Geology of selected sites along the northern margin of the Barberton Greenstone Belt Masters’ Dissertation Submitted by Nonkululeko Phumelele Mashele 28 September 2022 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. i Plagiarism Declaration I, Nonkululeko Phumelele Mashele (458861), declare that this dissertation is my own, unaided work. It is being submitted for the Degree of Master of Science at the University of the Witwatersrand, Johannesburg. It has not been submitted before for any degree or examination at any other University. __________ _________ Nonkululeko Phumelele Mashele 28th day of September 2022 in Windischeschenbach ii Abstract An increasing number of UNESCO World Heritage Sites are recognised for their geological significance. One such site forms part of the Barberton Greenstone Belt (BGB), which has been widely acknowledged as an outstanding site for Archean geology research. Though members of the geoscience community are familiar with the outstanding universal value of the Barberton Greenstone Belt, the people who live on it often do not. With the World Heritage Status (2018) comes an obligation to unpack the geological heritage of the Barberton region to local people and visitors, and to promote further research so the true value of the geological importance of the region may be better understood. For this reason, this project set out to interpret and contribute to concepts formulated from studies of BGB rocks to produce a thesis usable for generating various science communication media and topics for further research. Themes discussed in this thesis include granite-greenstone relationships, alteration of ultramafic rocks and the associated economic products, as well as deformation of various lithologies. Ease of access guided the selection of eight mostly roadside outcrops along the R38 secondary road linking Barberton to the southern gate of the Kruger National Park. These sites link up to form a geological trail along the northern margin of the Barberton Greenstone Belt, allowing visitors to easily interact with outcrops on their way to other popular attractions. By driving along the R38 with science communication media produced from this project, locals and visitors will learn how geology shaped their world. They will gain insight on how mineable concentrations of verdite, talc and magnesite occur in certain ultramafic bodies. They will also learn of different ways in which rocks respond to stress as a common perception of rocks among laypersons is that they are solid, stable and unchanging. iii Acknowledgements This project was funded by the Centre of Excellence for Integrated Mineral and Energy Resource Analysis (CIMERA) in collaboration with the South African national government Department of Science and Innovation (DSI) and the National Research Fund (NRF). I would like to extend my gratitude to my Wits supervisors: Professors Judith Kinnaird and Paul Nex, who conceptualised and supported this project in many ways. They recruited the external supervisor, Professor Christoph Heubeck (Friedrich-Schiller University of Jena, Germany), and Mr Tony Ferrar, who - as authors of the Barberton Makhonjwa Geotrail guidebook - provided multi-faceted support during this project. Landowners Mr Whiteman, Mr Van Wyk and Mr Bladen† are acknowledged for granting access to research sites. Amidst the 2020 global crisis, a community of people contributed to the completion of this project. Firstly, to members of the Wits School of Geosciences: Thank you to my supervisors for staying patient and positive through this unprecedented geotourism MSc project. Thank you to Professor Gillian Drennan and Dr Grant Bybee for granting fieldwork and lab access permits once the state allowed it. Also, to the thin section technicians and departmental vehicle access administrators, Mr Sam Tshabalala, Mr Caiphus Majola and Mr Matthew Kitchin† for their time and assistance. Also to Mr Marlin Patchappa for his assistance with bulk chemical analysis. Thank you to Professor Carl Anhaeusser for his keen interest in this project and sharp memory of locations he accessed in previous decades. Barberton locals Tony and Sandy Ferrar, Dion Brandt and Roland Jones are thanked for field assistance. Not forgetting