The influence of vegetative cover on runoff and soil loss - a study in Mukogodo, Laikipia District

Abstract

The influence of rainfall, vegetative cover and soil characteristics on runoff and soil loss were investigated on the semi-arid area of Mukogodo, Laikipia District, Kenya. This study was carried out between July 1992 and August 1993. Two site conditions with respect to vegetation cover and two management systems were the treatments. They were designated as Perennial Enclosed (PE), Perennial Open (PO), Bare Enclosed (BE) and Bare Open (BO). Three runoff plots (each 2 by 10 metres) were designed and set up in each of the above treatments. Rainfall parameters (amount and intensity) were measured. storm kinetic energies and erosivities were determined using the formula E = 11.9 + 8.7 * log I (Wischmeier et aI, 1958). Runoff and sediment yield from each plot were measured after each rainfall event. Soil moisture was monitored in each treatment using a Neutron probe up-to 75 cm deep. Vegetative cover and soil surface condition were also monitored in all the plots. Over the study period, the Bare plots (BE and BO) lost over 50% of rainfall as runoff. The Perennial plots (PE and PO) had runoff less than half the runoff from the Bare plots. Total runoff values for BE and BO were

not significantly different at p<O.Ol, and thus closure to grazing did not reduce rainwater loss from the Bare plots. There was a significant difference (p<O.Ol) between total values of runoff for the PE and PO plots. This showed that closure to grazing reduced rainwater loss from the Perennial plots. The total values of soil loss for the Bare plots were above the acceptable limit of 10 t/ha/yr (Hudson , 1981), while the Perennial plots recorded acceptable soil loss values. Closure to grazing did not lead to an improvement in vegetative cover in the Bare plots. A fast cover recovery was observed in the Perennial plots after the short rains. High crust strength values were recorded in this study. The Bare plots had crust strength values which were 15-30% higher than the crust strength values recorded in the Perennial plots. In all treatments, available soil water was between 25% of Available Water Capacity (AWC) and 2% of AWC most of the time. As compared to the Bare plots, the Perennial plots had higher total available soil water at the beginning of each season. Soil moisture results also showed that very little water (less than 2 mm) was recorded below the 50 cm depth for all the treatments. Multiple regression analysis revealed that rainfall,

erosivity and vegetative cover accounted for 65% of the variability in rainwater loss. Similarly, rainfall, erosivity, and runoff accounted for 72% of variation in the in the soil lost. Vegetative cover significantly (p<O.05) reduced runoff, while the impact on soil loss was not statistically significant. Important implications for developing a strategy to reclaim denuded land arise from the results of this study. The results show that vegetative cover reduces runoff. It can be expected that a reduced runoff will lead to a reduced soil loss. This means that vegetative cover should be encouraged as a means to improve denuded land. From the results, zones with some cover of perennial grasses will recover by closure alone. The high runoff from the Bare areas shows that easy recovery is impossible unless some measures are taken to get water into the ground