
annual income from hay.

But most other migrant villages have not been able to secure and rehabilitate
land. Even where such villages, with support of the project, started up land
registration procedures without consent of the landowning Sheikh, the higher
traditional leadership, who are ethnically related to landowning Sheikhs, blocked
the procedure at some point.

Many different cases are found in the project area and the following 5 profiles are
helpful to illustrate the differences.

1. El Jefil is an old village without immigrants and it is the cradle of the Jeleba
tribe. The sheikh has customary authority over the extensive El Jefil land as well
as some of the neighbouring villages, and he represents the inhabitants of this
large village. Also the higher leadership (Umda and Nazir) originates from El Jefil.
Decisions on land are taken with a high degree of autonomy.

2. About 100 years ago a group of people from Aloba village decided to settle in
an area they named El Goz. The present Sheikh of El Goz, a distant cousin of
the Sheikh of Aloba, is entitled to take important decisions on land issues without
reference to the sheikh of Aloba. The land boundaries with other villages, in case
of conflict, are determined by the higher leadership (Umda, Nazir) who originate


from elsewhere.

3. Baduga village was established 30 years ago, when 5 families split off from the
village of Higeina. The sheikh of Baduga may be considered an "assistant
sheikh" who cannot take decisions on important land issues, although he can
autonomously allocate land for agriculture to his people. The land of Baduga is
limited to whatever is taken into cultivation. Other land allocations as village
forest registration are only possible by consent of the Sheikh of Higeina, who
permitted the appropriation of 250na of good quality woodland by Baduga
through the village forest registration process.

4. Several villages of Haoussa origin were established about a hundred years
ago. They were allocated land by the landowning sheikhs which was limited to
the land taken into cultivation (the hakura, place for guests). In 1992 the project
suggested they establish village forests, and they requested land from their
respective sheikhs. Only one of them, Abunaanaa, succeeded (see box 1).

5. Several village forests are shared by distinct groups. Jebel Kordofan is shared
by three groups, the bigger village is represented by the landowning Sheikh, the
smaller two by assistant Sheikns. When the project approached the three
villages, the two smaller villages each wanted to establish their own forest, but
were prevented from doing so by the Sheikh. Jebel Kordofan is now a poorly

managed forest with much internal strife. Newa village is another example (Box

Box 3. Conflicting leadership in Newa village forest management

In 1994 the project started working in Newa village, North Kordofan. In 1996 the
project started supporting the Village Forest registration process, given that
Sheikh Ali of Newa, supported by others in the village, was in favour of
registration. A wooded area of 330ha was considered suitable for registration but
required permission of the representative of the Jellaba tribe, even though the
forest is located on Newa village land. The Jellaba representative in this matter,
Sheikh Hemiti, agreed with the registration of the forest in question by Newa
village and the registration process took off.

But later in 1996, after the applications had been approved by local leaders and
had gone to government level, serious problems started. It appeared that there
was both resistance from within Newa village, and from neighbouring Mehela
village. Project staff then investigated the social relations in some depth.

The area of which Newa is a part, is inhabited by two Arab tribes: the Jelleba and
the Dajo, with the Jelleba as the most numerous group and as the lineage
representing the first cultivator. The Jelleba Sheikh is consequently the most
powerful person in terms of resource tenure and in terms of people. Newa village
was only settled by the end of the 19th century by the two groups, Dajo and


Jelleba, of which the Dajo happen to be the by far most numerous group. The
Sheikh of the Dajo people in Newa, in terms of formal government procedure,
represents the inhabitants of Newa village, which is geographically one village
but consists of two distinct quarters.

The Jellaba as a minority group in Newa have their own leader and in early 1996
this was Hemiti, Sheikh of the Jelleba, who happened to live close to Newa. As
the landowning Sheikh he was entitled to decide that the woodland concerned be
registered as Newa village forest. But Hemiti died the same year, and the Jelleba
people decided that they did not want the eldest son, Yahya, who also lives in
Newa, to become the Sheikh, but instead asked his younger brother Mohammed
who lives in the main Jelleba settlement, Mehela, to become the successor.
Immediately upon the succession, Mohammed contested the village forest
registration through his representative in Newa, assistant Sheikn Yahya, his
eldest brother.