Dr Matthew Brayshaw and Matthew Hales for field discussions and Thomas Jones for acquiring drone images of the southern Stentor Pluton contact zone. Thank you to Mark Seady and Richard Bladen† for adding to my collection of printed books on Barberton geology, history and culture; and to the Barberton Museum for granting access to archives of historical photographs. Also, to Astrid Christianson of Barberton Tourism, Louis Loock, Rise Fm radio station, SANHU productions through SABC 3’s ‘Hectic On 3’, the University of Mpumalanga and the National Department of Tourism for providing platforms to fulfil the science outreach and geotourism promotion objectives of this project through radio and T.V. interviews, and a presentation to the local, regional and national tourism officials. Greatest gratitude goes to my family for field assistance and their unfailing support. †: Passed on Table of Contents Plagiarism declaration…………………………………………………………………………i Abstract………………………………………………………………………………………..ii Acknowledgements…………………………………………………………………………...iii List of Figures………………………………………………………………………………...iv List of Tables……………………………………………………………………………….....ix Introduction...…………………………………………………………………….....................1 Research question, aims, objectives, methods………….……………………………………...9 Chapter 1: Kaap Valley Tonalite…………………………………………...…………………13 Chapter 2: Verdite Mineralization……………………………...……………..……………...30 Chapter 3: Crenulated Phyllite……………….…………………………………………….…61 Chapter 4: Talc Mineralization…………………….…………………………………….…...70 Chapter 5: Moodies Basal Conglomerate………………….…………………………………80 Chapter 6: Lily Syncline………………….………………………………...………………...88 Chapter 7: Stentor Pluton Contact Zone…………………...……….………………………..106 Chapter 8: Magnesite Veins………………………………..…………………...………..….121 Conclusion…………………………………………………………………………..………151 References………………………………………………………………………………..…152 Appendices……………………………………………………………………………...…..171 iv List of figures Introduction Figure 1. Geological context of the Barberton-Makhonjwa Mountains World Heritage Site (BMM WHS: green filled polygon). ………….………………………………………….....…4 Figure 2. Selected roadside exposures along the R38 Geotrail………………………………..5 Figure 3. Simplified geological map of the R38 Geotrail in the context of the Barberton Greenstone Belt………………………………………….…………………………………….6 Figure 4. Chronological order (red numbers) of characteristic structures of main deformation events that affected the northern Barberton Greenstone Belt. Redrawn from Anhaeusser (1975). …………………………………………………………………………………………………8 Figure 5. Instruments used for analyses……………………...………………………………11 Figure 6. Analytical and data-processing instruments………………..……………………...12 Chapter 1 Figure 1.1 Main features of Site 1.…...………………………………………………………13 Figure 1.2 Kaap Valley Tonalite exposures at selected public and recreational localities in and around Barberton ……...……………………………………….…………………………….14 Figure 1.3 Distribution map of different tonalite phases of the Kaap Valley Tonalite….……………………………………………………………………………………18 Figure 1.4 Schematic diagram of the interleaved contact between the Kaap Valley Tonalite and metasomatised volcanics of the Onverwacht Group…………………….…….…………19 Figure 1.5 Exposures of the Kaap Valley Tonalite at localities that are easily accessed by locals………………………………………………………………………………………….21 Figure 1.6 Representative mafic enclaves of the Kaap Valley Tonalite form Site 1 of the R38 Geotrail, Tsekwane Pools (TP) and Caledonia Siding (CS)..………………...………………23 Figure 1.7 PPL photomicrographs of representative samples of the tonalite and enclaves from Site 1, sample numbers are shown in the bottom left corners………….……………………..24 Figure 1.8 Photomicrographs of hornblende textures in representative samples of the Kaap Valley Tonalite from Site 1……………………………………………….…………………..25 Figure 1.9 Element distribution maps of a representative sample of the tonalite (NP-001) shows an association of albite-forming elements (Na, Al, Ca and lesser K from potassic alteration)………………………………………………………………………………….....26 Figure 1.10 Element distribution maps of a representative sample of a fine-grained mafic enclave (NP-204A) shows an association of albite-forming elements (Na, Ca and lesser K)....27 v Chapter 2 Figure 2.1 Main features of Site 2.……………………………………………………….