The regional excavation of metamorphic rocks from mid and lower crustal levels is the consequence of erosion-driven denudation of mountain belts that are in permanent isostatic equilibrium. If the erosion process occurs at different rates in different parts of a topographically asymmetric mountain belt, then the flexural properties of the lithosphere need to be considered in order to describe the features of the isostatic rebound. Such a situation is given in plateaulike mountain ranges which are likely to erode only at their steeply sloping edges whereas erosion on top stagnates. This paper investigates the flexural isostatic response of a plate that is unloaded due to the continuous retreat of an erosional escarpment. The isostatic response is tested for two limiting model cases: (1) for a plate with a broken edge and (2) for a continuous plate with variable flexural rigidity. It is shown that for both cases the rate of topography development on the exposed plate is not linear with the retreatrate of the escarpment. This feture is utilized to predict the nature of the subsequent erosion of the newly formed topography and ultimately of the denudation of the plate. Geological limitations of the theoretical model and its relevance to plateaulike topographies in southern Africa and the eastern highlands of Australia are then discussed.

In this areas the origin of the apparently cyclic rebound forces creating ''stairlike'' topography at their continental margins and the old age of their present surface topography has previously been unexplained. Within the presented model these features can be explained satisfactorily. Two important conclusions may be drawn: (1) The distance of flexural swells from great escarpments and the origin of subsequent stair-like topographies may be explained by simple flexural considerations. (2) The flexural rebound forces uplifting a plate exposed by a retreating escarpment are cyclic. Excavation times of metamorphic rocks under such mountain belts may therefore be prolonged not only by the nature of erosion of their initial topography but also by the nature of the subsequent isostatic rebound and ultimately their denudation history. ¿1991 American Geophysical Union