Unravelling the evolution of a continental rift by a multi-proxy provenance study (Albertine Rift, Uganda)

Abstract

Sedimentary provenance analysis has rarely been applied to continental rifts, because source-rock terrains are assumed to be uniform and supply systems stable and closely connected to the formation of graben shoulders. Here, we test the suitability of multi-proxy sedimentary provenance analysis to unravel the evolution of a segment of the East African Rift System from its sedimentary record, where an extreme relief has been created and a continent-wide drainage system has been affected. Multi-proxy techniques include petrographic-mineralogical analysis (framework, heavy mineral spectra, zircon morphology, and colour), as well as single-grain garnet geochemistry, Zr-in-rutile thermometry, and geochronological methods (detrital zircon U–Pb ages). Published data are compiled and complemented by new U–Pb detrital zircon data of rift sediments. Because a comprehensive database of basement source-rock compositions from East Africa is missing, published modern river sand compositions have been added to provide endmember source fingerprints. The selected Albertine Rift is surrounded by four main source terrains within the East African Basement, which cannot be separated clearly by individual provenance indicators. Most useful were heavy mineral spectra, garnet composition, and with limitation Zr-in-rutile thermometry showing different metamorphic overprint. U–Pb detrital zircon spectra assisted well in detecting juvenile magmatism of different orogenic events, but repeated inheritance in younger units prevented a strict allocation. Petrographic maturity turned out to be more dependent on multi-cycle detritus as on climate shift. Applying endmember composition to rift sediments, three evolutionary stages of the Albertine Rift could be identified which coincide with basin-scale unconformities. Moreover, we could prove a large, southwest trending drainage network in East Africa during the Miocene, which was destructed step-wise by ongoing rifting. Extreme uplift of the Rwenzori Mountains up to 5109 m could be documented by a specific set of provenance indicators justified by modern river sand. Three main lessons can be drawn from this study: (1) basement units can be only reliably characterized by a combination of provenance-sensitive parameters, and here, the frequently used U–Pb detrital zircon ages are not sufficient; (2) it is possible to define subunits, e.g., individual fault-blocks within the rift; and (3) chemical weathering is unproblematic when applying a multi-parameter approach even in a tropical setting. Graphical abstract