…..30 Figure 2.2 Hand specimen of barbertonite (stichtite-2H), a lilac-to-purple chromium hydrotalcite mineral……………………………………………………………..……………31 Fig. 2.3 An example of related geological features that occur at different scales ……....…...32 Figure 2.4 Local Swati women carrying red ochre from Dumaneni, a pre-colonial iron mine in the northeastern margin of the Barberton Greenstone Belt…………………….……………..32 Figure 2.5 Background of the old Handsup Mine……………..…………………..…………36 Figure 2.6 Lithological context of the remaining in-situ verdite pods in the Handsup Quarry..40 Figure 2.7 Characteristics of the chlorite-muscovite-albite schist host of verdite in the Handsup quarry………………………………………………...………………………………………41 Figure 2.8 Mineral composition of the verdite-bearing samples and host rocks. …..………43 Figure 2.9 Lithologies in Nelspruit verdite quarry A and B (images E and G) labelled with sample names………………………………………………..………..………………………45 Figure 2.10 Widely unknown verdite deposits 6 km south of Nelspruit (Quarry A: 26 32 30.36 S; 30 57 52.50 E. Quarry B: 25 32 21.30 S, 30 57 46.26 E)…………………………………..46 Figure 2.11 Nelspruit verdite quarry A………………………………………...……………..47 Figure 2.12 Main minerals related to the formation of Nelspruit verdite from quarry A and B……………………………………………………………………………………………...48 Figure 2.13 Distribution maps of selected elements within a compositionally layered verdite sample from the Barberton Handsup Layered Ultramafic Complex (sample NP-004)…….…50 Figure 2.14 Distribution maps of selected elements within a compositionally layered verdite sample from quarry A of the Nelspruit mafic ridge (sample NBV1/N1V1)…………….…….51 Figure 2.15 Distribution maps of selected elements within key lithologies of the Nelspruit mafic ridge, sample names at top right corners. ……………………………..………………52 Figure 2.16 Phase maps of sample NP-008.. …………………………………….…………..53 Figure 2.17 Modified Jensen (1976) Al-Fe+Ti-Mg oxide plot of verdite samples plotted for comparison with direct host lithologies and a variety of alteration products of ultramafic rocks mainly from the Barberton Greenstone Belt….……………………………………………….55 Figure 2.18 Classification diagram for altered lithologies from Winchester and Floyd (1977), modified after Agangi et al. (2018)…………………..……………………………………….56 Figure 2.19 Modified Jensen (1976) Al-Fe+Ti-Mg oxide plots of selected ultramafic-mafic host rocks of the Barberton Greenstone Belt and their alteration products plotted against comparable ultramafic-mafic lithologies……………………………………………………..57 vi Chapter 3 Figure 3.1 Main features of Site …………..……………………………………….………..61 Figure 3.2 Kaap River gold mining during the 19th century..…………………………………62 Figure 3.3 Figure 3.3 Lithological context of the mafic schist at Site 3.....…………………...64 Figure 3.4 Equal-area stereonet plots of schistosity and crenulation lineation measurements from a sheared phyllite invaded by black quartz veins that accentuate the crenulation fabric. ………………………………………………………………………………………………..65 Figure 3.5 Two main directions of compressional forces recorded at Site 3…………..……66 Figure 3.6 Main mineralogy of the phyllite section affected by later sub-horizontal shortening (sample NP-212).…………………………………………... ………………………………..67 Figure 3.7 Element distribution map of the crenulated phyllite sample NP-212.……….…….68 Figure 3.8 Sequence diagram showing associated changes that result in the development of crenulated schists or phyllites…………………………...……………………………………69 Chapter 4 Figure 4.1 Main features of Site 4. (A): Simplified geology of the Barberton Greenstone Belt, rectangular border outlines part of the research area………………………………………….70 Figure 4.2 Simplified geological map of the northern BGB between Joe’s Luck Siding and Eureka Siding…………………………………………………………...……………………73 Figure 4.3 Schematic diagram of lithologies along line A-B in the previous map show the relationship between steep ridges of the Lily Syncline and nickel-sulphide mineralised zones. ……………………………….……………………………………………………………….74 Figure 4.4 Altered serpentinites of the Bon Accord Layered Ultramafic Complex at Scotia Talc Mine……….