The project staff pointed out that Yahya was in no position to oppose the ongoing
process, but it found that Sheikh Ali of Newa and the Dajo people were no longer
interested to continue registration, particularly now that it was obvious that
Sheikh Mohammed was clearly opposed to it. The project talked with Sheikh
Mohammed to try and settle the matter, but was told off: "The project will bring
blood on Newa if it continues forest registration". The project explained to Newa,
that under modern law, the village forest would belong to Newa villagers, not to
Jellaba or Dajo tribes. But the Dajo people of Newa are scared of tribal conflict.
But the registration process was now in the hands of government and continued.

In 1997, the registration and gazettement process had almost been completed
and by law, the forest should be managed as a reserve by the Newa village
forest committee. The Newa village forest committee is dominated by the
majority Dajo, whose Sheikh insists that the committee is the only competent
agency. But assistant Sheikn Yahya of the Jelleba states that he not the
committee, is the main guard of the forest. Sheikn Mohammed of neighbouring
Mihela states that neither Sheikh Ali nor the committee, but he and his people,
including Yahya and the 7 families he represents, decide on Newa village forest.

In the past, large tracts of woodlands existed outside the village forests so that
the pressure on the village forest was modest. But presently, most woodlands
around Newa except for the village forest are being cleared by a commercial
company. In the near further, the two ethnic groups of Newa and the large
population of neighbouring Mihela will only have Newa village forest as the major
source of forest products. Resource conflict and poor forest management should
be anticipated, unless the Dajo and Jelleba arrive at a mutually acceptable



In West African case studies, recently immigrant groups are often
underrepresented in local leadership. In the Samori forest in Mali, members of
recently established villages must show respect to the leaders of the founding
village. All migrant villages and the founding village together may have one forest
management committee, but in important issues, the founding village dominates.
Contacts between local people and outsiders such as government and project
staff are closely followed by representatives of the founding village, and
communication in their absence may be almost impossible. But cases of poor
forest management due to unequal representation have not been recorded by
the project. A main constraint to forest management in the Samori is the permit-
and-tax system of the forest service which allows outsiders to exploit local

In the rural firewood markets, the forest management system is occupied with
sustainable firewood production for urban areas. In most cases, a piece of
woodland is managed by the nearest village through a management committee,
under contract with the forest service. As a result of the subsectoral focus of the
markets, the committee members tend to be wood cutters and thus a specific
social group in the local community. In the available case studies, the
composition of the committee is not representative of the local community.

Baban Rafi forest in Niger is of major importance for livestock production. One
estimate puts the economic value of livestock production, most of which is
herded by the Peuhl, at three times that of firewood production through the
market system (Paris,P.,1996). The 22 rural firewood markets in the forest are
managed by committees in which the herders are poorly represented. Project
requirements demand inclusion of a representative of the Peuhl herders in the
committee, but very few of the relevant positions such as chairman and treasurer
in the 22 committees are occupied by Peuhl herders.

This is analysed in detail in one community, Tientiergou village, Niger. The
representative of the Peuhl herders is nominally in the committee, but is
effectively excluded. The woodcutters, who are all members of the Rimaibe
ethnic group, have monopolised power in the committee. Consequently, one
ethnic group within Tientiergou dominates forest management, relations with the
key government institution (the forest service) and the funds which have been
paid out for the benefit of community projects in the framework of the firewood
market. The researcher concludes that this is a socially explosive situation

The Takieta gazetted forest in Niger is traditionally used by the surrounding
farming communities of mixed ethnic origin and, during the rainy season, by the
Peuhl and by other herders. In the process of disengaging the forest service and
putting the user groups in charge, the project has encouraged both farmer and
herder groups to plan and implement Takieta forest management. A large
number of consultative meetings have been held which has led to two different


kinds of association: a general assembly of all interest groups which meets
occasionally and which sets out policy, and four committees which meet regularly
and which implement local management policy in the four geographic zones.
Transhumant herders were very keen to be involved in the consultation for
Takieta forest management, which is important in the annual movement between
dry season and wet season pastures. In an all important general assembly in
1997, they were numerically at par with the representatives of settled
communities. But they also expressed reservations about the frequent and time
consuming representation in the four committees responsible for day to day
management. This emphasises the need for measured participation and
representation: neither too little, nor too much, depending on the nature of forest
resource use.

