The Sedimentology and Stratigraphy of the Beaufort Group of the Karoo Supergroup in the Vicinity of Thaba Nchu, Central Free State Province. Anthony Brian Rutherford A dissertation submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Master of Science. Johannesburg, 2009 ii Declaration I declare that this dissertation is my own, unaided work. It is being submitted for the Degree of Master of Science in the University of the Witwatersrand, Johannesburg. It has not been submitted before for any degree or examination in any other University. _________________________________________________________ Anthony Brian Rutherford _____________________day of _______________________________ 2009 Abstract Different basin fill models have been proposed for the retroarc foreland basin within which the Karoo Supergroup of South Africa was deposited. The most recent of these proposals suggests that the Karoo Basin behaved as a partitioned entity and that reciprocal infill occurred either side of a hinge line in the centre of the basin. Detailed sedimentological, palaeontological, stratigraphic and geophysical information for the Beaufort Group in the eastern central Free State in the vicinity of Thaba Nchu is presented here enhances aspects of this model. It is shown that the fossils assignable to the Dicynodon, Lystrosaurus and Cynognathus assemblage zones occur within the sandstones and mudrocks of the Balfour, Katberg and Burgersdorp formations in the study area. Using this combined data set it is shown that parts of the Beaufort Group stratigraphic succession which are present in the southern part of the Karoo Basin are absent in the central Free State. This suggests a depositional hiatus at both the beginning of the Lystrosaurus Assemblage Zone and at the end of Cynognathus Assemblage Zone times. During these hiatuses deposition was restricted to the proximal sector i.e. south of the hinge line. This finding is in line with the proposal that the Karoo Basin behaved as a partitioned basin with reciprocal fill. Geophysical information defines a basement high north-east of Thaba Nchu that probably accounts for a coarse-grained unit with anomalous palaeocurrent directions within the study area. Acknowledgements This thesis could not have been completed without the support of a large number of people who gave generously of their time and effort. I would like to thank all those who gave me assistance over the past few years. In particular I would like to thank the following people: I am extremely grateful to my supervisors, Professor Bruce Rubidge (Bernard Price Institute for Palaeontological Research) and Dr John Hancox (formerly of the Geology Department, University of the Witwatersrand), for their guidance and support. Their suggestions and advice were always welcome and often used. In addition their practical and financial support of my research is acknowledged, without it I would not have been able to have undertaken a number of field visits. I would also like to thank Susan Webb (School of Earth Sciences, University of the Witwatersrand) for reading and commenting on the geophysical work. Her many suggestions were most useful. A debt of gratitude is owed to Professor Marianne Bamford who identified the fossil wood and prepared specimens of it for thin section analysis. Dr M Raath and B Zipfel (Bernard Price Institute for Palaeontological Research) assisted me by providing data from the Bernard Price Institute for Palaeontological Research fossil collection database. Ntaupo Ntiri (formerly of the Bernard Price Institute for Palaeontological Research) using his knowledge of the Thaba Nchu area helped me in collecting fossils, for which I am most grateful. Thanks to Dr R Damiani (formerly of the Bernard Price Institute for Palaeontological Research) who identified the Procolophonoid for me. To the many farmers, land owners and residents of the Thaba Nchu and Tweespruit areas who gave me free access to their land, thank you. I would like to specifically thank the Orchesen and Kemp families who provided me with accommodation for a number of my field trips and whose enthusiasm for all aspects of life and nature in the Thaba Nchu area sustained my own interest. The Council for Geoscience, where I was employed during part of this study, assisted me by providing transport and financial support to undertake the field work. I am particularly grateful to Patrick Cole and Edgar Stettler, the head and former head of the Geophysics Unit respectively, who promoted the project within their annual budgets and gave permission for the geophysical data to be used. I am indebted to Lorraine van der Merwe of the Council for Geoscience?s library who tracked down and copied many of the journal articles and often obscure sources of data. Leone Mare of the Geophyics Unit (Physical Properties Laboratory) assisted by providing me with the rock density data which was used in constructing the gravity model. Many thanks to Janine Cole who provided the gravity and magnetic data along with much needed advise. Dr Johan Neveling kindly identified the Procolophon specimens that were collected and along with Dr Linda Karny provided data from the Council for Geoscience?s fossil collection. J.C. Loocke (Free State University) kindly pointed out the existence of the Musgrave unit and other geological features in the vicinity of Bloemfontein. A number of companies provided data and or technical assistance. Geosoft, through Kevin Fischer, kindly allowed me the use of an Oasis Montaj license which was used to grid the geophysical data. Anglogold Ashanti, through Jeff Kennedy, provided me with Karoo borehole data and agreed to let me use it. Petro SA kindly agreed to release the Soekor data held by the Council for Geoscience. Personal support and encouragement has come from many people throughout all of my studies and research. In addition to these unnamed many I would like to specifically thank my wife Engel Rutherford, my sister Patricia Ruthven and my friend Karen Weissensee. Lastly, and most importantly, I would like to thank my children David and Nicholas, who allowed me to pursue a dream even though it meant my absence for long periods of time. To all who helped I think I can now say ?from nature?s infinite book of mystery a little I can read? (William Shakespeare, Antony and Cleopatra). Contents Abstract List of Figures List of Tables 1. Introduction???????????????????????????....1 1.1 Literature Review ???????????????????????.2 1.1.1 Geology ??????????????????????.......2 1.1.1.1 Lithostratigraphy of the Beaufort Group ??????????.2 1.1.1.2 Biostratigraphy ????????????????????.11 1.1.1.3 Depositional Environments??????????????.....13 1.1.2 Sedimentary Basin Setting ???????????????....15 1.2 Present Study ????????????????????????..16 1.2.1 Aims and objectives of the present study ??????????..?.16 1.2.2 Materials and Method????????????????........17 2. Geology????????????????????????????..20 2.1 Facies Descriptions??????????????????????21 2.1.1 Coarse Grained Facies ??????????????????.21 2.1.2 Sandstone Facies ???????????????????.....22 2.1.2.1 Sandstone Petrography????????????????...30 2.1.3 Fine Grained Facies ?????????????????..?31 2.2 Architectural Element Analysis ?????????????????34 2.2.1 Elements Formed within Channels?????????????...35 2.2.1.1 Channels (CH) ?????????????????..?.....35 2.2.1.2 Downstream Accreting Macroform (DA)?????????..37 2.2.1.3 Laminated Sandstone Sheets (LS) ?????????..?.?..39 2.2.2 Elements of the Overbank Environment...............................................40 2.2.2.1 Floodplain Fines ? Pedogenic Profile (FP)?????????.41 2.2.2.2 Flood Plain Channel (CHf)???????????????.42 2.2.2.3 Crevasse Splay (CS) ????????????????.?.43 2.2.2.4 Flood Plain Fines (FF)???????????????.?...44 2.3 Lithostratigraphy ?????????????????????...?45 2.3.1 Dubbledam Mudrock Unit ???????????????...?45 2.3.2 The Musgrave Grit Unit ??????????????...?..?..47 2.3.3 Middle Sandstone Units ?????????????????...48 2.3.3.1 Sepenare?s Hoek Sandstone ???????????.?..??48 2.3.3.2 Townland Fines ????????????????..?..?..48 2.3.3.3 Eden Sandstone ????????????????..??....49 2.3.4 Mohono Mudrocks ???????????????..?...?..?49 2.5 Paleocurrents ??????????????????????...?..50 3. Palaeontology ??????????????????????????52 3.1 Fossils of the Thaba Nchu Area ??????????????..??..52 3.1.1 Vertebrates???????????????????????52 3.1.1.1 Dicynodontia ?????????????????.???.52 3.1.1.2 Procolophonoidea?.?????????????????...55 3.1.1.3 Archosauramorpha..??????????????????56 3.1.1.4 Cynodontia?????????????????????..56 3.1.1.1Amphibians and Fish??????????????...???57 3.1.2 Fossil Wood ??????????????????????.57 3.2 Additional Fossils??????????????????..????58 3.3 Position of Fossils within the Local Stratigraphy??????..????58 4. Geophysics????????????????????????.???59 4.1 Geophysical Surveys???????????????????.??.59 4.1.1 Gravity Method?????????????????????.59 4.1.1.1 Principles??????????????????????.59 4.1.1.2 Gravity dataset????????????????????.61 4.1.2 Aeromagnetic Survey??????????????????....63 4.1.2.1 Principles ??????????????????????63 4.1.2.2 Aeromagnetic Dataset?????????????????.64 4.2 Seismic Survey, Regional Borehole and Outcrop Geology??????...66 4.2.1 Regional Borehole Data??????????????????66 4.2.2 Seismic Survey depths??????????????????..67 4.3 Data integration, interpretation and modelling???????????..68 5. Discussion???????????????????????????..74 5.1 Palaeoenvironments?????????????????????...74 5.1.1 Dubbledam Mudrock Unit?????????????????75 5.1.2 Musgrave Grit Unit???????????????????...78 5.1.3 Middle Sandstone Units??????????????????78 5.1.4 Mohono Mudrocks Unit??????????????????80 5.2 Stratigraphy?????????????????????????80 5.3 Basin Development??????????????????????83 6. Conclusion???????????????????????????.88 References Figures Tables Appendix List of Figures Figure 1.1: Location of the study area, showing the major cities and towns and towns mentioned in the text. Inset shows position of main map in southern Africa. Figure 1.2: Geological map showing the distribution of the rocks of the Karoo Supergroup in South Africa (after Rubidge, 2005). Note the informal term ?Stormberg Group?, refers collectively to the Molteno, Elliot and Clarens formations. Figure 1.3: Distribution of the upper three biozones of the Beaufort Group in the vicinity of Thaba Nchu (after Rubidge, 1995). Figure 1.4: Map showing the location of some of the previous studies referenced in the text. Note that Theron (1970) described the Beaufort Group throughout the entire Free State Province so is not included on this map. Figure 2.1: Possible thrust fault (arrow) on Thaba Nchu Mountain. The two thick sandstone units which are over thrust are in the Elliot Formation. Figure 2.2a: Clasts in facies Gt. The feldspar crystals are fresh and poorly rounded. Figure 2.2b: Fossil wood in facies Gt. Logs larger than these were used to provide paleocurrent information. Figure 2.2c: Trough cross stratification within facies Gt (bracketed). Figure 2.3a: Intraformational conglomerate (facies Sei )(bracketed). Note how the second unit follows the erosional scour and contains bedding. Figure 2.3b: Matrix supported mud pebble conglomerate (facies Sei). Note the size of the clasts in this image is relatively uniform and small in comparison with other localities where some clasts are much larger. Figure 2.4a: A unit of horizontally laminated fine grained sandstone (facies Sh). This facies forms the bulk of all the facies types in the area. Note that what appear to be fine beds are in fact thin laminae some of which erode preferentially. Dip is due to the presence of a dyke and is not a depositional remnant. Figure 2.4b: Parting lineation on the surface of a unit of facies Sh. Figure 2.5: Low angle cross stratified sandstone (facies Sl) in sharp contact with a unit of planar cross stratified sandstone (facies Sp). Both facies are rarely found within the area with the former difficult to identify. Figure 2.6a: Cosets of trough cross stratified sandstone (facies St) (bracketed). Figure 2.6b: Surface expression of trough cross bedding (facies St). Measured down the trough axis gives a paleocurrent direction (NNW) for this unit i.e. into the page. Figure 2.7: Ripple marks are found at a number of localities, typically occurring as a single layer above facies Sh. Here interference ripple marks are exposed on the top surface a unit of Sh. Figure 2.8: Ripple cross stratified sandstone (facies Sr) formed by the downstream migration of chains of ripples. The facies caps a unit of facies Sh. Figure 2.9: A unit of facies Sm (bracketed) above and below units of Sh. 2 Figure 2.10a: Photomicrograph of a fine grained sandstone from the study area. Typical features include the fine, rounded to subangular and sorted monocrystaline quartz. Figure 2.10b: Peloid within a fine grained sandstone. Note the typical rounded to sub angular fine grained quartz matrix. The peloid is rounded and contains black fragments that may be fossilized plant material. Figure 2.10c: Quartz and feldspar (microcline) fragments. Figure 2.10d: A rounded polycrystalline quartz grain. Some hematite staining around the edges of the grain is present (arrowed). Figure 2.10e: Quartz grain included within a feldspar grain forming a lithic fragment suggesting a metamorphic or igneous rock provenance. Figure 2.10f: Fine grained sandstone commonly forms the matrix to the mudpebble conglomerate. Here it encloses a mud pebble (10 mm diameter) consisting of replacive prismatic calcite. Figure 2.11: Facies Fl comprising interbedded sand-, silt- and mudstone horizons on a scale of a metre. Each rock type varies in thickness from 10-300 mm. Contacts, where they are visible, are abrupt. Note the blocky weathering. Figure 2.12: Facies Fsm consists of couplets of mud- and siltstone. This facies differs from facies Fl in that the thin units of fine grained sandstone are absent. 3 Figure 2.13: Facies Fm consists of massive mudstone which in the study area weathers rapidly and into fragmentary chips of rock. Colour varies from red, green to black and as in this photograph mottling occurs. No units thicker than 500 mm were observed but this may be an artefact of the rapid weathering. Rare example of bedding seen in facies Fm. Figure 2.14a: Calcareous nodules in a fine grained sandstone. Figure 2.14b: Fragment of fossilised bone (arrow) in a concretionary nodule. Figure 2.15: Location of the architectural elements discussed in the text. CHm = Major Channel, CHt = Stacked Sand Channel, CS = Crevasse Splay, DA = Downstream Accreting Macroform, LS = Laminated Sandstone Sheets, FP = Pedogenic Sequence, FF = Flood Plain Fines, CHf = Flood Plain Channel. Figure 2.16a: Lateral profile showing both architectural element CHm, cross cutting channels within a larger channel (top of ridge) and architectural element LS (lower outcrop on photograph/diagram). The numbers refer to bounding surfaces. All surfaces dip to the right but are artificially distorted due to camera angle. Stratigraphically the upper outcrop is Eden Sandstone whilst the element LS is within the Townlands fines. Outcrop is on Moroto Mountain. Exposure is orientated along 166o-346o, which is oblique to the palaeocurrent direction of 004o. The outcrop is exposed for 90m. Figure 2.16b: Close up view of one of the smaller channels in Figure 2.30a. Paleocurrent direction is into and to the right of the photograph (350o). Location is on Moroto Mountain and stratigraphically in the Eden Sandstone Unit. Figure 2.17: Stacked channels (element CHt) on the farm Abramskraal. Individual channels (defined by dotted lines in figure) have a mean thickness of 500 mm 4 and are built up of units of Sh-St-Sr. The larger channel in which they occur is laterally extensive with a width of at least 350 m but a mean thickness of 1.5 m. Paleocurrent direction is into the photograph. Figure 2.18a: Element DA ? downstream accreting macroform. Miall (1996) gives as some of the defining characteristics the slightly concave up 4th order bounding surfaces and a dip of less than 10o in the downstream direction. Paleocurrent direction is toward the left and obliquely out of the page. Note that the lower unit is built up almost entirely of Sh except for an erosional unit of Sei into the underlying unit of Fm. The other units have abrupt bases. Outcrop on Abramskraal. Figure 2.18b: Element DA. This bedform consists of wedge shaped units. Paleocurrent direction is obliquely out of the page to the left. Abramskraal Farm. Figure 2.19a: Pedogenic sequence. Note the continuous layers of concretionary nodules (arrowed). Stratigraphic section was measured along the line (Figure 2.32b). Location is on the farm Melden Drift 28 and stratigraphically within the Townland Fines. Figure 2.19b: Stratigraphic section showing pedogenic features. Figure 2.20: Flood plain channel (CHf). Note that the base of the channel is not the lenses of sandstone but rather a much finer grained silty unit beneath it. Other than the ripple tops or mud veneer, the lenses of sandstone are made up of facies Sh. Location is the farm Sepenare?s Hoek and stratigraphically within the Dubbledam mudrocks. Figure 2.21: The flood plain channel (CHf) at Thaba Nchu Townlands consists of overlapping lenses 3-5m wide and 100-500mm thick. Each sandstone unit is made up of either facies Sr or Sh and is capped by a thin unit of siltstone or mudstone (Fsl and Fm). The bases are thin units of Sei. Figure 2.22a: Crevasse splay (element CS) (main splay arrowed, smaller events bracketed) at Feloana. The strata dip to the right at between 4o-6o. The section through the outcrop was taken along the displayed line (Fig 2.22b). Stratigraphically this forms part of the Dubbledam mudrock unit. Figure 2.22b: Section through the out crop at Feloana. Note the large amount of sandstone. The main crevasse splay noted in Fig 2.22a is arrowed. Figure 2.23: Outcrop of flood plain fines (architectural element FF). Figure 2.24: Map showing locality of stratigraphic sections taken in the Thaba Nchu area. The detailed locality maps and the sections themselves are provided in Appendix A. Figure 2.25: Measured stratigraphic sections placed on a map to indicate the geographic distribution of lithological units. See Appendix A for stratigraphic sections. Figure 2.26: Map showing the location of farms mentioned in the text. Figure 2.27: Thick packages of facies Fsm makes up a large portion of the Dubbledam mudrock unit. Note that in this exposure the mudrocks are predominately green, however fragments of rock from a unit of red mudrocks at the top of the outcrop gives the illusion that it consists of purple/red facies Fsm. Figure 2.28a: Typical exposure of tabular sandstone unit in Dubbledam mudrock unit. Two sets of facies Sl are seen here with 40-80 mm of laminated silt- and very fine sandstone separating them. Figure 2.28b: Close up of a tabular sandstone unit. Two sets of facies Sl with laminated siltstone and very fine sandstone between. Note that the unit, and in particular the bottom sandstone, dips to the left. The outcrop is only 10 m long so it was not possible to determine if these are stacked sandstones. Figure 2.29: Colour mottling within the Dubbledam Mudrock Unit. Figure 2.30: Laminated siltstone and very fine grained sandstone above a tabular unit of fine sandstone in the Dubbledam mudrock unit. Note the small bar form (arrowed) to the right of the bag. Figure 2.31: Invertebrate traces on the surface of a tabular sandstone in the Dubbledam mudrock unit. Figure 2.32: Bedforms within the Dubbledam mudrock unit on the farm Ratabane 100. Figure 2.33: Smaller scale troughs filled with facies Gt and making up the Musgrave grit unit. This outcrop forms the upper layer at the Feloana Dam locality. Figure 2.34: Outcrop of the Sepenare?s Hoek Sandstone on the farm Sepenare?s Hoek 759. Figure 2.35: Townland Fines unit (bracket TF) with small channel sandstone (bracket CHf). Location is on Thaba Nchu Townlands. Figure 2.36: The Eden Sandstone Unit on the western slope of Mohono Mountain. View is looking toward the east, palaeocurrent direction is north-north-west. Figure 2.36A: Close up of inset A in Fig. 2.26. View is to the south along the edge of the outcrop. Figure 2.36B: Close up of some of the large channels forming the Eden sandstone. Note the thick unit of facies Sei above the hammer. View is to the east. Figure 2.37: Typical outcrop of Mohono mudrocks unit on Eden 96 (Mohono Mountain). Figure 2.38: Part of a channel within the Mohono sandstone unit on Thaba Nchu Mountain (Wilgeboomnek 29). Figure 2.39: Rhizocretions found within the Mohono mudrock unit. Figure 2.40a: Palaeocurrent directions from the Dubbledam mudrock unit. Mean current direction is 340 from18 readings. Readings were taken from parting lineation and ripple marks (trends) and sole marks (direction). Figure 2.40b: Palaeocurrent directions from the Sepenare?s Hoek and Eden sandstone units. Mean current direction is 342 from 78 readings. Readings were taken from parting lineation and cross bedding. Figure 2.40c: Palaeocurrent directions from the Mohono mudrocks unit. Mean current direction is 338 from only 17 readings. Readings were taken from Parting lineation, ripple marks, planar cross bedding (trend) and cross trough bedding (direction). Figure 2.40d: Palaeocurrent directions from the Musgrave grit unit. Mean current direction is 230 from 22 readings. Readings were taken from only two localities. Readings taken from cross bedding and fossilised logs. Figure 2.41: The stratigraphy in the Thaba Nchu area showing the predominant facies types, architectural elements and palaeocurrent directions for each unit. Figure 3.1: Important fossil localities mentioned in the text. Figure 3.2: Localities of all fossils collected during this study. Figure 3.3: Location of fossils previously collected and held in various collections and for which there is accurate locality data. Figure 3.4: Lateral view of a skull of a Dicynodon in grey-green mudrocks of the Dubbledam mudrock unit at Thaba Nchu Townlands. Figure 3.5: Anterior view of a snout of Lystrosaurus murrayi (BP/1/6138) collected at Abramskraal 65 within 2m of the base of the Sepenare?s Hoek sandstone unit. Figure 3.6: Kannemeyeria tusk (arrow) and both caniniform processes collected on Thaba Nchu Mountain within the Mohono mudrocks unit (BP/1/6774). Figure 3.7: Procolophonoid left lower jaw ramus in mud pebble conglomerate. Note anterior to posterior increase in tooth size indicating a more advanced form than Procolophon. Figure 3.8: Left lower jaw ramus of an erythrosuchid found on Mohono Mountain (Eden 96) within the Mohono mudrock unit, Lateral view (Specimen BP/1/5877). Figure 3.9: Lungfish tooth plate (arrowed) recovered from the Mohono mudrocks unit on Mohono Mountain (Eden 96) (BP/1/6169). Figure 3.10a: Fossilised and silicified wood. Figure 3.10b: Radial longitudinal thin section of silicified fossil wood, Agathoxylon africanum, showing biseriate alternate bordered pits on the radial walls of the tracheids. (400x magnification). Figure 3.11: The stratigraphic horizons from which fossils were collected. Figure 4.1: National Bouguer Gravity data set showing area from which the geophysical data subsets for this study were extracted. Red square indicates the study area. Figure 4.2: Bouguer Gravity data subset. Bouguer gravity data has had all known effects removed and shows the relative variation in gravity (density). Figure 4.3: Sun Shaded Bouguer gravity data (45o inclination, 45o declination). Figure 4.4: Map showing position of Bouguer gravity anomalies. Figure 4.5: Map showing an image of the aeromagnetic data and anomalies discussed in the text. Figure 4.6: Aeromagnetic data over Thaba Nchu, subset. This is a subset of the regional data set in figure 4.5. MF indicates the large dolerite dyke that runs across a portion of northern Thaba Nchu and which divides two important outcrops (i.e. the outcrops on the farm Chubani 9). Figure 4.7: Contour map showing depth to basement of the Karoo and the boreholes used to generate the contours. The closest boreholes to the study area, Sunnyvale 785 and Middelpunt 2029, are indicated. Figure 4.8: Contour map showing depth to basement (in metres above/below see level) from Soekor seismic map. Figure 4.9: Comparison between the contour maps produced from seismic data and gridded borehole data. Although in broad agreement the seismic map shows much more detail. Figure 4.10: Karoo Supergroup stratigraphy and depth to the Karoo basement overlain on the sun shaded Bouguer gravity map. There appears to be very little correspondence between the stratigraphy and the gravity data. Figure 4.11: Bouguer gravity map (anomaly C) with the Karoo Supergroup stratigraphy and borehole data overlain. Note the variation in depth across the anomaly as well as that the anomaly cross cuts the Karoo stratigraphy. Figure 4.12: Bouguer gravity anomalies G and I with the Karoo Supergroup stratigraphy and borehole data overlain. Note the variation in depth from north to south. Profile line is discussed below. Figure 4.13: Bouguer gravity map of the area in the vicinity of the study area showing depth to basement contours with values at boreholes. Profile lines are discussed below. Figure 4.14: Sun shaded Bouguer gravity map showing profile (section) lines, gravity anomalies, boreholes along the profile lines and Karoo Supergroup geology. Figure 4.15: Model of the topography of the floor of the Karoo Supergroup along profile line X, Y. The green dotted line are the actual data values recorded, whilst the solid black line is the gravity values calculated from the proposed model. Standard regional value of 144 mgals was removed. Figures in the white boxes within the bodies are relative density differences with the surrounding basement rock (white area).The base of body 1 (blue) represents the base of the Karoo Supergroup. Bodies 2-6 are modeled as denser than the surrounding basement rock. Orange arrows indicate interpreted palaeovalleys. Figure 4.16: Model of the topography of the floor of the Karoo Supergroup along profile line X1, Y1. The green dotted line are the actual data values recorded, whilst the solid black line is the gravity values calculated from the proposed model. Standard regional value of 132 mgals was removed. Figures in the white boxes within the bodies are relative density differences with the surrounding basement rock (white area).The base of body 1 (blue) represents the base of the Karoo Supergroup. Bodies 2-6 are modeled as denser than the surrounding basement rock. List of Tables Table 1.1: Historical biostratigraphic subdivision of the Beaufort Group. Table 1.2: Lithostratigraphic subdivision of the Beaufort Group according to previous studies. Wavy lines indicate the limit of the respective studies. Cf Fig. 1.4 for the location of some of the various studies. Table 3.1: Fossils collected during the course of this study from within the study area. Table 3.2: Fossils listed in the data bases of the Council for Geoscience (GHG), University of the Witwatersrand (BPI) and the National Museum (NM). Note the absence of co-ordinates for many of the specimens. Table 4.1: List of boreholes with collar co-ordinates and depth to the basement of the Karoo Supergroup. Table 4.2: Rock densities from some South African geological units (Mare and Oosthuizen, 2000). Note the large differences between the least to most dense. Table 4.3: Thickness he various groups and formations those boreholes that lie along the profile lines as well as the percentage of the various lithologies and their density contribution. 1 Chapter 1 Introduction The geographic area covered by this study lies approximately 60 km due east of the city of Bloemfontein in the Free State Province of South Africa (Fig. 1.1). The town of Thaba Nchu is the largest of a number of settlements servicing both a subsistence and commercial farming industry. Topographically the region is dominated by flat plains with a number of prominent hills to the west, east and south. Covered by a mixture of grasses and low bushes, the slopes of these hills provide most of the outcrop which comprises sedimentary rocks of the Karoo Supergroup along with younger dolerite dykes and sills. The Karoo Supergroup covers more than half of the land surface of South Africa, some 600000 km2 (Smith, 1995), and consists of a thick succession of mainly sedimentary rocks that represent an almost continuous record from the Late Carboniferous to the Middle Jurassic, a period of about 100 million years ago (Smith et al., 1993). It has been argued that the sediments accumulated within an intra-cratonic retroarc foreland basin and deposited under a number of climatic regimes and depositional settings (Smith, 1995; Hancox and Rubidge, 2001). Currently three groups (Dwyka, Ecca, Beaufort and Drakensberg) and three additional formations (Molteno, Elliot, Clarens and Drakensberg) make up the Karoo Supergroup (SACS, 1980) (Fig. 1.2). The lithological boundaries between the groups and formations manifest changes in depositional style as well as erosional contacts. A number of boundaries between the various groups and formations are diachronous, a common feature of the Karoo. Fossilized plant and animal remains are found in abundance within certain strata and in particular those of the Beaufort Group (Hancox and Rubidge, 1997). This has allowed for the Beaufort Group to be divided into biostratigraphic units (Rubidge, 1995). The biostratigraphic and lithostratigraphic record has helped in producing a detailed basin development model of the Karoo Supergroup (Catuneanu et al., 1998). 2 1.1 Literature Review 1.1.1 Geology 1.1.1.1 Lithostratigraphy of the Beaufort Group The 1963 edition of the 1:250 000 Bloemfontein (2926) Geological Map of the area east of Bloemfontein, in the vicinity of Thaba Nchu, shows the presence of the Beaufort Group and the Molteno, Elliot (Red Beds) and Clarens (Cave Sandstone) formations. The key to the map divides the Beaufort Group into three stages, a Lower, Middle and Upper Stage. The Lower and Middle Stages consist of sandstone, shale and mudstone whilst the Upper Stage is indicated as purple and green shale and thick sandstone beds (Theron, 1963). The Molteno Formation, which overlies the Beaufort Group, is described as being composed of feldspathic sandstone grit and green shale. Theron (1970) noted that this was the subdivision (of the Beaufort Group) suggested by du Toit (1954) which was made on both lithological and palaeontological grounds (Table 1.1). Based on palaeocurrent directions and provenance studies, Theron (1970) divided the Beaufort Group in the Free State Province into four formations. Within Theron?s succession a basal Lower Beaufort Formation of bluish-grey shale and mudstone, with minor fine-grained sandstone, is overlain by a coarse grained feldspathic sandstone and arkose unit. The lower unit has a palaeocurrent direction indicating a southerly provenance area. Theron (1970) termed the latter, the Northern Beaufort Formation and suggested a proximal provenance area consisting of a crustal block within the larger basin. These formations were followed stratigraphically by the Middle and Upper Beaufort formations, both having a southerly source and distinguishable from one another by their lithology. The Middle Beaufort Formation had the largest areal coverage of the four and was described by Theron (1970) as similar to the Lower Beaufort Formation, further suggesting that within the Free State this unit may be subdivided into two formations. The Upper Beaufort Formation, according to Theron (1970), is the most distinct (in the field) of the Beaufort formations and consists of red and purple mudstone with prominent sandstone units. 3 The South African Committee on Stratigraphy (SACS; 1980) subdivided the Beaufort Group into two Subgroups, a lower Adelaide Subgroup and an upper Tarkastad Subgroup (Table 1.1). The latter is distinguished from the former by a greater sandstone:mudstone ratio (SACS, 1980). According to SACS (1980), the boundary between the two subgroups can be followed throughout the Karoo Basin except possibly in the northern and central Free State. The Adelaide Subgroup is equivalent to the Lower Beaufort of Du Toit (1954) and previous workers (Hamilton and Cooke, 1960; Haughton, 1963). It includes the Koonap, Middleton and Balfour formations, whose stratotypes lie in the southern and south eastern part of the main Karoo Basin, four to five hundred kilometres south and south east of the central Free State (SACS, 1980). In the south east of the basin, five members are recognised in the Balfour Formation, the upper two being the Elandsberg Member (grey mudstone) and the Palingkloof Member (red mudstone). Strata attributable to the Balfour Formation have been recognised within the central-eastern Free State (P Bosch, pers comm.). In the north eastern Free State the Adelaide Subgroup comprises the Normandien and Estcourt Formations. Groenewald (1989) correlated the Normandien Formation with the Balfour Formation further to the south. Later Catuneanu et al. (1998) noted that the Normandien Formation correlates with only the upper portion of the Balfour Formation. Botha and Linstrom (1978) identified the Estcourt Formation as the lateral equivalent of the Balfour Formation in north-west KwaZulu-Natal. The Tarkastad Subgroup south of 31o 30? S is divided into two formations: the arenaceous Katberg Formation and the overlying argillaceous Burgersdorp Formation (SACS, 1980). The Burgersdorp Formation had previously been correlated with the then upper two vertebrate biozones of the Beaufort Group, viz. the Procolophon and the Cynognathus Zones (SACS, 1980). In the Free State SACS (1980) proposed that the Upper Beaufort of Theron (1970) is the lateral equivalent of the Tarkastad Subgroup south of 31o 30? S, and 4 because it is not possible to distinguish two formations in this region they should be regarded as a single formation, the Tarkastad Formation (SACS, 1980). In the north eastern Free State, Groenwald (1989) divided the Tarkastad Subgroup into a lower Verkykerskop Formation and an upper Driekoppen Formation, whilst Botha and Linstrom (1978) (north western KwaZulu-Natal), subdivided the succession in north western KwaZulu-Natal into an upper Otterburn and lower Belmont Formation. Groenewald (1990) correlated the Verkykerskop and Driekoppen formations in the north eastern Free State with the Katberg and Burgersdorp formations of the Eastern Cape. A similar correlation was made between the Belmont and Otterburn formations in KwaZulu- Natal. Since the work of Theron (1970) little work has been done on the Beaufort Group in the Free State and, in particular, the central eastern Free State (J.C. Loock, pers. comm.). Individual formations and subgroups have been examined by amongst others, Groenewald (1996) and Neveling (2002). Some work has been done on the Permo-Triassic boundary which many workers argue, lies on or close to the Adelaide/Tarkastad contact (Hancox et al., 2002). The Karoo Supergroup can also be geographically divided into distal and proximal dependant on the distance from the Cape Fold Belt, the original main provenance area (Johnson, 1976). The overall tectonic setting, stratigraphy and palaeontology of the Beaufort Group have recently been reviewed (Smith, 1990a; Smith et al., 1993; Johnson et al., 1997; Catuneanu et al., 1998; Smith et al., 1993; Catuneanu and Elango, 2001 and Rubidge, 2005). Recent works by Groenewald (1996), Hancox (1998a) and Neveling (2002) have added considerably to the understanding of the continental Triassic of South Africa. Very little work has been undertaken on the facies within the central portion of the basin i.e. the modern geographical centre of the Karoo Supergroup (M Johnson, pers. comm.) which Catuneanu et al. (1998) showed as a transition from a distal to a proximal depositional setting within a foreland basin (see below). 