…………………………………………………………………………………75 Figure 4.5 Photomicrographs of a talc-bearing carbonatised serpentinite from Scotia Talc Mine (Sample NP-STM)……………………………………………………………………...76 Figure 4.6. Phase maps of talc-bearing, carbonatised serpentinite from Scotia Talc Mine (Sample NP-STM)……………………………………………………………………………77 Figure 4.7. Elemental maps of selected elements within a carbonatised serpentinite (Sample NP-STM)....………………………………………………………………………………......78 Chapter 5 Figure 5.1 Main features of Site 5………………………………..…………………………..80 Figure 5.2 Geological context of the Moodies Basal Conglomerate from which >3.3 Ga potassic granite clasts and detrital zircon grains were liberated (Kröner et al., 2018 and references therein)………………………………………………...……………….…………84 vii Chapter 6 Figure 6.1 Main features of Site 6A………………………………..………………………..88 Figure 6.2 Nest of a bird of prey at Site 6A…………………….…………………………….89 Figure 6.3 Simplified geological map of the northern margin of the Barberton Greenstone Belt…………………….……………………………………………………………………..91 Figure 6.4 Cross section of the Lily and Eureka synclines near the western extent of the Lily Syncline………………………………..……………………………………………………..92 Figure 6.5 Lithologies of the Lily Syncline exposed across a north-to-south traverse through the Lily Ridge in the vicinity of the Honeybird Gorge (Site 6A). …………………………......96 Figure 6.6 Lithologies of the Lily Syncline exposed at Louiville from north to south…...…...97 Figure 6.7 Lithologies exposed along the roadside traverse at site 6B………………………..98 Figure 6.8 Mineral composition of markers in the Moodies Group core of the Lily Syncline………………………………………………………………………………………99 Figure 6.9 Element distribution maps of a representative sample of the altered feldspathic quartz-chert wacke marker (Sample NP-LS1C=BP)……..…………………………………101 Figure 6.10 Element distribution maps of a representative sample of the mafic schist marker (NP-LS2)…….……………………………………...………………………………………102 Chapter 7 Figure 7.1 Main features of Site 7…………………………………………………………..106 Figure 7.2 Characteristic rock fabrics at Honeybird Siding……….………………………..107 Figure 7.3 Painted boulders of the Stentor Pluton as seen from the R38. …………………..108 Figure 7.4 Range of compositions and relative proportions of felsic minerals of the Stentor Pluton granite sheets……………………………...…………………………………………110 Figure 7.5 Strain associated with Stentor Pluton emplacement at the Honeybird Siding locality…………………………………………….…………………………………….…..112 Figure 7.6 Phase map of the southern margin of the Stentor Pluton at the Honeybird Siding locality c. 6 km west southwest of Low’s Creek. ………….…………….………………….116 Figure 7.7 Characteristic lithologies of the Honeybird Siding contact zone………….……..117 Figure 7.8 Elemental maps of sample NP-10……………………………………………….118 Figure 7.9 Elemental maps of Sample NP-13……………………………………………….119 Chapter 8. Figure 8.1 Main features of Site 8……………..……………………………………………121 viii Figure 8.2 Simplified map of Kaapmuiden shows the intersection of rivers, roads and railway lines northeast of Kaapmuiden.…………… ………………………………………………..122 Figure 8.3 Old magnesite mines in in the Budd and Ship Hill layered ultramafic complexes………………………………………………………………………………...…124 Figure 8.4 Layered ultramafic complexes (LUC) of the Barberton Greenstone Belt……….127 Figure. 8.5 Deformation features associated with magnesite from the northern margin of the Barberton Greenstone Belt…………………………………………………………………..132 Figure. 8.6 Magnesite veins at site 8 (Budd deposit) occur as white layers shortened into microscale folds……………………………………………………………………………..133 Figure. 8.7 Main microscopic characteristics of magnesite…………………………..…….135 Figure. 8.8 Distribution of phases in a magnesite vein (sample NP-MJ2) from site 8 (Budd deposit)……………………...………………………………………………………………136 Figure. 8.9 Distribution maps of selected elements on sinusoidal magnesite veins from deposits on the northeastern (A) and southwestern (B) margins of the Stentor Pluton………..………………………………………………………………………….......137 Figure. 8.10 Distribution maps of selected elements in a magnesite vein from Site 8………………………………………………………………………………………….….138 Figure. 8.11 Distribution maps of selected elements on a magnesite vein from the Sugden deposit……………………………………………………………………………...……….139 Figure. 8.12 Element distribution maps of a sample from site 8 (Budd deposit)……………140 Figure. 8.13 Distribution of selected elements on a magnesite-bearing dunite from Strathmore Magnesite Mine……………………………….