Farmers' identification of erosion indicators and related erosion damage in the Central Highlands of Kenya

Most soil and water conservation planning approaches rely on empirical assessment methods and hardly consider farmers' knowledge of soil erosion processes. Fanners' knowledge of on-site erosion indicators could be useful in assessing the site-specific erosion risk before planning any conservation measures. The aims of this study were to identify erosion indicators based on the farmers' knowledge and assess relevance of these indicators in estimating soil erosion damage. Household, community and key informant surveys were carried out in an agriculturally high potential smallholder farming system in Central highlands of Kenya. Survey data were assessed based on consensus views, percentage frequency and descriptive statistics. Eleven erosion indicators were identified and described in local language, and their causes outlined, which closely agreed with scientific knowledge. Indicators were not only distinguished between current (splash pedestals, rills, sheetwash, sedimentation, root exposure) and past (stoniness, red soils, gullies, loose soils) but classified into erosion rates of high, moderate and low. Current and past erosion indicators were distinguished by the duration of rainfall event(s). Current indicators were perceived to be reversible in the sense that they were frequently obliterated through seasonal ploughing and weeding. They were therefore assumed to cause less soil damage thus signifying low to moderate soil loss rates. But continuous neglect of current indicators led to formation of past (irreversible) erosion indicators. Soils with such indicators were not easily restored to food crop production instead converted to other enterprises like stone crashing for construction materials and growing of mirraa (Catha edulis). The spatial distribution of the erosion indicators along the hillslope was described fitting the influence of runoff velocity with increasing gradient and slope length-steepness factors. Steep slopes were more severely damaged than the gentle and flat slopes. Farmers illustrated how relative erosion indicator weights; an index of soil damage, could be used to identify more severely eroded portions of fields or hillslopes that needed conservation attention. This approach did show that severity of erosion risk or extent of soil damage by erosion was dependent on the number and type of erosion indicators, and the total weight index of on-site erosion indicators. Consequently, widescale adoption of farmers' erosion indicators in estimating the rate of soil loss or soil erosion risk might be a rational approach by which land-users would undertake self evaluation of erosion status on their own farms. By this way, they would probably get convinced to implement conservation measures without external enforcement. (c) 2006 Elsevier B.V. All rights reserved.