Comparison of grass and soil conditions around water points in different land use systems in semi-arid South African rangelands and implications for management and current rangeland paradigms

Abstract

Globally, the determinants of rangeland dynamics and sustainability have been debated for decades. This study sought to explain the drivers of rangeland ecosystems in semi-arid South Africa by examining the vegetation and soil conditions around animal drinking water points, and their implications for sustainable management and paradigm debates. For this purpose, we selected three land use systems (continuous-communal livestock farms, rotational-commercial livestock farms, and continuous-game reserves), with the livestock farms having either high or low stocking rates. Soil organic carbon (OC), bulk density (p < 0.05), and soil compaction (p < 0.0001) varied greatly at land scales, being greater in the communal farms or game reserves than the commercial farms. Within the communal farms, sites with high stocking rates showed higher (p < 0.05) soil OC and bulk density than sites with lower stocking rates, but in the commercial farms, sites with high stocking rates had lower soil OC. Above ground grass biomass, and the abundance of highly desirable grass species were greater (p < 0.001) in the commercial farms and game reserves. Our results conclude that equilibrium elements, importantly animal density and diversity, seemed to play a major role in soil and vegetation dynamics. We also observed variations in edaphic factors and rainfall across the land use systems and over seasons (rainfall). We assume that these non-equilibrium elements may also influence rangeland ecosystems. To confirm this, however, we recommend future studies to include comparable measures of the vegetation and soil, and their relations with edaphic and temporal long-term rainfall variability across the land use systems.

Keywords:

• Continuous grazing
• equilibrium
• land scale
• non-equilibrium
• rotational grazing

Additional information

Funding

This publication is an output from a research project funded by the Govan Mbeki Research and Development Center of the University of Fort Hare, South Africa.