5 Balfour Formation The entire Adelaide Subgroup has a maximum thickness, in the south eastern Karoo Basin, of 5 000 m (Johnson et al., 1997). This decreases to 800 m in the central basin and tapers off to 100-200 m in the far north (Johnson et al., 1997). The Balfour Formation has a maximum thickness of 2 150 m in the Fort Beaufort area (Catuneanu and Elango, 2001) and thins westwards to 650 m in the Graaff-Reinet area (Visser and Dukas, 1979) (see Fig. 1.4 for the geographic locality of previous studies). Within the south eastern Karoo Basin, the Balfour Formation lies stratigraphically above the Koonap and Middleton formations. Together the Koonap and Middleton formations are considered to form a single fining-upward unit (Catuneanu et al., 1998; Catuneanu and Elango, 2001). To the west of the basin, the Koonap and Middleton formations grade laterally into the Abrahamskraal and Teekloof formations. A sub-aerial unconformity separates the Middleton and Balfour formations (Catuneanu et al., 1998; Catuneanu and Elango, 2001).The top of the Balfour Formation terminates the Adelaide Subgroup, above which lies the Tarkastad Subgroup. The Balfour Formation in the south eastern Eastern Cape (proximal basin setting) is dominated by alternating layers of grey, fine grained sandstone with greenish-grey or bluish-grey mudstones and shales (Hiller and Stavrakis, 1984). Individual layers along with the whole unit display a fining-upward profile (Hiller and Stavrakis, 1984; Catuneanu and Elango, 2001). At the type locality in the Eastern Cape, six fining-upward cyclothems have been identified within the Balfour formation (Catuneanu and Elango, 2001). If the entire Formation represents a second order sequence then the 6 cyclothems within it are third order sequences (Myers and Milton, 1996; Catuneanu and Elango, 2001). The Balfour Formation sandstones can be massive (lithofacies Sm) or contain internal sedimentary structures such as horizontal stratification (Sh), planar cross-stratification (Sp), trough cross-stratification (St) and ripple cross lamination (Sr) (Hiller and Stavrakis, 1984; Catuneanu and Elango, 2001). Typically individual sandstone beds are less than 10 6 m thick. The mudstones are massive (Fm), horizontally laminated (Fl) or ripple layered (Fr) (Catuneanu and Elango, 2001). Catuneanu and Elango (2001) note that the overall fining-upward profile of the formation may be ascribed to the lower cycles being placed nearer to source than the younger and, hence, more distal cycles. The Palingkloof Member is a distinctive argillaceous unit of red and maroon mudstone with occasional light olive to light grey sandstone located at the top of the Formation (SACS, 1980). The Member attains a maximum thickness of 100 m in the south-eastern part of the basin, and thins to 40-80 m towards the centre of the basin (Hancox, 2000; Hancox et al., 2002). Recent work has identified the Permian Triassic boundary within the lower part of this member (Smith 1995; Ward, 2001; Hancox et al., 2002). Estcourt Formation In north western KwaZulu-Natal the Estcourt Formation (distal basin) is conformable with the underlying Ecca Group. The Formation consists of layers of both dark mudstone and light coloured fine sandstone, overall the package is argillaceous (Botha and Linstrom, 1978). Typically the sandstones are fine grained but medium and coarse grained units are also present. The coarse grained units consist of fresh feldspar and large quartz grains and are attributed by Botha and Linstrom (1978) to uplift in the south-eastern source area. Twenty metres of red mudrock cap the Estcourt Formation. The Estcourt Formation has been correlated with the upper Balfour Formation and with Theron?s (1970) combined Lower and Northern Beaufort formations (Botha and Linstrom, 1978). Normandien Formation The Normandien Formation, present in the north-eastern Free State (distal basin), comprises three arenaceous and three argillaceous units (Groenewald, 1989). The basal Member of this Formation (the Frankfort Member) nonconformably overlies the Volksrust Formation of the Ecca Group and consists of upward coarsening coarse grained feldspathic sandstone and green shale (Groenewald, 1989). The remaining arenaceous members (Rooinek and Schoondraai members) are fining upward succesions of fine grained sandstone and mudstone. The Rooinek Member consists principally of lenses of very 7 coarse to very fine feldspathic sandstone, which pinch out laterally over short distances (Groenewald, 1989). Coarse grained feldspathic sandstone, containing pieces of silicified wood, forms the erosional base of stacked units of upward fining sandstone of the Schoondraai Member (Groenewald, 1989). The three argillaceous members succeed each of the sandstone rich members. The lower two consist primarily of green mudstone whilst the uppermost member of the Normandien Formation is a red mudstone package which Groenewald (1989) named the Harrismith Member. Rubidge et al. (1995) equated the Normandien Formation with the Balfour Formation in the north-eastern Free State. Botha and Linstrom (1977) and Csaky (1977) equated the Lower Beaufort and Northern Beaufort formations of Theron (1970) with the upper part of the Adelaide Subgroup. Johnson et al. (1997) argued for the absorption of the Estcourt Formation into the Normandien Formation. Palaeocurrent indicators for the upper Adelaide Subgroup vary across the basin. To the north of Fort Beaufort palaeocurrent data indicates that the palaeoslope dipped predominately to the north east (Cataneanu and Elango, 2001). Variations in drainage direction are attributed by Cataneanu and Elango (2001) to changes in the strike of the orogenic loading to the south of the basin. Johnson (1976) on the other hand found the main drainage direction was to the north north-west. Essentially both directions point to a source south of the respective study area. In the north-eastern Karoo Basin the palaeocurrent direction is from the south-east, which led Botha and Linstrom (1978) to conclude that the source area was to the south-east and east of the Karoo Basin from garnetiferous basement rocks. In the north eastern Free State (northern Karoo Basin) the palaeocurrent varies through the Normandien Formation. Palaeotransport was from the east during the deposition of the Frankfort Member, north-east for the Rooinek and south south-east for the Schoondraai Member (Groenewald, 1989). Within the Free State Province the Lower Beaufort Formation (of Theron, 1970) has a south south-westerly source whilst the Northern Beaufort Formation (of Theron, 1970) is almost opposite to this with a south westerly palaeocurrent direction (Theron, 1970). 8 Groenewald (1996) described the contact between the Balfour and the Katberg formations in the north-eastern Free State Province as ?a transition zone 100 m thick that is conformable, intertongued or abrupt.? Stravakis (1980) observed a paraconformity at the top of the Balfour in the Eastern Cape. Both Hancox (2000) and Catuneanu and Elango (2001) placed an unconformity between the Balfour and Katberg formations in the Eastern Cape. Katberg Formation According to SACS (1980) the Katberg Formation is the lowermost of two formations that comprise the Tarkastad Subgroup and equates to the Middle Beaufort of Theron (1970). Initial work on the Katberg Formation was undertaken by Mountain (1946) and Johnson (1966; 1976) in the Eastern Cape in the vicinity of Katberg Pass. Stavrakis (1980) and Hiller and Stavrakis (1984) added lithological detail to the description of the Katberg Formation in the Eastern Cape. Groenewald (1996) traced and described the Katberg Formation throughout the basin and considered the Tarkastad Subgroup in the north- eastern Free State to comprise a basal arenaceous Verkykerskop Formation and the overlying argillaceous Driekoppen Formation (Groenewald, 1989). SACS (1980) questioned the division of the Tarkastad Subgroup north of 31o S into two formations. However Neveling (2002), in the south-eastern Free State, found that the subgroup was represented by two distinct formations and favoured the retention of the terms Katberg Formation and Burgersdorp Formation. The Katberg Formation is predominantly arenaceous with minor intercalations of mudstone (Groenewald, 1996; Neveling, 2002). Light olive grey, greyish yellow green, light grey and light brownish grey lithic sandstones typically make up 60-80% of the Formation (Groenewald, 1996). Johnson (1976) noted variations in the percentage of sandstone across the Eastern Cape from 95% to as little as 30%. This variation lead Johnson (1976, p 246) to define the boundary of the Balfour and Katberg formations south of 32o S as ?the horizon above which sandstone is predominant? and the boundary north of 32oS as the ?horizon above which a sudden and marked increase in sandstone abundance takes place?. The remaining 10-70% consists of greyish-red silt and clay size mudstone 9 (Groenewald, 1996; Johnson, 1976; Stavrakis, 1980). Intra- and extraformational conglomerate with clasts of mudstone, porphyritic granite, opalised wood and sandstone are found at the base of the sandstone units on scour surfaces (Mountain, 1946; Groenewald, 1996; Neveling, 2002). Hiller and Stavrakis (1984) note that in the southern Eastern Cape Province, 95% of the ?Katberg Sandstone? is fine to medium grained sandstone with mudstones making up the difference. In the south-eastern region, the Katberg Formation varies in thickness from 1238 m to 370 m (Johnson, 1976). A northward decrease in the sandstone/mudstone ratio accompanies the decrease in thickness and indicates an increasing distance from a source in the south and south-east (Groenewald, 1996). According to Groenewald (1996), the top of the Formation is not preserved in the north-western part of the basin (north eastern Free State Province, south of the town of Memel), where he gives a thickness of 245 m to 83 m. A variation in the percentage of arenaceous material also occurs with values of between 47 and 74%. In the north-east of the basin, the Katberg Formation lateral equivalent, the Verkykerskop Formation, is highly arenaceous and in places forms an 80 m thick unit (Groenewald, 1996). In the south eastern Free State poor exposure makes it difficult to determine the sandstone/mudstone ratio but an estimate of 50:50 is given by Neveling (2002). A number of facies are recognised in the Katberg Formation including scour-fill sandstone (Ss) (or the clay-pebble conglomerate (Gc) of Groenewald, 1996), horizontally stratified sandstone (Sh), low-angle cross-stratified sandstone (Sl), planar cross-stratified sandstone (Sp), massive sandstone (Sm), ripple cross-laminated sandstone (Sr), laminated silt and mudstone (Fsl) and massive silt or mudstone (Fm) (Neveling, 2002). The facies elements together build architectural elements which include channel fills, downstream accretion macroforms, sandy bedforms and various flood plain deposit elements (Neveling, 2002). These elements indicate that the Katberg Formation was deposited onto a broad braidplain by low sinuosity bedload dominated fluvial systems (Hancox, 1998b). 10 Palaeocurrent data indicates that the source area lay to the south-east and south-west and that the streams drained over a north-westerly palaeoslope (Johnson et al., 1997; Hancox, 1998b). Arid to semi-arid conditions appear to have prevailed during the depositional period (Hancox, 1998b). Burgersdorp Formation The boundary between the Katberg and the Burgersdorp formations is defined as the level at which mudstone becomes more abundant than sandstone (Hiller and Johnson, 1990; Groenewald, 1996). This contact is conformable with a transition zone of about 100 m (Groenewald, 1996). However Neveling (2002), supported by palaeontological evidence, argues that in the distal sector of the basin there is an unconformity between the two formations. In the southeast of the basin the transition zone has been considered to be 400 m thick (Stavrakis, 1980) which implies that the Katberg and the lower Burgersdorp are laterally equivalent (Neveling et al., 1999). Mudrock comprises up 50-82 % of the Burgersdorp Formation and is typically greyish-red to greenish-grey. The argillaceous beds are mostly massive but with horizontally laminated, micro cross-laminated and wavy bedding and desiccation cracks. Individual units of interbedded sandstone and mudstone can be 80 m thick (Groenewald, 1996). In the south-east, the formation attains a maximum thickness of 1 000 m in a down-faulted block north of East London (Groenewald, 1996). Northwest of this it thins to 342 m and finally 80 m in the vicinity of Senekal. To the west (the eastern Free State), thin outliers exist with thicknesses varying from 54 to192 m. At Hobhouse the Burgersdorp Formation is 192 m thick, whilst at Excelsior (north of Thaba Nchu) it reaches 102 m (Groenewald, 1996). The Burgersdorp Formation is considered to have been deposited in a depositional environment of high sinuosity rivers and extensive floodplains. The presence of paleosols indicates periods of flooding with long periods of non-deposition (Hancox, 1998b) and the 11 climate appears to have changed from arid to more humid over the depositional history of the formation (Hancox, 1998b). Palaeocurrents indicate various sources to the south south- east, south, and south south-west (Hancox, 1998b). 1.1.1.2 Biostratigraphy As the Beaufort Group contains little or no intercalated volcanic rocks few radiometric (absolute) dates are known. The lower Ecca Group (Collingham Formation) is dated at 271 Ma (using a tuff within the group, Turner, 1999) and the Drakensberg lavas at 183 + 2 Ma (Duncan et al., 1997). These two dates broadly constrain the absolute age of the Beaufort Group. However, the Karoo rocks contains a rich diversity of tetrapod fossils, which have allowed the Beaufort Group to be divided into a number biostratigraphic assemblage zones (Rubidge, 1995) (Table1.1). At present eight biozones are formally recognised and a number of subdivisions have been proposed (Boonstra, 1969; Hancox et al., 1995; Rubidge, 1995; Neveling, 2002). The biostratigraphy of the Karoo Basin has received much attention over the past century and is of great use for stratigraphic correlation of strata (Hancox and Rubidge, 2001). This abundance of fossils has allowed the strata of the Karoo to be correlated with radiometrically-dated strata from other parts of the world (Rubidge, 2005). As the Beaufort Group contains few basin wide lithological marker horizons the biozones are useful both for stratigraphic correlation and for constraining time periods (Hancox and Rubidge, 2001). Many authors consider that the main Karoo Basin of South Africa represents the type section for the tetrapod biostratigraphy of the Early-Middle Triassic (e.g. Ochev and Shishkin, 1989; Lucas, 1998). In addition to the stratigraphic importance of the Karoo fossil fauna, they have also facilitated palaeoenvironmental and basin development interpretations within the Karoo Basin. Only the upper three vertebrate biozones of the Beaufort Group are present in the eastern central Free State Province viz. the Dicynodon, the Lystrosaurus and Cynognathus Assemblage zones (Rubidge, 1995; 1:250 00 Geological Map) (Fig.1.3). Additional research 12 since the work of Rubidge (1995) has allowed for some of the Assemblage zones to be sub divided. In particular the Cynognathus Assemblage Zone has been divided into three sub- zones (Sub-zones A-C) and a Procolophon subzone has been identified at the top of the Lystrosaurus Assemblage Zone (Neveling, 2002). According to Rubidge (1995) the Dicynodon Assemblage Zone occurs throughout much of the Balfour Formation terminating at the base of the Palingkloof Member. Smith (1995) however place the top of the Dicynodon Assemblage Zone near the top of the Palingkloof Member and at the Permo-Triassic Boundary. The Lystrosaurus Assemblage Zone is by definition the Assemblage Zone within which the there is an abundance of Lystrosaurus and the absence of Dicynodon lacerticeps (Groenewald and Kitching, 1995). This definition therefore initially placed the assemblage zone boundary at the base of the Palingkloof Member of the Balfour Formation but it has subsequently been shown that there is an overlap between the two index taxa and hence the change in the biozone boundary relative to the lithostratigraphy (Smith, 1995). Botha and Smith (2007) show that the overlap is limited to only two of four species of Lystrosaurus (L. maccaigi and L. curvatus), only L. curvatus is present both below and above the boundary. Historically an additional assemblage zone, the Procolophon Assemblage Zone (or Procolophon Bed) was placed between the Lystrosaurus and overlying Cynognathus Assemblage Zone (Broom, 1906) but was later disregarded (Hotton and Kitching, 1963). Neveling (2002) showed that some taxa from the Lystrosaurus Assemblage Zone extended beyond the last appearance datum of Lystrosaurus i.e. Lystrosaurus does not extend all the way through the Assemblage Zone. The uppermost part of the Lystrosaurus Assemblage Zone is dominated by Procolophon which in the northern Karoo Basin does not extend stratigraphically upward beyond the last sandstone of the Katberg Formation (Neveling, 2002). The boundary between the Lystrosaurus and Cynognathus Assemblage Zones is relatively poor in fossils (in the south east of the Karoo basin) but it has been shown that there is some overlap in the ranges of biozone defining taxa (Neveling et al., 1999). 13 Kitching (1995b) positioned the Cynognathus Assemblage Zone in the upper two thirds of the Burgersdorp Formation and noted a thinning of the Assemblage Zone from 600 m in the Queenstown area to 20 m in the north-eastern Free State. Hancox et al. (1995) subdivided the Cynognathus Assemblage Zone into three subzones. The lowest of these subzones (Subzone A) is characterised by the presence of the amphibian genus Kestrosaurus and some archosaurs. Whilst the Subzone A occurs across the entire basin, a thinning of subzones B and C zones occurs and the latter is restricted to the more proximal portions of the basin (Rubidge, 2005). The middle, B Zone, contains fossils of the dicynodont Kannemeyeria (Hancox et al., 1995). A preliminary biostratigraphy based on fossil wood has been suggested for the Beaufort Group (Bamford, 1999) which can be used as an additional tool in both litho- and chronostratigraphic correlation. 1.1.1.3 Depositional Environments Visser and Dukas (1979) demonstrated that the Balfour Formation (upper Adelaide Subgroup) in the south western portion of the main Karoo Basin, consists of a number of upward fining sedimentary cycles. Each of these cycles has, as its base, a perennial braided stream depositional environment which is followed by a much thicker unit of flood plain facies. The overlying Katberg Formation forms the base of the uppermost cyclic unit. Visser and Dukas (1979, p152) showed that the ?coarser grained and more feldspathic? Katberg Formation sandstones were deposited in a ?typical? ephemeral braided river environment. According to Catuneanu and Elango (2001) the various lithofacies present within the Balfour Formation form the basis for three different fluvial styles, viz., deep and shallow perennial sand-bed braided systems; sand-bed meandering systems and fine grained meandering systems. The depositional environment was interpreted from a combination of lithofacies assemblages and an understanding of the assemblage of depositional elements (architectural element analysis ? see Chapter 2). 14 Botha and Linstrom (1978) concluded that the sedimentary rocks of the Estcourt Formation were deposited as river mouth bars, deltas and ?swamps? on large flood plains. A similar but drier depositional environment is argued for the upper Otterburn Formation. The arenaceous Belmont Formation was considered to have been deposited by slow flowing braided rivers onto a flood plain under dry conditions (Botha and Linstrom, 1978). In the north-eastern Free State the Normandien Formation was interpreted to have been deposited in a meandering river environment (Groenewald, 1989). Groenewald (1996) proposed varying depositional environments for both of the formations of the Tarkastad Subgroup according to their geographic locality. Based on lithofacies assemblages Groenwald (1996) concluded that the Katberg Formation was deposited in a high energy fluvial environment (braided river) in the south east, south west, north eastern and eastern areas of the central Karoo Basin. In the western portion of the basin (approximately 100 km south east of the current study area) the Katberg Formation was largely deposited in a lower energy proximal lacustrine environment with some braided channels present. The overlying Burgersdorp Formation was deposited by a meandering river system within or adjacent to a lacustrine (western region), playa lake or flood plain depositional environment (Groenewald, 1996). Within the Free State Province Theron (1970) argued that the Beaufort Group was deposited initially within a deltaic (lower Beaufort Group) and later a fluvial environment. The fluvial ?conditions? (Theron, 1970, p200) were aggrading meandering streams, ?backswamps? and small basins within a vast coastal plain. The exception to this was the perennial braided rivers of the Northern Beaufort Formation (of Theron, 1970) flooded during annual snow melt. The upper Katberg Formation in the vicinity of Senekal (north eastern Free State) was deposited under relatively rapid flow by perennial, sand-bed, braided rivers (Neveling, 2002). The more proximal (southern) exposures of the Katberg Formation show a slight variation to the Senekal outcrops, with deposition initially occurring in deeper channels on a braid plain during annual flooding. The channel sizes decrease upwards indicating a 15 decrease in palaeoslope and discharge (Neveling, 2002). An anastomosed and ephemeral fluvial environment in an arid setting was proposed for the lower Burgersdorp Formation in the southern Free State and northern Eastern Cape (Neveling, 2002). Further north deposition occurred within large semi-perennial ponds and playa lakes as well as on extensive, dry floodplains (Neveling, 2002). The ponds and lakes were fed by small anastomosed channels which also deposited material onto the flood plains during large flood events (Neveling, 2002). 1.1.2 Sedimentary Basin Setting The Karoo Basin was formed as a retroarc foreland basin that developed to the north and in front of the Cape Fold Belt in southwestern Gondwana during the Late Carboniferous (Johnson, 1991; Smith, 1995; Catuneanu et al., 1998). The fill of the basin is highly asymmetrical being thickest in the south-west and thinning rapidly to the north-east. In the south of the basin a maximum cumulative thickness of 12 km is recorded. The southern portion of the basin, 31o S to 34o S, is considered proximal to the Cape Fold Belt while the northern basin is distal and contains a much thinner sequence of rocks. Earlier workers, notably Rust (1962), Turner (1975) and Cole (1992), considered that the Karoo rocks were deposited in a single unitary subsiding basin, within which a continuous base-level rise occurred. The various stratigraphic features were explained by making use of varying subsidence and sedimentation rates. The general northward thinning of the various units was considered to be a function of the distance from the depocentre and regional slope. In contrast Catuneanu et al. (1998) considered that the transition from a proximal to a distal depositional setting, as seen in the current geographical centre of the Karoo Supergroup, represents a tectonic hinge line generated by the flexural behaviour of the lithosphere in response to subduction, compression, collision and terrain accretion along its southern margin. The loading of a tectonic plate generates an adjacent area of subsidence (foredeep) as well as a more distal but smaller upward deflection (forebulge) and secondary area of sedimentation (foresag). Loading and unloading of the tectonic plate allows for the various components (fordeep, depocentre, forebulge and foresag) to geographically shift. In 16 particular, the loading and unloading within the Cape Fold Belt allowed for the depocentre, forebulge and foresag to advance and retreat through time. This, in turn, has generated a reciprocal or out-of-phase stratigraphy. This model holds that sediment deposition occurred within a foredeep in front of the Cape Fold Belt (proximal sector) as well as in the forebulge (the distal sector). The forebulge would have become the foresag during periods of unloading and, hence, generated areas of accommodation and deposition, non deposition or erosion relative to the stratigraphic hinge line. During periods of loading these areas would have shifted. Similarly the style of deposition would have varied according to this position (Catuneanu et al., 1998). 1.2 Present Study 1.2.1 Aims and objectives of the present study Given the above discussion it is obvious that a more detailed study of the Beaufort Group in the central portion of the Karoo Basin is required as it is a critical sector between the proximal and distal settings in the reciprocal model, and is where the flexural responses should be very pronounced. The Thaba Nchu region of the Free State Province is situated between the stratigraphic hinge line developed during Adelaide times and that of the later deposition of the Tarkastad Subgroup (Catuneanu et al., 1998). The 1:250 000 Geological Map (1963) shows that the area comprises rocks from both the Adelaide and the Tarkastad Subgroups as well as from the Molteno, Elliot and Clarens formations. Fossils from the Lystrosaurus, Dicynodon and Cynognathus assemblage zones are known to have been collected within this area. The Council for Geoscience has recently mapped the area at a 1:50 000 scale so that the areal extent of the various formations and groups are understood (P Bosch, pers. comm.). At present there are two different basin development models for the Karoo, i.e., the single basin model of Rust (1962), Turner (1999) and Cole (1992) and the model advanced by Catuneanu et al. (1998) of the flexural forebulge/backbulge basins. Whilst the stratigraphy of both the proximal and far distal sectors of the upper Beaufort Group is fairly well documented, it is important that the geology, stratigraphy and palaeoenvironment of the central part of the basin be examined as well. By determining the stratigraphy, 17 palaeoenvironments, palaeocurrents, provenance and source areas, thickness and lateral extent of facies and the position of lithological boundaries within a time line constrained by palaeontology, a comparison can then be made with similar studies from elsewhere in the basin. This study aims to make an appraisal of the sedimentological, stratigraphic and palaeoenvironmental setting of the Permian and Triassic strata in the vicinity of Thaba Nchu. The following will be addressed: a) The general geology of the Thaba Nchu area and in particular the sedimentary units of the Beaufort Group will be described and discussed. b) Using the descriptions of the sedimentology, establish the palaeoenvironmental setting in which the various stratigraphic units of the Beaufort Group were deposited. c) Establish which vertebrate biozones are present within the area and their temporal and spatial distribution. d) Contrast and compare the stratigraphy, lithology and thickness of the Karoo Supergroup within the study area to that found elsewhere within the Karoo Basin and so establish the overall tectonic setting within which the rocks were deposited. e) Using the comparison with the regional litho- and biostratigraphy, see how well the study area stratigraphy fits within existing basin development models. 1.2.2 Materials and Methods Field work was undertaken during the years 2002-2004 to investigate the geology, stratigraphy and sedimentary environments of the Thaba Nchu area. During this time the overall geology of the area was established, sedimentary facies types and architectural 18 elements (according to the scheme of Miall, 1996), were determined and fossils collected and their locality and stratigraphic position recorded. The geographic locality of all studied outcrops and fossils found were positioned using 1:50 000 topo-cadastral maps of the area and a hand held global positioning system (GPS). A Silva compass and clinometer was used to determine palaeocurrent directions or trends and the dip of the strata within the area. Because of the low angle of the dip of the strata (4o <) from which the readings were taken no dip correction was made to the palaeocurrent readings. Too few readings were taken for a full statistical analysis of the palaeocurrent readings but the individual readings were plotted onto a rose diagram and the mean palaeocurrent direction obtained. A number of vertical stratigraphic sections through the Beaufort Group were measured in the area using a Jacob Staff. These are presented in the Appendix A and discussed in Chapter 2. Sedimentary features (architectural elements) were analysed and then described using both their external form as well as their internal architecture. The description of the architectural elements was undertaken by tracing them on photographs and including information (such as thicknesses of individual beds, palaeocurrent direction, fossils present, dip of beds). A primary aim of the palaeontological fieldwork was the collection of identifiable fossils which would be useful for biostratigraphic determination. All fossils collected are curated by either the Bernard Price Institute of Palaeontological Research (BPI) at the University of the Witwatersrand or at the Council for Geoscience in Pretoria. Fossil preparation for taxonomic identification was undertaken by staff at the BPI Palaeontology. The database of the National Museum in Bloemfontein was utilised for additional fossil locality information. To determine the thickness of the Karoo Supergroup and to compare the local to the regional stratigraphy and depths to basement, previously drilled boreholes and seismic data were used. The borehole data was obtained from the database of the mining company 19 Anglogold Ashanti and from Southern Oil Exploration Company (Soekor) reports held by the Council for Geoscience. Seismic data was extracted from maps and reports produced by Soekor. An additional examination of the depth to basement as well as the regional geology was undertaken using the regional gravity and aeromagnetic surveys undertaken by the Council for Geoscience. A subset from the national aeromagnetic and gravity data base was extracted using the Thaba Nchu area as the centre point and the lateral extent of the Karoo basin to the north, west and east of this point as the boundaries of the data subset. The southern extent of the aeromagnetic and gravity data subset was chosen to include the most southerly of the basinal hinge lines, as proposed by Catuneanu et al. (1998), as these reflect the flexure behaviour of the lithosphere throughout the depositional history of the Karoo Basin. Processing of the geophysical data is discussed in Chapter 4. A number of lithological hand specimens were collected in the field, and thin sections were prepared from them at the Council for Geoscience?s petrology laboratory. These were examined and described using transmitted light microscopy and photomicrographs. 20 Chapter 2 Geology In this chapter a lithological description is provided for the sedimentary rocks of the stratigraphic interval encompassing the Dicynodon, Lystrosaurus and Cynognathus assemblage zones in the study area. This interval corresponds with the upper Adelaide- and the entire Tarkastad Subgroup of the Beaufort Group. Rocks of the Adelaide Subgroup exposed in the study area comprise yellow-green sandstone bodies alternating with blue- grey and dusky red mudrock and green grey siltstone units. The Tarkastad Subgroup is subdivided into the yellow sandstones and red and blue mudrocks of the Katberg Formation, and the maroon and red mudrocks of the Burgersdorp Formation. Other than the Karoo sedimentary sequence found in the area, a large number of dolerite dykes and sills of Jurassic age are present (Duncan et al., 1997). The intrusive rocks interrupt the sedimentary sequence either horizontally or vertically at many places. Some faulting of units occurs with dolerite intruding along the fault planes. Faults are usually normal with a high dip angle (70o-85o). On Thaba Nchu Mountain a thrust fault is present (Figure 2.1) which is an unusual feature in Karoo rocks in this part of the basin. The faulting, and dolerite intrusions on Thaba Nchu Mountain complicates stratigraphic work, but for the rest of the study area, the beds are horizontal with the dip at all stratigraphic levels never exceeding 8o, and for the most part being less than 4o. Dip direction varies and is influenced by the many dolerite intrusions present. Outcrop occurs largely on or at the base of a series of low mountains that occur to the east, north-east and south of the town of Thaba Nchu in the central Free State Province. The coarser units (sandstones) form a series of small cliffs whilst the finer or thinner units form slopes between these. The latter are commonly covered in scree. Vegetation surrounding the mountains consists of grassland and mountains are covered in a combination of trees, shrubs and grasses. The vegetation and scree cover makes lateral tracing of lithological 21 units difficult. Erosional gullies and dongas provide access to fresh outcrops of the finer grained units in which most vertebrate fossil material was found. 2.1 Facies Descriptions A lithofacies is a body of rock having the same physical and chemical properties (Reading, 1986). There are a finite number of lithofacies found in sedimentary units, these are in turn produced by a limited number of physical parameters, such as water depth, flow velocity and clast size (Miall, 1996). In order to make a coherent interpretation of the depositional environment it is useful to define a set of lithofacies that are found in the sedimentary sequence being investigated (Harms et al., 1982). The scheme of Miall (1996), in which lithofacies are primarily grouped according to the dominant grain size and sedimentary fabric is followed here. 2.1.1 Coarse Grained Facies Trough Cross-stratified Gritstone (Gt) A coarse-grained unit consisting largely of trough cross-stratified coarse sandstone to grit stone occurs at a very limited number of localities and within a single stratigraphic level (Musgrave grit unit ? see section 2.3 below for description of the local stratigraphy) in the northern half of the study area (Fig. 2.2a-c). The troughs are up to 3 m wide at one locality (on the farm Chubani 9) but at other localities smaller (<1000 mm). The foresets dip up to 26o and have a palaeocurrent direction almost perpendicular to any stratigraphic unit above or below it. Intra clasts of facies Gt consist of sub-angular fresh feldspar up to 40 mm in length, angular to sub rounded fragments of granite from 10 mm to 100 mm in diameter, coarse and subrounded grains of quartz, scattered pieces of silicified wood and fragments of fossil bone. In addition to the pieces of wood, large silicified trunks thicker than 350 mm, and at 22 least 3700 mm long lie perpendicular to the palaeocurrent trend and act as palaeocurrent indicators themselves (Fig. 2.2b). The matrix of this clast supported conglomerate consists of medium to coarse sub-angular to sub-rounded quartz grains. Facies Gt represents a high energy environment. The troughs would have been deposited in a similar manner to the more typical trough cross-stratified sandstone (see below) but under much higher flow regime conditions. The presence of large trunks of wood suggests that this may have been under flood conditions if the wood had undergone any transportation i.e. if they had not grown locally. The trunks were not found within a single locality but were scattered. No root stocks were found which may indicate that they did not originate locally. 