The purpose of this study was to develop and test a predictive understanding of the vegetation dynamics of the Lowveld of South Africa (30°35'E to 30°40'E and 24°00'S to 25°00'S). The study covered about 5000 km2 in Adjacent Private Protected Areas (APPA) adjoining the Kruger National Park (KNP). Data gathering (800 sites; 23 properties) commenced in 1989 and those recorded up to 2004 are reported here. The value, both ecological and economic, of the wildlife and tourism industry dependent on this savanna region is discussed in both historical and current perspectives. A range of land-use objectives and anthropogenic interventions were exposed. The properties ranged in size from 30 to 800 km2 and formed an effective and extensive manipulative experiment for investigating interaction of bush density, animal stocking, use of fire and landscape-scale processes. The first descriptive classification (at 1:250 000) of the area was developed using Inverse Distance Weighted interpolations. This confirms similar landscape/vegetation patterns in the KNP and Mocambique. The current mode of determining stocking density or carrying capacity was interrogated and indices suitable for complex multi-species systems developed. This was done in the context of equilibrial/disequilibrial paradigms. Application of the original indices resulted in drought-related declines in animal biomass of 4000 kg km2 over 20 years due to overestimation of carrying capacity. The model proposed here uses rainfall, animal type, biomass and vegetation parameters to determine stocking density for both coarse (regional) and ranch-specific scales. Principal driving determinants (rainfall, geology, soil type, tree density canopy cover, animal numbers, feeding classes and fire) of vegetation structure and their influence on the herbaceous layer were investigated. Groupings on ecological potential showed 'high' potential areas are less sensitive to animal impact than those classified as 'low' potential. Sustainability, embedded in a forward-looking component viz. Strategic Adaptive Management (SAM) with well-articulated endpoints viz. Thresholds of Potential Concern (TPCs) was used to study fluctuations in animal populations with Connochaetes taurinus (Blue wildebeest) as the case study. The TPC approach provides strong pointers for proactive management aimed at maintaining the system within bands defined by TPCs supporting operationally practical and periodically reviewed objectives. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2005.