Does the effectiveness of soil and water conservation techniques in reducing runoff and soil loss change over time?: a case study from the semi-arid Ethiopian highlands

Soil erosion by water is the most critical problem to be solved for a sustainable agricultural development in the north Ethiopian highlands. In Tigray region soil and water conservation (SWC) programs have started two decades ago under the umbrella of food for work program. Several studies were then conducted and these demonstrated that SWC reduces soil erosion. However the effectiveness and temporal evolution of the effectiveness of major SWC techniques are not quantified. The aim of this study is therefore, to better understand the temporal evolution of the effectiveness of SWC techniques on runoff and soil loss when installed on different land use types and slope gradients. Eight runoffmeasuring sites were selected at May Leiba catchment in the central zone of Tigray for the two major land use types: cropland and rangeland. For each land use type, three sites were identified based on slope gradients, i.e. gentle (5%), medium (12%) and steep (16%). Plot-measuring sites were replicated for the medium slope on both land use types. For each site on rangeland four runoff plots were installed: one control plot and three plots treated with stone bunds, trenches and stone bunds with trenches. Similarly, for each site on cropland three runoff plots were installed: one control plot and two plots treated with stone bunds and stone bunds with trenches. The results showed that seasonal runoff coefficient (RCs) and seasonal soil loss (SLs) were the highest on rangeland compared to cropland. RCs on rangeland were 2 to 2.5 times higher compared to RCs on cropland, whereas SLs were 3 to 5 times larger on rangeland compared to SLs on cropland. This difference is due to soil tillage on cropland, which creates ridges and increases surface roughness, and to vegetation cover. With increasing slope gradient, RCs and SLs tend to decrease, which is attributed to increased rock fragment cover. Introduction of SWCT leads to a 22 to 84% and a 43 to 90% reduction of RCs and SLs respectively relative to a control plot on both land use types. Over a three years period the change in effectiveness of SWC techniques with respect RCs and SLs on cropland were insignificant. On rangeland, however, the effectiveness of trenches and stone bunds with trenches declined rapidly due to infilling with sediments which reduces their storage capacity. Effectiveness of trenches declined from 74% to 28% and that of stone bunds with trenches from 77% to 55% for RCs during the three years study period. During the same period, mean SLs increased from 4 to 20 ton ha-1 for trenches and from 1.3 to 7 ton ha-1 for stone bunds with trenches. Therefore, regular maintenance of the SWC structures is important to preserve their effectiveness in reducing runoff and soil loss