Major drainage divides in southern Africa are interpreted to reflect lines
of epeirogenic flexuring of the subcontinent associated with the formation
of co-related basins. The Great Escarpment, which separates coastal and inl
and drainage systems, marks the locus of the Escarpment axis. It was initia
ted by Early Cretaceous rift flank uplift associated with the break-up Gond
wana. Geophysical studies suggest that subsequent erosion, coupled with sed
imentation on the continental shelf, would have resulted in progressive inl
and migration of this flexure. The divide between the Orange-Vaal River sys
tem and the Limpopo and Molopo-Nossib-Auob drainage basins is designated th
e Etosha-Griqualand-Transvaal (EGT) axis. Upper Cretaceous flexuring along
this axis disrupted old drainage lines, and initiated deposition of the Kal
ahari formation. The end-Cretaceous Ovamboland-Kalahari-Zimbabwe (OKZ) axis
forms the watershed between the Zambezi and Limpopo Rivers in Zimbabwe, an
d separates the latter river system from fossil endoreic drainage lines in
the Kalahari, which originally emptied into the Makgadigadi Pans system. In
the south of Botswana, this axis is defined by the Kalahari Schwelle, whic
h separates the fossil Kalahari drainages from the Molopo-Nossib River syst
em. Processes responsible for initiating the EGT and OKZ flexures are poorl
y understood. However, the inferred ages of both these two axes and the Esc
arpment axis correspond with episodes of alkaline volcanism in southern Afr
ica. This argues for a link between continental flexuring and volcanic acti
vity. Major Pliocene uplift occurred along a line intermediate between the
Great Escarpment and the present coastline in the east of the country (the
Ciskei-Swaziland axis). More subdued Plio-Pleistocene flexuring along a sou
thwest-northeast axis (designated the Bushmanland-Harts axis) traversing th
e interior of South Africa was responsible for the formation of major pans
('floors') in Bushmanland and the Orange Free State. There are a number of
subordinate lines of uplift (the Khomas, Otavi, and Zoutpansberg axes) whic
h are parallel to the Bushmanland-Harts axis. They are presumably related t
o the same stress field, and thus probably of similar age. These latter axe
s are all sub-parallel to active faults in northern Botswana which are inte
rpreted to reflect southwestwards migration of the east African rift system
, following lines of structural weakness. Sequential uplift along the axes
which have been identified provides a framework for interpreting the evolut
ion of drainages and erosion surfaces on the sub-continent.