Serge Guillobez

The changing distribution of two riverine tsetse flies over 15 years in an increasingly cultivated area of Burkina Faso

Changes in the distribution of two riverine tsetse flies, Glossina tachinoides Westwood and Glossina palpalis gambiensis Vanderplank are described in an agro-pastoral area of Burkina Faso subject to increasing human population pressure and land use change. Two similar entomological surveys (one trap every 100 m, 120 km of river) were conducted in 1981 and 1996. Changes in tsetse distribution were compared to land use changes through high resolution remote sensing imagery (LANDSAT, SPOT). There was a close relationship between proximity of crops relative to riverine forest and the density of Glossina. Where fields encroached on riverine vegetation, tsetse populations declined. Where the geomorphological structure was not well suited to agricultural activity, riverine vegetation and tsetse fly populations were relatively unaffected, even with intense agricultural activity nearby. In contrast, increased human activity and higher cattle densities in the surrounding savannah areas were associated with increased tsetse numbers. The results demonstrated a wide diversity of tsetse distribution and habitat within a few kilometres in an agro-pastoral landscape in West Africa.

Runoff and sorghum performance as affected by the spacing of stone lines in the semiarid Sahelian zone.

Crop yields are primarily water-limited in dryland production systems in semiarid regions. This study was conducted in a catchment located in the ‘‘plateau central’’ of Burkina Faso to assess the impact of the space between stone lines on runoff and crop performance. The experimental design consisted of four plots in which stone lines were installed. The spacing between the lines was 100 m in the first plot, 50 m in the second, 33 m in the third, and 25 m in the last plot. The soil was a Ferric lixisol and the slope, which is characteristic of the area, was about 1‐3%. Subplots placed at regular and fixed distances from the lines were used to monitor soil water content and crop yield. Runoff from all plots was measured using a water discharge recorder. It was found that 31% of rainfall was lost through runoff in plots without stone lines. The efficiency of stone lines in checking runoff and in improving soil water storage increased with reduced stone line spacing (runoff was reduced by an average of 5% on plots where the space between the lines was 33 m, but was reduced by 23% when the stone line spacing was 25 m). Soil water content decreased with increasing distance from the stone line. Sorghum (Sorghum bicolor (L.) Moench) performance was greatly affected by stone line and plant straw and grain yield were doubled in plots with stone lines compared with those of plots without stone lines. At an area of about 6 m from the stone lines (upslope), where organomineral sediments were collected, sorghum grain yields were 60% greater than that obtained at 19 m from stone lines. The stone line technique seems to be a sound option to mitigate water stress during dry spells. # 2000 Published by Elsevier Science B.V.

Effect of stone lines on soil chemical characteristics under continuous sorghum cropping in semiarid Burkina Faso

In the semiarid Sahel, farmers commonly lay stone lines in fields to disperse runoff. This study was conducted in northern Burkina Faso to assess the chemical fertility of soil under a permanent, non-fertilised sorghum crop, which is the main production system in this area, 5 years after laying stone lines. The experimental design consisted of four plots in which stone lines had been laid. The spacing between the lines was 100 m in the first plot, 50 m in the second, 33 m in the third, and 25 m in the fourth. To determine soil chemical characteristics in relation to the stone line spacing patterns studied, soil samples were taken from subplots at regular and fixed distances from the lines at the start of the trial and then 5 years later. Under the continuous non-fertilised sorghum cropping system, the beneficial effect of stone lines on soil fertility was limited. Five years after installing stone lines, soil organic C, total N, available P and Na concentrations and soil pH had decreased. Within the plots, these same variables were higher upslope than downslope of stone lines, probably because of water storage and sediment accumulation in front of the stone lines. In plots where stone lines were relatively close together (<33 m) the decrease in soil fertility was less. It is concluded that in Sahelian zones, stone lines alone are not sufficient to insure the conservation of soil fertility. Combining soil and water conservation techniques with soil fertility management practices are needed to sustain soil productivity.

Regionalized soil roughness indices

The soil roughness at the field level is an easy visually perceptible notion, but difficult to describe numerically. The objective of this paper is to propose a method to establish quantitative and descriptive soil roughness indices. A measurement system was developed using a laptop-computer and a laser cell. The elevation data are measured on the ground along a square array. They are treated with geostatistics methods. For a given direction, the semi-variogram shape, depending on the distance, is always comparable with ground relief, and reflects in a positive way the variations of this relief. New quantitative indices are proposed. They are derived from semi-variogram and they characterize the roughness in both vertical and horizontal planes. For the five plots the main results are: on crusted soil without soil tillage, the semi-variogram is nearly flat in isohypsic direction and has a parabolic shape in the slope direction. On an isohypsic bank the semi-variogram presents a shape with a hole effect upright to the bank. For an oriented knife soil tillage, the semi-variogram is bimodal, with a hollow effect in the axis perpendicular to the vine-rows. With a not clearly oriented soil tillage, the semi-variograms in all directions have a practically similar feature with a sill. The semi-variogram corresponding to the direction transversal to a gutter has a bell-slope.

Regionalised soil roughness indexes

The soil roughness at the field level is an easy visually perceptible notion, but difficult to describe numerically. A measure chain was developed around a laptop-computer laser cell. The elevation data are measured on the ground with a square grid, they are regionalised variables, treated with geostatistics methods. For a given direction, the semi-variogram shape, depending on the distance, is always comparable with ground relief. This shape reflects in a positive way, the variations of this relief. New quantitative indexes are proposed. They are derived from semivariogram and they characterize the roughness vertically as well as horizontally.<<ETX>>