Skip to main content
ARS Home » Research » Publications at this Location » Publication #285406
Title: A modeling approach to evaluate the impact of conservation practices on runoff and sediments in Sasumua watershed, Kenya

Author
item MWANGI, JOHN - Kenyatta University
item SHISANYA, CHRIS - Kenyatta University
item GATHENYA, JOHN - Jomo Kenyatta University
item NAMIREMBE, S - World Agriculture
item Moriasi, Daniel

Submitted to: Journal of Soil and Water Conservation
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/19/2014
Publication Date: 3/1/2015
Citation: Mwangi, J., Shisanya, C., Gathenya, J., Namirembe, S., Moriasi, D.N. 2015. A modeling approach to evaluate the impact of conservation practices on runoff and sediments in Sasumua watershed, Kenya. Journal of Soil and Water Conservation. 70(2):75-90.

Interpretive Summary: Proper watershed management is essential for the enjoyment of beneficial ecosystems services such as production of food, timber, fuel wood and fresh water and non-tangible benefits such as flood regulation, sediment retention and water purification. Conservation practices (CPs) are often used to prevent or minimize pollution from non-point sources within watersheds. However, it is important that detailed analyses be carried out in order to determine the most cost-effective CPs and implementation locations within the study watershed. The capacity to regulate flow and maintain water quality in Sasumua watershed, which is the source of 20% water supply to Nairobi city, Kenya, is affected by land degradation. In this study, the Soil and Water Assessment Tool (SWAT) model was used to evaluate the impact of agricultural conservation practices on runoff and sediments in the Sasumua watershed. SWAT was calibrated and validated for streamflow at the watershed outlet. Annual average erosion rates for cropped area and the sediment yields from previous studies in the study area were used to calibrate the model for sediment. The model performance was assessed using the coefficient of determination (R2), Nash-Sutcliffe efficiency coefficient (ENS) and percent bias (PBIAS) statistics on a monthly time-step in addition to time-series graphs. The calibrated and validated model was used to simulate sediment yield at the sub-basin level. The effects of filter strips, contour farming, parallel terraces and grassed waterways and their combinations on surface and sediments were simulated by adjusting the relevant model parameters. The model yielded monthly streamflow R2 values of 0.80 and 0.85, ENS values of 0.74 and 0.81, and PBIAS values of ±5% and ±6% during the calibration (1989-1992) and validation (1993-1996) period, respectively. The simulated annual average erosion rate for cropped areas and sediment yield were 18.4 tons ha-1 (7.4 tons ac-1) and 9.0 tons ha-1 (3.6 tons ac-1), respectively and within those reported in previous studies. Parallel terraces and contour farming reduced surface runoff by 20%, and 12%, respectively. A combination of 30 m wide filter strips and grassed waterway reduced sediments into the watershed outlet reservoir by 80%; parallel terraces, 10 m filter strips and grassed waterway by 75%; 10 m filter strips and grassed waterway by 73%; contour farming and grassed waterway by 66%; and grassed waterway only by 54%. These results will be used to devise future ecologically-sound Sasumua watershed development plans.

Technical Abstract: Degradation of agricultural watersheds often reduces their capacity to provide vital environmental services such as food production, clean potable water, water bodies for recreation and generation of hydro-electric power. Soil and water conservation practices on agricultural lands can enhance the capacity of watersheds to produce these services. The objective of this study was to evaluate the impact of agricultural conservation practices on runoff and sediments using the Soil and Water Assessment Tool (SWAT), a distributed hydrologic model. The study area was Sasumua watershed (107 km2; 43 mi2), source to 20% of water supply to Nairobi city, Kenya where land degradation has affected its capacity to regulate flow and maintain water quality. SWAT was calibrated and validated for streamflow at the watershed outlet. Annual average erosion rates for cropped area and the sediment yields from previous studies in the study area were used to calibrate the model for sediment. The model performance was assessed using the coefficient of determination (R2), Nash-Sutcliffe efficiency coefficient (ENS) and percent bias (PBIAS) statistics on a monthly time-step in addition to time-series graphs. The calibrated and validated model was used to simulate sub-basin sediment yields and sediment loads into outlet reservoir over the period 1970-2010. The effects of vegetative filter strips (VFS), contour farming (CF), parallel terraces (PT) and grassed waterways (GWW) and their combinations on surface runoff and sediments were simulated by adjusting the relevant model parameters. The model yielded monthly streamflow R2 values of 0.80 and 0.85, ENS values of 0.74 and 0.81, and PBIAS values of ±5% and ±6% during the calibration (1989-1992) and validation (1993-1996) period, respectively. The simulated annual average erosion rate for cropped areas and sediment yield were 18.4 tons ha-1 (7.4 tons ac-1) and 9.0 tons ha-1 (3.6 tons ac-1), respectively and within those reported in the previous studies. A 30 m VFS +GWW, PT + 10 m VFS + GWW, 10 m VFS +GWW, CF + GWW, and GWW reduced sediment loads into watershed outlet reservoir by 80%, 75%, 73%, 66%, and 54%, respectively. Implementation of specific conservation measures had minimal effect on surface runoff. The results give an insight into the hydrological implications of the present landuse management that can be used to devise future ecologically-sound Sasumua watershed development plans.