…………………………………………….141 Figure. 8.14 Raman spectra of magnesite sample MJS from the Sugden locality..............…142 Figure 8.15 Proposed events associated with the development of enterolithic folds observed in BGB magnesite..………………………..……………………………………………...……147 ix List of tables Chapter 2 Table 2.1. Compositions of verdite samples and selected host rocks. ………………………..54 Chapter 4 Table 4.1. Grade scheme of talc samples ranging from most to least valuable (Grade I-IV respectively). Information sourced from Halder (2013). ……………………………………..71 Chapter 5 Table 5.1 Variable clast deformation in the Moodies Basal Conglomerate of the Eureka Syncline (Anhaeusser, 1975). ………………………………………………………………..83 Chapter 6 Table 6.1. Decreasing flattening strain in a conglomeratic horizon along a 5m north-to-south traverse across the northern limb of the Lily Syncline at Site 6B. …………………………...98 1 0.1 Introduction 0.1.1 Barberton Granite-Greenstone Terrane Archean landmasses remain our primary source of information about early Earth processes. The volcano-sedimentary sequences and adjacent plutonic rocks contained in Archean landmasses such as the Barberton Granite-Greenstone Terrane (BGGT) are not only the best- preserved in quantity, continuity of sequences and in quality (Byerly et al., 2018 and references therein), they are also easily accessible. For this reason, the Barberton region is of outstanding universal value and worthy of world heritage status (World Heritage Committee, 2008). Through the years, researchers have studied various components of the BGGT, collectively producing plausible reconstructions of early Earth conditions and processes (Byerly et al., 2018 and references therein). Extensive work has been done covering such diverse topics as the evolution of the lithosphere and construction of continents (Anhaeusser, 1969, 1975; 1984; De Wit, 1992; Heubeck and Lowe, 1994; Van Kranendonk, 2010; Bedard, 2017), the origins of life (Barghoorn and Schopf, 1966; Walsh and Lowe, 1985; Ward, 1999; Schopf, 2006; Tice and Lowe, 2006; Kohler et al., 2013) and its emergence from marine to terrestrial environments (Noffke et al., 2006; Homann et al., 2015; Eriksson, 1977), the evolution of the atmosphere (Pflug, 1967; Lowe and Knauth, 1978; Reimer, 1980; Hessler et al., 2004; Lowe and Tice, 2004; Sleep and Hessler, 2006; Hessler and Lowe, 2006), the response of the early crust to regional stresses (Anhaeusser, 1969, 1972; Gay, 1969; Heubeck and Lowe, 1994b; Dirks et al., 2013) and the implications for ore- forming processes (Anhaeusser, 1969, 1975, 2012; Dirks et al., 2011; Gloyn-Jones and Kisters, 2018). Although many of these topics - or aspects thereof - remain debatable, the global significance of the BGGT is universally acknowledged. As a result, and for geoscientific reasons, a large part of the BGGT was recently (2018) declared a World Heritage Site. This study addresses the scientific outreach objective of both CIMERA and the management authority of the Barberton Makhonjwa Mountains World Heritage Site (BMM WHS) by collating, updating and disseminating geological knowledge about selected outcrops within the BGGT. To broaden the appeal of the Barberton Makhonjwa Mountains World Heritage Site as a tourist destination, the focus is on eight geologically significant and instructive sites along or near the R38 road, skirting the northern margin of the Barberton Greenstone Belt (BGB). This road links Barberton to the southern gates of the Kruger National Park. The sites represent contact zones of the BGB supracrustal sequence and surrounding granitoids, mineralisation 2 associated with ultramafic intrusives, highly strained conglomerates, and low-grade metamorphism. Objectives vary between original scientific work (including high-resolution geological mapping, petrographic and bulk chemical analysis, elemental mapping using µ- XRF) and synthesising pre-existing scientific data. By placing the selected sites in the context of modern BGB stratigraphy and tectonics as well as preparing learning objectives for laypersons, this study will produce both original geological literature as well as laying the base for educational products aimed at the general public, including adequately phrased pamphlets and guidebooks. 