Land use and erosion in the uluguru mountains, Tanzania

Land degradation by soil erosion causes much concern in Tanzania, greatly affecting both steeplands and semi-arid regions. Erosion plots were established in the high rainfall steepland area near Morningside in the Uluguru Mountains of Tanzania. Inherent geological characteristics, soils, high rainfalls, land clearing activities and the farming practices cause frequent mass wasting. Accumulation of clays, mica and graphite minerals in the intermediate horizons of weathering profiles creates plane of low shear strength on which slumps and earth flows occur. Such mass movements make a major contribution to river sediment loads but are not measured in erosion plot studies. The amounts of soil washed from the surface under the forest cover, secondary growth cover on fallows, on bare plots and on small terraced ridges were observed using 20 m2 experimental runoff plots on 300 average slope. The single rain season of 4 months produced losses under forest cover of 0.11 t ha“l, which is about 9 % of soil loss on the bare plot. Soil loss on fallow with secondary growth and standing maize stubble was about 83 % of the bare soil loss. There was no significance different between losses on the bare soil and fallow surfaces. Surface runoff depth recorded under forest was 17.2 mm, on fallow was 21.3 mm, on bare soil plot was 21.4 mm and on terraced ridges was 15.4 mm of 1,378 mm of rainfall received in the 4 months wet period. Influences of soil loss on field plots are rainfall energy, antecedent conditions before rains and the state of the ground cover. Significant correlations of soil loss occur with the maximum intensities at 15 and 30 minutes and the threshold energy of the rain storms in the open. i.e. Kinetic energy of the maximum 15 and 30 minutes rainfalls. There is nearly zero correlation between soil loss and total energy of the storm. No relationship between rainfall energy and soil loss under forest was found. Regression equations indicate that rainfall erosivity surface runoff and antecedent conditions are the best predictors of soil loss on bare soil; rainfall erosivity, surface runoff and the quantity of precipitation are the best predictors under fallow conditions, but only the erosivity index and runoff provide a good forecast of soil loss under forest. The erosivity index KE > 15 has the highest significant correlation with soil loss on the secondary growth cover and on the bare soil.