The South African Witwatersrand Basin, deposited on the Kaapvaal Craton, is the largest gold deposit in the world but the ore-forming process(es) is/are still not clear; the debate revolves mostly around the role of syngenetic (placer) versus epigenetic hydrothermal processes. However, it is clear that several fluid events have been recorded throughout the basin. Consequently, it is important to determine the scale of these fluid circulations.
For the purpose of this study, we sampled calcite-quartz veins in three different goldfields (Welkom, Klerksdorp and Evander) in close vicinity to gold-bearing reefs. In addition, we systematically sampled their host rocks (sedimentary or mafic rocks) in order to investigate the effects of rock-fluid interactions. We performed oxygen and carbon isotope analyses on both the veins and their host rocks, as well as a fluid inclusion study.
We show that the oxygen isotope composition of quartz in veins is buffered by the host sedimentary rocks. It is therefore not possible to determine the nature of the original fluid. Another important result is that carbon isotopes analyses in quartz-calcite veins hosted by mafic rocks present a range of values that suggest two distinct sources, regardless of their location within the Witwatersrand Basin. One end-member (δ13C lower than −13‰) is represented by organic matter found locally within the basin sedimentary rocks whereas the second end-member (δ13C higher than −3‰) has a marine signature. The most likely candidate for this marine signature is the overlying limestones from the Chuniespoort Group. This implies a basin-scale downward fluid circulation from the overlying rocks to the basin sedimentary rocks. This fluid must have circulated after ca. 2642 to 2432 Ma, which is the age of the Chuniespoort Group deposition.