The dramatic escarpment bounding the Drakensberg-Maluti mountainland has long been regarded as a classic example of a topographic feature formed by scarp retreat processes. However, this view has recently been challenged on the basis of apatite fission track (AFT) and cosmogenic isotope data, which are argued to be inconsistent with a uniform rate of scarp retreat from the original position of continental break-up. It was suggested rather, that the evolution and present position of the escarpment was controlled primarily by a pre-existing inland drainage divide. Numeric surface process models have been used to support this interpretation, although these rely on a large number of unconstrained variables, and thus do not provide unique solutions.
However, several lines of direct geological field evidence support the scarp retreat model for the evolution of the Drakensberg escarpment. The Drakensberg-Maluti Mountains in southern Africa are formed by a ~1000 m thick sequence of Karoo basalts, capping Karoo sediments. This mountainland has an approximately rectangular shape, surrounded by orthogonal escarpments, except in the southwest, where it is dissected by the deeply incised Orange River. In detail, these marginal scarps are everywhere double topographic features, reflecting two resistant layers that form prominent cliffs: the thick upper basalt flows, and the Clarens Formation sandstone at the base to the lava sequence. Moreover, in the Drakensberg-Maluti example, the presence of inland-facing escarpments reveals that escarpment formation is not exclusively related to processes that occur at or adjacent to the site of continental break-up. These direct geomorphological and geological observations demonstrate that resistant units are the dominant influence in escarpment formation. Headward retreat of large waterfalls on major rivers over considerable distances (10–100 km) provides clear field evidence that scarps formed by resistant lithologies will not invariably degrade as a result of the existence of an inland drainage divide, as has been argued on the basis of surface process modelling.
Several complementary factors provide a ready explanation for apparent inconsistencies in the scarp retreat model that were identified from the AFT and cosmogenic isotope studies. The Drakensberg-Maluti mountains are surrounded by a dense network of dolerite dykes that were contemporary with, and supplied, the Karoo basalts. The massive lava flows capping the modern escarpment therefore probably originally extended at least to the edge of the dyke network. The advance inland of an erosion front, initiated by continental break-up, would have slowed dramatically where such massive lava flows were encountered. Tectonic processes, climate changes, and possibly plant evolution may have compounded this decrease in erosion rates. Although these latter factors are not readily quantified, they were probably subordinate to the dominant lithological control that we invoke. Escarpments may thus form and persist over long periods, extending from the time of continental break-up to the present, independently of inland drainage divides. These field observations must be taken into consideration when assigning values to the numerous unconstrained variables used in surface process models.