2.1.2 Sandstone Facies Intraformational Conglomerate (Sei) In his original facies code Miall (1978) included two lithofacies Ss and Se to describe two similar units of coarse to very coarse sandstone above an erosional surface. The facies Se originated with Cant and Walker (1976) who included the underlying erosional surface in their description. Miall (1996) discarded the facies Se in favour of facies Ss and excluded the erosional surface as part of his definition of this lithofacies. Intraformational or mud- pebble conglomerates have been noted from a wide range of localities throughout the stratigraphic and geographic range of the Beaufort Group. Some of the localities from which the facies have been recorded include Jansenville in the south-eastern Karoo (Turner, 1981a), Laingsburg (southern Karoo) (Kruger, 1975), Queenstown (Johnson, 1984), Victoria West (central Karoo) (Le Roux and Keyser, 1988), Beaufort West (south western Karoo) (Kubler, 1977) and southern Free State (central Karoo) (Brynard et al., 1982; Hancox, 1998a; Neveling, 2002). All except Turner (1981a) merely note or give a brief description of the facies. Groenewald (1996) (north-eastern Karoo) described a clay pellet intraformational clast supported conglomerate set in a matrix of sandstone which he referred to as facies Sei. 23 Hancox (1998a) recorded a very similar facies found in both the lower and upper Burgersdorp Formation appearing both as a thin calcarenitic basal unit as well as a thicker unit at the base of thicker sandstones. He attributed the thinner units to accumulations of sheet wash in shallow flood plain depressions. Hancox (1998a) noted that the thicker accumulations are eroded cutbank material lining the base of channels and which he labels as facies code Sei. Neveling (2002), working in the northern Karoo Basin, used the code Ss to describe a facies consisting of mudstone and siltstone intraformational clasts set within fine-grained sandstone. The coding of Hancox (1998a) will be followed in this study. Facies Sei is an intraformational mud pebble conglomerate. It forms a small percentage of the total volume of sedimentary rock but has a high frequency of occurrence. It is found as fill in erosional scours and commonly directly overlies above erosional surfaces (Fig. 2.3a). The lithofacies definition does not include the scour or erosional surface itself, as per Miall (1996). In the study area the facies is largely clast-supported but at a few localities matrix supported units are found (see Fig. 2.3b). The clasts are discoidal and rounded to well- rounded when found in the clast-supported units, however rounded to well-rounded equant and discoidal clasts can be seen in the more matrix-dominated units. Clast sizes, measured along the long axis, vary in size from 5 to 70 mm. The mean size for the clast-supported unit is 35 mm while that for the matrix supported facies is 15 mm. Clasts consist largely of mudstone and siltstone with minor fossil bone, plant and calcareous nodules. Both matrix and clast supported units have fine-grained sandstone and/or calcite as their matrix. In thin section analysis the matrix commonly contains a high percentage of calcite (see Sandstone Petrography below). Thicker clast-supported conglomerates show some sedimentary structure, having crude horizontal bedding that can be attributed to clast size sorting and the preferential (horizontal to sub horizontal) orientation of discoidal clasts. Thinner units line the base of erosional scours and like the rarer matrix-dominated units show no internal structure. Bone fragments and plant stem casts and leaf impressions are found at a number of localities but 24 in general the unit is poor in both. Fossil bones include amphibian and other tetrapod fragments as well as rare completely preserved but disarticulated bones. Basal contacts of the intraformational conglomerate are always erosive whilst the upper contact is typically abrupt. In places gradational contacts are observed with overlying massive sandstone units. Facies underlying Sei include almost all the other facies types described below including horizontally stratified, trough cross-stratified and ripple cross- stratified sandstone, as well as the mudrock facies. Contacts not found are planar cross- stratified sandstone. This is most probably due to the scarcity of that facies. Upward in the sequence low angle cross-stratified sandstone or horizontally stratified sandstone commonly follows Sei. Rarely is it followed by ripple or trough cross- stratified sandstone. Set thicknesses vary from 20 mm to 1500 mm with the thinner units often lining the bottom of smaller channel scours. The calcite-rich, matrix-supported conglomerates are generally measured in decimetres but can be up to 1500 mm thick. Hydrodynamically most pebbles require high shearing force in order for them to be eroded and transported. The increase in shearing force comes about through an increase in viscosity of the surrounding fluid or an increase in its velocity. Typically, in in a fluvial channel pebble-sized clasts are transported as bed load and are the first to be deposited as flow wanes (Allen, 1994). Successive flood surges cause the pebble to be successively transported and abraded until they reach a site of final deposition where the clast lies in maximum hydrodynamic stability. The nature of mudrock clasts implies that this cycle is not fully developed and that the clasts found in facies Sei were derived from a local source (intraformational). The mud clasts would have been partially lithified either through pedogenic mineralization and/or drying (Fagerstrom, 1967) before being eroded and redeposited in the channel thalweg as a lag accumulation. Despite this hardening, repeated cycles of erosion and deposition would have soon worn the mud clasts to small fragments. Because of its cohesive clay content, mud-clasts may be entrained in a fluvial channel either by the erosion of mud-rich banks, from a mud coated surface or from existing mud fragments such as those found in desiccation cracks (Fagerstrom, 1967). Mud clasts, as 25 well as bone fragments, calcareous nodules and silt, may also become part of the sedimentary load of a river through the erosion and collapse of channel banks or levees due to the lateral migration of channels. Mud clasts can also accumulate in aeolian deflation lags. Horizontally Stratified Sandstone Facies (Sh) One of Miall?s (1977) original facies codes, this lithofacies type has also been recognised and described from a large number of localities and stratigraphic horizons throughout the Karoo (e.g. Stravakis, 1980; Turner, 1981b; Groenewald, 1996; Hancox, 1998b; Neveling, 2002). Facies Sh forms the bulk (in frequency and volume) of all the sandstone facies and is prevalent throughout the area. It comprises fine to very fine-grained sandstone which are horizontally stratified (Fig. 2.4a). In the study area facies Sh is either greenish grey (5GY6/1) or light yellowish brown (10YR4/6) in colour. The sandstone contains poorly to well-sorted fine quartz sand with a percentage of silt or clay (see below for the petrographic description). An exception to this occurs at a single locality (see the discussion on the crevasse splay below) where a medium to coarse grained sandstone contains horizontal bedding. Parting lineation is frequently present on the upper surfaces and is used as an indicator of palaeocurrent trend (see Fig 2.4b). Parting lineation, according to McBride and Yeakel (1963), is due to the orientation of the long axis of the individual grains such that the lineation is orientated in the same direction of the original depositing current and is thus useful as an indicator of palaeocurrent trend. Thicknesses of Sh beds vary from 200 to 2500 mm. It is commonly found above units of facies Sei. Facies Sh is also associated with low angle cross-bedded sandstone (facies Sl, see below), trough cross-stratified sandstone (St), planar cross-stratified sandstone (Sp) and the finer grained facies. Abrupt, erosive and gradational contacts occur. Basal contacts are gradational with facies Sei, abrupt with planar cross-stratified facies and low-angle 26 cross- stratified sandstone facies, and erosive with respect to trough cross- stratified sandstone and ripple cross-laminated sandstone facies. Upper contacts with ripple cross- stratified sandstone, planar cross-stratified sandstone and low-angle cross-bedded sandstone facies are abrupt. This facies usually occurs as part of a sequence of low-angle cross-stratfied sandstone, trough cross-stratified sandstone and ripple cross- stratified sandstone facies. Wood impressions, wood casts and black carbonaceous material are found along with fossilised tetrapod post-cranial elements. Carbonate concretions commonly occur in this lithofacies. These range in size from 20 mm to 2000 mm and are round or irregular in shape and consist of calcite-cemented clay, mud or silt. In places they can be seen to be displacing the horizontal lamination indicating that they were precipitated and cemented prior to the compaction of the sandstone. Miall (1996) noted two separate conditions for the production of horizontal laminae to form facies Sh viz. upper and lower flow regime conditions. The presence of parting lineation is an indication of upper flow regime conditions (Miall, 1996). Based on the results of laboratory experiments, Bridge (1978) postulated that the deposition of the sediment under upper flow regime conditions occurred as a result of the interaction between the turbulent boundary layer, bed shear and the sediment particles. Paola et al. (1989) in generating parallel-laminated deposits in flumes, suggested that the laminae are caused by deposition of extremely low amplitude bedforms produced under upper flow regime conditions. The varying thickness of the laminae is due to the difference in grain size. McKee et al., (1967), Sneh (1983) and Stear (1985) all recorded the production of thick units of facies Sh (upper flow regime) during a single flash flood event. Deposits of horizontally stratified sandstone formed under lower flow regime conditions consist of coarse to very coarse sandstone (Miall, 1996). Stavrakis (1980) recorded fine grained lower flow regime beds within the Beaufort Group. He argued that these were formed by particle settling as stream flow stops. Lower flow regime horizontally laminated 27 sandstone can be distinguished in the field by the absence of parting lineation, an uneven fracture and the frequent occurrence of mottling (Stavrakis, 1980). In the study area the frequency of parting lineation on the surfaces of facies Sh indicates that this facies was deposited predominantly under upper flow regime conditions. No distinction will be made between the two types but in interpreting the depositional environment the presence of parting lineation will imply the presence of upper flow regime conditions. Low-Angle Cross-Stratified Sandstone (Sl) This facies was first described by Cant and Walker (1978) and then by Rust (1978). Miall (1996) described facies Sl as being similar to facies Sh, with the presence of low-angle (<20o) cross-stratification distinguishing it from horizontally stratified sandstone. The cross-stratification is frequently asymptotic to the upper and basal bounding surfaces and dips at a low angle (Miall, 1996). Groenewald (1996), Hancox (1998a) and Neveling (2002) all recognised this facies within the Beaufort Group. Neveling (2002) stated that facies Sl together with facies Sh are the dominant sandstone facies in the Beaufort Group (Tarkastad Subgroup) in the south eastern Free State. Lithofacies Sl consists of fine-grained greenish grey (5GY6/1) or light yellowish brown (10YR4/6) sandstone (Fig. 2.5). The coarse fraction consists of fine quartz grains with quartz overgrowths and the matrix is siltstone or mudstone with minor calcite. The grains are sorted to poorly sorted and rounded to sub-rounded. Laminated to thinly bedded facies Sl occurs above gradational contacts with lithofacies Sei or sharp or gradational contacts with Sm and Sh. Occasionally facies Sl can be found between two units of Sh. Miall (1996) interpreted facies Sl as either plane beds deposited on initially dipping surfaces or as the internal form of some types of dunes. Trough Cross-Stratified Sandstone (St) In the study area facies St consists cross-stratified elongate scoop shaped sets (Fig. 2.6a). It occurs as fine-grained greenish grey (5GY6/1) or light yellowish brown (10YR4/6) 28 sandstone. The laminae or thin beds (<1cm) are concave upward and asymptotic to the underlying bounding surface. Facies St is found in single or multiple sets (cosets). Individual sets are between 30 mm and 1200 mm thick, with a mean thickness of 50 mm. Total thickness of cosets are between 1000 to 1500 mm thick and consist of two to three sets. Frequently only a single set is observed. Trough widths range from 1000 mm to 3500 mm with the average being 3000 mm and lengths from 500 mm to 5000 mm. Width to length ratios are important for calculating discharge. Trough cross-stratification is typically found within the thicker sandstone units (Fig 2.6b), usually in association with units of facies Sh or Sl. Basal contacts are typically sharp and directly above facies Sh. Occasionally the basal contacts are erosive into underlying Sh forming scooped hollows which are filled with St. The finer grained facies are frequently found above facies St but in most places the contacts are obscured by soil and scree. Where sandstone facies (Sh, Sl) occur above St they are usually erosive into it and plane off the top of the set. The exception is ripple cross-laminated sandstone, which is in sharp contact with St at all localities. Harms et al. (1975) and Miall (1996) noted that tough cross-stratification is formed by the migration of three dimensional dunes formed during the transition from upper to lower flow regime conditions. The laminae represent the in-filling of scour hollows beneath the lee face of the dunes. A number of different sedimentary environments produce this facies including braided rivers (Bristow, 1993), meandering rivers (Nanson, 1980) and flashy ephemeral streams in arid environments (Stear, 1985). Planar Cross-Stratified Sandstone (Sp) Facies Sp is rare in the study area and in the few places that this facies can be seen it comprises fine-grained greenish grey (5GY6/1) sandstone. Facies Sp has laminae or thin beds (<1.5 cm) that dip in a down current direction and intersect the upper and lower bounding surfaces at a steep angle (>20o) (Fig. 2.5). Units are thin, never more than 0.5 m in thickness, with the average thickness being 0.3 m. Cosets consisting of two sets of Sp 29 are very rare and these are present above facies Sei or massive sandstone (Sm). Upper and lower contacts are sharp and usually flat. Planar cross-stratification is formed by the migration of two dimensional straight-crested mid channel dunes along a planar surface (Harms et al., 1975; Miall, 1996). As the dune advances sediment transported up the stoss side avalanches down the lee side and accumulates forming the planar cross bedding surfaces. Facies Sp may also be formed by the lateral growth of stream banks (i.e. transverse bar) and on downstream side of point bars. Ripple Cross-Laminated Sandstone (Sr) Rippled surfaces are present throughout the study area at the top of sandstone packages (see fig. 2.7). Units thicker than just the top surface are not common and where present they consist of cross-laminated out of phase units 10 ? 50 cm thick (Fig 2.8). Facies Sr typically occurs in abrupt contact above facies Sh and St and is commonly draped by a mudrock veneer. The ripples have a wavelength of 100 ? 200 mm and amplitude of less than 40-mm. Both symmetrical and asymmetrical ripples occur and if planimetrically exposed they can indicate palaeocurrent direction or trend. Similarly both straight crested as well as sinuous crested ripple surfaces occur. Facies Sr is generated by the migration of chains of ripples (Allen, 1968; Miall, 1996) and in fluvial setting are commonly formed by sediment ?dumping? during waning flood phase. Massive Sand (Sm) Facies Sm was initially not recognised by Miall (1977) but has subsequently been recognised by him and a number of other workers (Miall, 1996; Martin and Turner, 1998; Jones and Rust, 1983). 30 A number of units in the study area appear to consist of apparently massive fine-grained sand i.e. without any bedding, lamination or other sedimentary structure (Fig. 2.9). The units are not laterally extensive and are never more than 10 m in width. Where true massive units occur they are found in abrupt contact with facies Sei in single units of 300 - 500 mm thickness. The upper contact is either abrupt or erosional below facies Sp or Sh. A few scattered mud-pebbles can occasionally be seen in fresh surfaces. These are less than 20 mm in diameter and are discoidal. There are numerous ways facies Sm may be formed including, mass transport of sediment under gravity flows, post-depositional dewatering and bioturbation (Miall, 1996). The scattered pebbles found within the few units in the study area supports the idea of mass transport and rapid dumping of sediment. 2.1.2.1 Sandstone Petrography With one exception all the sandstones in the study area are very fine to fine grained. The samples examined contained either very little matrix or were close to greywackes. A similar wide variation in the roundness of the grains was also observed. With quartz grains ranging from well-rounded to angular, but on the whole being rounded. Individual slides show variations from very angular to rounded and rounded to well-rounded. Typically the grains are quartz although both lithic fragments, peloids and feldspars were observed. The peloids are formed from lithified paleosoils and are well rounded. The quartz is predominantly mono-crystalline but some poly-crystalline grains were observed. Other grains consist of rock fragments, feldspars (largely plagioclase), heavy minerals and minor biotite. Quartz makes up 75-80% of the rock with the remaining percentage consisting largely of feldspars or rock fragments. Some quartz overgrowth was observed along with hematite staining. The matrix consists of organic material, silica and calcite. Some samples show quartz grains included within a feldspar grain and points to a granitic provenance (Fig. 10a-10f). 31 2.1.3 Fine Grained Facies Most of the Thaba Nchu area is covered by fine-grained sedimentary rocks which weather rapidly with the result that sedimentary structures within them are difficult to see. Furthermore they are often covered in scree from the overlying sandstone and sills. Hence fresh outcrop is limited and typically found directly beneath overhanging sandstone units, in erosion gullies or active stream beds after heavy rainfall. This study used a slightly different approach to that proposed by Miall (1996), an approach which has been successfully applied in other studies of the Beaufort Group (Hancox, 1998a; Neveling, 2002). In addition to the three fine grained facies (Fl, Fsm and Fm), pedogenic overprint (P) is included. The relative proportion of each of the three facies is difficult to determine in the area because of the nature of the outcrop but they form at least 65% of the total stratigraphic thickness for the area. Horizontally Stratified Sandstone, Siltstone and Mudstone (Fl) Facies Fl consists of thinly-bedded, horizontally and ripple-cross stratified units of fine or very fine sandstone, siltstone and mudstone in varying proportions (Fig. 2.11). As weathering produces discontinuous (vertically and laterally) outcrop the relationship between this and other facies cannot always be determined. Similarly internal contacts are difficult to recognise but where they can be distinguished they appear to be abrupt, with occasional erosive contacts of sandstone into siltstone and mudstone. The individual horizontally stratified sand-, silt-, and mudstones couplets of this facies occur in units of 10 to 300 mm, the units are intercolated with each other. Typically this occurs in a fining upward succession of sand-, silt-, and mudstone but some coarsening upward sequences are present. These are not as thick (centimetre scale) as the fining upward sequence (tens of centimetres). The sequence is in places interrupted by concretionary nodular layers (palaeocrete), which are often coated in mudrock but set within sandstone. Within these concretionary nodules several therapsid fossils were found. Pedogenic features are common, particularly mudstone cutins, rhizocretions and root traces. 32 Facies Fl is the coarsest of the overbank environment facies. The sand fraction represents deposition by traction currents whilst the mud and silt were deposited from suspension (Miall, 1996). Laminated Siltstone and Mudstone (Fsm) Similar to Fl, this facies consists of laminated mud- and siltstone but without the sandstone (Fig. 2.12) of facies Fl. Palaeopedogenic nodules, rhizocretions and burrows are present as well as vertebrate fossil material which occurs as isolated bones or occasionally as articulated skeletons. The low energy setting of this environment most probably means that it was deposited more distal to the channels that produced facies Fl (Miall, 1996). The poor nature of the outcrop makes the origin, traction currents or settling, difficult to establish. Massive Siltstone or Mudstone (Fm) Miall (1977) initially considered this facies to be the result of mud drapes in low-stage channel abandonment but subsequently included massive mudstone beds (Miall, 1996). Neveling (2002) included both massive mud- and siltstone in his description of facies Fm. The nature of the outcrop in the study area is such that mudstone and siltstone weather away quickly to fragmentary chips. Weathering obscures any possible bedding or laminae making it appear massive when thicker units are observed. Mud drapes occur on the upper surfaces of sandstones, particularly facies Sh. Thin (10-30 mm) beds of mudstone occur between units of facies Sh. The thickness of these thin beds pinches and swells. Where outcrop of thicker mud- or siltstone does occur, it is in units of up to 500 mm or possibly greater (Fig 2.13). The mudstone is a variety of colours including olive black (5Y2.5/1), dark blue green (2.5Y3/2), dusky red (5R3/4), greyish red purple (5RP4/2), greenish grey (5Y7/2) and rarely black. When thicker units of facies Fm are exposed in outcrop with other facies it is always facies Fl and Fsm. 33 According to Miall (1996) this facies represents the lowest energy sedimentary environment and hence the most distal from the main fluvial channel of all of the facies. Typically it forms in standing bodies of water such as small playa lakes and flood plain pools (Miall, 1996). Neveling (2002) argued that the facies was deposited in a very low energy environment distal from the main water-distributing channels and that suspension settling was the dominant depositional environment with bedload deposition playing a minor role. Pedogenic Features Several pedogenic features are present in the study area. The most obvious feature is the calcrete layers that form horizons which are more resistant to weathering than the fines (Fm, Fsm and Fl) in which they occur. Smith (1990b) identified a number of different pedofacies with different calcrete horizons and linked these to the distance from the channel and so proposed a sequence of paleosol maturity. Other pedogenic features include sand dykes, invertebrate burrows, tetrapod burrows, rhizocretions, cutins, root traces, concretionary nodules and cone in cone structures. Calcareous nodules are ubiquitous throughout the stratigraphic package that makes up the geology of the Beaufort Group in the vicinity of Thaba Nchu. They occur in a wide variety of sizes from 1cm to 1.5m diameter (Figure 2.14a) and occur in sandstone as well as the finer grained facies. Typically, but not exclusively so, they are rounded and smooth surfaced. The smaller nodules are usually more equant with some forming near perfect spheres. Large nodules are typically elongated and flattened. Calcite rich, they also contain clasts of sand and finer grained sediments. In thin section some contain fragments of fossilised bone (Figure 2.14b). Although calcareous nodules are here described under ?Pedogenic Features? it is likely that those found within sandstone are of diagenetic origin. At some localities their relationship with the enclosing sedimentary rock clearly shows the later to be displaced by the nodules post deposition and so indicating a diagenetic rather than pedogenic origin. It more likely 34 that the small spherical peloids and the nodules within facies Sei are of pedogenic origin. These clasts would have been entrained during bank collapse and deposited as flow waned. 2.2 Architectural Element Analysis Historically the vertical relationships between facies, within a stratigraphic unit, have played an important role in the interpretation of the depositional environments (Miall, 1996). Different fluvial environments may however produce similar vertical associations or a single environment may have different vertical associations (Miall, 1977). To overcome this limitation the concept of architectural element analysis has been developed (Miall, 1996). The basis of architectural analysis is the architectural element, which is characterised by its facies assemblage and internal and external bounding surfaces which define its geometry. The architectural element is the three dimensional building block of a sedimentary sequence which, either on its own or combined with other elements, helps in defining the fluvial style of the original depositional environment (Miall, 1996). Miall (1996) proposes a subdivision of fluvial architectural elements into those formed within channels and those of the overbank environment. These architectural elements consist mostly of macroforms, which are the end result of long periods of deposition and include point bars, side bars and other compound elements (Miall, 1996). The subdivision into channel and overbank environments is followed in this study. The relationship between any two vertically adjacent sedimentary units is dependant on the surfaces dividing them. Just as there is a hierarchy of depositional units (micro-, meso- and macroform) so there is a hierarchy of bounding surfaces (Miall, 1996). As the bounding surface increases in value so the scale of the depositional unit enclosed by the bounding surface increases and usually also the time period it represents. The classification of Miall (1996) is followed here, where first order surfaces enclose bedform sets, second order 35 surfaces define co-sets, third order surfaces define bar or macroform growth increments (this includes reactivation surfaces), fourth order surfaces define minor channel bases or the convex up tops of macroforms and fifth order surfaces the bases of major channels. A single architectural element, as with a single vertical profile, should not be used to define the fluvial style for any particular stratigraphic unit (Miall, 1996). The limited and discontinuous nature of the outcrop in the study does not allow for a complete description of the fluvial style of the various stratigraphic units using this technique. Nevertheless the description of those architectural elements present may assist, in the absence of other methods, in determining the fluvial style. Hence the inclusion below of those architectural elements found within the study area. Figure 2.15 shows the location of the various architectural elements described below. Figure 2.41 shows the position of the architectural elements in relation to the local stratigraphy. The architectural elements, along with the lithofacies and additional data will be used to interpret the depositional environment in Chapter 5. 2.2.1 Elements Formed within Channels These are mostly macroforms as defined by Jackson (1975) and are the cumulative effects of long term sedimentation (Miall, 1996). They include bars, channels and flood sheets. 2.2.1.1 Channels (CH) This element was erected to describe a channel fill that is simple and cannot be broken down further into other elements. There are a number of different channel styles within the study area, with both minor and larger channels being found. Minor channels are taken as being smaller than fifty meters in width while the larger compound channels are built up of a sequence of stacked sand filled smaller channels. Although flood-plain channels should fall within the description of minor channels they have their own particular geology and geometry and so are dealt with under ?Flood-Plain Elements? (see below). 36 Major Channel Complexes (Element CHm) Many of the thicker sandstone units in the study area are clearly large channels which are filled with thick sequences of facies Sh and to a lesser extent St. The basal fifth order surfaces include channel margins and bases that have eroded into the underlying finer grained sediments. The floors of the channels always comprise a unit of Sei which can be up to 50 cm thick and laterally extensive. Local erosional scours can be of a similar depth but overall the basal surface is relatively smooth with channel margins that dip up to 20o but more typically are less than 15o. The fill of these channels is, in most places, a sequence of massive units (2-3 m) of Sh followed by St. As one moves vertically within the channel fill it divides into smaller channels which usually overlap and cross cut one another. An example of a major channel fill is present on the eastern side of the ridge at Moroto Mountain (26o 59.23? E 29o 05? 17.00? S, Moroto 68). The outcrop is 150 m long and 10 m thick were most fully developed (Fig 2.16a-b). The outcrop comprises largely facies Sei, Sh, St and some Sl and Sr. Fragmented fossil bone and wood occur in the scour filling of facies Sei. The lower contact with the underlying fines is largely obscured; nevertheless the overall form of the channel can be followed. Channel margins, where seen, dip at between 10o and 20o toward the axis. The top 4 m of the channel is made up of a number of smaller channels that cross cut one another, as well as the underlying Sh and St packages. These channels decrease in size upward from 20 m to 10 m wide in cross-section. The upper, minor cross-cutting channels consist almost exclusively of facies St. They cut into the underlying sandstone and have little or no basal conglomerate, and also are not capped by mud or other fines. Multistoried Channels (Element CHt) A laterally persistent but thin (2 m) sandstone unit occurs just above the base of the Lystrosaurus Assemblage Zone in the study area. It outcrops in the vicinity of Thaba Nchu town (Thaba Nchu Townlands), near the settlement of Moroto, on the farm Abramskraal 37 65 (at 026o 51.266 E 29o 06.83 ?S) (Fig. 2.17) and in the south at Sepenare?s Hoek 759. At the Abramskraal 65 and Sepenare?s Hoek 759 localities a sequence of small stacked sandstone channels are found in thin but laterally extensive outcrop. Locally this sandstone is erosive into the underlying mudrock bearing units (Fsm) of the Lystrosaurus Assemblage Zone. The outcrop is capped by scree and in places facies Fsm forms the upper contact. Being east-west orientated it is nearly perpendicular to the palaeocurrent direction of the channels, which varies from 352o to 004o. The sandstone narrows and interfingers with finer grained units of facies Fsm and Fl to the east, and terminates against a dyke to the west. Individual channels are generally 3-10 m wide, with a maximum width of 25 m and shallow with a mean preserved thickness of 500 mm. Thicker and more massive channels are found in what is interpreted to be the thalweg or main channel. These larger channels contain water escape structures and concretions with a diameter of 400 mm. The channels are stacked in a brick-work like fashion on top of, and to the side of, each succeeding channel. Individual channels partially truncate and erode into the underlying channels with marginal slopes of between 5o and 15o. Most consist entirely of facies Sh or Sr, however the more developed channels contain a series Sh-St-Sr. A few channels are lined with a thin layer (<10 mm) of Sei at their base. The troughs on the top surface of facies St are less than 500 mm in width and are in the upper channels no more than large ripples. 2.2.1.2 Downstream Accreting Macroform (DA) Almost all sandstone elements formed within a channel complex will display evidence of having developed by both lateral accretion and vertical aggradation. Miall (1996) distinguishes between bar forms that accrete predominantly on their lateral margins from those that aggrade vertically. He further sub-divides the former into those in which accretion is predominantly downstream (Downstream Accretion Macroform) and those for which, third order, lateral accretion surfaces are the main method of bar form growth 38 (Lateral Accretion Macroform, see below). Bars do not simply grow in a single direction; instead they usually have both lateral and downstream development. The distinguishing feature, according to Miall (1996), is the relationship between accretion surface and cross- stratification. Where the two are within 60o of one another then the element is a downstream accreting form (Miall, 1996). In order to determine the direction of growth it is therefore necessary to view the element in three dimensions (Miall, 1996). When identifying a downstream accreting macroform, the internal bounding surfaces should dip at less than 10o in a downstream direction and enclose cosets of downstream orientated flow-regime bed-forms. These surfaces may display a slightly convex up profile. For the Thaba Nchu area the requirement of three dimensional outcrops presents a problem given the nature of the outcrop. Although these elements are believed to be ubiquitous in fluvial deposits (Miall, 1996) only a few examples from a single location (outcrop) were found in the study area. On the farm Abramskraal 65, stratigraphically above the outcrop of CHt described above, is a second outcrop of sandstone (Figure 2.18a and b) which is stratigraphically positioned above a series of Fsm and pedogenic horizons, within which specimens of Lystrosaurus were found. The outcrop is orientated almost perpendicular (340o-160o) to the lower lying Stacked Sand Channels and has a relatively constant NW-NNW palaeocurrent direction, but with some smaller channels at high angle to this. A series of small to medium sized (5- 10 m) sand filled channels and predominately downstream accreting bars combine to form a larger channel which has a fifth order surface as its base. The later elements are 5-10 m long in lateral profile. The nature of the outcrop is such that total width of the elements could not be measured, but they are at least 1.5m wide. At three separate localities downstream accreting bars were observed within the same palaeochannel. Two of these were nearly identical and consist of stacked units of Sh one metre thick with flat abrupt bases that onlap and cross-cut similar units further upstream (Fig. 2.18a). Some units have a thin (10 mm) base of Sei and occasional ripple tops but otherwise consist solely of facies Sh. The bars are ten metres long across the face of the outcrop, which is oblique to the palaeocurrent direction and so is not a true width (Fig. 39 2.18a). The downstream angle of the accretionary surfaces is a fairly constant 8o with steeper and flatter angles observed in places. The outcrop at these points is vertically thin with the two bars separated by a sandstone-filled channel. The top of the outcrop is eroded therefore not possible to determine the original thickness of the bars and hence channel depth. The third downstream-accreting bar found at Abramskraal 65 (see Fig 2.18b) is built up of wedge or lens-shaped units each consisting of the upward facies series Sei -Sh/Sl-St-Sr and each with flat erosive bases. Some units lack facies Sei but are still erosive into the underlying unit, whilst others consist solely of facies Sei with a few centimetres of medium to fine sandstone interstitial to, and overlying, the conglomerate. The units range in thickness from 200 to 600 mm and the bars are 4 to 6m in wide. The wedge shaped units dip downstream at an angle of less than 10o with the wedge pointing upstream and on lapping onto the underlying unit. The units accreted both vertically and in a down-stream direction with the fourth order surface of the element distinctly convex up. 2.2.1.3 Laminated Sandstone Sheets (LS) Thin but wide, stacked packages of tabular sandstone sheets are present between relatively thick sequences of finer grained facies (Fm and Fl) in the lower horizons of both the Middle sandstone and the Mohono mudrock units (see section 2.3 below). Never more than three metres thick, they extend for 100-120 m and terminate laterally in fine grained facies. The units are separated from one another by flat, second or third order surfaces as defined by Miall (1996). The base of the macroform rests on a fifth order surface. Where outcrop is sufficiently exposed, this fifth order surface is followed in a few places by a layer (<20 mm thick) of facies Sei. The rest of the macroform is made up of units of Sh (90%) followed occasionally by thin units (<200 mm) of St. Laminated sandstone sheets have been described by a number of workers e.g. Tunbridge (1981) and Stear (1985) from the rock record, and McKee et al. (1967) from modern environments. Stear (1985) compared an example from the Beaufort Group with a modern flood event in which multiple sand sheets were generated by a single flood event and found 40 a large number of similarities. The sandstone sheets within the study area closely resemble Miall?s (1996) description of laminated sandstones (LS) viz. tabular sheets of facies Sh with minor St gradationally capping the individual sequences. One of the better exposed examples of element LS occurs on the eastern slopes of the ridge (26o 59.25? E 29o 05.25 S; Moroto 68) that juts out of Moroto Mountain (Fig. 2.16a), southwards toward the village of the same name. The outcrop consists of stacked tabular sandstone sheets which are in turn predominately made up of facies Sh. The macroform consists of a series of stacked (vertically accreting) sand sheets, each one between 400-1100mm thick, which are separated by abrupt contacts and/or rare erosional surfaces. The individual sand sheets are laterally extensive stretching the length of the outcrop, 120m. The element is bounded at the bottom by a 5th order surface. Poor exposure does not show the presence of a channel margin or cutbank but given that the underlying facies are from the overbank environment it is possible that the entire element lies within a channel with a low angle margin. The entire macroform dips very gently down stream but the individual units do not accrete in that direction i.e. they accrete vertically. The presence of thick packages of parallel laminated sandstone (lithofacies Sh) is an indication that the sandstone sheets were deposited in an ephemeral stream within a semi- arid area as lithofacies Sh is not a feature of perennial streams (Tunbridge, 1981). It is likely that the deposition occurred during a single flood event or within a short time span of one another given the low preservation potential of ripple tops (Stear, 1985). 