Magister Scientiae - MSc / The prospect of unconventional shale-gas development in the Karoo Basin (South Africa) has created the need to obtain baseline data on natural radioactivity in Karoo groundwaters. The Karoo Basin groundwater radiological baseline developed through this study could serve as a reference to research potential future radiological contamination effects due to hydraulic fracturing. The major naturally occurring radioactive material (NORM) studied was radon (222Rn), in particular in-water activity concentrations; however, supplementary radium (226Ra and 228Ra) in-water activity concentrations and uranium (238U) in-water concentrations measurements were also made. A total of 53 aquifers across three provinces were sampled for groundwater and measured, with three measurement series from 2014 to 2016. The aquifers were categorized as shallow, mixed, or deep source. The radon-in-water baseline of the Karoo Basin can be characterised by a minimum of 0.6 ± 0.9 Bq/L, a maximum of 183 ± 18 Bq/L and mean of 41 ± 5 Bq/L. The radon-in-water levels from shallow sources (with water temperature < 20 °C) were systematically higher (40 Bq/L) than for deep sources (with water temperature > 20 °C). The natural fluctuations in radon-in-water levels were predominantly associated with shallow aquifers compared to almost none observed in the deep sources. The uranium in-water baseline can be characterised by a minimum of below detection level, a maximum of 41 μg/L, and the mean of 5.10 ± 0.80 μg/L. Similar to radon-in-water levels, uranium in-water levels for shallow sources were systematically higher than for deep sources. The limited (six aquifers) radium (228Ra and 226Ra) in-water activity-concentration measurement results were very low, with a maximum of 0.008 Bq/L (226Ra) and 0.015 Bq/L (228Ra). The 228Ra/226Ra ratio baseline were characterised by a minimum of 0.93, a mean of 3.3 ± 1.3, and a maximum of 6.5. The radium isotopes’ activity concentration ratio is an isotopic tracer for hydraulic fracturing wastewater. Pollution and contamination (radiological), due to unconventional shale gas development, in water resources has been noticed in the Marcellus Basin (United States). Consequently, developing and improving continuous baseline monitoring are of importance to study the environmental radiological effect of hydraulic fracturing.

>Magister Scientiae - MSc / The present research deals with the characterization of the Lower Ecca Group in terms of sedimentology, mineralogy and organic geochemistry. A field study was conducted in order to characterize the sedimentology and thereby determine the environments of deposition of the Prince Albert, Whitehill and Collingham Formations. In addition, shale samples were subjected to geochemical and mineralogical analyses so as to ascertain its source rock properties. The study utilized X-ray diffraction (XRD), scanning electron microscopy with an energy dispersive spectrometer (SEM-EDS), total organic carbon (TOC) and Rock-Eval pyrolysis to determine the mineralogy and organic geochemistry of shale from the formations under investigation. The sedimentological investigation revealed that the upper Prince Albert Formation is dominated by shale with thin beds of carbonate. These shales are interpreted to have been deposited by suspension settling in a marine environment which was occasionally interrupted by deposition of carbonates that form in a shallow marine environment. The overlying Whitehill Formation consists predominantly of carbonaceous shale with relatively more resistant shale beds also present. The fine sediments are interpreted to have been deposited from suspension settling under anoxic bottom conditions which would favor the preservation of organic rich material. Deposition of the Whitehill Formation was followed by the Collingham Formation which is dominated by rhythmic deposits of shale and sandstone that are occasionally interrupted by tuff layers. The clay size sediments are interpreted to have been deposited from suspension settling which are interbedded with low density turbidite current deposits in a marine environment. Based on the findings of the field study, it is apparent that the Prince Albert and Collingham Formations were deposited in marine environments, with the Whitehill Formation being deposited in an anoxic environment. These environments are known to be dominated by phytoplanktonic organisms and algal debris, and as such shales deposited in these environments contain predominantly Type I (derived from algae) and II kerogen (derived from plankton). It can therefore be postulated that Type I and II kerogens are the dominant constituents of organic matter in the Lower Ecca Group shales. Mineralogically, the shales consist chiefly of kaolinite, smectite and illite clay minerals, which are derived from a combination of weathering of feldspars, and the alteration of other clay minerals. The latter being inferred from the existence of albite in all the studied shale samples, pyroclastic material observed in the field, as well as the occurrence of alteration along clay mineral edges.