In forests which have been, and still are occupied by one ethnic group, questions
of ethnic representativity appear to be less important and are less likely to put
local forest management at risk. The Alamodiou forest management committees
are culturally and numerically dominated by the Dogon tribe. The land is owned
by the Dogon people and their culture dominates. The Peuhl herders are a
minority of 7 out of 50 villages, wno have been more or less incorporated in the
local economy and in the forest management committees, despite cultural
differences between the traditionally animist Dogon farmers and the Muslim
Peuhl herders.

The membership of the management committees often has important animist
attributes such as ceremonies when members are born and die. The Peuhl are
strict followers of Islam, but they may participate in the Alamodiou management
committees and Peuhl members may also participate in the ceremonies, be born
and die as members and enjoy spiritual blessings from the Alamodiou. But other
Peuhl propose a distinct Peuhl management committee, parallel to the
Alamodiou, which does not have the Dogon cultural characteristics alien to them.
The project has supported this initiative.

Across the Mali-Burkinabe border, in the forest of Banh, the Peuhl are the
traditional landusers and managers of the forest, with few immigrants to date
(Painter, T. and Sanou,S., 1996). Representativity is traditionally assured through
the lineage representative in the decision making body, the Sudu Baba.
However, people of other ethnic groups also live in this area. The concern of
enhanced representativity has been addressed through the constitution of an
association, the Walde Kawral Pulaka, which is intended to represent all



Ethnic divisions are obviously not the only social differences which have a
bearing on representativity in forest management. Sahelian societies have
greatly changed in the 20th century and the extended family has broken up in
many ways. Opportunities for market production and migration have increased
and individuals have become anxious to appropriate profits privately. As a
consequence, socio-economic differentiation has increased. This has been
observed for the Haussa in Northern Niger, the Dogon and the Dafing in Mali, the
Jelema in Kordofan and elsewhere (e.g. Mortimore,M.,1997; IIED,1998,
Chevenix-Trench,P. et al,1998; Kerkhof,P.,1998). Nevertheless, forms of
extended family solidarity are maintained but vary from place to place.

Seasonal and long term migration has become a general feature of the
landlocked Sahel, with an anticipated 12 million migrants by the year 2020
(OECD et al). Migration has induced a degree of specialisation in natural
resources management which probably did not exist previously. While
subsistence forest exploitation used to be a common feature of all members of
an age group, an increasing number of community members are involved in
income generating activities, often through seasonal migration. Their knowledge
of the local forest resource is very limited, and an increasing number of young
men have never been in their village forest.

From two studies of socio-economic stratification in forest exploitation it appears
that those who continue to exploit the forest increasingly do this for the market,
including sales to community members who practice seasonal or long term
migration. Forest exploitation is hard work and those who have sufficient
alternative income opportunities may prefer to buy forest products from other
community members. The knowledge and practice of forest exploitation is
subject to a differentiation which did not exist previously.

Specialisation may be geographic, if one village has alternative income
opportunities such as dry season water sources for gardening, and not it's
neighbour. Or specialisation may be defined by wealth class within a given
village; poorer members of the community are more active in forest exploitation
than the richer strata, and sell a part of their produce to the richer families.

Socio-economic differentiation appears to be less of a factor in success and
failure of local forest management than ethnic division. Those who are regularly
absent from the village due to migration are less well represented on committees.
In some institutions, committee membership is excluded for those who do
migrate regularly. But since migrants are less dependent on the forest than those
who stay behind, this may not be much of a problem.

Villages who do not use their forest much may enter into conflict with
neighbouring villages which lack forest resources, and which exploit their


resources. Case studies exist of villages where this has happened, and others
where conflict has not arisen. Conflict appears to surface more quickly if the
forest produce has a high market value and if villagers feel that the rate of
exploitation is not sustainable.

