Influence of slope and stone cover on tillage operations in the Ethiopian highlands

Human population pressure has forced smallholder farmers in some areas of the central plateau region of the Ethiopian highlands to cultivate plots located on slopes or hillsides having varying amounts of stone cover. The objectives of this pilot study were to determine if slope and(or) stone cover influenced tillage operations on smallholder farms in the Debre Birhan area of the Ethiopian highlands, and to examine the relationship between different soil/tillage-related factors and draft of the indigenous single-tined ard (maresha). This paper also briefly describes the traditional systems of cropping and soil management, and provides data on selected physical and chemical properties of soils found in the study area. A land classification (LC) system was developed to categorize plots according to the slope and stone cover. Surface stone cover of plots was calculated and a method of estimating subsurface stones was tested. Tillage measurements included plowing frequency or number (PN) per area per annum, soil moisture tension, plowing depth, furrow “wedge” of soil moved, area plowed, contour angle, and the speed of travel and pulling force exerted by oxen. Plots were located on slopes ranging from 0% to 23%. Degree of slope was a key factor in determining the acuteness of angle of the contour at which plowing is done prior to seed covering, but had no significant effect on maresha draft. Stone cover of plots ranged up to 27%, with individual stones moved by the ard weighing up to 2.5 kg. Plowing depth ranged from 10.1–15.3 cm for all PN across LC. Oxen travelled at speeds of 0.35–0.58 m s^?1. Mean values for draft of the maresha across LC ranged from 0.83–1.04 kN. Farmers plowed from 0.10–0.15 ha during a 5–6 h work period, irrespective of LC or PN. Soil moisture tension (range 16–33 kPa) was found not to have a significant effect on draft, due to tillage being carried out prior to the major and minor rains. Percent stone cover explained very little of the variation in implement draft (R²=0.10). Adding stone weight to the model increased R² slightly (0.22), but significance was low (P<0.16). Furrow wedge of soil displaced by the ard head accounted for most of the variation in draft of the maresha across all LC and PN (R²=0.75, P<0.008). The design and construction of the maresha allows it to be used equally well for tillage operations on plots having minimal or high stone cover which may be located on flat or gentle sloping land, or on steep hillsides.