The Main Karoo Basin of South Africa is renowned for its exceptional palaeontological record and while its vertebrate fossils have been extensively researched, Karoo floras have received considerably less attention. Poor yields of palynomorphs from the Beaufort and “Stormberg” groups have complicated the task of erecting a comprehensive palynozonation scheme for the Karoo Supergroup. For this study, 65 palynologically productive samples from the Dwyka, Ecca, Beaufort and “Stormberg” groups allowed for systematic descriptions of all palynomorphs, as well as the ranges of the different taxa through the entire Karoo stratigraphic succession. Taxa with restricted ranges are useful for biostratigraphic correlation and these palynomorphs were used to delineate microfloral zones for the Karoo basin. The Dwyka Group contains high numbers of acritarchs and is generally low in species diversity. Useful biostratigraphic taxa for the Ecca Group include Cannanoropollis, Hamiapollenites, Platysaccus and Striatopodocarpites. Aratrisporites is a marker for the Latest Permian / Early Triassic Beaufort Group, while Cyathidites, Dictyophyllidites, Equisetosporites and Uvaesporites are indicators of the Late Triassic / Early Jurassic “Stormberg” Group. Palynostratigraphic zones correlate largely with the Karoo vertebrate biozones and severe and sudden extinction events are recognised among Karoo palynomorphs in the upper Tapinocephalus and Dicynodon assemblage zones. The first comprehensive palynological biozonation scheme for the Main Karoo Basin is proposed and the study provides a broad overview of Gondwanan Carboniferous - Jurassic floras. This study demonstrates that palynology is useful in correlating age equivalent lithostratigraphic units in the proximal and distal sectors of the Karoo Basin. Microfloras from previous South African studies are integrated within the proposed palynostratigraphic scheme, and palynological signatures of the various Karoo formations are shown to be consistent. Despite the constraints of floral provincialism, South African microfloras can be correlated to selected Gondwanan biozonations from Australia, Africa, Antarctica, New Zealand and South America. Future studies should focus on sampling more intensively over smaller stratigraphic intervals, which will assist in the correlation of time equivalent lithostratigraphic units in the different sectors of the basin, thus aiding in refinement of basin development models. Key words: palynology, Karoo, vertebrate biozones, stratigraphy, Gondwana

M.Sc. (Geology) / The study was undertaken to investigate the stratigraphic and sedimentological character of the transition between the Ecca and Beaufort Groups in the northeastern region of the main Karoo basin. Detailed stratigraphic profiles demonstrate that the transition comprises a regressional sequence of dark basinal shale overlain by sandy delta-lobe deposits. This is overlain by lenticular sandstones and shale representing meandering stream deposits. The deltaic sedimentary rocks vary markedly along strike and record deposition in four different deltaic sUbenvironments, namely: (A) "Normal" central delta lobe areas characterized by distal mouth bar sandstone-shale rhythmites gradationally overlain by proximal distributary mouth bar sandstones. (B) Interdistributary bay areas characterized by stacked thin upward-coarsening shale-sandstone cycles. (C) Areas of strong distributary channel erosion where prodelta shale deposits are erosively overlain by distributary or fluvial channel sandstones. No distributary mouth bar deposits are present. (D) Storm-dominated areas comprising prodelta sedimentary rocks which are in turn overlain by storm and possible barrier sand deposits. Palaeocurrent directions indicate that sediment transport was from the north-west, north and north-east. The concentrations of the trace elements B, V, Zn and Ni in prodelta shales indicate the possible prevalence of both marine and freshwater conditions in the near-coastal waters of the Ecca basin. Lateral variation of depositional environments on the same stratigraphic level makes the placing of the Ecca-Beaufort contact based on genetic criteria unsuitable. It is therefore suggested that purely lithostratigraphic criteria be used. On this basis the Ecca-Beaufort contact is taken at the base of the first prominent sandstone which overlies the Volksrust Shale Formation.

