The topographic thresholds of hillslope incisions in southwestern Rwanda

Abstract

This article presents new quantitative evidence that land use in Rwanda contributes to the development of hillslope incisions. Two types of hillslope incisions can be distinguished in southern Rwanda. Incisions of the first type drain an area depending on the form and extension on the natural topography and geology. The Runyinya gully (25°) and the Rugabano soil slippage (39°) are two examples. On a logarithmic plot of critical slope inclination at the incision head versus drainage area towards the incision head, both incisions lay sensibly to the right of the Montgomery–Dietrich (M-D) envelope. The latter gives the range of these topographical thresholds for gully and mass-wasting incision in parts of North America. The Runyinya and Rugabano cases obey the linear equation: S cr =(±0.6)A −(±0.6) where S cr=critical slope gradient (tangent of slope in °) at the gully head or the scar and A=the area (ha) drained towards the incision head. Hillslope incisions of the second group rely on a run-on area larger than normal because they are localised at the ‘outlet’ of artificially runoff-collecting systems like roads, soil conservational contour trenches, tracks and other linear landscape elements. Such systems often drain a surface much larger in extension than the natural run-on area to the ‘outlet.’ These hillslope incisions, taking into account their artificially big drainage area, concentrate more or less along the line: S cr =(±0.3)A −(±0.6) This line is about in the center of the Montgomery–Dietrich envelope. If, however, only the natural drainage area of these ‘outlet’ incisions is taken into account, all points fall close to the left border or even to the left of the Montgomery–Dietrich envelope. This indicates a much higher probability for incision in those localities receiving supplementary runoff or interflow from outside the natural drainage area. In the case of a soil slippage at Rwaza Hill, detailed stability calculations show that the slope failure should be due to excessive water infiltration into the bottom of a trench. The digging of the trench provoked an increase in the area drained to the slippage head by a factor of 6. The phenomenon of ‘forward’ erosion is compatible with the existence of threshold combinations of slope and drained area. For slopes steeper than 7–8°, the phase of regressive erosion does often follow the forward incision event with a delay of several years or more. Finally, the scanty data set now available for Rwanda suggests that the drainage area critical to hillslope incision on the red-brown ferrallitic soils in Rwanda might be nearly twice as big as those in North America.