Evidence is presented to model drainage evolution across Zimbabwe since the Permian. This provides the framework to understand the marked difference in character of the rivers to the north and south of the modern central Zimbabwe watershed, which separates the Zambezi and Limpopo drainage basins. North-flowing tributaries of the Zambezi rising off this river divide have low gradients and senile characteristics. The northwest orientation of the upper sections of many of these rivers is unusual for tributaries of a major east-flowing drainage, but is in accord with the west-orientated fluvial system that deposited the Triassic sediments of the Karoo Supergroup in the Cabora Bassa basin of the Zambezi Valley. The modern drainage system to the north of the central Zimbabwe watershed is thus largely controlled by a surface that has existed since pre-Karoo times. Headwaters of the Zambezi tributaries were originally located well to the south of the modern divide, with high ground extending to the present-day Chimanimani and Nyanga mountainland in eastern Zimbabwe. This drainage system persisted until the late Triassic, when rifting, linked to the early disruption of Gondwana, initiated the formation of the modern Save and Zambezi river systems. The central Zimbabwe watershed represents a late Palaeogene (~43 to 33 Ma) asymmetric epeirogenic flexure, part of the Ovamboland-Kalahari-Zimbabwe Axis, which beheaded the headwaters of the early Zambezi tributaries. The resultant steeper gradients to the south of the watershed initiated the modern youthful south-flowing drainage system. A further disruption to the Zambezi drainage system occurred during Plio-Pleistocene arid episodes, when major dunes developed across dry river systems such as the Shangani, in the northwest of the country. Renewed flow in these rivers during subsequent wetter pluvial episodes resulted in them exploiting the inter-dune streets to develop new courses. Some, like the Shangani, incised their courses through the Kalahari sand cover to become superimposed drainages.
The landscape of much of Zimbabwe reflects the imprint of two major cycles of erosion (African and post-African) since the disruption of Gondwana. The African erosion cycle commenced with the disruption of Gondwana, while the ensuing post-African cycle of erosion was initiated by the late Palaeogene uplift along the line of the modern central watershed. This rejuvenated the river network, leading to removal of the carapace of deeply weathered saprolite that developed under the humid mid-Cretaceous climate of the earlier African cycle. The post-African surface is thus an etch surface, with the characteristic plain and inselberg topography marking the weathering base of the African erosion event. A very subordinate Plio-Pleistocene cycle is reflected by terraces immediately marginal to the major river systems. The confinement of the Save and Zambezi drainages to graben structures resulted in their evolution largely independently of the two major erosion cycles that moulded the landscape of the rest of the country.
The palaeo-drainage reconstruction has important implications for the dispersion of diamonds and associated pathfinder minerals from primary kimberlite sources. The Sese-Murowa kimberlites are inferred to be the primary source of hitherto unexplained alluvial diamonds in basal gravels of the Somabula Karoo outlier, located on the central Zimbabwe watershed, some 120 km to the northwest. The drainage evolution model also provides a framework to infer likely distal kimberlite sources for a number of major unexplained kimberlitic pathfinder mineral anomalies associated with the southern margin of the Kalahari Formation.