M.Sc. / The Late Carboniferous to early Jurassic Karoo Supergroup of South Africa witnessed two of the “big five” Phanerozoic mass extinction events, and the formation and subsequent break-up of the supercontinent Pangea. The closure of the Permian Period witnessed the greatest biotic crisis in the history of life. What is known about the Permian-Triassic boundary (hereafter referred to as the PTB) comes almost exclusively from marine successions in Europe and Asia. Although a major extinction event has been recognized in terrestrial successions, surprisingly little is known about its effects and timing. The exact placement of the PTB in the Karoo basin is not well constrained due to shortcomings of stratigraphic methods employed to date. This has made it extremely difficult to correlate the mass extinction events in the marine and non-marine environments; however, paleomagnetic studies could provide answers to both problems of absolute placement and correlation of the PTB in non-marine and marine successions. The PTB appears to lie within an interval of reversed polarity in many marine successions. A detailed magnetostratigraphic survey across the presumed PTB in the Karoo succession at localities in the north and south of the main Karoo Bain reveal two magnetic chrons, reversed followed by normal (with the boundary close to the reversal), which extends to slightly younger results from a previous study that identified an N/R pattern, thereby identifying a R/N/R pattern. The normal chron might correlate with the long basal Triassic normal polarity interval and the reversed polarity zones above and below it known from marine successions in the Alps, Russia, Pakistan and China. The PTB is thought to be situated coincident with the LAD of Dicynodon and the event bed of Ward et al. (2000), apparently above but not necessarily diachronous with a lithology change from predominantly green- to predominantly red mudstone. This placement falls within a normal polarity interval, but could conceivably have taken place at a time of reverse polarity due to delayed acquisition of magnetic remanence. The idea of an extraterrestrial impact as the cause of the end-Permian mass extinctions is strongly enhanced by a synchronous relationship between them. The configuration of the supercontinent Pangea during this time of earth history has been the matter of debate for the last three decades, with numerous alternative reconstructions to the classic Pangea A1 having been proposed for the time preceding the Jurassic. Paleomagnetic data from the Karoo allow for the definition of a new paleopole for West Gondwanaland, which prove a valuable tool for evaluating these various reconstructions. It is neither consistent with a Pangea B-type not C reconstruction for Pangea during this time interval, because of geological ambiguities. The most likely solution to the problem is that of a persistent non-dipole field contribution to the geomagnetic field during this time. Approximately 50 million years later Pangea was unambiguously in a classic Pangea A1 configuration, and life on earth suffered yet another set back. The end-Triassic mass extinction, which marks the sequence boundary between the Triassic and the Jurassic, has not received as much attention as the other four big Phanerozoic biotic disasters. In the Karoo a pronounced turnover in faunal assemblages from typical Triassic fauna to Jurassic Fauna (dinosaurs) is seen in the Elliot Formation. Magnetostratigraphic study of localities in the north and south of the Karoo Basin provided a magnetic zonation pattern for the Elliot Formation, a tool that has led to the constraining of the sequence boundary to the transition from the lower Elliot Formation to the middle Elliot and added to the hypothesis that the faunal turnover is globally synchronous. The determination of a paleolatitude for the Elliot Formation in combination with characteristically arid lithologies (eolian sandstones) provided the base for the evaluation of the paleoclimate that characterized Pangea during the Late Triassic to Early Jurassic. Key words: Karoo Basin, Magnetostratigraphy, Mass Extinction, Paleoclimate, Paleogeography, Paleomagnetism, Pangea, Permian-Triassic, Triassic-Jurassic

>Magister Scientiae - MSc / The Kalahari Karoo Basin is one of several basins in southern Africa filled with Late Carboniferous to Jurassic sediments that are primary targets for Permian aged coal. In order to determine the Coalbed Methane (CBM) potential of the Central Kalahari Karoo Basin, 9 exploration boreholes were drilled. Vitrinite reflectance (%Ro) and proximate analysis were conducted on cored coal intervals. Passey’s ΔLogR method used in this thesis employs the use of resistivity and porosity logs to identify and quantify total organic carbon (%TOC) in potential source rocks. Compared with lab measured %Fixed Carbon, the results showed that Passey’s ΔLogR method effectively identifies coal intervals as organic enriched. In terms of %TOC calculations, the method works poorly in coal metamorphosed by dolerite intrusions. These heat affected coal samples display %Ro from 0.77% to 5.53% and were increased in rank from primarily sub-bituminous to higher ranking volatile bituminous and finally to anthracitic coal. Their higher level of organic metamorphism (LOM), accompanying compositional changes and increased density associated with accelerated coal rank seem to have hindered the method in its estimations or lack thereof. Compositional changes in the coal were controlled by proximity to sill intrusion, with a decrease in fixed carbon and volatile matter, and increases in ash and moisture in the contact metamorphism zone (2-12m from sill). In heat unaltered coal that has undergone normal burial maturation characterized by %Ro of 0.44% to 0.65%, the method works very well even attaining accuracy in some samples. In unintruded boreholes CH1 and CH6, correlations between fixed carbon and generated %TOC curves indicate strong relationships with R2 from 0.70 to 0.83. Therefore, it was found that Passey’s ΔLogR method can be applied effectively on coal that has undergone normal burial maturation only.

