Abstract
M.Sc.
Bracketing the depositional age of the Gamagara/Mapedi to Lucknow and Olifantshoek succession in Griqualand West is important because it not only represents one of the oldest known red bed successions in the world but also hosts some of the first well preserved lateritic soil profiles and carbonates with heavy 13C values traditionally correlated with the so-called Lomagundi carbonate carbon excursion. In addition the ancient supergene very large high-grade hematite iron ore deposits of the Sishen-Postmasburg area on the Maremane dome are associated with the erosional unconformity at the base of the Gamagara Formation (a lateral equivalent of the Mapedi Formation). However, the depositional age of especially the Gamagara/Mapedi to Lucknow succession is under dispute because it has been considered a) correlative to the lower part of the Waterberg Group in the Transvaal area, with the implication that it is younger than the Bushveld Complex with an age of ~2,054 Ga, and b) correlative to the Dwaalheuwel-Magaliesberg succession of the pre-Bushveld Pretoria Group of the Transvaal Supergroup in the Transvaal area. The upper age limit of the Gamagara/Mapedi to Lucknow succession is defined by 1,92 Ga felsic volcanics in the overlying Neylan-Hartley succession of the Olifantshoek Group. The Hartley Lava Formation is overlain by Volop quartzites.
This study involves age determinations of detrital zircon populations extracted from the basal Doornfontein conglomerate member of the Gamagara/Mapedi succession, and quartzites of the Gamagara/Mapedi, Lucknow, Neylan, Hartley and Volop Formations at various localities in Griqualand West.
Based on field work, three unconformity-bound sequences are defined, namely the Gamagara/Mapedi-Lucknow, Neylan-Hartley and Volop sequences. Most interestingly quartzites of the Gamagara/Mapedi-Lucknow sequence contain abundant zircons with ages similar to that of the Bushveld Complex at ~2,054-2,06 Ga in addition to zircons as young as ~1,98-2,01 Ga. An exception is results on one sample of the Doornfontein Member analyzed so far (it is from the Rooinekke iron ore mine south of Postmasburg) that contains only zircons that are older than the Bushveld Complex with a rather prominent youngest population bracketed between 2,2 Ga and 2,32 Ga. The youngest detrital zircon populations in the Neylan-Hartley sequence are either slightly older than the Hartley lava or contain zircons with similar
age to Hartley felsic lavas at 1,92 Ga. This sequence thus appears to have developed immediately prior to and coeval with Hartley volcanism. The overlying Volop sequence contains abundant zircons as young as ~1,89 Ga.
The results clearly illustrate that the Gamagara/Mapedi to Lucknow succession is certainly not a lateral correlative of the pre-Bushveld Dwaalheuwel-Magaliesberg succession of the Pretoria Group. Rather it should be considered time-equivalent lower parts of the Waterberg Group in the Transvaal area. This implies that the heavy carbonate carbon excursion known from the Lucknow Formation is at least 100 my. younger than the one known from the upper part of the Silverton Formation along the contact with the overlying Magaliesberg Quartzite. There are thus at least three heavy carbonate carbon excursions, known from Paleoproterozoic cover successions of the Kaapvaal Craton in southern Africa, namely one in the ~2.35 Ga Duitschland Formation, a second in the ~2,1 Ga Silverton Formation of the Pretoria Group of the Transvaal Supergroup and the third in the ~1,98-1,92 Ga Lucknow Formation. It is further known that carbonates with normal open marine 13C values of close to zero occur in stratigraphic intervals between each of the heavy carbonate carbon excursions.
The only unit that may still be correlated with part of the Pretoria Group is the Doornfontein Member at the base of the Gamagara/Mapedi succession. The correlation of this unit with the base of the Dwaalheuwel Formation and the Hekpoort paleosol of the Pretoria Group thus remain possible but analyses of additional samples are needed to make sure that the conglomerate, and by implication the ancient supergene Sishen-type iron ore deposits, does not also postdate the Bushveld Complex.