2.2.2 Elements of the Overbank Environment The lower energy and unconfined nature of the overbank depositional environment generally results in most of the elements comprising finer grained facies (Miall, 1996). Scree and soil cover these facies at most localities making the identification of the architectural elements in the overbank environment difficult. Two of the four elements described below (flood plain channel and a crevasse splay) are formed largely from sandstone facies (Fl, Sh and St) which are better exposed. A number of small exposures 41 have examples of the third element of the overbank environment, pedogenic overprint. Only a single locality has sufficient exposure, both laterally and vertically, to allow for a proper description. 2.2.2.1 Floodplain Fines ? Pedogenic Profiles (FP) On the farm Melden Drift 28 an atypical outcrop occurs along the banks of a small stream alongside the road from Tweespruit to Rooikraal (26o 57.4?E 29o 08.93?). The outcrop consists of a 2-3 m thick series of horizontally laminated red and green fines (see Fig 2.19a & b). The thickest and best exposed portion of the outcrop is about 70 m long but the entire exposure stretches for about 200 m tapering off to the north-east further downstream. The sequence consists almost exclusively of facies Fsm but with occasional lenses of very fine sandstone. A major feature of the outcrop is the eight laterally continuous layers of palaeocalcrete that can be used as markers. Fossilised bone, both fragmentary and complete pieces, are found scattered within the calcrete layers. Pedogenic features include burrows and rhizocretions, coalesced concretionary calcrete layers and claystone cutins. The carbonate beds are largely horizontal with one or two units sub horizontal in places and having low angle dips (<10o). At the foot of the outcrop, and forming a platform, is a palaeopedogenic calcrete horizon similar to that described by Smith (1990). The platform is extensive and can be followed both across and down the stream (40 m x 20 m). It consists of hummocky red mud coated coalesced concretions with gaps between them in places forming a honey comb texture. Smith (1990) argues that these continuous sheets of calcrete formed just above the groundwater table in the zone of capillary rise and indicate an arid environment. The horizontal laminae and the fine-grained nature of the outcrop together with the pedogenic features strongly indicates an overbank environment, particularly one in which there was sufficient time for pedogenesis to occur. Mud cracks indicate the presence of some clay minerals and periods of drought. It is possible that the deposit was formed in a distal floodplain hollow in a wetter palustrine (pond), playa or marginal carbonate mudflat 42 environment. The quite setting of this environment would allow for the settling out of mud and clay sized particles whilst the mudcracks show periodic drying out of the environment. 2.2.2.2 Flood Plain Channel (CHf) Transportation of sediment onto the floodplain may occur through a number of mechanisms including progressive avulsion of the main river channel through minor channels on the floodplain (Smith, 1971). The size (laterally and vertically) of element CHf indicates that it was not part of the regional drainage system but was a more local feature, either as the local tributary drainage network or part of a distributary system. In either case this element would have been formed in an environment removed from the main channel. The exact extent of the ribbon and sheet sandstones within the study area cannot be properly determined. They appear to be largely confined to certain of the stratigraphic intervals of the study area viz. the Dubbledam mudrock and Townland fines units (see section 2.3 below). To the west of Thaba Nchu Mountain on the farm Dubbledam 688 (and at 26o 51.27?E 29o 15.46?S) a series of dongas expose outcrop of the Dubbledam mudrock unit comprising a thick succession of finer grained facies (Fl, Fsm and Fm). A number of ribbon and sheet sandstone channels are present, one of which is shown in Fig. 2.20. Here a sandstone channel cuts into an underlying overbank environment and accompanying fines which show a number of pedogenic features including rhizocretions, root traces and calcrete concretions. The exposed portion of the channel is 12 m in width and 3 m thick in the centre and consists of a sandstone cap and a relatively finer grained base. The base of the channel is green, fine grained and silty sandstone that erosively cuts into the underlying fines. Erosion hollows are up to 300 mm deep in places. 43 The overlying sandstone unit has a sharp basal contact and is slightly coarser grained. Consisting of Sh it pinches out laterally from an initial thickness of 120 mm in the centre. Above this unit of Sh, and in abrupt contact with it, is a series of thin units of facies Sl, Sh and Sr. A full sequence contains Sh/Sl-Sr, however typically only facies Sh with a ripple top surface is present. Both upper and basal contacts of each unit are abrupt and flat. Some basal surfaces contain sole marks and the upper surfaces a mud veneer. The units of channel fill do not stack directly on top of one another but rather have the centre of each channel unit staggered. On the hill positioned within the Thaba Nchu Townlands (at 26o 52.08? E 29o 12.78? S) is another small channel that outcrops from within scree covered fines (Fig. 2.21). Consisting of overlapping lenses 3-5 m in width and 100 ? 500 mm in thickness the outcrop is 30 m long and 1.8 m thick. Palaeocurrent indicators (flute casts) give a direction of 276o which is nearly 70o west of the palaeocurrent direction of the much thicker and more extensive sandstone outcrop above. The lenses consist of facies Sr or Sh capped by either Fsl or Fm and with a thin base of mud pebbles. Some units erode into the underlying units. 2.2.2.3 Crevasse Splay (CS) A good example of an overbank splay system is present along the Feloana/Koranna River below the dam of the same name (26o 45.18? E 29o 06.93? S). The outcrop is situated stratigraphically within the Dubbledam mudrock unit (see section 2.3 below). A series of thinly interlaminated medium to coarse grained sandstone (Sh) and mud ? or siltstone 5 m thick is exposed along a bend in the river over a combined distance of 250 m. The entire package rests erosively on a succession of facies Fm which here includes a black carbonaceous mudstone as well as carbonaceous nodules (Fig. 2.22a and b). It is overlain by over 2 m of very coarse grained sandstones of facies St and Sh. Interlamination is such that in any 200 mm of outcrop there will be up to four cycles of medium to coarse grained sandstone alternating with black or green mudstone and siltstone. 44 The sandstone layers are laterally continuous, thickening in places to form small channels which are 50-150 mm thick and 2.5 m wide. These channels have erosive bases which are overlain with a layer of facies Sei. Where they are thick enough they consist of facies Sh or St but for the most part are filled by facies Sm. The sandstone forming them is poorly sorted with subrounded clasts and is quartz rich. The tops of the channels are often veneered in green mudstone. The sandstone laminae and individual beds fine upwards from a coarse grained to medium grained. The upper 2 m of the package is dominated by facies Sr. The beds and laminae dip at a very low angle (<10o) in the palaeocurrent direction. Units of facies Sh, St and Gt overlie this sequence. The latter facies outcrops poorly and within a single stratigraphic interval. Both the underlying crevasse splay units and the coarse grained units overlying it have an easterly to south easterly palaeocurrent direction (see Chapter 5 for discussion on the palaeocurrent direction of the various lithostratigraphic units). 2.2.2.4 Flood Plain Fines (FF) Finer grained facies make up the bulk of the rocks in the study area. As has been noted previously they also weather rapidly and outcrop largely as erosional dongas. This makes the description of vertically continuous units difficult. The relatively larger exposures of this element occur on the farm Sepenare?s Hoek 759, Dubbledam 688, Thaba Nchu Townlands, Eden 96 and Chubani 9. The element is constructed from all three of the finer grained facies viz. facies Fl, Fsm and Fm. Element FF differs from element FP (pedogenic sequence) in that extensive pedogenic features are not present. Nevertheless, as with element FP, the sequences of Fl and Fsm persist laterally and contain the same dusky red and olive green sand-, silt- and mudstone of element FP in the lower stratigraphy (Dubbledam mudrock and Townland fines units, see below), becoming increasingly dusky red to greyish red purple higher up in the stratigraphy (Mohono mudrock unit, see below). 45 Within thicker outcrops this element consists of one to one and a half metres of facies Fl which then grades (vertically) into facies Fsm and finally facies Fm. Facies Fm is frequently absent or if present consists more frequently of silt- rather than mudstone (Fig. 2.23). Calcareous nodules occur intermittently both within individual layers and through the sequence. These are flatter and more irregular than those occurring in the sandstone units. 2.3 Lithostratigraphy An analysis of the vertical arrangement and relative position of the different facies can assist in an analysis of the palaeoenvironment. As part of this study a number of vertical lithological profiles (stratigraphic sections) were measured at several localities throughout the study area (Fig. 2.24, Fig. 2.25, Fig. 2.41 and Appendix Figs: A1-A7). A description of the lithostratigraphy in the study area is given in this section. The sedimentary rocks of the Beaufort Group in the Thaba Nchu area can be divided into three predominantly argillaceous units which are separated by two arenaceous units (Fig. 2.41). A distinctive coarse grained unit is found within the lower mudstone and is described separately. Figure 2.26 shows the location of the farms mentioned in the text. The units are named according to the farm or locality on which the best exposures within the study area occur. The exception to this is the Musgrave Unit which is named from a locality within the city of Bloemfontein. 2.3.1 Dubbledam Mudrock The lowest stratigraphic unit in the study area is an argillaceous package of red and green mudrocks which is informally referred to in this study as the ?Dubbledam mudrock unit?. Analysis and description of this unit starts at an outcrop along a stream south west of the town of Thaba Nchu on the farm Blijdschap 105 (E 026o 49.2? S 29o 16.2?). This is the lowest unit of the Beaufort Group (Dubbledam mudrocks) in the area consists mainly of a thick succession of facies Fm, Fsm and Fl (Fig. 2.27). Smaller, thinner packages of 46 sandstone outcrop intermittently throughout the Dubbledam mudrock unit (Fig. 2.28a and b). The best exposures of the Dubbledam mudrock unit occur to the south-west of the town of Thaba Nchu (on the farms Blijdschap 105, Sepenare?s Hoek 759, Morake?s Hoek 58 and the type locality, Dubbledam 688) and on the Thaba Nchu Townlands 608. At all the localities the unit outcrops in a series of erosional dongas and in channel beds of small streams. With the exception of good exposure at Feloana dam and at Ratabane 100, the only other outcrops of this unit are of the sandstone packages which are found in the beds of streams scattered throughout the western portion of the study area. These outcrops are poor representations of the unit as the widths of sandstones are limited to the width of the stream. The bulk of the mudrocks (facies Fm and Fsm) are dusky red (5R3/4), with greenish grey (5GY6/1) facies Fm and Fsm occurring lower in the unit and blue-green facies Fm found towards the top of the sequence. Colour mottling and banding of red and dark blue green (2.5Y3/2) mudrock occurs at both a small scale (<100 mm) as well at a larger scale (2-5 m) (Fig. 2.29). Small invertebrate traces are found occasionally within the mudrocks along with calcareous nodules and distinctive palaeosol horizons. This lower argillaceous unit is interrupted by tabular (1-4 m thick) (Fig. 2.28a & b) and thin lenticular (<3 m) sandstone packages. The thickness, lateral extent and frequency of the sandstone packages decrease upward in the sequence with the tabular sandstones more abundant lower in the stratigraphy. As a result of outcrop limitations it is not always possible to measure the lateral extent of individual sandstones, but it appears less than 100 m. The thicker sandstones are built up of two to three sets of facies Sh (and minor Sl and Sp) (Fig. 2.28a & b). Resting on very fine silty sandstone, the packages of facies Sh, making 47 up the tabular sandstone units, have abrupt bases and fine upward into horizontally laminated siltstone. At a few of the larger outcrop areas the sets of the tabular sandstones dip and overlap such that they appear as if they accreted laterally. The siltstone between units is typically 50-100 mm thick and may themselves contain thin lenses (10-20 mm thick) of fine grained sandstone. At one locality the siltstones and very fine sandstone packages above the tabular sandstones (exposed on the farm Blijdscap 105) is up to 2 m thick and contains small bar forms (Fig. 2.30) and tetrapod fossils (see Chapter 3). The tops of the tabular sandstones rarely preserve invertebrate trackways (Fig. 2.31). On the farm Ratabane 100 a number of bedforms can be seen which dip away from the palaeocurrent direction (Fig 2.32). Composed of sets of Sh (250-500 mm thick), each set truncates the set beneath it. The succeeding set dips at approximately 180o to the underlying one and erodes into it. Oblate calcareous nodules/peloids 100-150 mm in diameter line the base of some of the sets. 2.3.2 The Musgrave Grit Unit A prominent gritstone bed is present towards the top of the lower argillaceous unit (Dubbledam mudrocks). This bed is exposed at four localities (the farms Chubani 9, Middeldeel 709, Feloana Dam and Midland 325) and comprises facies Gt with fresh feldspar, angular fragments of granite and silicified wood. The silicified wood consists of both large pieces of trunk (up to 2 m long and 150 mm in diameter) and smaller branches. The larger are orientated perpendicular to the palaeocurrent direction. At all other localities where the unit is present, the outcrop is small and occurs mainly as small scale troughs which erode locally into the underlying units (Fig. 2.33). A similar coarse-grained sandstone unit at the same stratigraphic is known from other localities in the Free State and has been informally named the ?Musgrave unit? (J.C. Loock, pers. comm.) although Theron (1970) refers to it as the Northern Beaufort Formation. Conventionally stratigraphic units are named after geographic localities and hence the name Musgrave grit unit will be used. 48 The Musgrave grit unit appears to lie stratigraphically within the Dubbledam mudrock unit. However the fragmentary nature of the outcrop prevents a more specific determination of its exact stratigraphic location. 2.3.3 Middle Sandstone Units A prominent arenaceous horizon is present above the lower argillaceous unit and comprises three parts: a lower relatively thin but laterally extensive sandstone unit; a sequence of fine grained rocks (facies Fm, Fsm and Fl) with small channels and sandy bedforms; and a relatively thick continuous sandstone package consisting of a large erosive channel built up of smaller bed forms. For ease of reference the three units that make up this package are named after the farm on which they are best observed. The lowest package is here referred to as the ?Sepenare?s Hoek sandstone?; the middle more argillaceous package is called the ?Townland fines?; and the upper sandstone unit as the ?Eden sandstone?. 2.3.3.1 Sepenare?s Hoek Sandstone This unit is differentiated from the sandstone packages within the underlying massive Dubbledam mudrock unit by its great lateral continuity and tabular appearance (Fig. 2.34). The thickness of the package is between 1.5 m to 3 m. For the most part the fine yellow sandstone package is built up of facies Sh, St and minor Sp with a locally erosive base lined with facies Sei and topped with ripples. At some localities, such as at Abramskraal 65, the unit consists of a series of overlapping and cross cutting channels. These are described above as architectural element CHt (Fig 2.17). 2.3.3.2 Townland Fines This package consists of red and green-blue fine grained facies (Fm, Fsm and Fl) with coarser facies appearing in the form of architectural element CHf, laminated sandstones (architectural element LS) and down stream accreting macro forms (architectural element DA) (Fig. 2.18). The thickness and colour of this package varies across the area. At Thaba Nchu Townlands it is 35 m thick whilst at Moroto it is more than 45 m thick. The proportion of blue-green mudrock decreases towards the north. Forming approximately 49 45% of the mudrocks in the immediate vicinity of the town of Thaba Nchu, the blue-green coloured mudrocks decrease to less than 20% at Rakhoi 5 in the far north of the study area. At this locality red mudrocks form a larger part of the stratigraphic succession. 2.3.3.3 Eden Sandstone Above the Townland fines unit is a relatively thick unit of fine grained sandstone consisting largely of facies Sm and Sh. These facies are arranged into a series of large channels (Fig. 2.36). The centre of the channels is up to 20 m thick and channel widths are up to 300 m wide. The large channel in turn comprises a series of smaller channels and bedforms (architectural element DA) which overlap, erode into or overly underlying packages. Facies Sei occurs frequently at the base of individual sandstone units. As a result of outcrop constraints the full extent of individual channels is seen at only a few localities in the Thaba Nchu area viz. Mohono Mountain (farms Eden 96, Sekuthlong 632 and Waalhoek 116), Moroto mountain (adjacent to Moroto village) and at Rakhoi village. The most complete sequence of the Eden sandstone is seen on the border of the farms Eden 96 and Waalhoek 115. 2.3.4 Mohono Mudrock Unit This interval of maroon mudrocks overlies the Eden sandstone unit and comprises a thick series of facies Fm, Fsm and Fl (Fig. 2.37). Facies Fm, Fsm and Fl dominate the unit. As noted before, the nature of the outcrop of facies Fm and Fsm is such that it is difficult to discern any internal structure. Concretionary nodules are not found as frequently as in the underlying units. Rhizocretions were collected from within the lower portion of the unit (Fig. 2.39). Within a package of facies Fl in the lower portion of the unit is a pebble lag which contained fragmentary fossilised tetrapod bones (see Chapter 3). The Mohono mudrocks are capped by a medium to coarse grained white sandstone of the Molteno Formation. The contact between the Mohono mudrocks and the Molteno 50 Formation was not observed as it was covered in scree at all localities. Similarly the lower contact of the Mohono mudrocks is poorly exposed and was observed only at Mohono Mountain (Eden 96). The fine grained rocks of this unit contain overbank channels (architectural element CHf). These channels differ from those within the ?Dubbledam mudrocks? in that they contain facies Sp overlying facies Sh and a thin base of Sei. The channels are typically less than 3 m in width and less than 60m wide. Larger channels outcrop at Mohono, Rakhoi and on Thaba Nchu Mountain (Wilgeboomnek 29) (Fig. 2.38). An outcrop of a larger channel occurs on Thaba Nchu Mountain (farm Wilgeboomnek). Here smaller (lenticular) channels consisting largely of facies Sh with bases of Sei overlap and overlie one another to form a compound channel. The only other outcrops of significant sandstone occur on Mohono Mountain (farm Eden 96). As with the lower Dubbledam Mudrock Unit the sandstones are exposed only in dongas with the result that their margins are not exposed. They consist of both facies Sh and Sp. Facies Sh is far more abundant than facies Sp. 2.4 Palaeocurrents Palaeocurrent readings were taken from all available palaeocurrent indicators, but because of the relatively few measurements taken it was not feasible to undertake a proper statistical analysis of the readings. All readings are presented according to the stratigraphic unit from which they were obtained (Fig. 2.40a-d). Palaeocurrent data for the Dubbledam mudrocks unit was obtained mainly from parting lineation found within the packages of sandstone and indicates a mean palaeocurrent trend of 340o / 160o (Fig. 2.40a). Ripple and sole marks were used to determine palaeocurrent direction. 51 In the Musgrave grit unit readings were taken from the trough cross-stratification of facies Gt and from fossilised branches and trunks of wood and indicate a source to the north east which is unique to this stratigraphic interval (Chapter 5) (Fig. 2.40b). Measurements from the two sandstone units (Sepenare?s Hoek and Eden sandstones) as well as the Townland fines have similar palaeocurrent directions. Measurements were taken largely from parting lineation, on the surfaces of facies Sh and from facies St. The combined mean palaeocurrent direction for the two units is 342o (Fig. 2.40c). The Mohono mudrock unit yielded few palaeocurrent indicators. However from the limited data (parting lineation) a palaeocurrent direction of 338o was determined (Fig. 2.40d). Overall the predominant palaeocurrent direction for the rocks of the Beaufort Group in the study area is toward the north north-west. Anomalously the current direction for the Musgrave grit unit is toward the south-west. 52 Chapter 3 Palaeontology During the course of this study numerous fossils of both tetrapods and plants were recovered from the mudrocks and sandstones of the study area (Fig. 3.1, Fig. 3.2, Fig.3.3 and Tables 3.1 and 3.2) and their locality data and stratigraphic position were accurately recorded. In addition to this study a number of other workers have in the past collected fossils from the Thaba Nchu area and these are held by the Council for Geoscience (Pretoria), BPI Palaeontology (University of the Witwatersrand) and the National Museum (Bloemfontein) (Table 3.2, Fig. 3.3). Most of these specimens lack accurate (GPS) co-ordinates and therefore the exact locality from which they were collected is not known and their precise position in the stratigraphy is uncertain. However with better knowledge of the location of mudrock exposures in the study area a better resolution is possible. 3.1 Fossils of the Thaba Nchu Area 3.1.1 Vertebrates 3.1.1.1 Dicynodontia a. Dicynodon: Specimens of Dicynodon are not abundant and only a single skull (BP/1/ 5862) was found, during this study, from the Thaba Nchu Townlands. The skull was found in the dusky red and grey-green mottled mudrocks (facies Fm and Fsm) of the Dubbledam mudrock unit and approximately 8 m below the base of the Sepenare?s Hoek sandstone. The Dicynodon skull, along with some postcranial elements, were discovered in situ within the network of erosional dongas immediately to the east of the town of Thaba Nchu (Fig. 3.4). Some weathered residue of palaeocalcrete covered parts of the fossil which were embedded within mudrocks. 53 Four additional specimens are recorded in the Council for Geoscience and BPI Palaeontology database (Table 3.2). The precise locality of the BPI specimens is unsure as the catalogue gives the locality as the Thaba Nchu Commonage i.e. Thaba Nchu Townlands. This covers a large area and includes rocks of both the Adelaide and Tarkastad subgroups (and also rocks of the Molteno, Elliot and Clarens formations). The fossils from the Council for Geoscience have precise locality information and were collected in the western portion of the Thaba Nchu Townlands, and are therefore stratigraphically within the Dubbledam mudrock unit and below the Sepenare?s Hoek sandstone unit, within the same set of erosional dongas. b. Lystrosaurus: Thirty-eight specimens of Lystrosaurus have been collected from a number of localities in the study area (Table 3.1 and Table 3.2). Of these 27 were found and collected during this study from the farms Dubbeldam 688, Lubbersrust 747, Morago 40, Abramskraal 65, Sepenares Hoek 759 and Thaba Nchu Townlands. Most of the Lystrosaurus specimens collected during this study were collected from within the finer grained facies, particularly Fsm and Fm. The majority of specimens were found within erosional dongas, either in situ or having rolled into them from a weathered unit. The collected specimens are all skulls; however a large number of postcranial fossils were found at all the localities and were left in situ. On the farm Lubbersrust 747, 4 km to the east of the town of Thaba Nchu, the impression of a Lystrosaurus skeleton is preserved in fine grained sandstone which forms the bed of a stream. This was the only specimen found within sandstone. A number of specimens were collected at the Thaba Nchu Townlands locality. At this locality both the Dubbledam mudrock unit as well as all three sub units of the Middle Sandstone Unit are present in outcrop. At the 54 Thaba Nchu Townlands the Lystrosaurus specimens came from both the Dubbledam mudrock unit and the Townland fines unit of the Middle Sandstones. A specimen collected 10 m from the base of the Eden sandstone is stratigraphically the highest Lystrosaurus specimen from the Thaba Nchu area. The stratigraphically lowest collected Lystrosaurus specimen was within the series of dongas 5m below the Sepenare?s Hoek sandstone unit and within the Dubbledam mudrocks to the west of the prominent hill on the Thaba Nchu Townlands and 5m (stratigraphically) above the collected Dicynodon specimens. All the specimens, collected during this study, from the Thaba Nchu Townlands locality have been identified as Lystrosaurus murrayi (P.J. Hancox, pers. comm.) (Fig. 3.5). Specimens, recorded as Lystrosaurus declivis and Lystrosaurus oviceps (CHG149CGS and GHG153CGS respectively), were collected by the Council for Geoscience within the same series of dongas at this locality. Specimens of Lystrosaurus were noted and left in situ on the farm Chubani 9. These occurred in a series of dongas that lie to the north of a large dolerite ridge that divides the lower quarter of the farm. A number of specimens (L. murrayi, L. declivis and L. curvatus cf. Table 3.2) have been collected from this locality by both the National Museum (Table 3.2) and the Council for Geoscience prior to this study. The dongas from which they were collected expose rocks from the Dubbledam mudrock unit. On the farm Abramskraal 65 three Lystrosaurus murrayi specimens were collected immediately below and 3 m above an outcrop of the Sepenare?s Hoek sandstone unit. Two of the Lystrosaurus crania (BP/1/6138 and BP/1/6140) were collected from within an erosion gully in the red mudrocks of the Dubbledam mudrock unit whilst the third was collected from the scree covering the Townland fines unit (BP/1/6137). This specimen 55 together with those found at Morago 40 and Thaba Nchu Townlands were the only ones found within the Townland fines unit. At Dubbledam 688 and Sepenares Hoek 759, to the west of Thaba Nchu Mountain, seven specimens were found in a thick series of green and blue mudrocks (Fm and Fsl) cut by erosional dongas. The rocks here form part of the Dubbledam Unit and the fossils were found 5-10 m below the Sepenare?s Hoek sandstone that outcrops here. c. Kannemeyeria: During this study only a few pieces of cranial material of a single specimen of Kannemeyeria was collected (Fig. 3.6). The specimen, consisting of a tusk and caniniform process and parts of the cranium, was collected from within the Mohono mudrock unit on the northern slope of Thaba Nchu Mountain. The Mohono mudrock unit here consists of red- purple mudrocks which are highly weathered. The specimen was located south of the road running to the ?Telkom Tower? on Thaba Nchu Mountain and above a (relatively) thick sandstone channel, on the farm Wilgeboomnek 29. A number of specimens have been collected from the same locality prior to this study. The National Museum collection records two Kannemeyeria specimens collected close to the locality of the specimen found during this study. With one of them collected below the thick sandstone noted above. The 1:250 000 geological map lists Kannemeyeria sp on the farm Koele 97 which is on the northern slopes of Thaba Nchu Mountain closer to the town of Thaba Nchu (and adjacent to the farm Wilgeboomnek 29). 3.1.1.2 Procolophonoidea Three partial procolophonoid mandibles were collected during the course of this study from three different localities. All three were collected from 56 facies Sei. Two of these have been identified as belonging to the genus Procolophon (J Neveling, pers. comm.) whilst the third has been identified as belonging to a more advanced procolophonoid (R. Damiani, pers. comm.). The two Procolophon specimens were collected from within the scree at two separate localities. One of the specimens was collected on the farm Sekuthlong 632, below the ridge that skirts Mohono Mountain. Although not in situ, the specimen was collected from beneath an outcrop of Eden sandstone unit. The second Procolophon was collected from a small hill along the Thaba Nchu Excelsior road. This second specimen was not in situ. The more derived procolophonoid (Fig. 3.7) was found in situ in a unit of Sei within the Eden Sandstone at Moroto Mountain. 3.1.1.3 Archosauramorpha The partial lower jaw (Fig. 3.8), isolated vertebrae and a quadrate of an erythrosuchid (possibly cf. Garjainia)( BPI/1/5877) was collected from a series of red-maroon mudrocks of the Mohono mudrock unit on the farm Eden 96 on the western slopes of Mohono Mountain, 10 m above the top of the Eden sandstone unit which here forms a terrace on the mountain. The fossils were embedded in a bone and pebble lag which had weathered out of the base of a channel within the Mohono mudrock unit. 3.1.1.4 Cynodontia During this study a cynodont lower jaw fragment was collected from within the same bone and pebble lag in the Mohono mudrock unit as the archosaur specimen. The National Museum collection lists four cynodonts collected previously from within the study area. Two of these are identified as Cynognathus 57 (NM3197 and NM3198) and were collected from the farm Wilgeboomnek 29, 10 m above the ?Telkom Tower? road on the northern slope of Thaba Nchu Mountain (Table 3.2). The other two (NM886 and NM887 and listed as ?indeterminate cynodonts? cf. Table 3.2) were collected on the farm Chubani 9. No co-ordinates for the later two specimens are given. The 1:250 000 geological map lists Cynognathus on the Thaba Nchu Townlands on the slopes of Thaba Nchu Mountain. 3.1.1.5 Amphibians and Fish A number of temnospondyl fragments were recovered from the western slopes of both Moroto and Mohono Mountains. Only fragments of a breast plate and part of a lower jaw were discovered and generic distinction cannot be made. A single lung fish tooth plate was collected from the Mohono mudrocks on Mohono Mountain (Eden 96) (Fig. 3.9). 3.1.2 Fossil Wood Numerous pieces of silicified wood were collected or observed on the farms Middeldeel 709, Chubani 9 and Midland 325. The silicified wood on these farms was found in situ and consisted of fragments, branches and trunks (Fig. 3.10a). Thin sections were made for identification purposes. From the slides the wood was identified as Agathoxylon africanum (M Bamford, pers. comm)(Fig 3.10b). The silicified wood found on all three farms was collected from within the Grit Unit outcrops of which occur to the south of the dolerite ridge that crosses Chubani 9. In some sandstones within the Townland fines and in the Mohono mudrock unit plant impressions were recorded but were too poorly preserved for identification. 58 3.2 Additional Fossils The BPI Palaeontology, National Museum and the Council for Geoscience collections database list a number of fossils from the Thaba Nchu area which have stratigraphic importance. None of these fossils have precise geographical data and so their position within the stratigraphy outlined in Chapter 2 is not known. However as will be argued in Chapter 5, their presence in the Thaba Nchu area may be an indication of the completeness of the Thaba Nchu fossil record. These include the therocephalian Whaitsia platyceps (NMQR888), collected at Chubani 9 and a specimen of Emydops (NMQR1238) collected from a ?gravel sloot? at Thaba Nchu, an indeterminate therocephalian femur (collected at Thaba Nchu) and an ?indeterminate gorgonopsian? (BP/1/1491) (collected at Thaba Nchu Commonage -Townlands). The Thaba Nchu Townlands was also noted as a collection site for a specimen of Moschorhinus kitchingi (recorded on the 1:250 000 Bloemfontein geological map). 