Issues in representativity other than those of ethnic or socio-economic origin
have also been recorded. In one of the Alamodiou entities, for instance, historic
sensitivities of four ethnically closely related villages have lead to near collapse of
the management institution. The problems first arose when the French colonial
administration accorded priority to a settlement which was not the founding
village. Five years of project assistance in recent years has not had any
significant impact on the poor functioning of this Alamodiou.


Women tend to be underrepresented at the decision making level in
development. The general development policy has been to impose the condition
of female representation in project management committees. This has been the
practice many of the case studies, and some projects require that 50% of
members are women. Also, the traditional leadership, which is characterised by
all male membership, has been under pressure in some Gestion de Terroir and
firewood market projects. In the Niger market system and in the Gaya palm
groves, for instance, the village chief has been excluded from the key positions in
the management committee, although he is a member.

In all the case studies, women are formally members of the management
committees. Does this imply that they are well represented in forest management
decision making ? For the Kordofan case study, two indicators have been used:
the presence and the participation in the many meetings over the period March
1997 through to March 1999.

Apart from meetings specifically organised for women, in about two thirds of the
management committee meetings, no women were present. In the meetings
where women were present, they did not substantially participate in the
discussions. This is illustrated by a sample of the numerous cases:

- When Sheikh Ali was asked to enumerate the members of the Village Forest
Committee, he gave the names of men only. Then he was asked about possible
female members, and he responded that there were women members but he
could not name any of them.

- Upon request of the project, the Halfa committee was made up of equal
numbers of men and women, all present at the meeting. One of the five woman
frequently attempted to participate in the discussions, but she did not get a
chance and left halfway through the meeting. The other women did not speak a


- In Goz village, half of the committee members are women, as required by the
project. During the meeting they did not speak up.

- Sheikn Yahya confirmed that women hardly participate in matters concerning
the village forest. He and his councillors said that the women do not know the
management rules and that this does not matter, because the women do not use
the forest, which is situated about 4km from the village.

- The female head of the project extension team noted that women are not really
involved and don't even want to be involved in village forest establishment. For
the women, land issues are a matter for the Sheikh and the men in general. The
project has organised special meetings for the women to inform them in spite of
their wish not be involved.

In Mali, the Alamodiou committee has female members if only because part of
the members were destined by culture to become so even before they were born.
Women in these committees do speak out and they participate in forest patrol,
but in some important issues, such as conflicts with the forest service, only men
were represented. The Alamodiou leadership is uniquely male. None of the
research meetings held with Peuhl herders in the neighbouring Samori forest
over the period 1996-1999 had any presence of women.

Women seem to participate more in communities where Islamic traditions are
weaker, such as the Dogon villages. Separate meetings for women can obviously
be held by any research or development agency. Projects and policies may
enforce a formal committee composition which represents women at numerically
equal terms, but these formal committees may not be functional. In the long term,
important decisions are most likely to be made in accordance with local
traditions. Project staff cannot police committee meetings in a sustainable and
replicable manner, so that active representation of women in forest management
committees is likely to be a function of cultural norms in the long term.
Nevertheless, the presence of women may inform village women of the issues at
hand and gender policies may contribute to the wider issue of women's


Sahelian forest, woodland, bushland and wooded savanna is of high value to
livestock production, particularly during the time that cropland is not accessible to
animal husbandry. Much of the land categorised as forest is actually grassland.
The principle of mise en valeur or land development as a way of securing land
rights was codified by the colonial government as a tool to induce economic
development and continued to be applied in post-colonial tenure as a major
criterion of access and ownership (Elbow,K., 1997). This led to insecurity for the
multitude of traditional landusers who relied on mobile livestock production.