Fine-grained sedimentary rocks of the Lower Permian lower Ecca Group in the Main Karoo Basin (MKB) in southern Africa have been identified to form a possible unconventional hydrocarbon reservoir, the gas potential of which is still relatively unknown. The lower Ecca Group is comprised of flysch-like successions of the Prince Albert, Whitehill and Collingham formations, which were impacted by diagenesis as well as contact and regional and metamorphism. Studies of fine-grained sedimentary rocks are notoriously complex as mechanical breakdown and diagenesis result in mineralogical changes and loss of source rock information. Therefore, a variety of analytical techniques are required for their thorough investigation. In this study, we analyse Permian rocks from two boreholes (KZF-1 and KWV-1) that were drilled ~ 830 km apart in the western and eastern MKB, respectively, and compare their composition, thermal maturity and petrophysical characteristics to better assess the hydrocarbon potential of the lower Ecca Group. In addition, whole rock geochemistry is used to reconstruct the palaeo-environment and provenance settings. Lastly, we investigate the influence of shale composition, porosity, well pressure and temperatures on their geomechanical properties such as compressive strength and elastic moduli. Sedimentary rocks in both boreholes were affected by burial and regional metamorphism (linked to orogenic events), but the sedimentary rocks in BH KWV-1 were also altered by contact metamorphism due to dolerite intrusions in the Early Jurassic. Major mineralogical differences between the boreholes include: 1) Fesilicate (greenalite); carbonates (rhodochrosite and dolomite) and Mn nodules (birnessite) being found only in BH KZF-1; and 2) metamorphic minerals such as garnet, cordierite, staurolite being found only in BH KWV-1. The results show that these rocks are over mature as evidenced by the low quantity of free hydrocarbons (S1 peak; 0.02–0.06 mg/g) and potential to release hydrocarbons (S2 peak; 0.06– 0.14 mg/g) and low hydrogen index (HI) values (2.40–167 mg HC/g TOC) from Rock Eval pyrolysis accompanied with high vitrinite/bitumen reflectance (BH KZF-1 is VRo= 4 and BH KWV-1 is BRo= 5). The total organic carbon (TOC) content is the highest in the Whitehill Formation in both boreholes (BH KZF-1: 5.17 wt%; BH KWV-1: 4.87 wt%). Burial diagenesis significantly reduced interparticle porosity and most of the measurable porosity comes from intraparticle nano-pores confined to organic matter v and dolomite. Meso- and macro-porosity are limited to phyllosilicates, pyrite framboids and microfractures. The Chemical Index of Alteration (CIA) is highest for the Prince Albert Formation (72– 78) as the palaeo-environment changed from an ice-house to a green-house, postglacial period in southern Gondwana. Enrichment of trace elements in BH KZF-1 and BH KWV-1 are in order: Mn > Ba > Zn > Co > Cu > V > Rb > Cr > Ni; and for BH KWV-1: Mo > Ba > Cu > Zn > Mn > V > Co > Cr > Rb > Ni. Geochemical proxies used in determining levels of anoxia were: V/Cr, V(V+Ni), Ni/Co and Fe-S-TOC plots. Salinity was deduced using Rb/K ratios and with identification of potential benthic foraminifera and minerals such as apatite, birnessite and phosphate nodules, which are assumed to be syn-sedimentary in origin. Overall, the geochemical results indicate that the depositional conditions fluctuated with respect to anoxia, salinity levels, and that while the sedimentation rates were low and the bio-productivity level was high, at least episodically. Provenance was investigated using Discriminant Function Diagrams, Principle Component Analysis Results (PCA), Zr/Ti ratios and the Index of Compositional Variability (ICV). Results indicate that the Prince Albert and Whitehill formations in the western MKB comprises sediment that was mostly sourced from the Cambrian to Late Carboniferous Cape Granite Suite, Cape Supergroup, the Dwyka Group and possibly the Precambrian Kango Group (Saldania Belt) and/or the Namaqua-Natal Belt. The Prince Albert and Whitehill formations in the eastern MKB have a mafic signature, indicating that these sediments may have been derived from the PermoCarboniferous Dwyka Group, Precambrian Kaapvaal Craton, Natal Belt (Tugela Terrane), Natal Group and the Pan-African Mozambique and Maud Belts. The mudstone and wackes in the Collingham Formation were mostly sourced from the quartz-arenites and granites that were most likely associated with the Cape Fold Belt, the orogen immediately adjacent to and largely coeval with the MKB. Geomechanical results from BH KZF-1 show that the lower Ecca Group has a high proportion (~ 50–70 vol%) of mechanically strong minerals (e.g., quartz, feldspar, pyrite), ~ 30–50 vol% weak minerals (e.g., clay, organic matter) and up to ~ 0–50 vol%, highly variable, intermediate components (e.g., carbonates). Constant strain rate deformation experiments (T ≤ 100 °C; p ≤ 50 MPa) and compressional tests performed perpendicular and parallel to bedding show that the Prince Albert Formation is the strongest and most brittle followed by the Collingham Formation and then the Whitehill Formation. Triaxial compressive strength as well as static Young's vi moduli increased with increasing hard minerals and decreased with increasing mechanically weak minerals and porosity. On comparison with European and American shales, the shales in the lower Ecca Group are geomechanically stronger and more brittle, and thus may allow for fracture propagation without rock breakdown under pressure. The lower Ecca Group varies from the western to the eastern MKB in mineralogy, thermal maturity, palaeo-depositional environment and provenance. These observations directly challenge the previous assumptions that these black shales were deposited under predominantly anoxic conditions. Indeed, during deposition, the levels of anoxia appear to have been inconsistent across the MKB and were likely dependent on the changing primary bio-productivity levels, sedimentation rates and basin morphology. Neither borehole contained significant gas, likely due to degasification by dolerite intrusions in the Early Jurassic and structural deformation during syn- and post-Karoo times. In summary, this study quantitatively confirms that the lower Ecca Group has the potential to be an easily frackable, unconventional shale gas reservoir. However, this study also attests to the compositional and geomechanical complexity of these Karoo rocks, and this inevitably translates to a gas potential that is expected to vary significantly across the basin.

A thesis submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy Johannesburg, August 2015 School of Geosciences, University of the Witwatersrand / The possibility of extensive shale gas resources in the main Karoo Basin has resulted in a renewed focus on the basin, and particularly the Whitehill Formation. The main Karoo Basin has been the subject of geological studies since before the 1920s, but geophysical data provides an opportunity to shed new light on the basin architecture and formation. In this thesis, I use regional gravity, magnetic and borehole data over the basin, as well as vintage seismic data in the southern part of the basin. Modern computational capacity allows for more information to be extracted from these seismic data, and for these data to be better integrated with potential field data. The integration of datasets in a three-dimensional model (3D) has allowed for a better understanding of the shape of the basin and its internal structure, in turn shedding light on basin formation. A new depth map of the basin constructed using this extensive database confirms that the basin deepens from on- to off-craton. The basin is deepest along the northern boundary of the Cape Fold Belt (CFB), with a depth of ~4000 m in the southwestern Karoo and ~5000 m in the southeastern part of the basin. Sediment thickness ranges from ~5500 to 6000 m. The Whitehill Formation along this boundary reaches a depth of ~ 3000 m in the southwest and ~4000 m in the southeast. Despite limited boreholes in this region, the basin appears to broadly deepen to the southeast. These seismic and borehole data also allow for mapping of the Cape Supergroup pinch-out below the Karoo basin (32.6°S for the Bokkeveld and 32.4°S for the Table Mountain Group), with the basin reaching a thickness of around 4 km just north of the CFB. The gravity effect of these sediments in the south is not sufficient to account for the low of the Cape Isostatic Anomaly near Willowmore and Steytlerville. This ~45 mGal Bouguer gravity low dominates the central region of the southern Karoo at the northern border of the CFB. The seismic data for the first time show uplift of lower-density shales of the Ecca Group (1800 – 2650 kg/m3) in this region, and structural and seismic data suggest that these lower density sediments continue to depth of 11 to 12 km along normal and thrust faults in this region. Two-dimensional density models show that these shallow crustal features, as well as deeper lower crust compared to surrounding regions, account for the anomaly. These seismic and borehole data also allow for constraints to be placed on the distribution and geometry of the dolerite intrusions that intruded the basin after its formation, and in some cases impacted on the shale layer, to be constrained. The highest concentrations of dolerites are found in the northwest and east of the basin, pointing towards two magma sources. The region of