GRAZING INTENSITY AND ECOLOGICAL CHANGE IN EASTERN SENEGAL: IMPLICATIONS FOR THE MONITORING OF SAHELIAN RANGELANDS.

The objectives of this dissertation were to identify soil and vegetation changes due to grazing by livestock and to develop guidelines for monitoring the impact of livestock in the Sahelian zone of Africa. The study was conducted in eastern Senegal during 1983 to 1985. The basic methodology was a study of the process of degradation (retrogression) through the collection of data on soils and vegetation along a gradient of increasing livestock pressure. Two gradients were selected along piospheres. Each gradient was placed on a range site that had homogeneous state factors (climate, geology, and rainfall). The results show that the changes in the soil were confined to the top soil horizon. There was a negative correlation between livestock pressure and factors such as infiltration, percent nitrogen, and percent carbon, and a positive correlation between livestock pressure and the factors of bulk density, and phosphorous content. Changes in the vegetation community were dependent on the range site. On the sandy site, there was a linear, negative correlation between livestock pressure and the factors of plant density, plant cover and biomass. On the loamy range site, the changes in these factors along the gradient had a 2nd or 3rd order polynomial relationship. The effects of bush fires and short term droughts on the vegetation were also studied. Drought, in particular, has an impact that is similar to that of livestock pressure, and may compound the results. Multiple regression analysis and a Twinspan ordination program were used to select parameters and plant species that were highly correlated with the gradient, and could act as indicators of each stage of degradation. These indicators can be used to monitor the impact of livestock on rangelands. General parameters, such as total plant density, are less sensitive to drought-induced changes than the composition of indicator species, but the latter are easier to monitor. Both types of indicators can form the basis for a monitoring methodology that can be used in Africa at three management levels: the individual herder, extension agents and regional branches of the division concerned with rangelands, and policy makers at the ministerial level