The rates of soil erosion and land degradation in Ethiopia are frighteningly high. Crop production, livestock keeping and energy supply situations are at risk. The highlands are the most affected. Past rehabilitation efforts have been immense. Much labour, capital and trained staff have been mobilized to correct the situation, but the outcome has not been encouraging. There are a number of reasons for the failure. Methodical and technological problems are evident. Exclusion of farmers and their indigenous knowledge at all levels of planning and implementation, the use of uniform and 'foreign' soil conservation and reforestation technologies, mistrust between farmers and facilitators, farmers' bias to production over conservation, miss-use of food-for-work programmes in conservation works, lack of conducive land tenure and tree usufruct have all contributed.
Success in the effort calls for construction of an approach by which the traditional soil and water conservation and agroforestry knowledge of farmers can be studied, adapted and used. Therefore, farmers themselves were necessarily involved in the study, adaptation, implementation and evaluation of the rehabilitation work. The conceptual framework and the research questions were designed to reflect these issues and concerns.
The research has been farmer-participatory. It is an action research, which is conducted both at community/catchment-level and at a household/farm-level. Farmerparticipatory trust building, socioeconomic diagnosis, enviromnental assessment, planning, implementation and evaluation methods are researched. The research is administered in a soft-system approach. The outcome of the research is a participatory agroforestry approach by which soil conservation is benefiting. It is composed of 'six sub-processes' that each are inter-linked in a logical order.
· The first sub-process deals with methods of approaching the farmer(s) and finding appropriate extension methods that can secure genuine trust and acceptance of development facilitators. Indigenous means of approaching farmers and trust building are devised.
· The second sub-process deals with a comprehensive social diagnostic approach. Appropriate methods of development extension, identification of production desires, social limitations and production potentials, indigenous land husbandry technologies are studied and identified.
· The third sub-process deals with methods of environmental assessment. The application of GIS output maps for synthesizing the information and enhancing the participatory research work on site diagnosis and relative land potential assessments is presented.
· The fourth sub-process focuses on methods of reconciling the findings in the human sector with the site factor. A methodology by which traditional soil and water conservation and reforestation skills, production desires and targets of farmers are understood and used in planning is devised. It involves preparing prescriptions for each of the planning units developed in the sub-process.
· The fifth sub-process focuses on adapting catchment level plans and prescriptions to farm and farmer-level situations, technology appropriations and implementations. A methodology by which these activities of the sub-process are conducted is devised.
· The six sub-process dwells on conducting tests on sustaining land quality, cost effectiveness and adoption ease of the implemented agroforestry development options at farm level. A methodology by which the participatory evaluation can be conducted is developed.
Each of the six sub-processes of the approach are constructed and tested under Tikurso catchment conditions. In the process, the researching steps that were initially conceptualized in their abstract form are transformed into more illustrated sub-processes. As a result, an agroforestry approach whose nested sub-processes are connected to form a multi-loop approach is evolved. The constructed approach assumes that circumstances change in either the social sector or the enviromnental setting or both through time. For generation and usage of this expectedly new information, the approach has three alternate avenues. Choice among the alternative avenues is made depending on the knowledge-gap identified during the evaluation and feedback sub-process of the approach. The results from land sustaining quality, cost-benefit analysis and adaptability tests confirm that the approach has significant benefits to soil and water conservation.
The evolved sub-processes indicate that the overall approach is nested in that the various discrete data and information generated in the sub-processes are hierarchical and built one in congruence with the other. The subsequent sub-processes make use of the conclusions and are guided by the information obtained in the preceding exercises. The approach is further characterized by an intimate bondage of the farmers' knowledge with the facilitator's knowledge.
In addition to its methodical aspect, construction of agroforestry and soil and water conservation intervention technologies for adoption by farmers is contained within the approach itself It is further realized that locally understood land quality grading variables can be defined, adapted and used for determining agroforestry and soil and water conservation planning units in broad and detailed intervention categories. For the moist Weyna-dega agro-climatic zone situation, eleven agroforestry and soil and water conservation intervention categories are defined. Eight of them occur within the adapted cropping limit while the remaining three are devised for those lands that are out of the cropping limit. For each of the intervention categories, agroforestry and soil and water conservation components are defined and implementation issues are prescribed. Catchment level interventions and prescriptions serve as a guide to farm-level agroforestry and soil and water conservation planning and implementation undertakings. The development of the eleven interventions has resulted in realization of agroforestry as a land use option by which soil and water conservation measures are combined with woody perennials and non-woody components on the same piece of land simultaneously. In this regard, PAA is an approach that contributes to remedying the methodical and technological shortcomings of land rehabilitation in Ethiopia.
The conditions for its application on large-scale are studied by considering 38 peasant associations from four administrative sub-districts. Site reconnaissance, individual and group interviews with the farmers and the government authorities concerned as well as literature search were the methods of the study. The study shows that the conditions are partly non-conducive. The recommendations indicate the actions that need to be taken for improving applicability of the approach itself and the conditions for its application on largescale.