lowest concentration is in the south-central part of the basin. Here the intrusions are confined to the Beaufort Group, ~1000 m shallower than the shale reservoir, suggesting it should be the focus of exploration efforts. These dolerite sills are shown to be between 5 and 30 km wide and are saucer-shaped with ~ 800 m vertical extent, and dips of between 2° and 8° on the edges. The sheets in the south of the basin extend for over 150 km, dipping at between 3° and 13°, and are imaged down to ~ 5 km. This change in dip of the sheets is linked to deformation within the Cape Fold Belt, with greater dips closer to the belt, although these sheets do not appear to intrude strata dipping at more than 15 to 20°. In order to understand the shape of the Karoo basin and construct a 3D model of the basin, an understanding is needed of the underlying basement rocks. The Beattie Magnetic Anomaly (BMA) that stretches across the entire southern part of the basin forms part of the basement Namaqua-Natal Belt. Filtered magnetic data confirm that the Namaqua and Natal Belts are two separate regions with different magnetic characteristics, which is taken into account during modelling. The BMA is shown to be part of a group of linear magnetic anomalies making up the Natal Belt. The anomaly itself will therefore not have an individual effect on basin formation, and the effect of the Natal Belt as a whole will have to be investigated. An in-depth study of outcrops associated with one of these linear magnetic anomalies on the east coast of South Africa suggest the BMA can be attributed to regions of highly magnetic (10 to 100 x 10-3 SI) supracrustal rocks in Proterozoic shear zones. Along two-dimensional magnetic models in the southwestern Karoo constrained by seismic data, these magnetic zones are modelled as dipping slabs with horizontal extents of ~20-60 km and vertical extents of ~10-15 km. Body densities range from 2800- 2940 kg/m3 and magnetic susceptibilities from 10 to 100 x 10-3 SI. These, as well as other geophysical and geological constraints, are used to construct a 3D model of the basin down to 300 km. Relatively well-constrained crustal structure allows for inversion modelling of lithospheric mantle densities using GOCE satellite gravity data, with results in-line with xenolith data. These results confirm the existence of lower density mantle below the craton (~3270 kg/m3) that could contribute to the buoyancy of the craton, and an almost 50 kg/m3 density increase in the lithospheric mantle below the surrounding Proterozoic belts. It is this change in lithospheric density along with changes in Moho depths that isostatically compensate a large portion of South Africa’s high topography (<1200 m). The topography higher than 1200 m along the edge of the plateau, along the Great Escarpment, are shown to be accommodated by an asthenospheric buoyancy anomaly with a density contrast of around 40 kg/m3, while still mimicking the Bouguer gravity field. These findings are in line with recent tomographic studies below Africa suggesting an “African Superplume” or “Large Low Velocity Seismic Province” in the deep mantle. The basin sediment thickness maps were further used to investigate the formation of the main Karoo Basin. This was accomplished by studying the past flexure of the Whitehill Formation using north-south two-dimensional (2D) profiles. Deepening of the formation from ~3000 m in the southwest to ~4000 m in the southeast is explained using the concept of isostasy, i.e., an infinite elastic beam that is subjected to an increasing load size across the Cape Fold Belt. Load height values increase from 4 km in the southwest to 8 km in the southeast. This larger load is attributed here to “locking” along a subduction zone further to the south. The effective elastic thickness (Te) of the beam also increases from around 50 km over the Namaqua and Natal Belts in the southwest to 80 km over the Kaapvaal Craton and Natal Belt in the southeast. The changes in Te values do not correlate with changes in terrane, i.e., a north to south change, as previously though. The large extent and shape of the Karoo basin can therefore, in general, be explained as a flexural basin, with the strength of the basement increasing towards the southeast. Therefore, while factors such as mantle flow could have contributed towards basin formation, reducing the load size needed, it is no longer necessary in order to account for the large extent of the basin. This flexure model breaks down further to the southeast, most likely due to a very high Te value. This could be the reason for later plate break in this region during Gondwana breakup. It is inferred that this increase in Te is linked to the buoyancy anomaly in the asthenospheric mantle.

