GIS and RRA in Local -Level Land Use Planning A case study in Sri Lanka

GIS and RRA in Local -Level Land Use Planning A case study in Sri Lanka

M.Mari, P.Bitter
Dept. of Geography
University of Zurich-Irchel
Switzerland

Abstract
Land use planning, here defined as the process of identifying the optimal and sustainable use for a given piece of land under given environmental, social and economic conditions, has been one of the key issues in rural development for many year. Two main schools of thinking on how this process should be accomplished can be found today: The 'pechnocaratic school' argues that modern (western) science is needed to do the job, whereas the 'participatory school' argues that he farmers know the problems and their solutions best and, therefore, land use planners should ask them. There have been arguments between the two schools for years; however, very little research has been done to actually compare the results of the two methodologies. The authors present such a case study from Sri Lanka and outline potential links to combine the strengths of both approaches.

1. Introduction

1.1 The GIS approach
Agricultural land use planning activities are usually started with the objectives to increase productivity,and to improve the sustainability, i.e. to conserve the natural resources. There are various scientific approaches to the goal:

Land capability evaluation, e.g. according to FAO guidelines
Process and simulation models
Process and simulation models
Parametric indices, e.g. the Storie Index

The point here is not discus the advantages and disadvantages of those approaches. We would like to draw the attention on two important characteristics which they all have in common:

They combine a (variable) amount of information on natural resources and conditions (soils, rainfall, terrain, etc.).This information is typically represented in map form, i.e. it is spatial data.
They attempt to utilise knowledge on the relations between environmental conditions, land management practice, and reaction of the land to that practice. This knowledge can stem from experience (trial and error), from experts, or other sources. The crucial point is that this knowledge has been generated elsewhere. The 'scientific' land use planning then tries to transfer this exogenous knowledge by analogy to the area in question.

Since spatial data is used and extrapolation of knowledge is attempted, it seems quite obvious to use a GIS to help in the process. In Sri Lanka, the probably first attempts at using computers in land use planning have been made more than 10 years ago [DENT 1986]. Since then, further attempts with better software and more powerful computers have been launched by various organizations and projects on various scales (loal / District / National); some met with more and some with less success (see for instance DENT 1993, p. 18], [TURNER], [IIMI]. Indeed, the idea to be able to process huge amounts of data, to simulate alternative land use options, and eventually to determine the best land use at the push of a button hasn't lost it's fascination.

The application of these scientific methods has also drawn quite a bit of criticism. We would like to outline two major challenges that are raised:

Two data which is actually available, especially in developing countries, rarely meets the requirements of the models that are being used. Deficiencies can be the total absence of certain data (e.g. soil maps, climatic data), non-comparable classification schemes (also often the case with soil maps), inadequate scales, or insufficient geometric and thematic accuracy of the maps [dent 1993]. Thus statements on crop suitabilities and sustainability are made which are not justifiable even by the models. Additionally, the models often tend to focus on 'hard data' on the physical environment (soils, climate, topography), but have no established place for data on socio-economic features (such as access to inputs and markets, availability of labour, land title security, cultural preferences) and for local knowledge.
the second line of criticism argues that the use of computers re-inforces tendencies of top-down planning. Scientific land use planning has been a largely centralized, sectoral, and top-down affair even before the advent of computers [DENT 1993, P.5]. The interests, needs, and also the knowledge of the farmers as actual land users have been completely neglected. The philosophical reason for this may be that the central state with its institutions is perceived as the only grantor of development, and consequently, also as the custodian of the natural resources. The practical reason is that the application of scientific methods requires scientifically qualified staff which is available only at central levels.

However, the capacities of any Government for planning and, even more severely, implementation of the plans, are limited. Thus land use plans have often remained a piece of paper, if they came to that stage al all. If can be assuemed that the introduction of computers raises new hopes in Government institutions that they can now perform their Herculean task without changing the basic concept. Moreover, the operation of computers requires even more specialized staff which can be found only in central places, and the communication gap between planners and land users widens once again.