The relationships between different traditional landusers have greatly changed
since precolonial times. The complementarity of farming and nomadic livestock
production which used to be a common feature has been transformed at least
partly into a relationship of competition. Not only have pastures and livestock
corridors been taken into cultivation, but farmers have also diversified their
economy through livestock production. This is now an important strategy in
dealing with the risks of rainfed agriculture in the Sahel. The implication is that
settled communities and mobile users increasingly compete for the same natural

Many livestock owning farmers have reduced access to forage and water
resources for traditional livestock keepers. The millet stalks in Mali's Seno plains
are removed as soon as the millet is harvested, and put into large stacks
protected by thorn fences. Farmers prefer to keep these resources for their own
livestock and take the risk of increased conflict. Such conflict arises when hard
pressed Peuhl herders remove the thorn fence. In Kordofan, farming
communities increasingly resent the annual arrival of the Baggara herders, since
"they finish all our grass in a short time", at least in years of poor rainfall. Deadly
conflicts between farmers and herders appear to be on the rise in the Sahel.

The description of enforcement in Kordofan villages usually includes cases such

- Nomads have cut a large number of Acacia seyal for small livestock and have
been fined.

- Nomads were found staying inside the village forest and were fined.

- A nomadic group had established camp inside the forest and had cut trees to
erect fences, refused to pay a fine and fled before arrival of the police.

In Kordofan, traditional usufruct (Afaa) by mobile herders of the village forest has
not been recognised in the registration procedure and no form of compensation
has been made, despite legal provisions in the Forest Act, 1989 and the Civi
Transactions Act, 1984. Once in the registration procedure, such forests are
legally no longer accessible to the nomadic herders unless by specific permission
from the village committee. By law, only villagers are represented in the fores
management committee, so that pastoralists are excluded.

A review of forest offences suggests that, at least in some cases, discrimination
is made between forest offences of nomadic and settled herders. In these cases,
herders from local villages got away with a warning, herders from other ethnic
groups have had to pay a significant fine. Many signs of frustration on the side of
Baggara herders have been noticed, such as destruction of all forest boundary
posts, plain refusal to accept directives from committee members and projec
staff and sometimes more violent events occur.

In the forests around lake Fitri, pastoralists are traditional users of the Acacia


forests. However, the increased gum arabic prices have encouraged the Bilala
farmers to appropriate natural Acacia senegal groves and to increasingly exclude
pastoralists (Tidiane,A.D. and others, 1999). Changing economic circumstances
have evidently changed the rules of access.

But In many cases, farmers tend to maintain relationships with specific nomadi
groups in which farmers’ livestock may be kept by the nomads under variou
arrangements. Also other forms of interdependency exist so that local fodder an
water resources continue to be negotiated. In Bankass, the Assaga And
Alamodiou committee and Peuhl leaders negotiate the pruning of Faidherbi
albida which is pollarded in years of need, but not more than once in three years.
After negotiations have been concluded, the herders are entitled to clip the trees
of any owner on the territory for a specified number of days against a payment to
the committee.


At the level of project organisation, pastoral groups are also underrepresented. In
none of the case studies, until 1998, had projects recruited any nomadic or
transhumant herders as team members. Gender is generally a criterion for
project staff recruitment, but otherwise the teams uniquely consist of members of
settled groups and some staff may have nomadic ancestors but without present

The conclusion is that mobile herders, despite the important stakes they have in
Sahelian forests, are poorly represented or totally excluded from forest
management in the majority of cases. Institutional innovations to enhance their
representation exist but are of a very recent nature. In the meanwhile, conflict
between mobile users and settled forest managers is on the rise.


The organisational characteristics of local forest management institutions seem
to be as different as their number. No two institutions appear to be identical in the
way they are organised. The hugely variable environmental, socio-economic and
political conditions in the Sahel are probably the main reason for the
heterogeneity. Furthermore, the legal definition of local management institutions
is poor in most countries.

In the Kelka, village level committees deal with rule making for their forest and for
day to day forest management. The rules vary from village to village, are
unwritten and appear to be ever changing (3). The 13 villages together have a
union of institutions called the Walde Kelka, with two representatives from each
village, which deals both with problems between villages and with problems
between villages and authorities or non-member villages. The Walde Kelka also
functions as a recourse for judgement passed by the village committee. The
ultimate recourse for anyone is the administration, but fines tend to be much
higher than under local arrangement.