The Gemsbok Sub-basin is situated in the south-western corner of the Kalahari Karoo Basin and extends south from the Kgalagadi District of Botswana into the Northern Cape (South Africa); and west into the Aranos Basin (southeast Namibia). The Sub-basin preserves a heterogeneous succession of Upper Palaeozoic to Lower Mesozoic sedimentary and volcanic rocks of the Karoo Supergroup. Because the succession is largely covered by the Cenozoic Kalahari Group, the stratigraphy of the succession is not as well understood as the Main Karoo Basin in South Africa. Most research in the Gemsbok Sub-basin is based on borehole data. This study focuses on the intrabasinal correlation, depositional environments and provenance of the Karoo Supergroup in the Gemsbok Sub-basin in Botswana and Namibia. Based on detailed sedimentological analyses of 11 borehole cores of the Karoo Supergroup in the Gemsbok Sub-basin of Botswana and Namibia, 8 facies associations (FAs) comprising 14 lithofacies and 2 trace fossil assemblages (Cruziana and Skolithos ichnofacies) were identified. The facies associations (FA1 to FA8) correspond to the lithostratigraphic subdivisions (the Dwyka Group, Ecca Group, Beaufort equivalent Group, Lebung Group [Mosolotsane and Ntane formations] and Neu Loore Formation) of the Karoo Supergroup. Sedimentological characteristics of the identified facies associations indicate the following depositional environments: glaciomarine or glaciolacustrine (FA1, Dwyka Group), deep-water (lake or sea) (FA2, Ecca Group), prodelta (FA3, Ecca Group), delta front (FA4, Ecca Group), delta plain (FA5, Ecca Group), floodplain (probably shallow lakes) (FA6, Beaufort Group equivalent), fluvial (FA7, Mosolotsane and Neu Loore formations) and aeolian (FA8, Ntane Sandstone Formation). The Dwyka Group (FA1) forms the base of the Karoo Supergroup in the Gemsbok Subbasin and overlain by the Ecca Group deposits. Three types of deltas exist within the Ecca Group: fluvial-dominated; fluvial-wave interaction and wave-dominated deltas. The Gemsbok Sub-basin was characterised by rapid uplift and subsidence and high sediment influx during the deposition of the Ecca Group. Petrographic and geochemical analyses of the Ecca Group sandstones revealed immature arkose and subarkose type sandstones dominated by angular to subangular detrital grains, sourced from transitional continental and basement uplifted source areas. The sandstones of Ntane Sandstone Formation are classified as subarkoses and sourced from the craton interior provenances.

A presente dissertação de mestrado apresenta uma descrição detalhada do primeiro dicinodonte reportado para o Permiano da América do Sul. O material (PV 0226) é composto por um crânio parcial e mandíbulas associadas, proveniente da Serra do Cadeado, estado do Paraná, Formação Rio do Rasto. O espécime foi descrito preliminarmente na década de 70 e atribuído ao gênero Endothiodon. Tal classificação implicava em uma correlação direta com algumas das consagradas biozonas do Grupo Beaufort da África do Sul, já que este táxon é encontrado nesse depósito. Assim, o principal objetivo deste trabalho era verificar a identidade taxonômica do material brasileiro, e a partir da confirmação ou da proposição de outra identificação para o espécime, discutir brevemente as implicações bioestratigráficas e paleobiogeográficas da presença deste fóssil no Brasil. O exemplar em questão foi comparado com espécimes de Endothiodon da coleção do American Museum of Natural History, e uma extensa revisão bibliográfica foi realizada para obter informações sobre espécimes depositados em outras instituições. Por fim, o material brasileiro foi confirmado como pertencendo ao táxon Endothiodon, devido à presença de: forame pineal situado em uma bossa, protuberâncias em forma de bulbo localizadas na face antero-lateral dos dentários, bossa situada na margem ventral do jugal, grande número de dentes inseridos na superfície dorso-medial do dentário, sulco do dentário lateral à fileira de dentes na mandíbula, porção mais anterior da mandíbula em forma de um bico curvo e pontiagudo, entre outras características. Entretanto, não foi possível atribuir a este exemplar uma das quatro espécies do gênero, visto que a atual separação das espécies é dada basicamente por diferenças no comprimento do crânio. Esta feição pode refletir meramente o estágio ontogenético dos indivíduos e não diferenças em nível específico, exceto para E. mahalanobisi que aparenta ser uma forma de pequeno