1.2 The participatory approach
Participatory planning has been promoted as an alternative to top-down land use planning. There are various degrees of participation; we understand it here as an active involvement of the land users in the identification of needs, planning, and implementation. The fundamental belief of participation advocates is that the farmers know their situation best and thus should play a key role in land use planning. Another objective, sometimes also on the hidden agenda of donors, is the objective of policically empowering local communities through participatory planning.

PRA /PRA
RRA is a set of methods to extensively understand problematic situations in certain areas and to provide foundations for development planning. The starting point is the knowledge of the local people. The aim of RRA is the creation of a framework for communication and learning. Some important tools in RRAs are sketch mapping, transect walking, listing and ranking of properties, group discussions, semi - structured interviews, analysis of seasonal and historical trends. While RRA initially developed mainly as an alternative to information collection by questionnaire surveys, PRA also leaves the planning responsibility. I.e. the decision-making, to the villagers. Outsiders are supported to act as catalysts [CHAMBERS].

The main criticisms on participatory planning include the following points:

There is a risk of de-linking local from regional and national planning. The integration of local plans within a wider framework (e.g. agricultural development policies, subsidy schemes, conservation plans) is technically and institutionally plans) is technically and institutionally difficult, and, ultimately, local development efforts may collapse through lack of higher level support [DENT 1993, p. 104]. The underlying problem is that local interests are often different from regional / national interests; a problem which is known in every society. What further compounds the problem in the contest of development cooperation is the lack of institutionalized platform for negotiating between the diverging interests. If outsiders unilaterally focus on the local development interests and leave after their project ends, there will be no base for continuation.
The results of RRA / PRA sessions often resemble 'shopping lists; of individuals or groups rather than strategic plans for a sustainable development of potentials. People tend to expect instant solutions to their immediate problems, if outsiders, supposedly with funding enter the village, Accordingly, they will tell what they suppose the outsiders expect to hear [LEACH]

2. The case study in Weeraketya

2.1 land use planning with GIS
The Sri Lanka Land Use Policy Planning Division [LUPPD] issued guidelines for the preparation of land use plans for Divisional Secretary Divisions, which are units of decentralized administration in Sri Lanka. The guidelines are similar to the FAO guidelines for land use planning. Based on these guidelines, we implemented a model for land ue planning on commercial GIS software (Arc/Info).

The intermediate results of this model GIS are planning maps for the various produce, e.g. coconut, paddy etc. They consist of two parts: the present and the calculated, ideal area of cultivation for each crop. To achieve this objective, one has to assess certain variables which restrict the cultivation of plants, e.g. water availability, nutrition, temperature regime, erosion hazard, ease of land use. The classification and combination of these restriction or suitability variables leads to suitability maps and to the ideal habitats of the crops (fig. 1).

Figure 1 GIS land use planning model
The first step in the GIS -model is the production of the base converages of different suitability variables L range, slope, soil rockiness, rainfall annural mean and rainfall with 75 per cent expectancy. The next step is to classify the variables for each crop separately into most (S1), moderately (S2), marginally (S3), and not suitable (S4). This classification, which we termed 'crop suitability coverage' in fig. 1, is based on [DENT 1985].

The final step is the combination of above factor-specific crop suitability coverages into an overall suitability which expresses the ideal area of cultivation of a particular cop. Another combination with the present area of this crop will result in the crop planning maps, e.g. for coconut (see fig. 2); or paddy etc.

Figure 2 Planning map for coconut
The result shows a relatively small area (dark grey) where the cultivation of coconut is ideal and actually being practiced. Large areas (light grey), principally in the north, are suitable but not is use. Mainly in the south coconut is growing n many places (middle grey) although it is unfavorable.

