Abstract
The Gondwana successions still cover very large parts of the reconstructed Gondwana. Most of these ‘basins’ merely represent erosional relicts of depositional systems or depositories that may originally have been much larger, and many of the current separated outcrop or basin areas may even have been interconnected. The main objective of this thesis was to investigate where all the sediment contained in the various Gondwana successions came from, and whether the strata were deposited in isolated basins or formed part of a much wider regional depositional system.
Gondwana successions from three types of Gondwana basins were sampled, namely foreland, cratonic sag and fault-bounded basins, in South Africa (Main Karoo Basin), Mozambique (Mid-Zambezi Basin), Brazil (Paraná Basin), Argentina (Sauce Grande Basin), Falkland Islands (Falkland Basin) and India (Bokaro and Jharia coal basins). In total, 6407 detrital zircon grains were extracted and analysed by LA-ICP-QMS, of which 3446 grains have an age concordance that exceeds 90%. With reference to the question whether the strata in the Gondwana basins were deposited in isolated basins or formed part of a much wider regional depositional system, the U-Pb age data allowed for a comparison of source region ages for the Gondwana strata on a continental and intercontinental scale.
A commonality of all samples studied from the various Gondwana basins, is that certain age fractions are common, while other age fractions appear absent. The 270 Ma age fraction is one of the most persistent in post glacial strata of the foreland and craton sag basins, with the ~550 Ma, ~1000 Ma, ~1300 Ma, ~1600 Ma age fractions less well defined. By bringing these ages into the context of the ages of primary basement source areas that were available in Gondwana at the time of deposition, it becomes apparent that the glaciogenic diamictites and strata of the fault-bounded basins must dominantly have been sourced from interior continental source terrains, typically older than 550 Ma, whereas the overlying post-glacial sediments must have received abundant sediment from the Permian volcanic arc that developed along the southern margin of Gondwana. This scenario fits well with the concept that after the amalgamation of Gondwana, up to the Late Carboniferous and Early Permian times, Gondwana had a southern passive continental margin and only later did this passive margin, develop into an active orogenic margin during amalgamation of the ~270 Ma Gondwanide Volcanic Arc.
The zircon grains in the diamictite of the foreland and cratonic sag basins, were mainly sourced from the Late to Early Neoproterozoic terrains, with the Sauce Grande Formation and Dwyka Group having secondary input from the Late Paleoproterozoic and Early Mesoproterozoic terrains. The thick Ordovician-Silurian successions, such as the Cape Supergroup of South Africa and correlatives in South America, which overlie these basement rocks with a marked erosional unconformity, also likely contributed significant proportions of detritus to the diamictites across the Gondwana basins. There is a virtual absence of zircon grains of Early Paleoproterozoic and Archean ages in the samples, which despite the large areas covered by basement rocks of Paleoproterozoic and Archean ages in the interior of Gondwana, is considered due to the sediment being recycled from the older sedimentary units already devoid of Paleoproterozoic and Archean ages.
The thesis concludes that due to similarities revealed in the age distribution patterns among the foreland basins in South America and Africa, coupled with the similarities with the age distribution patterns in the interior cratonic sag basins of Mozambique and India, that the Gondwana basins merely represent erosional relicts of depositional systems or depositories that were originally much larger. Furthermore, it is well recognised that the Late Carboniferous to Jurassic Gondwana sedimentary successions show certain sequence stratigraphic features that can be broadly correlated across the entire middle to southern region of Gondwana including all the foreland, cratonic sag and fault-bounded basins. As a result of the changes in the tectono-sedimentary conditions and climate, a clear mark was left on the sedimentary record, thereby providing a common thread linking the sedimentary fill of the basins that developed across the Gondwana group of continents between the Late Carboniferous to Jurassic. The interpretation that depositional systems or depositories were originally much larger is therefore, not only supported by the detrital zircon age distribution patterns revealed by the samples, but also by the fact that the Gondwana sequences can be correlated from one basin to another, on both a continental and intercontinental scale.
Regarding timing of deposition, the detrital zircon ages obtained for the basal glaciogenic beds of the Gondwana successions are much older than their depositional ages as suggested in literature. Although they do not contradict these proposed depositional ages, they rather represent the age of the youngest source area available during the time of their deposition. The youngest zircon ages for the Ecca and Beaufort groups of the Main Karoo Basin, the Tunas
Formation of the Sauce Grande Basin, the Tatui Formation of the Paraná Basin and the Brenton Loch Member of the Falkland Basin, all have ages younger than the depositional ages suggested in literature. A reason for these younger than expected ages could be that although these grains appear concordant, they are not regarded as a true reflection of the maximum depositional age because it is difficult to assess Pb-loss for very young zircon grains i.e., the Permian age grains analysed in this study. In more detail, if a zircon younger than ~500 Ma years has suffered Pb loss during one of the uplift events associated with African land surface evolution, which took place during the Cretaceous, then such Pb loss might not be easily detected within this age range of the Concordia diagram. The Discordia line associated with such zircons that have undergone Pb loss will be very close to the Concordia curve in this age range of the Concordia diagram. Thus due to the uncertainty regarding the authenticity of ages younger than their suggested depositional age, coupled with the youngest zircon grains in the diamictites being for the most part older than their proposed depositional age, the U-Pb age data obtained in this study does not necessarily allow for a more stringent cross-basin correlation.
A recommendation developed from the study is that further detrital zircon U-Pb geochronological investigations are required on the Gondwana successions across the Gondwana group of continents, and more specifically in combination with Hf-isotope analyses to further constrain the source of the sediment, to improve data and variations in provenance patterns in the Gondwana basins.