3.3 Position of Fossils within the Local Stratigraphy. The stratigraphic horizons of all fossils collected during this study are presented in Fig. 3.11, and will be discussed in Chapter 5. 59 Chapter 4 Geophysics Geophysical techniques are used to define the regional geology below the current surface of the Earth over large areas (Telford et al., 1990). When combined with known geological data, from outcrop and borehole studies, geophysical techniques are useful tools in understanding both regional and local geological features. In this chapter regional geophysical and borehole data are used to define the thickness of Karoo strata as well as the topography and structure of the basement. The importance of both the thickness and basement structure of the Karoo Supergroup in the study area will be discussed in Chapter 5. 4.1 Geophysical Surveys The geophysical techniques of reflection seismics, gravity and aeromagnetics may facilitate observation of physical changes in rock properties over large areas. The data are then integrated into a gravity model and the various geological features observed. Seismic reflection data from Soekor exploration and the national gravity and aeromagnetic data sets are used to determine the major three dimensional structures in the study area. 4.1.1 Gravity Method 4.1.1.1 Principles The gravity technique maps lateral relative variations in the gravitational field of the Earth which are due to changes in the density of the underlying rocks. By accounting for factors other than rock density variation in gravity will mirror the variation in density i.e. gravity data readings must be corrected to compare with the value they would have on an equipotential surface (Telford et al., 1990). 60 Factors influencing the gravitational field of the earth at any given point (and the corrections that need to be made to eliminate their influence) are: gravitational attraction of the sun and moon (the earth-tidal correction); distance from the equator (latitude correction); height above sea level (free air correction); subtraction or addition of excess mass between a given point and a reference plane (Bouguer correction); gravitational attraction of topographic features (terrain correction); and the tectonic equilibrium or disequilibrium of the underlying crust and mantle (isostatic correction) (Telford et al., 1990). This may be expressed by the following equation: GB = GOBS ? Gt + (?GL + ?GFA - ?GB + ?GT) ??????. (1) (where GB is the Bouguer anomaly, GOBS is the observed value, Gt is the theoretical gravity, ?GL is the latitude correction, ?GFA is the free-air correction, ?GB is the Bouguer correction and ?GT is the terrain correction) (after Telford et al., 1990). The objective of these corrections is to compare the gravity reading at a particular point with its theoretical value at that point. Therefore any variation between the two levels is the result of variations in the underlying density of the rock. In determining the theoretical gravity position on the Earth a formula, the International Gravity Formula, is typically applied (Telford et al., 1990): ) 90130066943799.01 86390019318513.01(7803267714.9 2 2 ? ? Sin SinGt ? + = ????????.(2) (where Gt is theoretical gravity and ? is the geographical latitude) (Ahern, 2004). Variations in gravity at the surface of the Earth are not particularly useful unless these variations are used to model or interpret the underlying geology. The interpretation of gravity data is made more difficult by the fact that different bodies at different depths may produce the same gravity response at the surface. This is because the gravity data at any 61 point is the result of the density and shape of a body as well as the distance from that body, all of which can change independently. Therefore, in interpreting the data, some pre- existing knowledge of the underlying geology and rock densities can help to constrain the geological interpretation (Telford et al., 1990). It is not only the near surface geology that has an effect on the gravity at the surface of the Earth but also deep seated and regional scale bodies. These often mask the anomalies at the surface or elevate the values recorded. For example, craton wide features situated in the lower crust will contribute to the gravity at the surface, however these features may have little to do with the geological formation of shallower features such as local variations in a sedimentary basin. When interpreting the data this regional trend is typically removed (Telford et al., 1990). 4.1.1.2 Gravity dataset The data used in this study consists of a subset of the national gravity database which was acquired in the last few decades by the Council for Geoscience. Gravity measurements were acquired on a 4 x 3 km grid (Stettler et al., 2001, R. Stettler, pers. comm.). A mean crustal density of 2670 kg/m3 was used in the calculation of Bouguer anomaly values. The maximum error for any measurement for this data set has been calculated as + 1 mgal (Stettler et al., 2001). In addition to the data obtained by the Council for Geoscience a separate survey was undertaken in Lesotho (Burley et al., 1982). These data were then incorporated into the Council for Geoscience data set. To produce a continuous set of values across the area the data were gridded, i.e. points between the 4 x 3 km measurements were mathematically interpolated. A minimum curvature algorithm and a cell size of 1000 m were used to grid the data (Stettler et al., 2001). The complete, gridded national Bouguer gravity dataset for South Africa is shown in Figure 4.1 and a subset of this data set, centred on the Free State, was then extracted for this study. The relative range in values, of the data subset, is 227.5 mgals with a mean value of -134 mgals. As noted above, this represents the relative change in the gravitational field and hence the relative variation in the density of the underlying rock column. 62 The gridded subset of the national data were used to produce a coloured image using the computer software package ?Oasis Montaj?(Fig. 4.2). This image represents the variation in gravity with the warmer (red, yellow) colours representing areas of relatively higher gravity (density)(maximum of 37.4 mgal) and the cooler colours (blue, green) areas of relatively lower density (minimum of -193.4 mgal). In order to highlight structural and other linear features in the basement of the Karoo Basin a second image, in addition to the coloured unfiltered image, was produced using sun shading (Fig. 4.3). Sun shading is a technique whereby a light source (?sun?) is shone from a chosen elevation and azimuth onto the data set. The amount of light reflecting off the data surface is dependant on the gradient of the surface at that point. The effect of sun shading is to produce an image wherein the gravity highs and lows appear as the topographic features on a relief map. Both declination and inclination of the light source was set at 45o. At these angles north-west/south-east trending features will be emphasised whereas north-east/south-west trending features will be suppressed. This allowed the north eastern rim of the Karoo Basin to be highlighted. The anomalous palaeocurrent directions recorded in the Musgrave grit unit (see Chapter 2) demonstrate a source to the north east of Thaba Nchu. Thus illuminating from the north east should highlight gravity anomalies in this area. Nine areas of distinct gravity lows and highs (anomalies) are evident from the dataset and are highlighted in Fig. 4.4. It is not the aim of this study to give a comprehensive description and discussion of all the gravity anomalies across the entire Karoo Basin. Instead geological features which may have influenced the depositional environment within the study area and or which are crossed by the profile lines used to produce the gravity model (see below) are described and discussed below. Nine different gravity anomalies are highlighted in Figure 4.4. The most prominent gravity high (labelled A) is situated in the south- western corner of the dataset and is visible as a circular to ovoid feature with a diameter of 50 to 75 km. The highest gravity values occur 63 toward the centre of the anomaly. A second circular feature (anomaly B), consisting of a series of concentric rings which open out toward the south east, occurs in the northern Free State. The ring structure of the anomaly is better seen in the sun shaded image (Fig. 4.3).This anomaly is enclosed by a gravity high (anomaly D) that stretches in an arc from south of Welkom into southern Gauteng. Anomaly H lies to the south of anomaly D and 20 km north of the study area. It is the smallest of the discussed anomalies and appears to be connected to anomalies D and E. However it is treated separately here as, as will be shown below, the two larger anomalies are likely due to two different sources. The largest anomaly (E) is situated on the western quarter of the map. It appears as a series of interconnected bodies through which a north- south trending lineament passes. The highest gravity values (+35 mgals) are recorded in anomaly F in the south eastern portion of the map. This anomaly lies on the eastern edge of the Karoo Supergroup and can be seen in the national data set (Fig. 4.1) to stretch along the eastern coast of South Africa from the Eastern Cape to northern KwaZulu Natal. To the north-east of the study area is a gravity high (anomaly C) which has a general trend north to south and is 190 km in length. Within this anomaly the data shows a secondary trend viz. north-north-west which is highlighted in the sun shaded image. The remaining gravity high (anomaly G) divides the large gravity low (anomaly I) in the centre of the data set. Anomaly G is not as prominent as the other gravity highs and is divided by the edge of the Kaapvaal Craton. The study area lies within this anomaly (I). The large gravity low (anomaly I) has a north south axis 300 km long and an east-west diameter of 250 km. Although divided by the gravity high (anomaly G) it is treated as a single feature. 4.1.2 Aeromagnetic Survey 4.1.2.1 Principles Magnetic surveys record and map variations in the Earth?s magnetic field largely due to changes in magnetic properties of the underlying rock. The presence of ferromagnetic 64 minerals, principally magnetite, determines the magnetic susceptibility of a rock and hence the amplitude of a magnetic anomaly (Telford et al., 1990). Ferromagnetic minerals have the ability to retain a magnetisation (strength and direction) in the absence of an external magnetic field, this is termed remanence or permanent magnetisation and can effect the amplitude of a magnetic anomaly (Blakely, 1996). Remanently magnetised rocks, such as ancient basement complexes affected by multiple stages of deformation and intrusion, will produce a magnetic field that is a sum of all these parts. This may have the effect of obscuring the underlying geology (Telford et al., 1990). In general non volcanogenic sedimentary rocks, in contrast to crystalline rocks, have a low average range of magnetic susceptibility (Telford et al., 1990). Despite the low magnetic susceptibility of sedimentary rocks magnetic surveys may still be of use in identifying structural features and intrusive rocks within sedimentary rocks or assisting in describing variations in basement rocks beneath sedimentary cover. The aeromagnetic data set was analysed to assess whether any prominent magnetic features are present in the study area which could have effected the deposition of sediments within the area. A prominent local feature, in the vicinity of Chubani 9, was examined to determine if faulting along it could be observed. 4.1.2.2 Aeromagnetic Dataset Data used in this study were collected as part of the national aeromagnetic survey by the Council for Geoscience of South Africa in different survey blocks conducted between the 1960?s and 1990?s (J. Cole, pers. comm.). Data were collected at a line spacing of 1 km and a flying height of roughly 150 m ? 15 m. Magnetic data is usually collected using a magnetometer which is set to collect periodically (the cycling time). For the national survey a cycling time of 1 second was used. Combined with the average aircraft velocity of 250 km/h this translates into a magnetometer reading every ~ 63 m (J. Cole, pers. comm.). These data were corrected for the 1 July 1975 magnetic epoch (J. Cole, pers. comm.). Survey data for Lesotho were not available and it is not known if any data has been collected over that country (J. Cole, pers. comm.). 65 A subset of data, similar in extent to the gravity data, was extracted from the national aeromagnetic dataset (Fig. 4.5). A second and smaller subset of data specifically over the study area was also extracted to examine a local magnetic high linear feature in the north of the study area (Fig. 4.6). The aeromagnetic data shows much higher variability across shorter distances than the gravity data. This is the result of both the larger data density and the higher variation in the magnetic susceptibility of the underlying geology. The low magnetic susceptibility of the sedimentary rocks of the Karoo Supergroup can be contrasted with the much higher susceptibility of the dolerite (Karoo dykes and sills). The former rock type has an average magnetic susceptibility more than one hundred times less than that of dolerite (Telford et al., 1990). The effect of this obscures variation in the magnetic field of the sedimentary rock cover in areas where there are dolerite intrusions within the sedimentary rocks. A number of regional magnetic anomalies surround the Thaba Nchu area. To the south east of Thaba Nchu, and extending eastwards to the Lesotho border, is an area of increasing amplitude and decreasing frequency (anomaly Am, Fig. 4.5). The highest amplitude areas lie adjacent to the Lesotho border and are assumed to straddle it. The northern extreme of this anomaly underlies the town of Thaba Nchu and the area of Thaba Nchu Mountain (Fig 4.5). A series of anomalies are located in the vicinity of and to the south of Bethlehem in the north eastern Free State (Figure 4.5; area Cm). The two to three separate anomalies straddle the Lesotho border and it is again assumed they continue into that country. Three other distinct anomalies are seen in the magnetic image. The first is the 60 km diameter circular feature to the south west of the study area (anomaly Dm) (Fig. 4.5). The second is a linear feature (labelled as LM1, Fig. 4.5) that runs from the south western corner of the data set eastwards across Aliwal North. A second linear feature (labelled LM2) trends north-south in the western third of the data set. 66 The Thaba Nchu subset of data may be divided into two distinct areas. The north west sector and the rest of the area (Fig. 4.6). The north western magnetic low is associated with the large regional anomaly that extends northward for approximately 50km. The rest of the Thaba Nchu area is largely covered by a series of magnetic anomalies (highs) that extend south eastward. A particularly prominent dolerite ridge occurs in the northern third of the area and runs from east to west across the farm Chubani 9. This ridge has a magnetic signature which is labelled as MF in Fig. 4.6. The discontinuous nature of the magnetic signal is due to the aeromagnetic line spacing. No faulting could be seen in the magnetic data. However this maybe due to the consistent magnetic response of the sedimentary rocks on either side of the dolerite dyke. 4.2 Seismic Survey, Regional Borehole and Outcrop Geology 4.2.1 Regional Borehole Data The depth to the basement of the Karoo Supergroup from 352 boreholes was obtained from the database of Anglogold Ashanti (Table 4.1) and from Soekor well completion reports (Pyke, 1973; Roux, 1972; Roux, 1974). The distribution of the boreholes is biased toward the northern Free State Goldfields with a few boreholes in the eastern and southern Free State Province and north-western KwaZulu Natal Province (Fig. 4.7). No boreholes were recorded in central KwaZulu Natal or within the Thaba Nchu study area. The closest boreholes to the Thaba Nchu area lie to the north-west on the farms Sunnyvale 785 and Middelpunt 2029 (Fig. 4.7). A contour map of the depth to basement was produced using all the borehole data( Fig. 4.7). The outline of the Karoo Supergroup along with inliers of pre-Karoo rocks were treated as additional data points (depth value = 0). Because the Karoo Supergroup was deposited within a basement and the data vary as a function of distance, Kriging was 67 chosen as the gridding method. The method recognises that the further one data point is from another the less correlation there will be (Swan and Sandilands, 1995). This is appropriate for a sedimentary basin within which the depo-centre as well as forebulge shifted with time. The gridded data were then plotted as a contour map (Fig. 4.7). The software program Oasis Montaj was used in the gridding and plotting of the data. The contour map produced shows the thickest part of the present Karoo Basin lies beneath southern Lesotho (Fig. 4.7). The depth to basement decreases northwards and outwards from this point. Between Bloemfontein and Welkom a ridge and valley can be seen trending north north-west. A similar valley can be seen west of the town of Bethlehem and east of the study area. The area in the north-eastern Free State is relatively featureless with no noticeable highs or lows. This may be due in part to the fact that there are relatively fewer boreholes in this area. The boreholes in the vicinity (north-west) of Thaba Nchu, on Sunnyvale 785 and Middelpunt 2029, show the basement to be at a depth of 800 m and 877 m respectively (Kingston et al., 1961). A borehole from the Anglogold Ashanti database has the depth to basement as 968 m. Total dolerite thickness (for the Sunnyvale 785 borehole) is shown to be 186 m which implies the total thickness of sedimentary rocks is 614 m. The area in which the boreholes lie is mapped on surface as lower Beaufort Group (Adelaide Subgroup or Dubbledam mudrock unit in this study) and so the thicknesses represent the thickness of the Dwyka and Ecca Groups and the remaining (un-eroded) portions of the Adelaide Subgroup. 4.2.2 Seismic Survey depths As part of the search for petroleum resources in the Karoo Basin, Soekor, the state oil exploration company, undertook a number of 2D seismic surveys in various parts of the Karoo Basin. Seismic lines were used by Soekor to develop a depth to basement map of the northern Free State (Fig. 4.8). These data were further constrained by existing boreholes at the time of the surveys. The seismic survey data are limited to the northern and central Free State. 68 These data provide a more detailed map of those areas it covers than the depth to basement contour map produced by the gridding of borehole data (Fig. 4.9). However the general trend of a thickening towards the south and western Free State is confirmed by these data. In addition the valleys and ridges noted in the borehole data appear more prominently. One such valley extends south from Welkom passing the study area to the east. A second valley lies to the west of the study area whist a ridge extends to just north of the area. Approximately 40 km east of the town of Bethlehem a valley can be seen running south- west for 50 km to edge of the seismic data set. To the north north-east of the same town is a palaeohigh. 4.3 Data integration, interpretation and modelling A number of the gravity and magnetic anomalies noted above overlie well known geological features. Most of these predate the formation of the Karoo Basin. However when modelling the gravity data it was necessary to take them into account. Gravity anomaly A and magnetic anomaly Dm overlie the Trompsburg Complex (Mar? and Cole, 2006) (Fig. 4.4). This Palaeo-Proterozoic layered intrusion is covered by approximately 1100 m of Karoo Supergroup rocks. Consisting of a suite of igneous rocks the complex intruded rocks of the older Transvaal Supergroup (Mare and Cole, 2006). The Trompsburg Complex is in contact with the Karoo Supergroup and may have provided material for its sediments. The second circular gravity anomaly (B) is the Vredefort Structure (Reimold and Gibson, 1996). This Palaeo-Proterozoic feature is the result of a meteorite impact (Reimold and Gibson, 2006) which lifted the existing strata and so produced the concentric rings visible in both the gravity and magnetic data. It is partly overlain by rocks of the Karoo Supergroup in the south-east and it is argued that it is unlikely to have provided sediments to the Beaufort Group within the study area. 69 The gravity high (anomaly D) that extends around the Vredefort Complex is caused by the Archean Witwatersrand Supergroup (McCarthy, 2006). The denser quartzites and shales that make up the bulk of the Witwatersrand Supergroup account for the higher gravity values. Rocks belonging to the Witwatersrand Supergroup are believed to have had no impact on the deposition of the Beaufort Group within the study area. However Cousins (1950) notes the presence of deep glacial valleys that cut through the Witwatersrand Supergroup strata south of Welkom and in the vicinity of Virginia. The valleys trend north- west and are according to him due to Dwyka glaciation. When comparing the seismic data and the gravity data it is evident that one of these valleys divides anomaly D from H. It is argued that anomaly H represents the western ridge of this glacial valley. Gravity anomaly E is underlain by sedimentary rocks of the Griqualand West Supergroup and the lavas and sediments of the Ventersdorp Supergroup (Tinker et al., 2002). These units are believed to have had no influence on the deposition of the Beaufort Group within the study area but they are crossed by the profile line X, Y (see below). This is also true of gravity anomaly F. It is argued that this feature does not represent a geological unit but is a result of isostatic disequilibrium due to recent erosion along the coastal plains and escarpment. The Karoo Supergroup overlies the whole of gravity anomaly C (Fig. 4.10 and 4.11). The geological feature that generates this anomaly is not known (M. Andreoli, pers. comm.). It is considerably larger than the magnetic anomaly Cm and it is therefore unlikely the two are related (Fig. 4.5). Both the seismic data as well as the borehole data show that there is a thinning of the Karoo cover in this area. Rocks of the Karoo Supergroup overlie the gravity low (anomaly I) and the gravity high (anomaly G) that divides it (Fig. 4.10 and 4.12). The borehole data indicate that the depth to the Karoo Supergroup basement increases from north to south but with some local variation. Boreholes drilled in the vicinity of Ladybrand show a depth to basement of the Karoo Supergroup of 1700-2000 m. North-east of the study area the thickness varies by more than 170 m over a distance of less than 20 km. 70 A gravity high is observed to the north west of the study area but does not overlie the basement ridge that extends towards the study area (Fig. 4.10 and 4.13). The boreholes drilled to the north-west of the study area show depth variations of up to 100 m over less than 10 km. This shows the large variation in the local topography of floor of the Karoo Basin. The two linear features are the edge of the Kaapvaal Craton and the north-south trending Colesberg Magnetic Lineament (Stanistreet and McCarthy, 1991). Both features can be seen in the Bouguer gravity map as well as the aeromagnetic anomaly map (Fig. 4.10). The Colesberg Magnetic Lineament is believed to be the boundary between two older terranes which joined during the Archean through a compressional event (Stanistreet and McCarthy, 1991, De Wit et al., 1992). Two geological models were constructed using the Bouguer gravity data to demonstrate the variation in the topography of the basement to the Karoo Supergroup. Models were produced using the software package ?Grav2dc?, a gravity modelling software package developed at the University of the Witwatersrand (Cooper, 2004). Where available density values obtained from the Council for Geosciences Physical Property database were used to constrain the data along with the borehole and geological data discussed above (Table 4.2). As only certain units have had measurements taken (L Mar?, pers. comm.) an average value for the various sandstones, mudrocks and other lithologies was taken (Table 4.2). To construct the models, data profiles (similar in principal to topographic section lines, except that the data are the gravity data not topographic height) were taken across the Bouguer gravity map (Fig. 4.14). The first profile (Line X, Y) lies across the central portion of the map in an east-west direction, traversing the entire northern Karoo Basin. The second line (line X1, Y1) runs towards the north-north-east from the vicinity of Aliwal North to north east of Bethlehem (Fig. 4.14). 71 The profile lines were chosen so as to cross through or close to the study area as well as to demonstrate the change in the Karoo Basin floor from north to south and east to west. Gravity anomalies that could not be explained using current geological knowledge, and which may have impacted on the depositional model, were covered as well. Where possible the lines passed through or close to borehole localities which allowed the model to be constrained by the depth and rock type (Fig. 4.14) (Table 4.3). The mean densities for each of the borehole columns was calculated by taking the percentage of sandstone, mudrock (siltstone and shale) and dolerite for each borehole and calculating the contribution of each using the density data for each lithology (Tables 4.2 and 4.3). Data from along the profile line was then input to the Grav2dc gravity modelling software package. Grav2dc works by comparing the input (field) data to data calculated from an initial proposed geology (Cooper, 2004). The geological model is then adjusted until the calculated data match the input data and the geological constraints. Models can be adjusted to fit the input data by changing a number of factors, these are: the regional value, the number, size, shape and density contrast of geological units and the strike length of the units. From the available borehole data it can be seen that the basement to the Karoo Supergroup consists of a variety of rock types (Table 4.3) (Kingston et al., 1961; Leith and Trumpelmann, 1967; Battrick, 1973; Roux, 1972, Roux, 1974; Mare and Cole, 2006). It is beyond the scope of this study to model every variation within the basement. In modelling the data the basement was set as the background and the Karoo Supergroup and the larger anomalies were then modelled relative to this background (Table 4.2 and, 4.3). The first model was constructed from data extracted along profile line X, Y (Fig. 4.14). Profile X, Y traverses from west to east across the northern portion of the study area. Starting in gravity anomaly E, it crosses the Colesberg Lineament, anomaly I and G and ends in central KwaZulu Natal. Three boreholes, situated along or close to this traverse, are used to constrain the geological model (Fig. 4.14). A regional gravity value of 144 mgals was removed from the data. This value was used as it was the mean of the data and 72 graphically was a best fit to the data. The modelling program, Grav2dc, allows for not only the length of the feature along the profile line (first strike length) but also its width perpendicular to the profile line (second strike length) to be modelled. For both models this second strike length was modelled as 100 km wide. Model X, Y shows the thickness of the Karoo Supergroup at Thaba Nchu as approximately 900 m (Fig. 4.15). This is consistent with the boreholes drilled to the north of the study area (Kingston et al., 1961). East of Thaba Nchu the basin deepens with two deep palaeovalleys increasing the depth of the Karoo Supergroup ? Basement contact in places to over 3000 m. These valleys are extensions of those seen in the seismic data. The western palaeovalley is possibly an extension of that described by Cousins (1950). Towards both the eastern and western edges of the basin the thickness of the Karoo Supergroup decreases gradually. The contact with the basement shows variations in topography particularly over some of the other modelled bodies. Anomaly G has been modelled (body 2 in Fig. 4.15) as a denser body within the basement. The density differences between this body and the basement is consistent with that found between some metamorphic rocks and some granites (Table 4.2). The anomaly is modelled as the Kaapvaal Craton boundary and not, as proposed by Burley et al. (1982), as a basement high. The craton boundary is modelled as metamorphic rock. As can be seen from the model the Karoo Supergroup undergoes no thinning in this area. Anomaly E has been modelled as a series of slightly denser bodies (4 and 6) that lie parallel to the Colesberg Lineament (body 5 in Fig. 4.15). An intrusion accounts for the anomaly above body 3. The basement beneath anomaly E contains rocks of the Transvaal Supergroup. These were not modelled separately but the density differences included in the modelled bodies. The second model was constructed from data extracted along profile line X1, Y1 (Fig. 4.16). The line transects the Bouguer gravity map from south to north, crossing the Kaapvaal Craton boundary in the south and gravity anomaly C in the north. Passing to the 73 east of the study area, the direction of line X1,Y1 was chosen so as to include the Kaapvaal Craton boundary and gravity anomaly C. The model produced from the data extracted along profile line X1, Y1 shows the Karoo Supergroup becoming thinner from south to north (Fig. 4.16). This is consistent with known geology as well as the contour, seismic and borehole data. A number of boreholes lie along the line and these were used to constrain the depth in the north and south. The Kaapvaal Craton boundary is modelled as a series of bodies with a relative density contrast to the basement of between 0.085 to 0.386 g/cm3. This is slightly higher than that for the proposed boundary within Lesotho. Anomaly C is similarly modelled as two different bodies (4 and 5 in Fig. 4.16) with varying density contrasts. The model shows the Karoo Supergroup thinning dramatically over the northern portion of anomaly C. A borehole along the profile line shows that the depth of the northern portion of the anomaly is 617 m. The relative difference in density between these bodies indicates that body 4 is possibly granite (Table 4.2). To the north of anomaly C the Karoo Supergroup thickens and then thins over a horizontal distance of less than 60 km. 74 Chapter 5 Discussion The aims of this study were to describe the sedimentary geology and stratigraphy of the Beaufort Group in the central Free State and to use the lithostratigraphy, biostratigraphy and geophysical data to understand the basin setting and depositional environment of the Beaufort Group in the central Free State. Both the location of the study area within the Karoo Basin as well as the discontinuous outcrop of the area has meant that the placement of the strata within a regional lithostratigraphic context is not possible without recourse to the biostratigraphy. In this chapter the sedimentary geology and the depositional environments of the various stratigraphic units will be considered in light of the facies and architectural elements that occur in the study area. The stratigraphy of the Thaba Nchu area will be discussed using both litho- and biostratigraphical data. The development of the Karoo Basin during this time will be examined using the stratigraphic and regional geophysical data. 5.1 Palaeoenvironments The five different stratigraphic units (viz. Dubbledam mudrock, Sepenare?s Hoek sandstone, Townland fines, Eden sandstone and Mohono mudrock units) are a reflection of the varying and changing depositional palaeoenvironments through time in this part of the Karoo Basin (Fig. 5.1). Historically fluvial environments have been defined in terms of fluvial style (Miall, 1996). Palaoenvironmental interpretation concentrated on attempts to define ancient sediments in terms of four channel styles viz. meandering, braided, straight and anastomosing (Miall, 1996). As noted in previous chapters, current discussion of fluvial depositional environments focuses on the architectural elements within a depositional unit, as well as the facies, to better understand the palaeoenvironment. 75 Within the study area, with one exception, all sandstones are built up of fine to medium grained clasts (see Chapter 2). The presence of intraformational mud-pebble conglomerate in the form of facies Sei demonstrates that the fluvial environment within which the sediments were deposited had sufficient energy to carry coarser grained material but that only finer grained material was available (Neveling, 2002). Therefore the provenance area for all but the Musgrave grit unit lies far from the study area. As the palaeocurrent direction for all the stratigraphic units, except the Musgrave unit, is northward, it suggests that the provenance area for all but the Musgrave unit is to the south. 