2.2 Main results of RRA in villages
The aim of the RRAs in two villages in Weeraketya in southern Sri Lanka was to collect information from peasant about their strategies and reasons for a certain land usage and land use planning and thus to cross-check the results of the GIS application. The perspective of the land users is in the center. However, since we didn't have the possibilities to implement any plans, we tried to avoid disappointment on the side of the farmers by not going to the extent of actually planning for land use changes.

We investigated four main areas:

Present land use Current land use
Cultivation criteria and -decisions
Crop requirements
Agricultural input
Importance of crops for household
Land tenure Seasonal differences Seasonal land use
Task distribution
Other incomes Land use changes Recent and historical changes
Changes in land tenure
Cultivation practices
Influence on land use
Effect, impact of the changes Wishes for future land use Problems, needs, expectations
Alternatives
Realization, implementation One of the RRA results are sketch maps of village areas that were drawn by the local people in the first evening meeting. Besides getting to know each other and forming faith between locals and outsiders, the most important geographical information and the local problems of land use were discussed in this way. Other RRA activities were interview with caretakers, peasants, big and interviews with caretakers, peasant, big and owners, smalholders as well as transect walks and group discussions.

Another outcome are the requirements of several plants, e.g. coconut, banana pepper, teak in relation to the suitability variables of the GIS model (see chapter 2.1). By means of listing and ranking, local people express their knowledge about physical needs of the plants, the relative important of certain requirements, advantages and disadvantages of different soil types for certain crops.

Further results are various sketch maps from peasants that show he personal property with the present land use or plans for future land use

Finally, a lot of information about he socio-eco-nomic and ecological situations in these communities has been acquired.

3. Potential links be between the two approaches

3.1 Complementing GIS with RRA
Verification and completion of based data The coconut planning map (see fig. 2) shows large areas where the cultivation of this crop would be ideal but is ideal but is apparently not being practiced. Most of those reqions are mapped as homesteads. The RRA fieldwork has revealed that that in fact a lot of coconut cultivation is taking place in these regions. However, these are "home gardens' whereas the class "coconut" in the land use map refers to estates only.

Such information could be documented by photographs and stored in the GIS. Other results from RRA sessions (sketch maps, problematic issues) can be photographed and included in the GIS to complete the picture of natural environment and society.

Adjustment of models
In some places coconut is grown although it appears unfavorable. Based on discussions on crop requirements and other RRA activities, the GIS model, especially the weighting of suitability variables, was adjusted, differentiated and complemented. The result of the revised planning map for coconut presents a more realistic picture of the situation. The unsuitable areas are below the 1300 mm "75% rainfall level". This corresponds with the experiences from the RRA fieldwork: rainfall is the most important constraint.

Gaining insight into functional relationships
The GIS model considers physical factors only and offers no explanation for the present land use situation. The relationship between environment, social and economic conditions of the land users are certainly an important dimension in land use planning. We found that RRAs are an excellent method to provide insight into the way of thinking and reasons for acting of different stakeholders.

Detail planning
Our GIS model is essentially based on data at a scale of 1:50'000. This is already better than can be expected in many cases, but still too small for village level planning. Thus GIS models can reasonable be used at higher planning levels to define policies and broad structure plans. On the other hand, PRA could be used to "put the meat on the bones", that is, to work out detail plans at village level.

3.2 Complemening PRA with GIS?

Documentation of indigenous knowledge
The potential links between both approaches in the sense of a completion of RRA through GIS can be found at different stages of the bottom-up land use planning process.

The basis of this planning is the indigenous knowledge of the target group. The reception, discussion and assessment of this knowledge are essential steps of bottom-up planning. GIS could be used to systematically document and integrate such knowledge, e.g. local classification and usage systems for soils [MULLER-BOKER], [BLIEK]. However this raises the question who will benefit from such a collection.

It is also thinkable to try to use GIS to homogenize information on indigenous knowledge, i.e. to bring to into comparable formats, or to interpolate results from spatially dispersed RRA's. However this approach is an intrinsic contradiction to the concept of indigenous knowledge, which is essentially local in nature.

