The Paleo-environmental significance of the iron-formations and iron-rich mudstones of the Mesoarchean Witwatersrand-Mozaan Basin, South Africa

M.Sc.The Mesoarchean Witwatersrand and Pongola Supergroups of South Africa are the oldest, well preserved supracratonic successions worldwide. Various banded iron formation (BIF) and iron-rich mudstone units occur within the West Rand Group of the Witwatersrand Supergroup and the Mozaan Group of the Pongola Supergroup. A granular iron formation (GIF) occurs in a single unit in the Nconga Formation of the Mozaan Group. The Witwatersrand Supergroup and Mozaan Group have been lithostratigraphically correlated and are interpreted to have been part of the same sedimentary basin. The studied BIF units occur in two associations: shale-associated and diamictiteassociated BIF. The GIF seem to have been deposited in shallower environments with greater hydrodynamic activity. The iron-rich mudstone shows a similar stratigraphic setting to that of the shale-associated BIF. The lithostratigraphic setting of the Witwatersrand-Mozaan basin BIFs are similar to what is seen for Superior-type ironformations, with the mudstones and associated BIFs marking marine transgressions. Various mineralogical facies of BIF were identified, including oxide, carbonate and silicate facies BIF, as well as mixed facies between these end members. The GIF is a unique facies and shows abundant petrographic evidence for biological activity. The iron-rich mudstone has been subdivided into iron-silicate rich, magnetite-bearing, carbonate-bearing, magnetite-carbonate-bearing and garnet-bearing subtypes. BIF, GIF and iron-rich mudstone have been subjected to lower greenschist facies metamorphism with some occurences of localized contact metamorphism. The abundance of magnetite shows that oxidation played an important part in BIF deposition, whereas the occurrence of 12C-enriched iron-rich carbonates suggests post depositional reduction of the deposited oxidized iron-rich minerals by organic matter. Al-bearing minerals are rare in the BIFs xxi and abundant in the iron-rich mudstones. Apatite and rare earth element (REE)- phosphates occur throughout. The major element geochemistry shows an inverse proportionality for Fe and Si in all the studied samples. BIFs show slightly higher Fe- and lower Si- and Al-concentrations compared to iron-rich mudstones which show higher Si- and Al- and lower Feconcentrations. The studied BIFs show major element geochemical attributes intermediate to those of Superior- and Algoma-type iron-formations. Provenance studies on some of the iron-rich mudstones illustrate that they were sourced from a mixture of mafic and felsic sources. The rare earth element (REE) geochemistry suggests strong hydrothermal input into the units, and positive correlation with the Fe-concentrations suggests that the Fe was introduced by high temperature hydrothermal fluids. The majority of the REEs are hosted by apatite and the REE-phosphates monazite and xenotime. The REEs were reconcentrated into these phosphates during diagenesis. A comparison of the studied lithostratigraphically correlatable units between the Witwatersrand Supergroup and Mozaan Group makes it possible to construct a depositional model for basin-wide BIF deposition in the Witwatersrand-Mozaan basin. Shale-associated BIF was deposited during the peak of transgression when reduced Ferich hydrothermal bottom waters were introduced into shallow ocean water that was either oxygenated or filled with anoxygenic phototrophic bacteria. Diamictite-associated BIF, in contrast, was deposited during interglacial periods when the melting of glacial ice introduced sunlight, nutrients and oxygen to the reduced, hydrothermally influenced Ferich ocean water. GIF was probably deposited in shallow, above wave base waters cut off from clastic input, and then washed into deeper depositional environments. Iron-rich mudstone was deposited in a similar setting as the shale-associated BIF, but in environments that were not completely cut off from detrital influx. The study shows that it is impossible to construct a general depositional model for Precambrian BIFs, since the lithostratigraphic and depositional settings vary between different examples of BIF