5.1.1 Dubbledam Mudrock Unit The Dubbledam mudrock unit consists largely of facies (Fm, Fsm, Fl) and architectural elements (FP, CHf, CS, FF) (cf. Ch 2). The predominance of the finer grained facies within this unit reflects a low energy environment. Within a continental (fluvial) depositional system this suggests a floodplain, lacustrine or playa lake environment. Whilst the possibility of playa lakes within the larger overbank environment cannot be excluded, the presence of defined channels indicates a floodplain environment. Tabular sandstones represent the major palaeochannels and smaller channels (CHf) formed part of a distributary system within the floodplain. The width of the sandstones when compared to their thickness indicates broad and shallow channels. Stear (1983), working in the lower Beaufort Group in the western sub-basin of the Karoo, described a series of shallow but wide sandstones which were vertically stacked but having some degree of lateral accretion. He attributed these to the periodic localization of a channel within the floodplain. Stear?s (1983) sandstone units contains multiple storeys of these sandstones, whilst the sandstones within the Dubbledam mudrocks typically have two or at most three storeys. The difference is attributed to the more frequent shifting of the channel across the floodplain. The presence of facies Sh (with parting lineation) indicates that deposition took place under upper flow regime conditions. Miall (1996) describes a fluvial model, which he 76 terms ?ephemeral, sand bed meandering river? (p219), and for which the ubiquitous presence of facies Sh is one of the key indicators of flashy ephemeral flow. North and Taylor (1996) argue that the dominance of this facies distinguishes ephemeral-fluvial sandstones from perennial, humid region rivers. However Miall?s (1996) other requirement that each successive storey erode into the underlying unit is only seen infrequently in the Dubbledam mudrock unit tabular sandstones. The thin layers of mudstone and siltstone that lie above some of the sandstones may have prevented the erosion into the underlying sand. The mud and silt layer also possibly indicates that water pooled on top of the sand bed allowing for the deposition of the suspended load. The small bar present toward the base of the Dubbledam mudrock unit gives some indication that the water depth appears to have been somewhat deeper earlier in the depositional history of this unit. Invertebrate traces on the tops of the sandstones are further proof of an interval of non-deposition. Architectural element CHf which is found within the Dubbledam mudrock unit is both thinner and narrower than the tabular sandstones (cf. Chapter 2). These lenticular features occur less frequently than the tabular sandstones. Smith (1980) describes similar sandstones within the western Karoo Basin (lower Beaufort Group) which he considers to represent levee deposits. The initial overbank flow produces sediments that that have flat bottomed scour surfaces followed vertically by horizontal lamination, ripple cross lamination and then into siltstone and mudstone (Smith, 1980). These levee deposits are underlain by calcareous nodules which Smith (1980) argues are developed by rapid and repeated draining of calcium enriched water within an arid environment. Architectural element CHf is both underlain by calcareous nodules and has the Sh to Sr sequence recorded by Smith (1980). Reineck and Singh (1973) note that rivers with rapidly changing channels have poorly developed levees. It is likely that CHf represents the more proximal, and hence more channelized, result of overbank flow. The architectural element CV represents a crevasse splay off a larger channel. According to Smith (1980) the crevasse splays within the western Karoo Basin are tabular, 77 horizontally laminated sandstones. The crevasse splay element in the study area is stratigraphically positioned low down in the Dubbledam mudrock unit succession, along with the bar forms described and discussed in Chapter 2. The bulk of the finer grained facies were deposited between channels within topographically low areas. These topographic lows would have been inundated during periods of flooding. This would produce sequences of facies Fl and Fsm as the flood waned. The more distal the deposition of the sediments, the more common facies Fsm becomes when compared to facies Fl. In order to develop substantial units of facies Fm, playa lakes or ponds would have to contain water for substantial periods of time (Smith, 1980). Units of facies Fm do occur infrequently within the study area. The paucity of playa lakes (standing bodies of water) may be a function of either an increasingly arid climate or a lack of topographic lows able to hold large amounts of water. The red and green variation in colour and the colour mottling has been ascribed to seasonal drying up of the floodplain and the lowering of the water table allowing for the oxidation of the surface layers (Smith, 1980). Groenewald (1990), in the north eastern Free State and north western Kwa-Zulu Natal, argues that the upper two members of the Normandien Formation represent the transition from a sandy meandering environment to dry floodplain with meandering channels (see below for a discussion on the stratigraphy). Botha and Linstrom (1978) found that the Estcourt Formation in western KwaZulu Natal was deposited initially as river mouth bars and in small pans and swamps on an extensive floodplain which became increasingly drier over time. Visser and Dukas (1979) found similar palaeoenvironmental settings to Groenewald (1990). The sedimentary rocks of the Dubbledam mudrock unit in the Thaba Nchu area displays similar characteristics to the Normandien Formation and hence represent a similar pattern of depositional environment from meandering sandy channels to arid floodplain environment. 78 5.1.2 Musgrave Grit Unit The very limited outcrop of this unit precludes any decisive conclusions being drawn as to the type of fluvial system within which it was deposited. The very coarse grain size of the extraformational clasts, the troughs that form the depositional features of this unit, along with the logs of fossilised wood, indicate that it was deposited within a high energy depositional environment. The freshness of the feldspar crystal clasts and the south- westerly palaeocurrent direction suggests a proximal provenance area to the north-east. 5.1.3 Middle Sandstone Units As noted in Chapter 2, the distinguishing feature of the Middle sandstone units is the lateral continuity of the Sepenare?s Hoek sandstone and the subsequent overall increase in the amount of sandstone in the rest of the unit when compared to the underlying Dubbledam mudrock unit. The stacked sandstones channels (architectural element CHt) of the Sepenare?s Hoek sandstone unit, together with the lack of channel margins suggest a braided river depositional environment for this unit. The presence of facies Sei (intraformational conglomerate) at the base of the stacked sandstones together with the predominance of facies Sh indicates a high energy depositional environment. Using the scheme of Bristow and Best (1993) the channels were deposited in a slowly aggrading system with a high frequency of channel migration. The channels migrated across their own floodplains collecting dried mud to form their bases of facies Sei. The channel sandstones were deposited by relatively shallow but fast flowing water. This can be deduced from the thickness of the individual channels and the units of facies Sh and St. The stacking of facies Sei -Sh-St-Sr (full sequence) with no internal surfaces indicates deposition from a single event with each succeeding facies type representing a decrease in water velocity. Water escape structures in some channels point to a rapid dewatering, whilst the presence of concretionary nodules demonstrates repeated cycles of wetting and drying (Smith, 1980). 79 The overall increase in thickness of sandstones packages as well as the type of architectural element indicate a gradual increase in the water depth from the Sepenare?s Hoek sandstone unit into the overlying Eden sandstone unit. This together with architectural element DA and CHm, indicates deeper and larger channels which carried and deposited the bulk of the sediment. The smaller channels (architectural element CHf) within the Townland fines unit represent branches off the main channel that cut across the adjacent floodplain. The fill of these channels consist almost entirely of facies Sh indicating deposition in a high energy environment. Blue-green and red mudrock suggests that parts of the floodplain were below and above the water table allowing for both a reducing as well as an oxidising environment (Botha and Linstrom, 1978). Within the Townland fines unit deposition of the larger sandstone packages occurred during flash floods. The laminated sand sheets (architectural element LS) are products of flash floods (Miall, 1996). The frequency of facies Sh (with parting lineation) is in an indication that most of the sandstone was deposited in upper flow regime conditions within both confined floodplain channels and open floodplain channels (Sneh, 1983). The sandstone that constitutes the Eden sandstone unit consists of compound channel deposits (architectural element CHm) and are the result of higher channel aggradation (Bristow and Best, 1993). In addition to the higher rate of aggradation the channels continued to migrate across the floodplain. The migration of the channels can be seen by the presence of intraformational conglomerate in facies Sei which was produced when the channels collected mud and bone fragments as it cut into the floodplain. Thicker individual channels, along with sets of facies St, indicate that the channels were deeper and contained more water than those found in the lower units. Neveling (2002) describes the Katberg Formation in the vicinity of Senekal (120km north north-west of the current study area) as consisting of sandy bedforms with a tabular geometry of mutually eroding sand sheets. He notes that this, together with the lack of 80 channel margins, implies that the Katberg here represents a perennial, sand-bed, braided river. Groenewald (1990) argues that the Verkykerskop Formation is a braided river deposit. Botha and Linstrom (1978) found the Belmont Formation to have been deposited by slow flowing braided rivers on extensive floodplains. The sediments that produced the rocks of the Middle Sandstone Unit (Katberg Formation) also show similar characteristics to the Katberg Formation. These include tabular sandy bedforms with mutually eroding sandsheets. Therefore it can be concluded that the sedimentary rocks of the Middle Sandstone unit were deposited by braided rivers that occupied an extensive floodplain. 5.1.4 Mohono Mudrock Unit As noted in Chapter 2 the Mohono mudrock unit presents very little outcrop and thus does not allow for a full palaeoenvironmental interpretation. Rocks of the Mohono mudrock unit represent, for the most part, a much lower energy environment than the underlying Middle sandstone units. The thick sequence of mudrocks (facies Fsm, Fl) represents an overbank environment. Neveling (2002), in the distal portion of the basin, north-east of the study area, describes the palaeoenvironment of the finer grained units of the Burgersdorp Formation as representing the deposits of playa lakes, semi-perennial ponds and floodplains. The amphibian fossils collected from within the Mohono Mudrocks along with the mud- siltstone couplets indicates that standing bodies of water occurred, thus supporting the argument of Neveling (2002). 5.2 Stratigraphy Rubidge (1995) maps the present study area as covering only the Dicynodon and Lystrosaurus Assemblage Zones. During the course of this study fossil tetrapods indicative of both these biozones were collected, but in addition fossils characteristic of the Cynognathus Assemblage Zone (subzones A and B (Hancox, 1998)) were collected as well. 81 The limited number of fossil localities in the study area, constrained by the paucity of good outcrop, hampers detailed biostratigraphic mapping in the area. However, by locating the faunal assemblage within a particular lithostratigraphic unit it has been possible to better define the geographic distribution of the biozones. Recognition of stratigraphic ?entities? through the combined use of bio- and lithostratigraphic data has enabled refinement of the stratigraphy of the Thaba Nchu area. The Dicynodon Assemblage Zone is defined by the presence of Dicynodon and Theriognathus (Kitching, 1995a). Dicynodon Assemblage Zone fossils were found only within the Dubbledam mudrock unit and up to10m below the Sepenare?s Hoek sandstone- Dubbledam mudrock unit boundary (Fig. 5.1). In the proximal sector the last appearance datum of Dicynodon is in the Palingkloof Member of the Balfour Formation (Smith and Ward, 2001). The Palingkloof Member has not been recognised in the Thaba Nchu area. In the distal sector the last appearance datum of Dicynodon is in the Normandien Formation which is considered to be the lateral equivalent of the Balfour Formation (Groenewald, 1990). According to Groenewald and Kitching (1995, p35) the overlying Lystrosaurus Assemblage Zone is defined by ??..absence of Dicynodon lacerticeps ?? whilst the lower boundary of the assemblage zone is ?characterised by the first appearance of Lystrosaurus... ? (p38). As noted previously in Chapter 1 it has been found that there is an overlap in the occurrence of Lystrosaurus and Dicynodon elsewhere in the Karoo basin (Smith, 1995; Macleod et al., 1999; Botha and Smith, 2007). Therefore, if there is an overlap within the study area, it is necessary to adjust the definition of the boundary to take this into account. Elsewhere in the basin the overlap has been limited to only certain species of Lystrosaurus (L. maccaigi and L. curvatus (Botha and Smith, 2007)). All the Lystrosaurus specimens collected during this study were Lystrosaurus murrayi which are considered to occur above the overlap zone (Botha and Smith 2007). The Lystrosaurus Assemblage Zone in the study area encompasses rocks of the Dubbledam mudrock unit, the Sepenare?s Hoek sandstone unit and the Townland fines 82 unit. Specimens of Lystrosaurus murrayi were collected from the upper regions of the Dubbledam mudrock and within the Townland fines unit, and upwards to the base of the Eden sandstone unit. Neveling (2002) noted the last appearance datum of Lystrosaurus to be beneath the last prominent sandstone of the Katberg Formation in the proximal sector but that in the distal sector (northern main basin in the Senekal district) Lystrosaurus does not occur above the first prominent sandstone of the Formation. He further demonstrated that Procolophon occurs above the last appearance datum of Lystrosaurus in both the proximal and distal sectors. In the proximal sector Procolophon is present above the Katberg Formation and extends into the lowermost Cynognathus Assemblage Zone (Neveling, 2002). In the distal sector (north-eastern Free State) Procolophon is restricted to the Katberg Formation and does not overlap with the ranges of the Cynognathus Assemblage Zone (Neveling, 2002). In the Thaba Nchu area Procolophon is present in the Eden sandstone which is above the strata containing Lystrosaurus (viz. Townland fines unit, Sepenare?s Hoek sandstone and Dubbledam mudrocks). Procolophon does not co-occur with faunal elements of the Cynognathus Assemblage Zone which is similar to the situation in the distal sector as reported by Neveling (2002). Cynognathus Assemblage Zone faunal elements were collected only above the Eden sandstone unit within the Mohono mudrocks. No overlap of fauna with that of the Lystrosaurus Assemblage Zone was recorded. Neveling (2002) records that in the proximal sector a number of species from the Cynognathus Assemblage Zone overlap with the fauna from the Lystrosaurus Assemblage Zone. Within the distal sector those specimens belonging to a Cynognathus Assemblage Zone fauna occur exclusively in the Burgersdorp Formation, which is the same as that reported for the distal sector by Neveling (2002). Faunal elements from both the Cynognathus A as well as B subzones were collected during this study. The Cynognathus Assemblage Zone subzone B fossils were collected close to the contact with the Molteno Formation. It is therefore unlikely that the Cynognathus Assemblage Zone subzone C occurs in the study area. This is in agreement with Rubidge (2005). However Welman et al. (1991) who collected Cynognathus Assemblage Zone vertebrate fossils from Thaba Nchu Mountain argued that the 83 Burgersdorp Formation corresponds only with the upper strata of the Burgersdorp Formation in the south of the Karoo Basin. It is thus evident that there are sections of the statigraphic succession, present in the proximal sector, which are missing in the study area indicating that these parts of the Beaufort stratigaphic succession were never deposited around Thaba Nchu. These are: The biostratigraphic overlap interval of Lystrosaurus mccaigi and Dicynodon lacerticeps and the Permian Triassic boundary. This is marked by the absence of the Palingkloof Member of the Balfour formation. The stratigraphic overlap between faunal elements of the Lystrosaurus Assemblage Zone and Cynognathus Assemblage Zone which is reported for the proximal sector, but is absent in the distal sector and in the study area. The uppermost levels of the Cynognathus Assemblage Zone (Subzone C) as reported for the proximal sector by Hancox (1998). 5.3 Basin Development It is generally agreed that the rocks of the Karoo Supergroup were deposited in a retroarc foreland setting (Rust, 1975; Johnson, 1991; Cole, 1992). Catuneanu et al. (1998) suggested that the Karoo behaved as a partitioned basin with infilling occurring in a reciprocal pattern between the proximal and distal portions of the basin as a response to uplift (loading) and erosion (unloading) in the Cape Fold Belt. A problem in correlating lithological units across the basin, is the lack of reliable radiometric data for the rocks of the Beaufort Group. The abundant tetrapod fossils in these rocks have led to improved time resolution for correlating basinal events (Hancox and Rubidge, 1997, 2001). The study area is in the distal sector of the Beaufort basin and the only vertebrate biozones present are the Dicynodon, Lystrosaurus and Cynognathus assemblage zones. These three biozones are considered to represent the Changhsingian to the Anisian (+254 ? 237 ma) 84 (Rubidge, 2005). For the under filled phase of the Karoo, recent basin development research has shown that within the Late Permian (Changhsingian) continual fluvial deposition took place (as the lower to mid Adelaide Subgroup, Eodicynodon to Cistecephalus Assemblage Zones) in the southern part of the basin (proximal sector) while contemporaneous deposition of subaqueous Ecca Group rocks occurred in the distal sector (Catuneanu et al., 2002). This is reflected in the position of the Dicynodon AZ, which occurs half way up the Beaufort Group stratigraphic succession in the proximal sector but at the base of the Beaufort Group the distal sector (Rubidge, 2005). By the end of the Permian (Dicynodon Assemblage Zone) the last remnant of the former Ecca ?sea? was covered by terrestrial (?Beaufort?) deposits and continental fluvial deposition occurred in both sectors. Various authors have pointed out that the Beaufort Group sedimentary infill thins dramatically as one proceeds north ward in the basin (Johnson, 1976; Cole, 1992) and this is due to a number of factors. The loading of the adjacent orogenic belt in the proximal sector was the principal cause of this northward thinning. A secondary cause was the position of the forebulge, which migrated within the foreland system and had the potential to generate areas of non deposition. Catuneanu et al. (1998) postulated that at the end of the Permian the forbulge was positioned immediately south of the present Thaba Nchu. Neveling (2002) theorised that the Katberg sandstone in the southern part of the basin comprises three units of which the lowermost and uppermost units are restricted to the south of the basin. The middle (Swartberg Member) is present in both the north and south. The depositional hiatus in the north of the basin, both above and below the Katberg Formation, thus supports the idea of a reciprocal stratigraphy. In the Thaba Nchu area the absence of Lystrosaurus maccaigi and the lack of a stratigraphic overlap between Lystrosaurus and Dicynodon indicates that this stratigrigraphic interval, which is present in the proximal sector (Botha and Smith, 2007), is absent in the study area. Similarly the stratigraphic overlap between Procolophon and faunal elements of the Cynognathus Assemblage Zone, which has been reported in the 85 proximal sector (Neveling, 2002), is not present in the study area and further north (Neveling, 2002). This suggests that while the Dicynodon and Lystrosaurus assemblage zones and the Procolophon and Cynognathus Assemblage Zone overlap periods were present in the proximal sector, they were absent in the entire distal sector. This corroborates the idea of Neveling (2002) that the lower and upper units of the Katberg Formation occur only in the proximal sector. Geophysical research carried out for this study (Chapter 4) has indicated that the Thaba Nchu area is positioned to the west of a series of palaeovalleys. The presence of Ecca Group rocks (Kingston et al., 1961) to the west of these valleys (i.e. on a palaeohigh) implies that these valleys had filled with sediment by Beaufort Group times. It is shown that neither of these valleys nor the other undulations in the floor of the Karoo Supergroup impacted on the deposition of Beaufort Group sediments within the study. These valleys and undulations lie well beneath the depositional surface upon which the sedimentary rocks of the study area, now exposed, were deposited. The large gravity anomaly to the east of the study area (Anomaly G, Fig. 4.4) is explained by the presence of the Kaapvaal Craton boundary. This feature has a similar gravity profile to the south of the study area. The western gravity high is explained by the westward thinning of the Karoo strata and the presence of basement rocks at surface. The overall north-westerly palaeocurrent direction of the Beaufort Group (see chapter 2) suggests that the main provenance area was in the south and was most probably the Cape Fold Belt, which is consistent with the findings of previous workers (eg. Johnson, 1976; Botha and Linstrom, 1977; Hiller and Stravakis, 1984). By contrast the south westerly palaeocurrent direction of the thin Musgrave grit unit is at right angles to the overall basinal trend. As discussed in Chapter 1 various authors have argued for a variety of provenance areas for the coarse grained sediments of the Musgrave grit unit. The suggestion that intrabasinal Archean granite to the west of Thaba Nchu could have been the source (Veevers et al., 1994) is improbable as the palaeocurrent direction is 86 toward the south-west. However the geophysical and borehole data indicate the presence of a palaeo-high toward the north-east (Chapter 4, Anomaly C Fig. 4.14), which is sufficiently close to account for the immature nature of the Musgrave unit. This unit is present as a marker bed at a particular stratigraphic level in the Balfour Formation at several localities during Dicynodon Assemblage Zone times (Chapter 2) and suggests activity of the basement high relative to the southern Cape Fold Belt source for a short period of time at the beginning of the Triassic or end of the Late Permian periods. 87 Chapter 6 Conclusion This study has shown that within the central Free State, which lies along the stratigraphic hinge line described by Catuneanu et al. (1998), the Balfour, Katberg and Burgersdorp formations of the Beaufort Group are all present. The Balfour Formation is mainly argillaceous but includes the very coarse Musgrave grit unit. The tripartite Katberg Formation includes a basal arenaceous Sepenare?s Hoek sandstone, a middle argillaceous Townland fines unit and an upper Eden sandstone member. The overlying Burgersdorp Formation consists of the Mohono mudrock unit. Biostratigraphic evidence has shown that the upper three biozones of the Beaufort Group (Dicynodon, Lystrosaurus and Cynognathus assemblage zones) are present within the study area. The Cynognathus Assemblage Zone has been shown to occur within this area of the Free State Province. Furthermore the presence of subzones A and B of the Cynognathus Assemblage Zone are revealed. Subzone B specimens were found within 30 m of the bottom of the Molteno Formation. However there is no evidence that subzone C, of the Cynognathus Assemblage Zone, is present in the area. The absence of Subzone C is as expected as it is only found in a narrow zone in the extreme south of the basin (Hancox, 1998). The Lystrosaurus Assemblage Zone is restricted to the lower part of the Katberg Formation (i.e. Sepenare?s Hoek sandstone and lower Townland fines units) with no overlap in the fauna of the Lystrosaurus and Dicynodon assemblage zones occurring. In addition no stratigraphic overlap between faunal elements of the Lystrosaurus and Cynognathus assemblage zones, as is recorded in the south of the basin (Neveling, 2002), occurs in the study area. It is thus evident that parts of the stratigraphic succession at the base and top of the Lystrosaurus Assemblage Zone as well as the top of the Cynognathus Assemblage Zone, 88 which are recorded in the south of the basin, are not present in the study area as is the situation in the most northerly regions of the Beaufort basin. The absence of these portions of the stratigraphic interval suggests that that no sediment deposition took place in the study area in the time interval at the beginning and also at the end of the Lystrosaurus Assemblage Zone, and also at the end of the Cynognathus Assemblage Zone. Catuneanu et al. (1998) suggested that the Karoo foreland basin behaved as a partitioned basin with reciprocal infilling. The absence of the base and top of the Lystrosaurus Assemblage Zone and subzone C of the Cynognathus Assemblage Zone in the study area, which was positioned just north of the forebulge, shows that during those brief times deposition in the Karoo basin was restricted to the proximal sector and that no deposition took place north of the hinge line. This finding is in line with that of Catuneanu et al. (1998). It is significant that all palaeocurrent directions throughout the stratigraphic interval studied indicate a south-easterly source area except for the Musgrave unit, which demonstrates a north-eastern source was active for a short period in late Dicynodon Assemblage Zone times. From the geophysical data presented it is clear that the basement floor, far from being smooth and homogenous, is heterogeneous and contains areas of substantial of topographic highs. Periodic exposure of these features would have provided a north-eastern provenance area for the Musgrave unit and a temporary alternative to the southerly source that supplied most of the sediments that formed the Beaufort Group within the Thaba Nchu area. 89 References Ahern, J.L. 2004. International Gravity Formula(e). University of Oklahoma, www.geophysics.ou.edu/solid_earth/notes/potential/igf.htm Allen, J.R.L. 1968. Current Ripples. North Holland, Amsterdam. 433pp Allen, J.R.L. 1994. Fundamental properties of fluids and their relation to sediment transport processes. In: Pye, K. (ed.). Sediment Transport and Depositional Processes, Blackwell, London. 397pp. Bamford, M. 1999. Permo-Triassic fossil wood from the South African Karoo Basin. Palaeontologia africana, 35, 25-40. Battrick, J. 1973. Geological Well Completion Report of BV 1/73, Soekor, Johannesburg, 21pp. Blakely, R.J. 1996. Potential Theory in Gravity and Magnetic Applications. Cambridge University Press, Cambridge. 441pp. Bonnet, L.J. 1977. Die Geologie van die Gebied 2827C Suidoos van Winburg Oranje ? Vrystaat. Unpublished MSc thesis. University of the Orange Free State. Bloemfontein. Boonstra, L.D. 1969. The fauna of the Tapinocephalus Zone (Beaufort beds of the Karoo): Annals of the South African Museum, 56(1), 1-73. Botha, B.J.V. and Linstrom, W. 1977. Paleogeological and paleogeographical aspects of the upper part of the Karoo sequence in north-western Natal. Annals of the Geological Survey South Africa. 90 Botha, J. and Smith, R.M.H. 2007. Lystrosaurus species composition across the Permo- Triassic boundary in the Karoo Basin of South Africa. Lethaia, 40, 125-137. Botha, B.J.V. and Linstrom, W. 1978. A note on the stratigraphy of the Beaufort Group in north-western Natal. Transactions of the Geological Society of South Africa, 81, 35-40. Bridge, J.S. 1978. Palaeohydraulic interpretation using mathematical models of contemporary flow in meandering channels. In: Miall, A.D. (ed.) Fluvial Sedimentology. Memoir Canadian Society of Petroleum Geologists, 5, 723-724. Bristow, C.S. 1993. Sedimentology of the Rough Rock: a carboniferous braided sandstone in northern England. In: Best J.L. and Bristow, C.S. (eds.). Braided Rivers. Geological Society of London Special Publication, 75, 291-304. Bristow, C.S. and Best, J.L. 1993. Braided rivers: perspectives and problems. In: Best, J.L., Bristow, C.S. (ed). Braided Rivers. Geological Society Special Publication, 75, 1-11. Broom, R. 1906. On the Permian and Triassic faunas of South Africa. Geological Magazine, Decade 5(3), 29-30. Brynard, H.J., Jakob, W.R.O., Le Roux, J.P. 1982. The sedimentology, mineralogy and geochemistry of the Mooifontein deposit, Orange Free State. Report of the Nuclear Development Corporation of South Africa. Burley, A.J., Kimbell, G.S., Patrick, D.J., Turnbull, G., Kashambuzi, R.. 1982. A Gravity Survey of Lesotho. Overseas Geology and Mineral Resources, Institute of Geological Sciences, No. 60, Her Majesty?s Stationary Office. 27pp. Cant, D.J. and Walker, R.G. 1976. Development of a braided-fluvial facies model for the Devonian Battery Poin Sandstone, Quebec. Canadian Journal of Earth Sciences, 13, 102- 109. 91 Cant, D.J. and Walker, R.G. 1978. Fluvial processes and facies sequences in sandy braided South Saskatchewan River, Canada. Sedimentology, 25, 625-648. Catuneanu, O., Hancox, P.J., Rubidge, B.S. 1998. Reciprocal flexural behaviour and contrasting stratigraphies: a new basin development model for the Karoo retroarc foreland system, South Africa. Basin Research, 10, 417-439. Catuneanu, O. and Elango, H.N. 2001. Tectonic control on fluvial styles: the Balfour Formation of the Karoo Basin, South Africa. Sedimentary Geology, 140 (3/4), 291-313. Catuneanu, O., Hancox, P.J., Cairncross, B., Rubidge, B.S. 2002. Foredeep submarine fans and forebulge deltas: orogenic off-loading in the underfilled Karoo Basin. Journal of African Earth Sciences, 35, 489-502. Cole, D.I. 1992. Evolution of the Karoo Basin. In: de Wit, M.J. and Ransome, I.G.D. (eds.). Inversion tectonics of the Cape Fold Belt, Karoo and Cretaceous Basins of South Africa. Balkema, Rotterdam, 87-99. Cooper, G. 2008. Geophysical Software. University of the Witwatersrand, www.wits.ac.za/Academic/Science/Geosciences/Research/Geophysics/GordonCooper. Cousins, C.A. 1950. Sub-Karroo Contours and notes on the Karoo succession in the Odendalsrus area of the Orange Free State. Transactions of the Geological Society of South Africa, 53, 229-242. Csaky, A.V. 1977. Stratigraphy and Hydrocarbon Potential of the Dwyka, Ecca and the Beaufort Groups in the Northern Karroo. Unpublished report No. 1977-0099. Geological Survey. Based on an unpublished Soekor report by Csaky A.V. and Wachsmuth W. 1971. 126pp. 92 De Wit, M.J, Roering, C., Hart., C., Armstrong, R.A., De Ronde, C.E.J., Green, R.W.E., Tredoux, M., Peberdy, E., and Hart, R.A. 1992. Formation of an Archean continent. Nature, 357, 553-562. du Toit, A.L. 1954. Geology of South Africa. Oliver and Boyd, London, 332pp. Duncan, R.A., Hooper, P.R., Rehacek, J., Marsh J.S., Duncan A.R. 1997. The timing and duration of the Karoo igneous event, southern Gondwana. Journal of Geophysical Research. 102(8), 18127 - 18138. Fagerstrom, J.A., 1967. Development, flotation, and transportation of mud crusts ? neglected factors in sedimentology. Journal of Sedimentary Petrology, 13(1), 73-79. Groenewald, G.H.. 1989. Stratigrafie en sedimentologie van die Groep Beaufort in die Noordoos-Vrystaat. Bulletin of the Geological Survey of South Africa, 96. Pretoria. 49pp. Groenewald, G.H.. 1990. Gebruik van paleontologie in litostratgrafiese korrelasie in die Beaufort Groep, Karoo Opeenvolging van Suid Afrika. Palaeontologia africana, 27, 21- 30. Groenewald, G.H.. 1996. Stratigraphy of the Tarkastad Subgroup, Karoo Supergroup, South Africa. Unpublished Ph.D. thesis, University of Port Elizabeth, South Africa. Groenewald, G.H., and Kitching, J.W. 1995. Biostratigraphy of the Lystrosaurus Assemblage Zone. In: Rubidge, B.S. (ed.). Biostratigraphy of the Beaufort Group (Karoo Supergroup). South African Commission for Stratigraphy, Biostratigraphic Series, 1, 35- 39. Hamilton, G.N.A. and Cooke, H.B.S. 1960. Geology for South African Students, 4th ed., C.N.A. South Africa, 441p. 93 Hancox, P.J. 1998a. A stratigraphic, sedimentological and palaeoenviromental synthesis of the Beaufort-Molteno contact in the Karoo Basin. Unpublished Phd thesis, University of the Witwatersrand, Johannesburg. Hancox, P.J. 1998b. The Beaufort-Molteno contact revisited: ramifications for the development of the Karoo retro-foreland basin during the Triassic. Journal of African Earth Sciences, 27(1A), 103-104. Hancox, P.J. 2000. The continental Triassic of South Africa. Zentralblatt f?r Geologie und Palaontologie, Teil I (11-12), 1285-1324. Hancox, P.J., Brandt, D., Reimold, W.U., Koeberl, C., and Neveling, J., 2002. The Permo- Triassic boundary in the Northwest Karoo Basin: Current stratigraphic placement, implications for basin development models, and the search for evidence of impact. In: Koerberl, V. and MacLeod, K.G. (eds), Catastrophic Events and Mass Extinctions: Impacts and Beyond, 429-444 (Geological Society of America, Special Paper, 356). Hancox, P.J., Rubidge B.S. 1997. The role of fossils in interpreting the development of the Karoo Basin. Palaeontologia africana, 33, 41-54. Hancox, P.J., Rubidge B.S. 2001. Breakthroughs in the biodiversity, biogeography, biostratigraphy and basin analysis of the Beaufort group. Journal of African Earth Sciences, 33 (3/4), 563-577. Hancox, P.J., Shishkin, M.A., Rubidge B.S., Kitching, J.W. 1995. A threefold subdivision of the Cynognathus Assemblage Zone (Beaufort Group, South Africa) and its palaeogeographical implications. South African Journal of Science, 91, 143-144. Harms, J.C, Southard, J.B., Spearing, D.R. and Walker, R.G. 1975. Depositional environments as interpreted from primary sedimentary sequences. Society of Economic Paleontologists and Mineralogists, Short Course Notes 2, 161pp. 94 Harms, J.C, Southard, J.B., Walker, R.G. 1982. Structures and sequences in clastic rocks. Society of Economic Paleontologists and Mineralogists Short Course Notes 9, 249pp. Harvey, J.D, de Wit, M.J., Stankiewicz, J, Doucoure, C.M. 2001. Structural variations of the crust in the Southwestern Cape, deduced from receiver functions. South African Journal of Geology, 104, 231-242. Haughton, S.H. (1963). Stratigraphic history of Africa south of the Sahara. Oliver and Boyd, Edinburgh, 365p. Hiller, N. and Stavrakis, N. 1984. Permo-Triassic fluvial systems in the southeastern Karoo Basin, South Africa. Palaeogeography Palaeoclimatology Palaeoecology, 45, 1-21. Hotton, N. and Kitching, J.W. 1963. Speculations on Upper Beaufort deposition. South African Journal of Science, 59, 254-258. Jackson, R.G. 1975. Hierarchical attributes and a unifying model of bed forms composed of cohesionless material and produced by shearing flow. Bulletin of the Geological Society of America., 86(11), 1523-1533. Jones, B.G. and Rust, B.R.. 1983. Massive sandstone facies in the Hawkesbury Sandstone, a Triassic fluvial deposit near Sydney, Australia. Journal of Sedimentary Petrology, 53, 1249-1261. Johnson, M.R. 1966. Stratigraphy and Sedimentology of the Cape and Karoo Sequences in the Eastern Cape Province. Unpublished M.Sc. thesis, Rhodes University, Grahamstown, South Africa. 76p. 95 Johnson, M.R. 1976. Stratigraphy and Sedimentology of the Cape and Karoo Sequences in the Eastern Cape Province. Unpublished Ph.D. thesis, Rhodes University, Grahamstown, South Africa. 336p Johnson, M.R. 1984. The Geology of the Queenstown Area. Explanation of Sheet 3126. Geological Survey. Pretoria. Johnson, M.R. 1991. Sandstone petrography, provenance and plate tectonic setting in Gondwana in the context of the south-eastern Cape-Karoo Basin. South African Journal of Geology, 94, 137-154. Johnson, M.R., van Vuuren, C.J., Visser, J.N.J., Cole, D.I., Wickens, H. de V., Christie, A.D.M., and Roberts, D.L. 1997. The Foreland Karoo Basin, South Africa. In: Selley, R.C. (ed.), African Basins. Reprinted from Sedimentary Basins of the World, 3. Elsevier, Amsterdam, 269-317. Kingston, J., Woodward, J.E., Malan, S.P. and Jennings, R.A. 1961. Geology and Petroleum Prospects of the Karoo Basin. Unpublished Report, Soekor. Kitching, J.W. 1995a. Biostratigraphy of the Dicynodon Assemblage Zone. In: Rubidge, B.S. (ed.), Biostratigraphy of the Beaufort Group (Karoo Supergroup). South African Commission for Stratigraphy, Biostratigraphic Series, 1, 29-34. Kitching, J.W. 1995b. Biostratigraphy of the Cynognathus Assemblage Zone. In: Rubidge, B.S. (ed.), Biostratigraphy of the Beaufort Group (Karoo Supergroup). South African Commission for Stratigraphy, Biostratigraphic Series, 1, 40-45. Kruger, F.J. 1975. Paleostream channels in the Beaufort Group near Laingsburg ? Cape Province. Petros, 6, 44-51 96 Kubler, M. 1977. The sedimentology and uranium mineralization of the Beaufort Group in the Beaufort West-Fraserburg-Merweville Area, Cape Province. Unpublished MSc thesis, University of the Witwatersrand, Johannesburg. Le Roux, G.G. and Keyser, A.W. 1988. Die Geologie van die Gebied Victoria-Wes. Toeligting van Blad 3122. Geological Survey. Pretoria. Leith, M.J. and Trumpelmann, F. 1967. Well Completion Report for Southern Oil Exploration Corporation (Pty) Limited of WE 1/66, Soekor, Johannesburg, 19pp. Lucas, S.G. 1998. Global Triassic tetrapod biostratigraphy and biochronology. Palaeogeography, Palaeoclimatology, Palaeoecology, 143, 347-384. Macleod, K.G., Smith, R.M.H., Koch, P.L., Ward, P.D. 1999. Timing of mammal-like extinctions across the PT boundary in South Africa. Geology, 28, 227-230. Mar?, L.P. and Oosthuizen, B.C. (2000). Physical properties of South African rocks, version II, vol. IV, 19pp., Council for Geoscience, Pretoria. Mare, L.P. and Cole, J. 2006. The Trompsburg Complex, South Africa: A preliminary three dimensional model. Journal of African Earth Sciences, 44, 314-330. Martin, C.A.L. and Turner, B.R. 1998. Origins of massive-type sandstones in braided river systems. Earth Science Reviews, 44, 15-38. McBride, E.F. and Yeakel, L.S. 1963. Relationship between parting lineation and rock fabric. Journal of Sedimentary Petrology, 33(3), 779-782. McCarthy, T.S. 2006. The Witwatersrand Supergroup. In: Johnson, M.R., Anhaeusser, C.R. and Thomas, R.J. (eds). The Geology of South Africa. Geological Society of South Africa, Johannesburg / Council for Geoscience, Pretoria. 