porte. Na realidade, outras características (região interorbital larga, presença de sínfise do dentário grácil ou robusta) apontadas como diagnósticas para o táxon são problemáticas e deveriam ser evitadas ou substituídas por outras mais informativas. Além disso, a forma de Endothiodon do Brasil apresenta uma estrutura denominada “dentary table”, recentemente identificada neste gênero. Em relação à bioestratigrafia, as correlações propostas para as faunas de tetrápodes da Formação Rio do Rasto ressaltam a semelhança dessas com associações faunísticas do Meso e Neopermiano da África do Sul e do Leste Europeu. Porém, até o momento, essas correlações são tentativas, pois poucos elementos dessas faunas são conhecidos para o Brasil e sobre estes, faltam informações relativas aos níveis estratigráficos em que foram coletados. Este último fator gera incertezas em relação à contemporaneidade dos fósseis encontrados em uma mesma localidade. Sobre a aparente escassez de vertebrados para esta formação, quando comparada a outros depósitos de mesma idade, esta parece advinda de um artifício de coleta, ou seja, o material conhecido até o momento não reflete a diversidade que essas localidades devem abrigar. O estudo detalhado de PV 0226 resultou em um artigo submetido para publicação em um periódico de cunho paleontológico. / The present master´s thesis presents a detailed description of the first dicynodont reported for the Permian of South America. The material (PV 0226) comprises a partial skull and associated lower jaws, collected in the Serra do Cadeado locality, Paraná state, Rio do Rasto Formation. The specimen was described preliminary during the 1970´s and assigned to the genus Endothiodon. This identification implied a direct correlation with some of the well-established biozones of the Beaufort Group, South Africa, since this taxon is found in this deposit. Thus, the main objective of this work was to verify the taxonomic identity of the Brazilian material and based on the confirmation or proposition of another classification for the specimen, to discuss briefly the biostratigraphic and palaeobiogeogrpahic implications of the presence of this fossil in Brazil. The specimen studied herein was compared with Endothiodon material housed in the collections of the American Museum of Natural History, and an extensive review of the literature was carried out in order to collect information about specimens housed in other institutions. Finally, it was confirmed the material from Brazil belongs to Endothiodon, due to the presence of: pineal foramen situated on a boss, prominent bulbous swellings on the anterolateral sides of the dentary, boss situated on the ventral margin of the jugal, extensive number of teeth on the mid-dorsal surface of the dentary, dentary sulcus lateral to the lower teeth row, anterior portion of the lower jaw prolonged into upwardly curved and pointed beak, among other characteristics. However, it was not possible to assign the material to one of the four species of the genus, as the current distinction of them is based mainly on differences in the skull lenghth. This feature may reflect merely the ontogenetic stage of the specimens and not differences at species level (except from E. mahalanobisi, which seems to be a small form). In fact, other characters (wide interorbital region, slender or robust dentary symphysis) pointed as diagnostic for this taxon are problematic and should be avoided or substituted for others more informative. Besides, the Brazilian form of Endothiodon bears a dentary table, recently indentified in this genus. Regarding biostratigraphy, the correlations proposed for the tetrapod faunas of the Rio do Rasto Formation highlight the similarities between these associations and others from the Mid and Late Permian of South Africa and Eastern Europe. But, until now, these correlations are tentative because few elements of these faunas were recovered from Brazil and most of them lack data about the stratigraphic levels from where they were collected. The latter leads to uncertainty concerning the age of the fossils found in the same locality. About the apparent poverty of vertebrates for this formation when compared to other deposits of the same age, it seems to be a collecting artifact, i.e. the amount of material recovered from this formation until the present does not reflect the potential diversity of it. The detailed study of PV 0226 resulted in an article submitted to a palaeontological journal.