Soil and water conservation project in two sites in Somalia : seventeen years later

soil and water conservation project in northwest Somalia to increase low crop yields in a cultivated watershed is assessed 17 years after the project's end. In the major intervention, about one-half of the watershed was bunded (surrounded with contoured earthen dikes) to catch spring rains. This virtually doubled grain yields in the early years, in comparison with non-bunded lands, although these increases have been reduced by about 50% over the years due to spotty maintenance by farmers over the years. Also, planners ignored the agropastoral nature of food production and the possible effects of having the best watered part of the farm permanently cultivated while the rest was given to perennial grazing by farm livestock. In addition, bunding even small farms has proved too labor-expensive to be economically justifiable for farmers. Finally, bunding had only a modest demonstration impact - this is contrast to an add-on subproject to build dams and diversion structures to provide flood-irrigation to a religious farming community. Without any systematic training inputs from the project, this community replicated the project at four sites, enabling it to satisfy its grain needs, produce a cash crop, and purchase farm machinery and chemical inputs. Lessons learned are that: (1) agropastoral subsistence depends more on the pastoral than on the agricultural base; (2) projects to enhance a productive subsector of less importance to producers should avoid dependence on substantial community inputs; (3) projects to enhance a weaker productive subsector should ensure that such will not occur at the expense of the more valuable subsector; (4) projects to enhance food security should consider all of the target area's nutritional and productive resources and problems and how these may be brought into congruence; (5) "appropriate" rural technology fits into existing social organization and does not depend on social tinkering by outsiders; (6) local communities invest themselves according to their own, not donors', priorities; and (7) the current trend towards involving rural people in research, planning, and resource allocation processes should be encouraged.