Infiltration and surface runoff responses To land use and tillage in Awach Kano Watershed, Lake Victoria Basin.

Abstract

Understanding of runoff sources and sinks coupled with development of rubost models that provide rapid and less expensive assessment of the impacts of land management on surface runoff is critical in the management of soil and water resources. In this study, a stratified sampling procedure was used to set up 51 runoff plots on representative land use types on the three slope zones in the landscape. On each slope zone a completely randomized design was utilized. Runoff measurements were done for 120 rainfall events involving two rain seasons. Three rain gauges with one gauge on each slope zone were used for rainfall measurements. Infiltration measurements were also conducted on all the land use types using two single ring infiltrometers. Topsoil samples (20 em deep) were collected from each runoff plot for laboratory soil spectral analysis. One way analysis of variance was utilized to investigate the effects of land use types on surface runoff in each slope zone. Calibration of soil spectral reflectance to mean seasonal runoff was done using partial least squares (PLS) technique. Results of the study showed that land use significantly (p=0.001) influenced runoff generation in on all slope zones. There were differences in seasonal runoff generation patterns in the lowlands due to the presence of saturated overland flow in the long rainy season. Results show that runoff generation mechanisms significantly (p=0.001) influenced surface runoff in the lowlands in the long rainy season. There were no statistical differences at 5% level of significance among tillage treatments in the uplands and midlands slope zones. However, runoff from conventionally tilled areas showed the highest variation (Cv=109%). There were good correlations (R2>0.70) between surface runoff and soil spectral reflectance for both rain seasons. However, the model validation of the relationship for short rains, 2001 had a coefficient of determination ofr2 = 0.54. x