DESERTIFICATION IN THE SEMIARID KAROO, SOUTH-AFRICA - REVIEW AND REASSESSMENT

The concept of desertification has recently been re-evaluated with two major outcomes. Firstly, the term itself has been reviewed, with most authors concluding that the emotive implications of 'desertification' (i.e. advancing deserts per se) are inappropriate in many cases and t hat the general term 'dryland degradation' is a more accurate descript ion of reality. Secondly, it has been argued that the extent, nature a nd apparent irreversibility of the process has been over-estimated and that degradation in some areas may only be temporary, consequent upon an unfavourable combination of climatic and anthropogenic factors. In southern Africa, a reassessment of the concept of desertification is long overdue; the traditional view has prevailed to the extent that gr azing management policy in the semi-arid parts of the sub-continent is essentially founded on the premise that widespread, irreversible degr adation has occurred in post-colonial times. In this paper, we examine this position in the light of the available published evidence. It is suggested that some changes in the vegetation of the semi-arid Karoo occurred prior to the onset of European colonization, a function of ph ases of climatic aridification and the impacts of both Kho San hunter- gatherers and Khoi Khoi herders. This argument contextualizes the alle ged historical impacts, which are shown to be interpreted along the li nes of three competing models accounting for the pattern of assumed ve getation change. The literature concerning Karoo degradation is review ed and it is revealed that much of the research throws little light on which, if any, of the three competing models most closely approximate s reality. Instead, it becomes clear that the recognition of dryland d egradation in the Karoo requires further investigation, and several po ssible approaches are presented by which the hypothesis of irreversibl e change may be tested and the pattern of vegetation change identified . It is concluded that a more systematic monitoring programme, includi ng the investigation of vegetation and other environmental variables, should be instigated in South Africa.