Soil water movement and nutrient cycling in semi-arid rangeland : implications for patterns of vegetation change and system resilience.

Dougill, A. J. and Heathwaite, A. Louise and Thomas, D. S. G. (1998) Soil water movement and nutrient cycling in semi-arid rangeland : implications for patterns of vegetation change and system resilience. Hydrological Processes, 12 (3). pp. 443-459. ISSN 1099-1085

Abstract

Recent decades have seen rapid intensification of cattle production in semi-arid savannah ecosystems, increasingly on formalized ranch blocks. As a result, vegetation community changes have occurred, notably bush encroachment (increased bush dominance) in intensively grazed areas. The exact causes of this vegetation change remain widely debated. Previous studies have suggested: (i) increased leaching of water and nutrients into the subsoil in intensively grazed areas provides deeper rooting bush species with a competitive advantage for soil water and nutrients, and (ii) nutrient leaching may be exacerbated by nutrient inputs from cattle dung and urine. Our research in the Eastern Kalahari showed that in infertile sandy soils both the magnitude of soil water and concentration of soil nutrients leached into the subsoil is largely unaffected by the ecological and biochemical effects of increased cattle use. We found that despite the high soil hydraulic conductivity ( &greaterno;12 cm h−1), relatively high subsoil moisture contents and the restriction of water movement to matrix flow pathways prevent leaching losses beyond the rooting zone of savannah grass species. No significant differences in patterns of soil water redistribution were noted between bush dominant and grass dominant sites. We also found that the low nutrient status of Kalahari soils and leachate movement as matrix flow combine to allow nutrient adsorption on to soil particles. Nutrient adsorption ensures that nitrogen and phosphorus cycling remains topsoil dominated even following the removal of vegetation and direct nutrient inputs in cattle dung and urine. This conclusion refutes environmental change models that portray increases in the leaching of soil water and available nitrogen as a major factor causing bush encroachment. This provides a possible explanation for the now widely cited, but hitherto unexplained, resilience of dryland soils. We suggest that infertile sandy soils appear resilient to changes in soil water distribution and nutrient availability caused by increased cattle use. Hence, soil characteristics contribute to the resilience to permanent ecological change that is increasingly recognized as an attribute of semi-arid rangelands.