155-186. 97 McKee E.D., Crosby E.J. and Berryhill H.L. 1967. Flood deposits of Bijou Creek, Colorado, June 1965. Journal of Sedimentary Petrology, 37, 829-851. Miall, A.D. 1977. A review of the braided river depositional environment. Earth Science Reviews, 13, 1-2. Miall, A.D. 1978. Lithofacies types and vertical profile models in braided river deposits: a summary. In: Miall, A.D. (ed.), Fluvial Sedimentology. Canadian Society of Petroleum Geologists, Memoir, 5, 597-604. Miall, A.D. 1996. The Geology of Fluvial Deposits, Sedimentary Facies, Basin Analysis, and Petroleum Geology. Springer, Berlin. 582pp. Mountain, E.D. 1946. The geology of an area east of Grahamstown: Explanation Sheet 136 (Grahamstown) Geological Survey South Africa, 56pp. Musset, A.E. and Kahn, M.A.. 2000. Looking Into the Earth, an Introduction to Geological Geophysics.Cambridge University Press. Cambridge. 470pp. Myers, K.J. and Milton, N.J. 1996. Concepts and Principles. In Emery, D. and Myers, K.J. (ed.), Sequence Stratigraphy. Blackwell Science, London, 296 p. Nanson, G.C. 1980. Point bar and floodplain formation of the meandering Beatton River, northeastern British Columbia, Canada. Sedimentology, 27, 3-10. Neveling, J. 1998. Fossil tetrapod ranges across the Lystrosaurus and Cynognathus Assemblage Zone contact (Beaufort Group, Karoo Supergroup) in the Early Triassic of South Africa. Journal of African Earth Sciences, 27(1A), 143-144. 98 Neveling, J. 2002. The Biostratigraphy and Sedimentology of the Contact Area Between the Lystrosaurus and Cynognathus Assemblage Zones (Beaufort Group: Karoo Supergroup). Unpublished Phd thesis, University of the Witwatersrand, Johannesburg. 232p. Neveling, J., Rubidge, B.S. and Hancox, P.J. 1999. A lower Cynognathus Assemblage Zone fossil from the Katberg Formation (Beaufort Group, South Africa). South African Journal of Science, 95, 555-556. North, C.P. and Taylor, K.S. 1996. Ephemeral-fluvial deposits: Integrated outcrop and simulation studies reveal complexity. American Association of Petroleum Geologists Bulletin, 80(6), 811-830. Ochev, V.G. and Shishkin, M.A. 1989. On the principals of global correlation of the continental Triassic on the Tetrapods. Acta Palaeontologica Polonica, 34(2), 149-172. Paola, C., Wiele S.M. and Reinhart, M.A. 1989. Upper regime parallel lamination as a result of turbulent sediment transport and low amplitude bedforms. Sedimentology, 36, 47- 59. Pyke, W.J. 1973. Geological Well Completion Report of WI 1/68. Soekor. Johannesburg. 18p. Reading, H.G. 1986. Sedimentary Environments and Facies. 2nd ed. Blackwell. Oxford. 330p Reid, A.. 1980. Aeromagnetic survey design. Geophysics, 45, 973-976. Reimold, W.U. and Gibson, R.L. 1996. Geology and evolution of the Vredefort impact structure. Journal of African Earth Science, 23, 125-163. 99 Reineck, H.E. and Singh, I.B. 1973. Depositional Sedimentary Environnments. Berlin, Springer Verlag. 439pp. Roux, H.J. 1972. Geological Well Completion Report of LA 1/68. Soekor. Johannesburg. 17p. Roux, H.J, 1974. Geological Well Completion Report of FI 1/72. Soekor. Johannesburg. 14p. Rubidge, B.S. 1995 (ed.), Biostratigraphy of the Beaufort Group (Karoo Supergroup). South African Commission for Stratigraphy, Biostratigraphic Series, 1, 45 pp. Rubidge, B.S. 2005. Re-uniting lost continents-fossil reptiles from the ancient Karoo and their wanderlust. South African Journal of Geology, 108, 135-172. Rust, I.C. 1962. On the sedimentation of the Molteno sandstones in the vicinity of Molteno, Cape Province. Annals of the University of Stellenbosch, 37, Series a, No.3, 165- 236. Rust, I.C. 1975. Tectonic and sedimentary framework of Gondwana Basins in southern Africa. Third Gondwana Symposium, Australia, 5, 554-564. Rust B.R.. 1978. Depositional models for braided alluvium. In: A.D. Miall (ed.), Fluvial Sedimentology. Canadian Society of Petroleum Geologists, Memoir 5, 605-625. Smith, N.D. 1971. Transverse bars and braiding in the lower Platte River, Nebraska. Geological Society of America, Bulletin 82, 3407-3420. Smith, R.M.H. 1980. The lithology, sedimentology, and taphonomy of flood-plain deposits of the lower Beaufort (Adelaide Subgroup) strata near Beaufort West. Transactions of the Geological Society of South Africa, 83, 399-413. 100 Smith, R.M.H. 1990a. A review of stratigraphy and sedimentary environments in the Karoo Basin of South Africa. Journal of African Earth Sciences, 10, 117-137. Smith, R.M.H. 1990b. Alluvial paleosols and pedofacies sequences in the Permian lower Beaufort of the southwestern Karoo Basin, South Africa. Journal of Sedimentary Petrology, 60(2), 258-276. Smith, R.M.H. 1995. Changing fluvial environments across the Permian-Triassic boundary in the Karoo Basin, South Africa and possible causes of tetrapod extinctions. Palaeogeography Palaeoclimatology, Palaeoecology, 117, 81-104. Smith, R.M.H., Eriksson, P.G., and Botha, W.J. 1993. A review of the stratigraphy and sedimentary environments of the Karoo-aged basins of Southern Africa. Journal of African Earth Sciences 16 (1/2), 143-169. Smith, R.M.H. and Ward, P.D. 2001. Pattern of vertebrate extinctions across an event at the Permian Triassic boundary in the Karoo Basin of South Africa. Geology, 29, 1147- 1150. Sneh, A. 1983. Desert stream sequences in the Sinai Peninsula. Journal of Sedimentary Petrology, 53, 1271-1280. South African Committee for Stratigraphy (S.A.C.S.). 1980. Stratigraphy of South Africa, pt 1 (compiled by L.E. Kent). Lithostratigraphy of the Republic of South Africa, South West Africa/Namibia, and the Republics of Bophuthatswana, Transkei and Venda. Handbook Geological Survey South Africa, 8. 690p. Soekor Plan A2 1969. North Karroo Seismic Survey, Depth Contour Map of Base of the Karroo System. Unpublished map, author unknown. Plan held by the Council for Geoscience. Sageolit number 90484897. Soekor Report Collection. 101 Stavrakis, N. 1980. Sedimentation of the Katberg Sandstone and adjacent formations in the south eastern Karoo Basin. Transactions of the Geological Society of South Africa, 83, 361-374. Stanistreet, I.G. and McCarthy, T.S. 1991. Changing tectono-sedimentary scenarios relevant to the development of the late Archean Witwatersrand Basin. Journal of African Earth Science, 13, 6-32. Stear, W. 1983. Morphological characteristics of ephemeral stream channel and overbank splay sandstone bodies in the Permian Lower Beaufort Group, Karoo Basin, South Africa. Special Publication International Association Sedimentologists, 6, 405-420. Stear, W.M. 1985. Comparison of the bedform distribution and dynamics of modern and ancient sandy ephemeral flood depostits in the southwestern Karoo region, South Africa. Sedimentary Geology, 45, 209-230. Swan, A.R.H and Sandilands, M. 1995. Introduction to Geological Data Analysis. Blackwell Science, London. 446p. Telford, W.M., Geldart, L.P., Sheriff, R.E. and Keys, D.A. 1990. Applied Geophysics. (2nd ed.) Cambridge University Press, Cambridge. 770pp. Theron, J.C., 1970. Some Geological Aspects of the Beaufort Series in the OFS. Unpublished Ph.D. thesis, University of the Orange Free State, Bloemfontein, South Africa. Theron, J.C., 1982. A note on a pre-karoo outcrop south-east of Kroonstad, O.F.S. Transactions of the Geological Society South Africa, 85, 59-60. Tucker, M., 1988. Techniques in Sedimentology. Blackwell, Oxford. 102 Tunbridge, I.P. 1981. Sandy high-energy flood sedimentation ? some criteria for recognition, with an example from the Devonian of SW England. Sedimentary Geology, 28, 79-96. Turner, B.R. 1975. The stratigraphy and sedimentary history of the Molteno Formation in the main Karoo Basin of South Africa and Lesotho. Unpublished Ph.D. thesis, University of the Witwatersrand, Johannesburg, South Africa. Turner, B.R. 1981a. The occurrence, origin and stratigraphic significance of bone bearing mudstone pellet conglomerate from the Beaufort Group in the Jansenville district, Cape Province, South Africa. Palaeontologica Africana., 24, 62-73 Turner, B.R., 1981b. Possible origin of low angle cross-strata and horizontal lamination in Beaufort Group sandstones of the southern Karoo Basin. Transactions of the Geological Society of South Africa, 84, 193-197. Turner, B.R. 1999. Tectonostratigraphical development of the Upper Karoo foreland basin: orogenic unloading versus thermally-induced Gondwana rifting. Journal of African Earth Sciences, 28(1), 215-238. Veevers, J.J., Cole, D.I., Cowan, E.J. 1994. Southern Africa: Karoo Basin and Cape Fold Belt. Geological Society of America, Memoir 184, 223-279. Visser, J.N.J. and Dukas, B.A. 1979. Upward-filing fluviate megacycles in the Beaufort Group, north of Graaf-Reinet, Cape Province. Transactions of the Geological Society of South Africa, 82, 149-154. Welman, J., Groenewald, G.H., Kitching, J.W. 1991. Confirmation of the occurrence of Cynognathus Zone (Kannemeyeria-Diademodon Assemblage-zone deposits (upper Beaufort Group) in the northeastern Orange Free State, South Africa. South African Journal of Geology, 94(2), 245-248. 103 Figures: Chapter 1 104 Figure 1.1: Location of the study area (hatched), showing the major cities and towns mentioned in the text. Inset shows position of main map in southern Africa. 105 Figure 1.2: Geological map showing the distribution of the rocks of the Karoo Supergroup in South Africa (after Rubidge, 2005). Note the informal term ?Stormberg Group?, refers collectively to the Molteno, Elliot and Clarens formations. The red square delineates the study area. 106 Figure 1.3: Distribution of the upper three biozones of the Beaufort Group in the vicinity of Thaba Nchu (after Rubidge, 1995). 107 Figure 1.4: Map showing the location of some of the previous studies referenced in the text. Note that Theron (1970) described the Beaufort Group throughout the entire Free State Province so is not included on this map. 108 Figures: Chapter 2 109 Figure 2.1: Possible thrust fault (arrow) on Thaba Nchu Mountain. The two thick sandstone units which are over thrust are in the Elliot Formation. Figure 2.2a: Clasts in facies Gt. The feldspar crystals are fresh and poorly rounded. Centimetre scale tape measure. 110 Figure 2.2b: Fossil wood in facies Gt. Logs larger than these were used to provide paleocurrent information. Tape measure is extended to 370 mm Figure 2.2c: Trough cross stratification within facies Gt (bracketed). Tape measure is 80 mm. 111 Figure 2.3a: Intraformational conglomerate, Facies Sei (bracketed). Note how the second unit follows the erosional scour and contains bedding. Hammer for scale is 270 mm. Figure 2.3b: Matrix supported mud pebble conglomerate (facies Sei). Note the size of the clasts in this image is relatively uniform and small in comparison with other localities where some clasts are much larger. Coin for scale is 21 mm. Sei 112 Figure 2.4a: A unit of horizontally laminated fine grained sandstone (facies Sh). This facies forms the bulk of all the facies types in the area. Note that what appear to be thin beds are in fact laminae some of which are preferentially. Dip is due to the presence of a dyke and is not a depositional remnant. Markings on staff are 100 mm long. Figure 2.4b: Parting lineation on the surface of a unit of facies Sh. Arrow shows palaeocurrent trend. Pen for scale is 140 mm in length. 113 Figure 2.5: Low angle cross stratified sandstone (Facies Sl) in sharp contact with a unit of planar cross stratified sandstone (facies Sp). Both facies are rarely found within the area with the former difficult to identify. Hammer is 270 mm. Figure 2.6a: Cosets of trough cross stratified sandstone (facies St) (bracketed). For scale markings on the staff are 100 mm long. Sl Sp 114 Figure 2.6b: Surface expression of trough cross bedding (facies St). Measured down the trough axis gives a paleocurrent direction (NNW) for this unit (i.e. into the page). Hammer for scale is 300 mm. Figure 2.7: Ripple marks are found at a number of localities, typically occurring as a single layer above facies Sh. Here interference ripple marks are exposed on the top surface a unit of Sh. Hammer for scale is 270 mm. 115 Figure 2.8: Ripple cross stratified sandstone (facies Sr) formed by the downstream migration of ripples. The facies caps a unit of facies Sh. Lens cap for scale is 55 mm. Figure 2.9: A unit of facies Sm (bracketed) above and below units of Sh. Hammer for scale is 300 mm. Sm 116 Figure 2.10a: Photomicrograph of a fine-grained sandstone from the study area. Typical features include the fine, rounded to subangular and sorted monocrystaline quartz. Figure 2.10b: Peloid within a fine grained sandstone. Note the typical rounded to sub angular fine grained quartz matrix. The peloid is rounded and contains black fragments that may be fossilized plant material. 117 Figure 2.10c: Quartz (Qtz) and feldspar n(Fspar) (microcline) fragments. Figure 2.10d: A rounded polycrystalline quartz grain. Some hematite staining around the edges of the grain is present (arrowed). 0.2 mm 0.2 mm Fspar Qtz Qtz Qtz Qtz 118 Figure 2.10e: Quartz grain included within a feldspar grain forming a lithic fragment suggesting a metamorphic or igneous rock provenance. Figure 2.10f: Fine grained sandstone commonly forms the matrix to the mudpebble conglomerate. Here it encloses a mud pebble (10 mm diameter) consisting of replacive prismatic calcite. Quartz Feldspar Calcite 0.1 mm 119 Figure 2.11: Facies Fl comprising interbeded sand-, silt- and mudstone horizons on a scale of a metre. Each rock type varies in thickness from 10-300 mm. Contacts, where they are visible, are abrupt. Note the blocky weathering. Hammer for scale is 300 mm. Figure 2.12: Facies Fsm consists of couplets of mud- and siltstone. This facies differs from facies Fl in that the thin units of fine grained sandstone are absent. Hammer for scale is 300 mm 120 Figure 2.13: Facies Fm consists of massive mudstone which in the study area weathers rapidly and into fragmentary chips of rock. Colour varies from red, green to black and as in this photograph mottling occurs. No units thicker than 500 mm were observed but this may be an artefact of the rapid weathering. Rare example of bedding seen in facies Fm. Hammer for scale is 270 mm. Figure 2.14a: Calcareous nodules (bracketed) within a pedogenic sequence. Coin for scale is 20 mm 121 Figure 2.14b: Fragment of fossilised bone (arrow) in a concretionary nodule. 122 Figure 2.15: Location of the architectural elements discussed in the text. CHm = Major Channel, CHt = Stacked Sand Channel, CS = Crevasse Splay, DA = Downstream Accreting Macroform, LS = Laminated Sandstone Sheets, FP = Pedogenic Sequence, FF = Flood Plain Fines, CHf = Flood Plain Channel. Figure 2.16a: Lateral profile showing both architectural element CHm, cross cutting channels within a larger channel (top of ridge) and architectural element LS (lower outcrop on photograph/diagram). The numbers refer to bounding surfaces. All surfaces dip to the right but are artificially distorted due to camera angle. Stratigraphically the upper outcrop is Eden Sandstone whilst the element LS is within the Townlands Mudstones. Outcrop is on Moroto Mountain. Exposure is orientated along 166o-346o, which is oblique to the palaeocurrent direction of 004o. The outcrop is exposed for 90 m. Element CHm Element LS LS Townland fines Eden sandstone CHm 124 Figure 2.16b: Close up view of one of the smaller channels in Figure 2.30a. Palaeocurrent direction is into and to the right of the photograph (350o). Location is on Moroto Mountain and stratigraphically in the Eden Sandstone Unit. Hammer (circled) for scale is 270 mm. Figure 2.17: Stacked channels (element CHt) on the farm Abramskraal. Individual channels (defined by dotted lines in figure) have a mean thickness of 500 mm and are built up of units of Sh-St-Sr. The larger channel in which they occur is laterally extensive with a width of at least 350 m but a mean thickness of 1.5 m. Paleocurrent direction is into the photograph. Hammer for scale is 270 mm. 125 Figure 2.18a: Element DA ? downstream accreting macroform. Miall (1996) gives as some of the defining characteristics the slightly concave up 4th order bounding surfaces and a dip of less than 10o in the downstream direction. Palaeocurrent direction is toward the left and obliquely out of the page. Note that the lower unit is built up almost entirely of Sh except for an erosional unit of Sei into the underlying unit of Fm. The other units have abrupt bases. Outcrop on Abramskraal. Hammer for scale 270 mm. Figure 2.18b: Element DA. This bedform consists of wedge shaped units. Palaeocurrent direction is obliquely out of the page to the left. Abramskraal Farm. Sh Figure 2.19a: Pedogenic sequence. Note the continuous layers of concretionary nodules (arrowed). Stratigraphic section was measured along the line (Figure 2.19b). Location is on the farm Melden Drift 28 and stratigraphically within the Townland Fines. For scale section line is 3.3 m. 127 Figure 2.19b: Stratigraphic section, along line in Fig. 2.19a, shows pedogenic features. See p 194 for legend. C u m ulativ e Thick n ess m Thick n ess m m Typ e of C o ntact F acies C od e M ud sto n e Siltsto n e Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate C olo u r/N otes 2.2 200 Concretions are coated in green mud 2 200 Scattered and isolated 5cm diameter concretions Mud and silt pebbles also scattered through (1-2cm) 200 Interlocking, irregular shaped nodules 200 Thin veneer of mud at top 1.5 200 Red/brown 200 1 200 Top forms continuous Surface of knobbly 200 30cm concretions Concretions (capped with purple mudstone ) 200 0.5 Rhizocretions, cutins 200 purple 0.2 200 Silty sand, green 128 Figure 2.19b cont.: Stratigraphic section, along line in Fig. 2.19a, showing pedogenic features. C u m ulativ e Thick n ess m Thick n ess m m Typ e of C o ntact F acies C od e M ud sto n e Siltsto n e Fin e S and sto n e M ediu m C o arse S and sto n e C o nglo m erate C olo u r/N otes Scree for + 2m, then Sei 3.3m 200 3m Purple and green 200 Purple mudstone Flat topped, not continuous 200 2.5m Continuous, smooth 200 Laminated silt and mudstone (red) Hummocky bottomed, coalesce into units Figure 2.20: Flood plain channel (CHf). Note that the base of the channel is not the lenses of sandstone but rather a much finer grained silty unit beneath it. Other than the ripple tops or mud veneer, the lenses of sandstone are made up of facies Sh. Location is the farm Sepenare?s Hoek and stratigraphically within the Dubbledam mudrocks. 130 Figure 2.21: The flood plain channel (CHf) at Thaba Nchu Townlands consists of overlapping lenses 3-5m wide and 100-500mm thick. Each sandstone unit is made up of either facies Sr or Sh and is capped by a thin unit of siltstone or mudstone (Fsl and Fm). The bases are thin units of Sei. Figure 2.22a: Crevasse splay (element CS) (main splay arrowed, smaller events bracketed) at Feloana. The strata dip to the right at between 4o-6o. The section through the outcrop was taken along the displayed line (Fig 2.22b). Stratigraphically this forms part of the Dubbledam Mudrock Unit. For scale the section line is 5.5m. 132 Figure 2.22b: Section through the outcrop at Feloana. Note the large amount of sandstone. The main crevasse splay noted in Figure 2.22a is arrowed. See p 196 for legend. C u m ulativ e Thick n ess m Thick n ess m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate C olo u r/N otes F o rm atio n 5.5m Angular to sub-rounded granite and f/spar in fine sst matrix. 5m Sh White quartz rich, units cut into and overlap one another laterally Khaki silty sst intercolated with fine 0.8 Fl to medium sand which becomes more abundant upward 4m Sh White with orange stain- ripple surfaces 0.4 Fl White sand with grey silt Sh White (orange stain) Fine white sand with 0.3 Fl Sr orange stain on parting and grey siltstone 3m White with orange stain on partings 0.5 Sh S Fl Siltstone with lenses of sandstone 0.22 Sh Sp White with orange stain partings 2m 0.05 0.02 0.04 S Black carbonaceous with mica flakes 0.03 Fsm Laminated siltstone Dark grey mudrock 0.02 Fm Laminated blue/khaki siltstone 1m 1 Fsm Laterally discontinuous changes to Fm Fm Blue/grey D ubbled a m M ud ro ck U nit Main Crevasse Splay 133 Figure 2.23. Outcrop of flood plain fines (architectural element FF). Hammer for scale is 270 mm. 134 Figure 2.24: Map showing locality of stratigraphic sections taken in the Thaba Nchu area. The detailed locality maps and the sections themselves are provided in Appendix A. 135 Figure 2.25: Measured stratigraphic sections placed on a map to indicate the geographic distribution of lithological units. See Appendix A for stratigraphic sections. Other features as per Figure 2.24. 136 Figure 2.26: Map showing the location of farms mentioned in the text. 137 Figure 2.27: Thick packages of facies Fsm makes up a large portion of the Dubbledam mudrock unit. Note that in this exposure the mudrocks are predominately green, however fragments of rock from a unit of red mudrocks at the top of the outcrop gives the illusion that it consists of purple/red facies Fsm. Figure 2.28a: Typical exposure of tabular sandstone unit in Dubbledam Mudrock Unit. Two sets of facies Sl are seen here with 40-80 mm of laminated silt- and very fine sandstone separating them. Figure 2.28b below is a close up view of the unit. Bag for scale bag is 500 mm in height. 138 Figure 2.28b: Close up of a tabular sandstone unit. Two sets of facies Sl with laminated siltstone and very fine sandstone between. Note that the unit, and in particular the bottom sandstone, dips to the left. The outcrop is only 10 m long so it was not possible to determine if these are stacked sandstones. Figure 2.29: Colour mottling within the Dubbledam mudrock unit. Hat for scale 340 mm across. 139 Figure 2.30: Laminated siltstone and very fine grained sandstone above a tabular unit of fine sandstone in the Dubbledam mudrock unit. Note the small bar form (arrowed) to the right of the bag. Bag for scale is 500 mm in height. Figure 2.31: Invertebrate traces on the surface of a tabular sandstone in the Dubbledam mudrock unit. Coin for scale is 20 mm in diameter. 140 Figure 2.32: Bedforms within the Dubbledam mudrock unit on the farm Ratabane 100. Hammer for scale is 270 mm. Figure 2.33: Smaller scale troughs filled with facies Gt and making up the Musgrave grit unit. This outcrop forms the upper layer at the Feloana Dam locality. Hammer for scale is 300 mm. 141 Figure 2.34: Outcrop of the Sepenare?s Hoek sandstone unit on the farm Sepenare?s Hoek 759. Hammer for scale is 270 mm. Figure 2.35: Townland fines unit (bracket T.F.) with small channel sandstone (bracket CHf). Location is on the Thaba Nchu Townlands. CHf T.F. 142 Figure 2.36: The Eden sandstone unit on the western slope of Mohono Mountain. View is looking toward the east, palaeocurrent direction is north-north-west . Figure 2.36A: Close up of inset A in Fig. 2.26. View is to the south along the edge of the outcrop. Hammer for scale is 300 mm. Figure 2.36B: Close up of some of the large channels forming the Eden sandstone. Note the thick unit of facies Sei above the hammer. View is to the east.Hammer for scale is 270 mm. A B 143 Figure 2.37: Typical outcrop of Mohono mudrocks unit on Eden 96 (Mohono Mountain). Hammer for scale is 300 mm. Figure 2.38: Part of a channel within the Mohono sandstone unit on Thaba Nchu Mountain (Wilgeboomnek 29). Hammer for scale is 300 mm. 144 Figure 2.39: Rhizocretions found within the Mohono mudrock unit. 50 mm 145 Figure 2.40a: Palaeocurrent directions from the Dubbledam mudrock unit. Mean current direction is 340 from 18 readings. Readings were taken from parting lineation and ripple marks (trends) and sole marks (direction). Figure 2.40b: Palaeocurrent directions from the Musgrave grit unit. Mean current direction is 230o from 22 readings. Readings were taken from only two localities. Readings taken from cross bedding and fossilised logs. 146 Figure 2.40c: Palaeocurrent directions from the Sepenare?s Hoek and Eden sandstone units. Mean current direction is 342 from 78 readings. Readings were taken from parting lineation and cross bedding. Figure 2.40d: Palaeocurrent directions from the Mohono mudrocks unit. Mean current direction is 338 from only 17 readings. Readings were taken from Parting lineation, ripple marks, planar cross bedding (trend) and cross trough bedding (direction). 147 Figure 2.41: The stratigraphy in the Thaba Nchu area showing the predominant facies types, architectural elements and palaeocurrent directions for each unit. C u m ulativ e T thick n ess M ud ro ck S a nd sto n e G ritsto n e P red o m in a nt F a cies Typ es A rchitectu ral Elem ents P ala eo cu rrent D irectio n (N o rth = top of p ag e centre L o cal Stratig raph 220m 200m 180m Fl, Fsm, Fm, FF, CHt Sh, Sp 338o 160m 140m Mohono Mudrocks 120m St, Sh, Sl, Sei CHm 342o Eden Sandstone 100m Sh, Fsm, Fl, P DA, LS, FF, FP 80m Townland Fines 60m Sh CHt 342o Sepenare?s Hoek Sandstone Gt 40m 230o Musgrave Grit 20m Fm, Fsm, Fl, Sh, P CHf, CS, FF, FP 340o 0m Dubbledam Mudrocks 148 Figures: Chapter 3 149 Figure 3.1: Important fossil localities mentioned in the text. See Fig 3.2 for location of fossils. 150 Figure 3.2: Localities of all fossils collected during this study. 151 Figure 3.3: Location of fossils previously collected and held in various collections and for which there is accurate locality data. 152 Snout Tusk Figure 3.4: Lateral view of a skull of Dicynodon in grey-green mudrocks of the Dubbledam mudrock unit at Thaba Nchu Townlands. Lens cape is 55 mm in diameter. Figure 3.5: Anterior view of a snout of a Lystrosaurus murrayi (BP/1/6138) collected at Abramskraal 65 within 2m of the base of the Sepenare?s Hoek sandstone unit. 153 Figure 3.6: Kannemeyeria tusk (arrowed) and both caniniform processes collected on Thaba Nchu Mountain (Wilgeboomnek 29) within the Mohono mudrocks unit (BP/1/6774). Scale bar is 20 mm. Figure 3.7: Procolophonoid left lower jaw ramus in mud pebble conglomerate (BP/1/6573). Note anterior to posterior increase in tooth size indicating a more advanced form than Procolophon. Scale bar is 50 mm. 154 Figure 3.8: Left lower jaw ramus of an erythrosuchid found on Mohono Mountain (Eden 96) within the Mohono mudrock unit, Lateral view (Specimen BP/1/5877). Scale in centimetres. Figure 3.9: Lungfish tooth plate (arrowed) recovered from the Mohono mudrock unit on Mohono Mountain (Eden 96) (BP/1/6169). Scale in centimetres. 155 Figure 3.10a: Fossilised and silicified wood (tape is extended for 370 mm) Figure 3.10b: Radial longitudinal thin section of silicified fossil wood, Agathoxylon africanum, showing biseriate alternate bordered pits on the radial walls of the tracheids. (400x magnification). 156 BP/I/5862 Figure 3.11: The stratigraphic horizons from which fossils were collected. Numbers are a reference for the fossils listed in Table 3.1. C u m ulativ e T thick n ess M ud ro ck S a nd sto n e G ritsto n e D icyn od o n Lystro su a ru s P ro coloph o n P ro coloph o n oid A rch o sau r K a n n em eyeria F o ssil W o od L o cal Stratig raphy 220m 200m 180m 160m 140m Mohono Mudrocks 120m Eden Sandstone 100m 80m Townland Fines 60m Sepenare?s Hoek Sandstone 40m Musgrave Grit 20m Dubbledam Mudrocks Molteno Fm 1 2, 12, 13, 29 3 4 5 6 - 11 14 -28 33, 34 32 30 35 157 Figures: Chapter 4 158 16?0'0"E 17?0'0"E 18?0'0"E 19?0'0"E 20?0'0"E 21?0'0"E 22?0'0"E 23?0'0"E 24?0'0"E 25?0'0"E 26?0'0"E 27?0'0"E 28?0'0"E 29?0'0"E 30?0'0"E 31?0'0"E 32?0'0"E 33?0'0"E 34?0'0"E 35?0'0"S 35?0'0"S 34?0'0"S 34?0'0"S 33?0'0"S 33?0'0"S 32?0'0"S 32?0'0"S 31?0'0"S 31?0'0"S 30?0'0"S 30?0'0"S 29?0'0"S 29?0'0"S 28?0'0"S 28?0'0"S 27?0'0"S 27?0'0"S 26?0'0"S 26?0'0"S 25?0'0"S 25?0'0"S 24?0'0"S 24?0'0"S 23?0'0"S 23?0'0"S 22?0'0"S 22?0'0"S Figure 4.1: National Bouguer Gravity data set showing area from which the geophysical data subsets for this study were extracted. Red square indicates the study area. 159 Figure 4.2: Bouguer Gravity data subset. Bouguer gravity data has had all known effects removed and shows the relative variation in gravity (density). 160 Figure 4.3: Sun Shaded Bouguer gravity data (45o inclination, 45o declination). 161 Figure 4.4: Map showing position of Bouguer gravity anomalies. 162 Figure 4.5: Map showing an image of the aeromagnetic data and anomalies discussed in the text. 163 Figure 4.6: Aeromagnetic data over Thaba Nchu, subset. This is a subset of the regional data set in figure 4.5. MF indicates the large dolerite dyke that runs across a portion of northern Thaba Nchu and which divides two important outcrops (i.e. the outcrops on the farm Chubani 9). 164 Figure 4.7: Contour map showing depth to basement of the Karoo and the boreholes used to generate the contours. The closest boreholes to the study area, Sunnyvale 785 and Middelpunt 2029, are indicated Sunnyvale 785 and Middelpunt 2029 boreholes 165 Figure 4.8: Contour map showing depth to basement (in metres above/below see level) from Soekor seismic map (after Soekor Plan A2, 1969). 166 Figure 4.9: Comparison between the contour maps produced from seismic data and gridded borehole data. Although in broad agreement the seismic map shows much more detail (seismic depth contours from Soekor Plan A2, 1969). 167 Figure 4.10: Karoo Supergroup stratigraphy and depth to the Karoo basement overlain on the sun shaded Bouguer gravity map. There appears to be very little correspondence between the stratigraphy and the gravity data. 168 Figure 4.11: Bouguer gravity map (anomaly C) with the Karoo Supergroup stratigraphy and borehole data overlain. Note the variation in depth across the anomaly as well as that the anomaly cross cuts the Karoo stratigraphy. Profile line is discussed below. 169 Figure 4.12: Bouguer gravity anomalies G and I with the Karoo Supergroup stratigraphy and borehole data overlain. Note the variation in depth from north to south. Profile line is discussed below. 170 Figure 4.13: Bouguer gravity map of the area in the vicinity of the study area showing depth to basement contours with values at boreholes. Profile lines are discussed below. 171 Figure 4.14: Sun shaded Bouguer gravity map showing profile (section) lines, gravity anomalies, boreholes along the profile lines and Karoo Supergroup geology. 