Provision of additional information into PRA
The combination of indigenous and scientific knowledge promises a greater success in hand use planning. GIS with their analysis, modeling and visualization tools can bring scientific knowledge into participatory local planning exercises. This could be information on p potentials for non-agricultural development, distributions of services, changes and trends etc. Moreover, the excellent possibilities of GIS to create customized visuals can be used to communicate villages relevant information on regional conditions, structure plans or scenarios. Such visuals could easilty be provided by Government officials as well as NGO professionals-provided they get access to the necessary data. Thus, communication between the planning levels can be improved, and the risk of detaching local from regional / national planning which is inherent in PAN can be minimized.

4. Conclusions
we compared two approaches to local 0level and use planning : the 'technocratic' approach which applies 'modern' science and GIS v. the participatory approach which is based on traditional indigenous knowledge. Although there are certain shortcomings in the present study (e.g. we did not consider the implementation of land use plans, and 'local level' was interpreted differently in the two cases), we would like to draw some preliminary conclusions on the methodologies.

After some field checking, the GIS model meets the ground situation surprisingly well
With the GIS model it was possible to prepare an indicative land use map for a relatively large area with very little effort (not accounting for the collection of base data). However, the scale of available data is too small for land use planning on the base of single plots.
From the first two conclusions it follows that GIS can be a suitable tool in the definition of policies and broad structure plans at a regional level.
Applying PRA for land use planning implicitly assumes the existence of a village community what a mandate for land use planning. However, we found no such communities, farmers act largely as individuals.
RRA/ PRA appears to be a very effective method to build awareness on both sides and to improve communication between locals and outsiders, i.e. farmers and professional planners.
Therefore we conclude that GIS and RRA / PRA are complementary rather than alternative methods. They both have their role to play at different stages and different levels of land use planning. The debate between paradigms of planning, and not between methods. The underlying question is L Who controls the land resources - central Government, local communities, or individuals - and who is in charge of land use planning ? this is a political question, not a methodical one.
We identified a few potential entry points to link the two methodologies more closely. Whether a better integration of methods will help to bring about a resolution of underlying political conflicts has yet to be seen. It would be interesting to undertake more comprehensive research in this direction.

Acknowledgments
The authors wish is to thank the people of Kuda Bedigama and Muruthawela, the Sri Lanka Land use Policy Planning Division, the Center for Remote Sensing at the Survey Department, and Mrs. Julie van der Bliek and Mr. Mahinda Seneratne for their support in this research.

References

BLIEK J. van der (1995) : Land Planning : land users' Perspective, Summary Information Kuda Bedigama Colombo, (unpublished)
CHAMBERS, R (1992): Rural Apprasisal : Rapid, Relaxed and participatory Discussion Paper 311, Institute of Development Studies, University of Sussex, UK
DENT, D L & R B Ridgway (1986): Land use planning handbook for Sri Lnaka UDP / Ministryof Lands and Land Development, Colombo
DENT, D L & K P Goonewrdene (1993): Resource Assessment and land use Planning in Sri Lanka : A Case Study International Institute for Environment and Development, London
IIMI, International Irrigation management institute (1994): GIS for Watershed Resource Planning SCOR Monitor Vo; 1 / No 2 &3
MULLER - BOKE, U (1995): Ethnookologie. Ein Beitraz zur geographischen Entwicklungs forchung (in German) Geographische Rundschau Vol. 47 No. 6: 375-
LEACH, M& J Fairhead (1994)L Natural Resource Management : The production land use of Environment Misinformation in Guinea's Forest-Savannah Tranistion Zone IDS Bulletin Vol. 25 No 2, Institute of Development Studies, University of Sussex, UK
LUPPD (1993) : Guidelines for the preparation of land use plans for Divisional Secretary Division Ministry of Lands, Colombo
TURNER, S T D & R White (1994) : Geographical Information Systems for natural resource management in South East Asia Environment and Forest Conservation Division, Mahaweli Authority of Sir Lanka