172 Figure 4.15: Model of the topography of the floor of the Karoo Supergroup along profile line X, Y. The green dotted line are the actual data values recorded, whilst the solid black line is the gravity values calculated from the proposed model. Standard regional value of 144 mgals was removed. Figures in the white boxes within the bodies are relative density differences with the surrounding basement rock (white area).The base of body 1 (blue) represents the base of the Karoo Supergroup. Bodies 2-6 are modeled as denser than the surrounding basement rock. Orange arrows indicate interpreted palaeovalleys. Distance (horizontal scale) is in metres. Note the borehole Sunnyvale 784 does not lie along the profile line but is projected 12 km south onto the line. 1 2 3 4 5 6 Thaba Nchu Anomaly G Borehole Basement 773 m Colesberg Lineament Borehole Basement 301 m Kaapvaal Craton Boundary X Y Palaeovalleys Borehole Sunnyvale 784 Basement depth 877 m 173 Borehole WE 1/66 Basement depth 3732 m Borehole Basement 2030 m Y1 Kaapvaal Craton Boundary Anomaly C Borehole Basement 617 m 6 2 Figure 4.16: Model of the topography of the floor of the Karoo Supergroup along profile line X1, Y1. The green dotted line are the actual data values recorded, whilst the solid black line is the gravity values calculated from the proposed model. Standard regional value of 132 mgals was removed. Figures in the white boxes within the bodies are relative density differences with the surrounding basement rock (white area).The base of body 1 (blue) represents the base of the Karoo Supergroup. Bodies 2-6 are modelled as denser than the surrounding basement rock. Distance (horizontal scale) is in metres. X1 1 3 4 5 4 Ladybrand Borehole Mara 1 Basement depth 820 m 174 Figures: Chapter 5 175 Figure 5.1. : Local and regional stratigraphy along with the stratigraphic location of fossil collected from the study area. C u m ulativ e T thick n ess M ud ro ck S a nd sto n e G ritsto n e D icyn od o n Lystro su a ru s P ro coloph o n P ro coloph o n oid A rch o sau r K a n n em eyeria F o ssil W o od L o cal Stratig raphy R egio n al Stratig raphy Bio stratig raphy 220m 200m 180m Mohono Mudrocks 160m B u rg ersd o rp F o rm atio n 140m Cyn og n ath u s A ssem blag e Z o n e 120m Eden Sandstone 100m 80m Townland Fines 60m K atb erg F o rm atio n Lystro sa u ru s A ssem blag e Z o n e Sepenare?s Hoek Sandstone 40m Musgrave Grit 20m Dubbledam Mudrocks 0m B alfo u r F o rm atio n D icyn od o n A ssem blag e Z o n e 176 Tables: Chapter 1 Table 1.1:Historical biostratigraphic subdivision of the Beaufort Group (after Rubidge, 1995; Groenewald, 1989; Neveling, 2002). Beds Zones Assemblage Zones Broom (1906) Watson (1914a, modified 1914b) Kitching (1970, 1977) Keyser and Smith (1977-78) Keyser (1979) SACS (1980) Rubidge (1995) Hancox et al. (1995) Neveling (2002) Stratigraphy Northern Free State (Groenewald, 1989) Subzone C New, unspecified biozone Subzone B Cynognathus Driekoppen Formation Cynognathus Cynognathus Cynognathus Kannemeyeria Kannemeyeria-Diademodon Cynognathus C y n o g n a t h u s Subzone A Kestrosaurus Procolophon Abundance Zone Verkykerskop Formation Procolophon Procolophon Lystrosaurus Lystrosaurus Lystrosaurus-Thrinaxodon Lystrosaurus Lystrosaurus Daptocephalus Dicynodon lacerticeps Dicynodon lacerticeps - Whaitsia Dicynodon Normandien Formation Kistecephalus Cistecephalus Aulacephalodon baini Aulacephalodon- Cistecephalus Cistecephalus Endothiodon Endothiodon Cistecephalus Tropidostima- microtrema Tropidostima- Endothiodon Tropidostima Pristerognathus / Diictodon Pristerognathus / Diictodon Pristerognathus Pareiasaurus Tapinocephalus Tapinocephalus Dinocephalian Dinocephalian Tapinocephalus Eodicynodon Volksrust Formation 178 Table 1.2: Lithostratigraphic subdivision of the Beaufort Group according to previous studies. Wavy lines indicate the limit of the respective studies. Cf Fig. 1.4 for the location of some of the various studies. SACS 1980 (South of 31oS) SACS 1980 (North of 31o S) Csaky (1977) Theron (1970) Johnson (1976) Botha and Linstrom (1978) Visser and Dukas (1979) Groenewald (1989) Neveling (2002) G r o u p S u b g r o u p Formations Formations Formations Formations Formations Formations Formations S u b g r o u p Formations S u b g r o u p Formations Molteno Molteno Molteno Molteno Molteno Molteno Molteno Molteno Molteno Burgersdorp Upper Beaufort Burgersdorp Otterburn Driekoppen Burgersdorp T a r k a s t a d Katberg Tarkastad Middle Beaufort Upper Beaufort Katberg Belmont Katberg T a r k a s t a d Verkykerskop T a r k a s t a d Katberg Middle Beaufort Normandien Northern Beaufort Balfour Balfour Balfour Balfour Normandien Middleton Middleton Middleton Middleton B e a u f o r t A d e l a i d e Koonap A d e l a i d e S u b g r o u p Koonap Lower Beaufort Lower Beaufort Koonap Estcourt A d e l a i d e Volksrust A d e l a i d e Ecca Group 179 Tables: Chapter 3 Table 3.1: Fossils collected during the course of this study from within the study area. Collection No. Description Current Identification Locality Name Co-ordinates Geology: Formation Number reference for Fig. 3.11 BP/I/5862 Skull Dicynodon Thaba N'chu townlands 29 13.05S 26 51.61E Dubbledam mudrock unit / Balfour Fm 1 BP/I/5873 Partial skull without lower jaw Lystrosaurus Thaba N'chu Townlands 29 12.87S 26 51.98E Dubbledam mudrock unit / Balfour Fm 2 BP/I/5875 Dicynodont skull Lystrosaurus Water tower at Morago - Houtnek 29 01.93S 26 49.24E Dubbledam mudrock unit / Balfour Fm 3 BP/I/5877 Archosaur lower jaw cf. Garjainia Eden, 96 - Moroto Mountain 29 10.54S 26 55.34E Mohono mudrock unit / Burgersdorp Fm 4 BP/I/5878 Gorgonopsian lower jaw Indeterminate Gorgonopsian Eden, 96 - Mohomo Mountain 29 10.54S 26 55.34E Dubbledam mudrock unit / Balfour Fm 5 BP/I/5930 Skull without lower jaw Lystrosaurus Thabanchu Townlands 29 12.87S; 26 51.98E Townland fines unit / Katberg Fm 6 BP/I/6127 Two pieces of skull in fine sandstone Lystrosaurus Thaba N'chu Townlands, east of Thaba N'chu Mt 29 12.927S; 26 52.142E Townland fines unit / Katberg Fm 7 BP/I/6128 Skull in concretion Lystrosaurus Thaba N'chu Townlands, east of Thaba N'chu Mt 29 12.927S; 26 52.142E Townland fines unit / Katberg Fm 8 BP/I/6129 Skull in concretion, in 3 pieces Lystrosaurus Thaba N'chu Townlands 29 12.927S; 26 52.142E Townland fines unit / Katberg Fm 9 BP/I/6130 Skull fragment Lystrosaurus Thaba N'chu Townlands 29 12.927S; 26 52.142E Townland fines unit / Katberg Fm 10 BP/I/6131 Fragments of bone in calcareous nodule Lystrosaurus Thaba N'chu Townlands 29 12.927S; 26 52E Townland fines unit / Katberg Fm 11 BP/I/6132 Two sections of skull Lystrosaurus Thaba N'chu Townlands 29 12.927S; 26 52.142E Dubbledam mudrock unit / Balfour Fm 12 BP/I/6133 Postcranial elements Lystrosaurus Thaba N'chu Townlands 29 13.125S; 26 51.791E Dubbledam mudrock unit / Balfour Fm 13 BP/I/6134 Skull Lystrosaurus Morago 40 29 01.910S; 26 49.472E Townland fines unit / Katberg Fm 14 BP/I/6135 Skull in red mudstone matrix Lystrosaurus Morago 40 29 01.910S; 26 49.472E Townland fines unit / Katberg Fm 15 BP/I/6136 Skull in weathered concretion Lystrosaurus Morago 40 29 01.910S; 26 49.472E Townland fines unit / Katberg Fm 16 BP/I/6137 Skull in concretion, showing calcite crystals Lystrosaurus Border between Paradys 358 / Abrahamskraal 65 29 06.944S; 26 51.326E Townland fines unit / Katberg Fm 17 BP/I/6138 Snout with tusks Lystrosaurus Border between Paradys 358 / Abrahamskraal 65 29 06.944S; 26 51.326E Townland fines unit / Katberg Fm 18 BP/I/6139 Partial skull in concretion Lystrosaurus Morago 40 29 01.910S; 26 49.472E Townland fines unit / Katberg Fm 19 BP/I/6140 Fragment of skull in weathered concretion Lystrosaurus Border between Paradys 358 / Abrahamskraal 65 29 06.944S; 26 51.320E Townland fines unit / Katberg Fm 20 BP/I/6141 Weathered skull in concretion - only tusk showing Lystrosaurus Morago 40 29 01.910S; 26 49.472E Townland fines unit / Katberg Fm 21 BP/I/6142 Partial skull Lystrosaurus Morago 40 29 01.910S; 26 49.472E Townland fines unit / Katberg Fm 22 BP/I/6143 Skull Lystrosaurus Morago 40 29 01.910S; 26 49.472E Townland fines unit / Katberg Fm 23 BP/I/6144 Skull in concretion Lystrosaurus Mafane 683 29 15.303S; 26 50.624E Townland fines unit / Katberg Fm 24 BP/I/6145 Skull with lower jaw in concretion Lystrosaurus Mafane 683 29 15.303S; 26 50.624E Townland fines unit / Katberg Fm 25 BP/I/6147 Skull in mud matrix Lystrosaurus Mafane 683 29 15.303S; 26 50.624E Townland fines unit / Katberg Fm 26 BP/I/6148 Partial skull in red mudstone matrix Lystrosaurus Mafane 683 Townland fines unit / Katberg Fm 27 BP/I/6149 Fragments of skull in fine green sandstone matrix Lystrosaurus Mafane 683 29 15.404S; 26 51.081E Townland fines unit / Katberg Fm 28 BP/I/6150 Skull in concretion Lystrosaurus Mafane 683 29 15.303S; 26 50.624E Dubbledam mudrock unit / Balfour Fm 29 BP/I/6168 Bone / mud-pebble conglomerate with temnospondyl amphibian remains Indeterminate temnospondyl amphibian Village of Moroto 68 29 05.486's 26 59.403'E Mohono mudrock unit / Burgersdorp Fm 30 BP/I/6169 Tooth plate Lung fish Mohono Mountain, Eden 96 Mohono mudrock unit / Burgersdorp Fm 31 BP/1/6573 Left lower ramus with teeth Procolophonoid Moroto Mountain 29 05.18S; 26 59.19E Eden sandstone / Katberg Fm 32 (Field No.) AR01/04/2 Lower jaw Procolophon Small hill on road between Rooikraal and Excesior 29 09.65 S; 26 53.434 E Eden sandstone / Katberg Fm 33 (Field No.) AR02/04/2 Lower jaw Procolophon Northern slope Mohono Mountain 29 09.549S; 26 57.102E Eden sandstone/Townland fines / Katberg Fm 34 BP/1/6774 Tusk, caniniform process Kannemeyeria Thaba Nchu Mtn, Wilgeboomnek 29 Mohono mudrock/Burgersdorp Fm 35 181 Table 3.2: Fossils listed in the data bases of the Council for Geoscience (GHG), University of the Witwatersrand (BPI) and the National Museum (NM). Note the absence of co-ordinates for many of the specimens. (* = data not recorded) Collection No. Description Current Identification Locality Name Co-ordinates/Locality description Geology: Formation BP/I/401 Skull indeterminate Dicynodont Thaba N'chu Commonage * * BP/I/491 Lower jaw indeterminategorgonopsian Thaba N'chu Commonage * * BP/I/492 Skull indeterminate Thaba N'chu * * BP/I/2241 Skull Lystrosaurus mccaigi Thaba N'chu Commonage * * BP/I/3238 Postcranial: femur indeterminate Therocephalian Thaba N'chu * * BP/I/3239 Skull Lystrosaurus sp. Thaba N'chu * * BP/I/3240 Skull Lystrosaurus sp. Thaba N'chu * * BP/I/4277 Large skull Dicynodon Thaba N'chu Commonage * * BP/I/4278 Skull (juvenile) Dicynodon ('juvenile') Thaba N'chu Commonage * * GHG144CGS Dicynodon lacerticeps Dicynodon THABA'NCHU 26 51 38E;29 12 59S ADELAIDE GHG153CGS Lystrosaurus declivis Lystrosaurus THABA'NCHU 26 52 11E;29 13 06S ADELAIDE GHG140CGS Dicynodon lacerticeps Dicynodon THABA'NCHU 26 51 34E;29 13 11S NORMANDIEN GHG141CGS Unidentified Unidentified THABA'NCHU 26 51 34E;29 13 11S NORMANDIEN GHG142CGS Dicynodon lacerticeps Dicynodon THABA'NCHU 26 51 35E;29 13 01S ADELAIDE GHG143CGS Unidentified Unidentified THABA'NCHU 26 51 35E;29 13 01S ADELAIDE GHG145CGS Unidentified Unidentified THABA'NCHU 26 51 38E;29 12 59S ADELAIDE GHG149CGS Lystrosaurus oviceps Lystrosaurus THABA'NCHU 26 52 09E;29 12 59S HARRISMITH 22G147CGS Dicynodon Dicynodontinae THABA 'NCHU TOWNLANDS A 26 51 44E;29 13 57S * GHG146CGS Unidentified Unidentified THABA 'NCHU TOWNLANDS A 26 51 44E;29 13 57S ADELAIDE GHG148CGS Moschorhinus kitchingi Moschorhinus THABA 'NCHU TOWNLANDS A 26 51 44E;29 13 57S ADELAIDE NM594 Lystrosaurus Chubani * * NM595 Lystrosaurus Chubani * * 182 Collection No. Description Current Identification Locality Name Co-ordinates/Locality description Geology: Formation NM596 Lystrosaurus Chubani * * NM598 Lystrosaurus Chubani * * NM683 Lystrosaurus New York * * NM861 Lystrosaurus Chubani 9 * * NM883 Indet. Dicynodont Sepani spruit on novo * * NM885 Lystrosaurus Chubani 9 * * NM886 Indet. Cynodont Chubani 9 * * NM887 Indet. Cynodont Chubani 9 * * NM888 Whaitsia platyceps Chubani 9 * * NM924 Indet. Dicynodont Lesaka 81 * * NM925 Indet. Dicynodont Lesaka 81 * * NM926 Indet. Dicynodont Lesaka 81 * * NM927 Indet. Dicynodont Lesaka 81 * * NM941 Indet. Dicynodont Bultfontein 113 * * NM942 Indet. Dicynodont Bultfontein 113 * * NM943 Indet. Dicynodont Bultfontein 113 * * NM946 Indet. Dicynodont Border of Palapala Novo 442 * * NM2917 Lystrosaurus Chubani 9 Very weathered outcrop of red mudstone in banks of stream very close to homestead. Extremely rich in bone fragments. NM2918 Lystrosaurus Chubani Very weathered outcrop of red mudstone in banks of stream very close to homestead. Extremely rich in bone fragments. NM2919 Lystrosaurus Chubani 9 Very weathered outcrop of red mudstone in banks of stream very close to homestead. E xtremely rich in bone fragments. NM2920 Lystrosaurus Chubani 9 Very weathered outcrop of red mudstone in banks of stream very close to homestead. Extremely rich in bone fragments. NM2921 Lystrosaurus Chubani 9 Very weathered outcrop of red mudstone in banks of stream very close to homestead. Extremely rich in bone fragments. NM2922 Lystrosaurus Chubani 9 Very weathered outcrop of red mudstone in banks of stream very close to homestead. Extremely rich in bone fragments. NM2923 Lystrosaurus Chubani 9 Very weathered outcrop of red mudstone in banks of stream very close to homestead. Extremely rich in bone fragments. NM2924 Lystrosaurus Chubani 9 Very weathered outcrop of red mudstone in banks of stream very close to homestead. Extremely rich in bone fragments. NM2925 Lystrosaurus Chubani 9 "Very weathered outcrop of red mudstone NM2926 Lystrosaurus Chubani 9 Very weathered outcrop of red mudstone in banks of stream very close to homestead. Extremely rich in bone fragments. NM3197 Cynognathus Wilgeboomnek 29 "In silty sandstone nodules NM3198 Cynognathus Wilgeboomnek 29 Out of context in nodule. 183 Collection No. Description Current Identification Locality Name Co-ordinates/Locality description Geology: Formation NM3199 Kannemeyeria Wilgeboomnek 29 In greenish mudstone deposited in overbank environment. NM3200 Kannemeyeria Wilgeboomnek 29 Out of context near QR3199. NM3341 Lystrosaurus Uitkomst 558 Gholfbaan.From the overlap zone of the Lystrosaurus and Dicynodon zones. NM1238 Emydops Gravel sloot * NM1715 Lystrosaurus Marokkoskop * NM1757 Lystrosaurus Marokkoskop * NM1873 Lystrosaurus Chubani Found in donga (blue mudstone) Table 3.2 continued. 184 Tables: Chapter 4 185 Table 4.1: List of boreholes with collar co-ordinates and depth to the basement of the Karoo Supergroup. Hole ID Latitude Longitude Collar (amsl) Depth to Base of the Base Karoo Depth 10933 30.8221069 26.8317380 1535.00 3732.00 2451 30.5250216 25.0594850 1270.64 1462.24 2450 30.5189229 25.0369810 1295.78 1403.00 11344 30.5003191 25.8591185 1319.16 1831.57 9968 30.2827506 25.3242667 1380.96 1444.00 9035 30.0122634 25.7572030 1400.30 1560.00 9020 29.9498983 25.6290010 1399.00 1142.00 9021 29.9400919 25.6137590 1388.10 1125.00 9023 29.9245410 25.5896280 1380.40 1108.00 9018 29.9178785 25.6752590 1355.10 1129.00 9026 29.9001657 25.6015330 1375.00 1068.00 9025 29.8939365 25.5925540 1384.10 1142.00 7105 29.7867294 25.7207730 1380.00 1100.00 1157 29.7627295 25.4202510 1440.00 1025.00 7528 29.7001389 25.6976700 1420.00 1081.00 2674 29.6156770 26.0931770 1429.55 1100.00 8191 29.5554594 26.0853720 1433.40 1090.00 5281 29.3669617 25.3882750 1360.00 571.00 9010 29.3324424 25.6667790 1370.00 733.00 7347 29.2545954 25.3700520 1310.00 465.00 2730 29.2308693 26.0218770 1330.00 1123.00 9967 29.2264615 25.1094330 1240.00 909.00 71535 29.1701287 27.3917370 1792.14 2030.78 5215 29.1114929 25.5022100 1290.00 585.00 5161 29.1021321 25.1067330 1190.00 295.00 71255 29.0832630 27.4770440 1597.00 1710.00 11095 29.0469076 25.4546280 1240.00 333.00 3939 28.9600341 26.3271680 1284.00 773.00 952 28.9201575 26.7638360 1462.00 968.70 6217 28.9132544 26.8605460 1430.00 877.20 10844 28.9102240 26.7493080 1473.00 900.00 812 28.9050042 26.7328640 1479.00 800.00 46433 28.8938568 27.8480350 1572.00 1873.00 5159 28.7832855 25.0716890 1235.00 301.00 149655 28.7083093 27.6888680 1695.00 1512.00 11013 28.7077924 26.2084190 1335.90 569.40 2421 28.6980431 25.9099940 1467.00 430.07 2211 28.6825667 26.4952360 1420.00 859.50 6463 28.6766469 29.3871950 1155.00 1056.00 2002 28.6656904 26.7402560 1390.00 618.00 6045 28.6521436 25.9849700 1306.08 525.50 8156 28.6306650 26.3818210 1367.00 620.00 832 28.6157105 25.9757754 1309.23 465.10 2225 28.6096803 26.6799780 1360.00 685.00 2220 28.5769482 26.4264940 1402.00 805.00 820 28.5638679 25.9427820 1321.45 421.69 834 28.5576602 26.0031940 1340.46 445.00 2652 28.5560835 25.4498560 1248.00 198.00 186 Hole ID Latitude Longitude Collar (amsl) Depth to Base of the Base Karoo Depth 6547 28.5302622 28.0788720 1677.41 1510.00 884 28.5155121 26.7741010 1360.00 665.70 821 28.5017626 25.8990290 1289.48 190.00 5024 28.4851248 25.9192010 1296.17 206.70 823 28.4654438 25.9406180 1311.40 260.00 6897 28.4639134 28.0904110 1657.81 1402.25 5906 28.4499984 26.7833360 1432.56 640.70 825 28.4400423 25.8968030 1329.71 201.00 6561 28.4310477 28.4006020 1779.69 1634.00 2635 28.4124773 26.5137180 1300.00 337.51 2593 28.4072459 26.7522860 1356.80 509.40 4428 28.4050434 25.4730640 1261.68 200.00 4306 28.3975182 26.9142970 1432.56 785.80 6519 28.3912816 28.0019510 1716.49 1275.00 7306 28.3843573 26.2771520 1364.00 453.50 6925 28.3789576 28.0830480 1725.05 1303.79 11123 28.3768004 26.3927360 1402.08 675.40 10620 28.3637145 25.4962600 1300.00 192.63 20281 28.3618194 28.7831370 1700.00 1374.00 2650 28.3606376 26.6899290 1420.00 568.41 828 28.3588462 25.9234940 1328.95 211.00 2628 28.3484518 26.7172160 1447.92 549.48 6533 28.3441688 28.2153330 1763.90 1373.00 2639 28.3424626 26.8867950 1432.00 684.10 2638 28.3406055 26.9500270 1420.00 664.40 9494 28.3232917 26.7118840 1460.40 574.49 32531 28.3185693 27.0097650 1410.00 581.10 2664 28.3170910 26.7476110 1416.12 333.00 3332 28.3168390 26.3946280 1385.09 485.00 7722 28.3150318 26.9133460 1398.11 643.25 3958 28.3130442 25.5329840 1290.00 199.34 3337 28.3115223 26.4128030 1359.46 456.69 2648 28.3104030 26.6561030 1393.00 402.40 9394 28.3097804 26.7725070 1436.98 648.23 9431 28.3023062 26.7554070 1409.65 460.00 2631 28.2995654 26.5316640 1300.00 285.90 132211 28.2982778 28.2112500 1740.00 1435.48 10843 28.2833297 26.8000030 1432.56 628.50 9480 28.2832835 26.7653180 1405.10 503.00 9418 28.2812002 26.7012210 1384.73 510.70 154821 28.2732051 28.1078793 1658.20 670.00 23039 28.2265532 27.0112060 1356.00 502.00 98793 28.1962962 31.3297830 676.80 149.00 79165 28.1874752 29.2791410 1850.00 1037.00 73439 28.1673566 27.0024440 1311.00 444.00 114935 28.1607931 27.3118930 1431.58 697.00 98807 28.1601966 31.5129710 338.48 218.00 149151 28.1441344 27.1819640 1407.00 636.00 147121 28.1295279 27.6132440 1528.00 766.00 139197 28.1140028 27.1754510 1442.00 562.00 146197 28.1074293 27.0467090 1341.00 299.00 187 Hole ID Latitude Longitude Collar (amsl) Depth to Base of the Base Karoo Depth 49891 28.1052712 27.0075300 1393.00 347.00 28681 28.0852507 28.7796870 1740.00 1110.00 22577 28.0716104 27.0184130 1385.00 297.00 24327 28.0700417 27.1926120 1455.00 563.00 144405 28.0594132 27.0660150 1431.00 421.00 58067 28.0567997 27.0582830 1423.00 126.00 37039 28.0562407 27.5528180 1524.00 768.70 145987 28.0540081 27.0423130 1408.00 243.00 67223 28.0408279 27.0561390 1402.00 376.00 151265 28.0393023 27.0285780 1384.00 267.00 146323 28.0386255 27.1149190 1448.00 607.00 67027 28.0364081 27.0509500 1396.00 254.00 24999 28.0349734 27.0099150 1370.00 342.00 151251 28.0293736 27.0376250 1384.00 235.90 25867 28.0282043 27.0189140 1372.00 270.00 57311 28.0280075 27.1215380 1430.41 490.00 1378128.0279183 27.0675210 139.0 4.0 2521 28.024560 27.457820 1482.0 659.28 108159 28.021704 27.0340640 1378.0 21.0 10607 28.0138560 27.017590 1369.0 287.0 14709 28.0124052 27.047050 1382.0 321.0 8259 28.014074 27.2307980 1481.0 549.0 147107 28.08972 27.03601 1376.0 340.0 24985 27.96703 27.1826820 1439.0 502.60 146701 27.986107 27.1560 1408.0 28.0 6959 27.978426 27.156520 1436.0 328.90 93 27.9780623 27.036830 1380.0 30.0 5687 27.9678056 27.052760 1385.09 319.0 541 27.9584309 27.06901 1385.9 540.0 621 27.9576581 27.439080 1493.0 514.0 1523 27.952909 27.079260 1387.0 20.90 871 27.948913 27.581230 1530.0 762.0 5735 27.932578 27.03770 142.3 387.0 32685 27.93153 27.141780 140.0 31.0 153981 27.9386780 27.045130 142.0 38.13 71269 27.91597 27.03980 1536.0 748.0 1045 27.91486 27.5421580 152.90 59.0 14327 27.904918 27.193090 147.0 48.10 126317 27.9074032 29.14670 1306.0 390.0 1047 27.9069183 28.1940970 165.4 947.0 74391 27.8964304 27.1902 1507.18 693.67 153715 27.895928 27.520870 1524.0 538.90 24761 27.895756 27.376803 148.0 561.0 359 27.8908574 27.482372 1481.86 59.0 9381 27.84291 31.5607920 578.23 378.24 15853 27.874585 27.1704 1439.0 20.0 13023 27.861420 29.8964120 1284.0 310.0 5085 27.8396482 27.1630 147.0 246.0 94705 27.839259 29.51630 2086.0 961.0 89469 27.831243 27.54160 1493.0 5.0 137867 27.835891 28.347350 1670.0 867.0 188 Hole ID Latitude Longitude Collar (amsl) Depth to Base of the Base Karoo Depth 97197 27.8320663 27.8566880 1524.00 693.00 150915 27.8308782 27.2557910 1402.00 452.00 93347 27.8247493 27.5221050 1452.00 463.91 1731 27.8241208 27.1936680 1433.00 238.00 121459 27.8170800 27.4554810 1463.00 574.35 92955 27.8159003 27.2993960 1365.00 313.80 27911 27.8030401 27.0501300 1358.00 240.00 93515 27.8010320 27.3653140 1426.00 186.00 82315 27.7872727 27.2129720 1410.00 209.00 93823 27.7867330 27.4382080 1412.00 477.45 92969 27.7854302 27.2840930 1360.00 139.60 73621 27.7731050 27.2647860 1341.00 272.00 92815 27.7721019 27.3844930 1416.00 322.90 93067 27.7715746 27.5510070 1410.00 402.86 93319 27.7702460 27.6095340 1426.00 454.61 151153 27.7649011 27.5171530 1390.00 419.00 93207 27.7621729 27.6948580 1450.00 477.00 65711 27.7587857 27.1856840 1426.40 157.00 2893 27.7542244 27.2881200 1429.00 132.60 93767 27.7494994 27.6394810 1484.00 528.00 151349 27.7489664 27.6663760 1506.00 525.00 93739 27.7489202 27.8255090 1557.00 621.15 44109 27.7429725 27.3428080 1359.00 220.00 80747 27.7416303 27.3732300 1359.00 332.00 93263 27.7409384 27.2846880 1427.00 215.56 78115 27.7407354 27.4027400 1380.00 349.00 93837 27.7399551 27.3884270 1376.00 230.30 93417 27.7392394 27.5927700 1436.00 466.43 89343 27.7385776 27.5313160 1438.00 376.00 99437 27.7362833 31.7778020 378.82 566.00 66985 27.7352329 27.2482490 1356.78 384.00 94117 27.7351305 27.6242060 1478.00 518.00 91513 27.7332365 28.2357380 1650.00 854.00 67475 27.7286989 27.8130760 1553.00 609.00 93193 27.7275771 27.8073790 1547.00 626.30 101691 27.7264393 31.1011000 1055.13 359.00 94201 27.7263005 27.5795370 1484.00 508.90 93935 27.7254055 27.3301790 1360.00 233.90 125113 27.7229180 27.7298860 1509.00 515.00 56219 27.7197899 27.1491680 1382.05 433.00 51039 27.7166684 28.7833320 1592.00 796.48 80047 27.7128038 27.7464570 1505.06 486.00 93781 27.7125421 27.6101860 1466.00 451.60 148493 27.7108463 27.5405220 1434.00 395.63 41925 27.7078120 27.6084410 1473.93 490.00 92899 27.7065142 27.2595550 1397.00 427.60 117301 27.7064966 27.4713550 1455.00 457.00 82777 27.7062456 27.3029490 1359.00 285.00 40987 27.7061771 27.6368620 1477.28 526.57 72207 27.7060955 27.7128530 1548.00 459.00 115047 27.7002525 29.4024690 837.56 189 Hole ID Latitude Longitude Collar (amsl) Depth to Base of the Base Karoo Depth 93025 27.6995702 27.5838970 1465.00 413.20 93053 27.6990671 27.6930190 1561.00 570.36 92885 27.6981905 27.4161680 1407.00 427.60 138315 27.6969112 27.4214350 1414.00 461.80 93697 27.6956084 27.8266680 1547.00 620.61 93221 27.6954429 27.0969180 1392.00 313.31 95965 27.6887319 27.5417310 1475.00 317.00 7471 27.6871977 28.5516990 1675.00 800.00 119933 27.6811015 27.7998830 1582.59 352.00 99045 27.6796335 31.7476560 342.16 733.00 93865 27.6790544 27.4465040 1428.00 437.20 82301 27.6777792 27.9888140 1620.00 611.60 42513 27.6720227 27.5677770 1422.31 205.00 69729 27.6716419 27.6347400 1496.00 520.30 92983 27.6709640 27.4069430 1448.00 404.30 122257 27.6700034 27.9324460 1530.00 660.00 113885 27.6674928 27.7841640 1562.35 398.00 28247 27.6664845 27.4812170 1448.00 364.63 65473 27.6634577 27.5432280 1496.00 124.00 93403 27.6623648 27.8018930 1600.00 443.55 975 27.6606624 27.5995840 1465.83 424.00 102349 27.6600194 27.6728280 1533.00 509.00 113857 27.6580146 27.8035540 1614.45 442.00 149683 27.6532279 27.7837410 1555.83 411.00 94285 27.6506198 27.4200420 1459.00 462.70 44207 27.6435987 27.4288810 1466.00 612.00 44193 27.6433550 27.4347070 1466.00 529.00 93011 27.6430244 27.6751460 1518.00 466.70 91499 27.6425512 28.1771250 1680.00 811.00 94761 27.6407654 27.6473790 1493.00 308.00 149725 27.6395206 27.7963530 1605.79 639.00 66831 27.6391057 27.2932610 1378.40 325.00 40959 27.6374551 27.5903760 1477.31 339.00 94747 27.6372592 27.6541470 1499.00 394.00 92913 27.6361751 27.3206900 1396.00 332.20 107249 27.6360789 27.6060730 1472.00 400.00 94775 27.6329169 27.6468060 1490.00 281.00 92871 27.6307271 27.4326300 1466.00 441.80 41589 27.6270213 27.5907490 1492.44 329.60 93571 27.6176679 28.5940040 1682.00 835.07 25923 27.6173478 27.1351430 1379.00 334.00 92997 27.6167858 27.4173800 1466.00 381.90 50115 27.6166284 27.5639660 1499.00 300.00 92843 27.6107734 27.2993430 1390.00 325.00 46069 27.6091670 27.5098360 1475.00 446.22 72515 27.6050681 27.6290380 1518.00 358.80 93431 27.6027468 28.6827500 1615.00 778.05 92773 27.6007884 27.7789190 1615.00 617.00 41561 27.5975347 27.6083450 1498.51 239.00 42555 27.5908311 27.6342110 1519.20 278.00 44249 27.5846559 27.4471240 1451.00 426.00 190 Hole ID Latitude Longitude Collar (amsl) Depth to Base of the Base Karoo Depth 30445 27.5841660 27.6973790 1536.00 444.00 83099 27.5836812 27.6675010 1542.00 500.00 93641 27.5830745 27.8025510 1564.00 569.00 92941 27.5818736 27.3088860 1448.00 382.80 41841 27.5800944 27.4567160 1430.00 420.00 44095 27.5797176 27.3877140 1424.74 356.00 41659 27.5773953 27.4344370 1424.68 395.00 46013 27.5736821 27.3311410 1438.00 470.30 101971 27.5734929 27.8739610 1505.00 526.10 148843 27.5716552 27.0288600 1384.00 330.10 33917 27.5702374 27.4628040 1424.99 350.00 89329 27.5700345 29.9285340 1242.36 210.00 134633 27.5698532 27.0009110 1408.00 385.30 93137 27.5696869 27.8429570 1539.00 571.74 44473 27.5645667 27.6004290 1501.47 176.00 41533 27.5620167 27.4733120 1435.18 442.00 94257 27.5599748 27.6424500 1550.00 135.00 44515 27.5596047 27.6385610 1547.09 213.00 97771 27.5583924 27.4879330 1446.00 397.00 66887 27.5509042 27.4180420 1435.00 308.00 92927 27.5483621 27.3589070 1410.00 196.90 93305 27.5462498 27.4628730 1438.00 258.24 93529 27.5351843 28.9204420 1660.00 707.39 42569 27.5331990 27.6540990 1490.23 429.30 24047 27.5287360 27.6257690 1478.00 452.00 49345 27.5225879 27.1351280 1369.00 312.00 92857 27.5225614 27.6341350 1498.00 462.60 93095 27.5158864 27.3465340 1397.00 166.04 138203 27.5110911 30.3444250 2113.00 1019.00 50675 27.4999110 27.0010120 1341.00 259.40 24005 27.4942215 27.0323820 1326.00 207.00 133961 27.4592700 28.5859060 1590.00 686.00 93949 27.4551446 27.2668750 1399.00 112.75 144419 27.4502434 27.7823090 1487.00 375.00 64955 27.4473751 27.4789470 1432.00 135.00 58319 27.4472116 30.1816440 1748.00 643.00 122243 27.4430408 30.3188740 1855.00 653.50 45803 27.4379050 27.6452810 1472.00 440.13 69407 27.4376123 27.0865570 1348.18 125.50 23641 27.4039725 27.0313450 1341.00 170.00 41939 27.3984802 27.4991600 1426.00 224.60 134115 27.3885277 28.6057210 1548.00 696.00 107627 27.3735988 27.2554390 1360.00 342.00 45789 27.3718389 27.6825100 1463.31 373.00 35583 27.3580359 29.8641380 1700.00 573.00 32979 27.3569220 29.7691400 1700.00 598.40 50395 27.3565132 27.5658270 1411.92 174.70 75497 27.3489200 27.9125630 1600.00 479.00 121473 27.3481200 27.6527440 1440.00 295.00 54455 27.3466309 27.8782950 1516.53 439.01 113619 27.3452583 30.0699850 1874.00 729.60 191 Hole ID Latitude Longitude Collar (amsl) Depth to Base of the Base Karoo Depth 33469 27.3425394 27.1363440 1384.00 152.00 79403 27.3257304 28.8685970 1477.00 600.46 148339 27.3204665 27.8886900 1530.00 561.70 70429 27.3169217 28.3189390 1540.00 595.58 134157 27.3142970 28.5323340 1563.00 617.00 64353 27.3077710 27.2962530 1377.00 244.00 93627 27.3054040 28.7156030 1594.00 563.92 113689 27.2980403 27.7749770 1429.00 337.00 24775 27.2886237 27.8502080 1463.00 287.00 134185 27.2811793 28.0087120 1575.00 512.06 6393 27.2803433 27.8554470 1485.00 512.00 146393 27.2796082 27.7625280 1478.00 289.00 112471 27.2753247 27.1219610 1371.60 172.00 94957 27.2693192 27.3383140 1371.00 142.00 94985 27.2672861 27.3393170 1371.00 178.00 95013 27.2657043 27.3403220 1371.00 151.00 134451 27.2571425 27.9530090 1520.00 556.56 39027 27.2492594 29.1878070 1670.00 518.46 69631 27.2353789 27.2796550 1341.00 173.00 69687 27.2349241 27.2814210 1341.00 157.99 69715 27.2340129 27.2857090 1341.00 173.00 34533 27.2245772 27.2649910 1341.00 69.49 34561 27.2236723 27.2662510 1341.00 63.40 34575 27.2218639 27.2680130 1341.00 207.57 54427 27.2080589 28.2240860 1555.94 513.50 50955 27.2078987 28.7822150 1615.00 469.70 149067 27.1959460 27.9139390 1463.00 287.10 111715 27.1917999 28.1741720 1560.00 439.00 54385 27.1878654 28.1896890 1537.33 502.00 69547 27.1874568 27.8428800 1464.00 338.00 134325 27.1815686 28.2989960 1550.00 475.00 54399 27.1717857 28.1609260 1550.73 379.50 143621 27.1655524 27.2118810 1417.00 57.91 143649 27.1637458 27.2128870 1417.00 79.86 143663 27.1637365 27.2189400 1417.00 90.53 91443 27.1626060 27.2204520 1417.00 101.50 111701 27.1578934 28.1634050 1540.00 401.00 54441 27.1184325 28.1126610 1569.04 613.64 54497 27.1036069 28.1539470 1517.14 560.10 135 27.0837396 28.2286100 1520.00 450.00 49863 27.0787917 27.1997410 1410.00 66.45 49807 27.0778497 27.1957470 1411.00 90.83 49905 27.0766677 27.2018840 1408.00 82.91 49947 27.0757601 27.2056638 1415.00 72.54 111659 27.0699856 28.0828110 1520.00 334.00 95321 27.0191194 28.2072270 1460.00 269.00 112415 27.0184075 27.8992380 1480.00 258.00 95279 27.0162085 28.2274500 1490.00 326.00 45845 27.0021114 29.2749050 1561.00 308.80 192 Table 4.2: Rock densities from some South African geological units (Mare and Oosthuizen, 2000). Note the large differences between the least to most dense. Rock Type Density [kg/m3] Average [kg/m3] Adelaide Arenite 2449 Balfour Arenite 2460 Tarkastad Arenite 2408 Ecca Arenite 2557 Ecca Arenite 2617 2498 Ecca Lutite 2 2475 Tarkastad Lutite 2650 2562 Cape Granite 2639 Halfway House Granite 2700 Mesklip Granite 2707 Nebo Granite 3189 Alkali-feldspar Granite (Int) 3069 2860 Metamorphic Rocks 2830 Nelshoogte Gneis 2720 Concordia Granite (metamorphic) 2660 Metamorphic Rock (Inlandsee) 2910 Onverwacht Metamorphic Rocks 2880 2800 Gabbro 2831 Novengilla Gabbro 2908 2869 Steynskraal Amphibolite 2925 2925 Timeball Hill 2780 Timeball Hill Lutite 2815 Campbell Rand Dolomite 2870 Chuniespoort Dolomite 2864 Chuniespoort Iron Formation 3210 2907 193 Table 4.3: Thickness he various groups and formations those boreholes that lie along the profile lines as well as the percentage of the various lithologies and their density contribution. Boreholes WE1(Alliwal North) Mara 1 Novo 1 Sunnyvale 784 LA 1/68 Middelpunt 2029 Total depth 3732 820 854 800 1710 877 Thickness of Stormberg Fm [m] 47 0 0 0 193 0 Thickness of Beaufort Group [m] 2425 453 460 571 1066 571 Thickness of Ecca Group [m] 1060 346 439 300 447 300 Thickness of Dwyka Group [m] 206 21.301 5 15 0 14 Dolerite Thickness [m] 954 51 164 209 270 170 Basement Lithology biotite gneiss Not known Not known metasediments quartzites Ventersdorp lava (50 m) then granite Dolerite % 27 7 19 22 16 18 % Sandstone 22 59 54 43 Not Known 44 % Shale & siltstone 51 34 28 36 Not Known 39 Mean density for borehole column 2630.81 2545.73 2611.39 2627.64 - 2614.72 Mean density for Karoo Supergroup 2606 kg/m3 194 Appendix 195 Key to Vertical profiles Facies Sh Cutins Facies Sm Root traces Facies Sl Mottling Facies Sr Pedogenic features (for larger scale sections) P P P Facies St Dolerite Facies Sp Mud Cracks Facies Sei Scree ? not to scale, missing thickness is indicated. Facies Sei ? scour fill Contacts: Abrupt Gradational A G Facies Fl Sharp Erosional S E Facies Fsm Facies Fm Facies Fm for pedogenic section (p 127) Rhizocretions Burrows Calcareous nodules: isolated, smooth, round Calcareous nodules: flat or dicoidal, smooth, flat topped Scree ? to scale 196 Figure A1a: Map showing the traverse of the Houtnek Section Start End 197 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 22m 2 Sh n/s 4.5 Red (with occasional Lystrosaurus sp. green mottles) Fragments of fossil bone n/s 10 Sh 1+ Red Extensive series of dongas Lystrosaurus M iddle S a nd sto n e U nit Figure A1b: Houtnek stratigraphic section 198 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 44m Scree but with occasional patches of red mud rock 22 M iddle S a nd sto n e U nit Figure A1b continued 199 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 88m some dolerite exposed 1 G St 10.5 Sh Fossil bone Fragment s (<5cm) White/yel low Sh/S t highly weathere d M iddle S a nd sto n e U nit Figure A1b continued 20m Scree 200 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 66m Some outcrop of dolerite 10 Red (with occasional Bone fragments Mottles of green) n/s Sh yellow M iddle S a nd sto n e U nit Figure A1b continued 201 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 96.2m Sh 4.2 St 230o Sh M iddle S a nd sto n e U nit Figure A1b continued Scree and dolerite to top of hill 202 Figure A2a: Map showing the traverse of the Mohono Section 203 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 22m Fm Red 20m G 1.5 0.3 Fsm 0.15 0.3 0.3 A A Sh Fm Sh 0.1 0.6 G E Sp Sei Fm 3 Patches of red Fm amongst scree 10m Sh 252o Green Plant impressions , 1 G Fsm Green 0.6 0.25 0.2 A E Fl Sei Fsm 1.5 G Sh Green (orange impression s) Plant impressions 0.25 0.4 G E G Sei Sm (Sei) Plant impressions Bioturbated 0.15 E Sei Sh n/s +2m Fl/ Fsm Red with green mottles M iddle S a nd sto n e U nit Figure A2b: Stratigraphic section on Mohono Mountain (Eden 96) 204 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 44m Fm Red 40m 1.2 n/s Fsm Green Red n/s 1.8 Fsm Green Green 0.4 G Sh Sl 0.03 1.8 E Ripple topped surface 30m 0.1 E Sh Sei Fm Red 11 M iddle S a nd sto n e U nit Figure A2b continued 205 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 66m Sh 0.6 Fl 0.5 G G Sr 330o 274o 291o Green 60m 7 Sh 0.25 0.25 G A Sm 0.5 0.3 G Fsm Sm 0.5 E Sei 3.75 Sh G 50m 0.2 0.6 0.3 E E Sm Sei Trough- cross 0.1 0.75 E Sei Sl M iddle S a nd sto n e U nit Figure A2b continued 206 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 88m Red 0.5 n/s Fsm green 4 Fm Red G Fsm Green 80m C St E 2.8 Sh n/s 5.3 Fl Large percentage of green silt 70m n/s 262o 1.6 Large 1.2 0.25 E Fsm concretions (1-2m diam.) oblate 0.25 Fm Sh M oh o n o M ud ro ck s U nit M iddle S a nd sto n e U nit Figure A2b continued 207 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 110m Fossil of Amphibian breast plate 11.2 Fm Red Fossilized bone fragments Scree covered in places Rhozocreti ons 100m Sh Small outcrops of Fm (red) off the section 0.5 Fsm 90m Fossil bone fragments A Zone M oh o n o M ud ro ck s U nit Figure A2b continued 208 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 132 16m 130m Red mudstone with green fine sst and Fl 120m 248o (single reading) 4.5 Sh 2-3cm diam. rounded concretions M oh o n o M ud ro ck s U nit Figure A2b continued 209 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 154m 12 150m Sh Light orange 1 Fsm 4 Fm Red 130m 0.75 Bottom of sets lined with Sei 0.2 1.6 Sh Green Silicified wood fragments M oh o n o M ud ro ck s U nit Figure A2b continued 210 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n Sugary white 163 M olten o F o rm atio n 160m 8m Sh M oh o n o M ud ro ck s U nit Figure A2b continued 211 Figure A3a: Map showing the traverse of the Rakhoi Section 212 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) N ote/C olo u r F o ssils F o rm atio n 22m Sh Yellow 0.5 S E Sei Yellow Sh 6 S 9m S Fl 7 Sh Yellow 1.5m E Sm Sei Sh 0.5 + E E Sei Sh Fsm Yellow/black Bioturbated M iddle S a nd sto n e U nit Figure A3b: Stratigraphic section on Rakhoi Mountain 213 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m C o arse C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) N ote/C olo u r F o ssils F o rm atio n 44m Brown/GreyStr eaked G Green G Red brown 3.5 Fl/P Green (brown concretions) 30m 3 Sh Yellow 0.5 St 0.5 S Sh St Sl A G E St Sh Sei 20m 13.5 Sh M iddle S a nd sto n e U nit Figure A3b continued 214 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m C o arse C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) N otes/C olo u r F o ssils F o rm atio n 66m Sm Yellow 7 54m 13 Fm Red brown 50m 0.9 G Sei Mudpebble s, Rhizocretio Rhizocre tions 0.8 Fl Red with grey streaks M oh o n o M ud ro ck U nit M iddle S a nd sto n e U nit Figure A3b continued 215 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m C o arse C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) N ote/C olo u r F o ssils F o rm atio n 88m yellow Sh 1.1 Sei Wood imprints 78m A Sh Sh 70m S St 0.1 S E Sp Sei Sh S S Sl Sm M oh o n o M ud ro ck U nit Figure A3b continued 216 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) N ote/C olo u r F o ssils F o rm atio n 127m 9 m 118m 4 Green fine sandstone, green and red brown silt and Fl Red brow mud: highly 114m weathered and poorly exposed 18 m 96m 2 St A 16 Sh 89m M oh o n o M ud ro ck U nit Figure A3b continued 217 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) N ote/C olo u r F o ssils F o rm atio n + 15m thick White, thickley bedded M olten o F m 176m 38 m 138m St Sh 134.4 m St S Sh S yellow 1.2 S St E S Sh St S Sh Sp 127m Sh M oh o n o M ud ro ck U nit Figure A3b continued 218 Figure A4a: Map showing the traverse of the Thaba Nchu Townlands Section 219 Figure A4b: Stratigraphic section Thaba Nchu Townlands. C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 22m 1.4 20m 0.4 E Sm Cocretions, flat,smooth 0.5 1 Sh Green, scattered mud pebbles Post cranial/cranial Silty green Fragments (<25cm) In purple mud lens 1.3 Sm Grey/green Silty sand Blocky weathering 10m 0.5 St 336o Thins laterally to north A G 1 m Sh Sm 306o 328o sc ou r Concretions ? brown, d=30- 100cm Organic matter and root traces 2.2 Sand is green Blocky weathering 0.70 3m 3 Biotubated/root traces Sh 0 Thick succession of red Fm below Dicynodon sp in red Fm M iddle S a nd sto n e U nit D ubbled a m M ud ro ck U nit 220 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 44m 1.5 Fl Brown concretions Caliche layers red 1.8 Fsm Interlinked 1.5 concretions (carbonate) layer of brown 40m 0.6 Sl Carbonate with green veneer 0.2 0.2 Sm Fm 0.5 0.15 0.2 Sm 30m 1 Ripple topped Brown concretions- 4.3 with green mud veneer Fl Grey and green silt and sand, purple mud M iddle S a nd sto n e U nit Figure A4b continued 221 Figure A4b continued C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 61m Sh Yellow white 1 E Green 0.75 0.15 Sh Sei Green 0.3 E E Sp 1.8 E Sh Green 0.1 1 A Sr M iddle S a nd sto n e U nit Scree but exposures of green Fm off section 50m and small channels Brown concretions 0.75 0.5 Sr Soft sediment deformation, Sr in sand units Scree, dolerite and +1m thick layers of fine sst out crop until to the top of the hill (25m) 222 Figure A5a: Map showing the traverse of the Thaba Nchu Mountain Section 223 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 22m 1.2 Sh 3.8 Sh 1.6 Fm red 10m 2 Fm red A Fossilise d tetra pod foot prints +3 Sh yellow/gre en M iddle S a nd sto n e U nit Figure A5b: Stratigraphic section on Thaba Nchu Mountain 224 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 44m 40m 30m 1.2 Sh Small patches of red M oh o n o M ud ro ck s U nit Figure A5b continued 225 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 66m 60m 22 50m M oh o n o M ud ro ck s U nit Figure A5b continued 226 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 88m 2.1 Sh Erosive contact Divides two units of Sh Sh 80m 70m 2 Sh M oh o n o M ud ro ck s U nit Figure A5b continued 227 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 110m Fsm 2.3 Ripple top surface 2.5 St 100m S Sh 6 E sh 1.1 E Sm 90 M oh o n o M ud ro ck s U nit Figure A5b continued 228 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 132m 130m 1 G Fsm 1.1 St 0.45 S S Sh 2 Fsm 120m G Sh 1.2 0.25 Sei 0.5 0.25 E Sl Fsm G Sh 0.4 Sei 0.9 Sh 0.8 Sei M oh o n o M ud ro ck s U nit Figure A5b continued 229 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 154m 0.7 0.4 A St Sh 0.4 0.1 0.5 A E Sp Sei St E 3.2 Sh 150 E 1 E St 1.4 Sh 140 M oh o n o M ud ro ck s U nit Figure A5b continued 230 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n 176m 170m 0.3 St Sh 0.4 0.2 Fsm 1.4 Sl 160m St M oh o n o M ud ro ck s U nit Figure A5b continued 231 C u m ulativ e Thick n ess H eight in m Typ e of C o ntact F acies C od e M ud Silt Fin e S and sto n e M ediu m sand sto n e C o arse S and sto n e C o nglo m erate P aleo cu rrent D irectio n (N o rth is V ertical) C olo u r/N otes F o ssils F o rm atio n Off the section line Fm (red 182m 180m 1.1 Sh St 0.25 0.1 0.5 E Sh Sei St 1 n/s Sh 0.6 n/s Sm 1 n/s Sl M oh o n o M ud ro ck s U nit Figure A5b continued 36m of scree and poorly exposed red mudrock, capped by medium to coarse grained white sandstone of the Molteno Formation