0.1.2. Barberton-Makhonjwa Mountains World Heritage Site The town of Barberton was developed in 1884 during a gold rush that predates the Witwatersrand goldrush by two years (see guidebook in appendices). The discovery of world class gold deposits during the 1880s brought the world’s attention to the town of Barberton. Once again, the town gains international attention as nearby mountains have become a UNESCO World Heritage Site (2018). This status applies to a site inscribed as the Barberton- Makhonjwa Mountains World Heritage Site (BMM WHS). It incorporates c. 40% of the Barberton Greenstone Belt (BGB) to the exclusion of areas associated with mining rights or those occurring in Eswatini. With the aim to boost geotourism activity around the then tentatively inscribed BMM WHS, the Barberton-Makhonjwa Geotrail was developed (2014). Located along the R40 road between Barberton and the Bulembu in the western part of Eswatini, the R40 Geotrail intercepts stratigraphic sequences of the BGB. Along this road, visitors can view rocks and information panels that describe the early Earth, local flora and fauna, as well as mining and cultural heritage of the region. It is currently the only geotourism physical product within the Barberton region and is the only road through the BMM WHS that intercepts the widest variety of lithologies. The R40 Geotrail has potential to gain more visitors if more attractions are created in the vicinity thereof. To explore that potential, the present study lays the foundation for extending the existing R40 Geotrail to the N4 national road connecting Pretoria to Mozambique. This extension will be done by developing a Geotrail on the tarred R38 road between the existing R40 Geotrail and the N4 road passing through Kaapmuiden. Along the R38 road are several geosites, most of 3 which are exposed as roadcut surfaces. These geosites display structures and minerals formed through multiple deformation events and metasomatism. Though R38 geosites present interesting features, laypersons may need more than just the geoheritage content to enjoy a geotourism tour. During the experience of the author in guiding laypersons to BGB geosites or giving talks about Barberton geoheritage, it became apparent that geotourism is better received by laypersons when coupled with other interests. This approach has the potential to work well in the Barberton region because of the following reasons. The region has a long history of human occupation spanning from the stone age to the present day (Liebenberg, 2016). It also has an active heritage society that conducts excursions to many geosites. There are documented sites rich in biodiversity (Emery et al., 2016) which are visited by birding clubs and ecology researchers who continue to make discoveries (Oosthuizen, 2021). The mining history extends for more than a century, and even longer when pre-colonial mines exploited during the iron-age are included (Delius, 2007; Liebenberg, 2016). Various natural or cultural heritage sites - published or those known only by locals - are often linked to rocks. For instance, a number of granitic bodies of the region preserve rock art (local knowledge; Liebenberg, 2016) or contain warm-water springs associated with mafic dykes (Visser et al., 1956). Other geosites double as cultural sites. This is the case for the painted boulders in Low’s Creek and ironstone exposures mined for red ochre since pre- colonial times (Myburgh, 1949; Liebenburg, 2016). For this reason, geology chapters of this thesis, which are numbered according to the sequence of the proposed R38 Geotrail sites from Barberton to Kaapmuiden, include cultural and/or historical attractions that will help tourism service providers attract groups of various interest to the R38 Geotrail and the associated BMM WHS. 4 0.1.3. Proposed R38 Geotrail Sites The R38 road winds along the northern edge of the Barberton Greenstone Belt, cutting across two granite bodies, various greenstones and good outcrops in two river systems. From about 7 km outside Barberton, the Kaap River can be seen winding in the valley adjacent to the Geotrail, providing wonderful scenery all year round. This river has been included in close-up maps of different sites. By spotting features of the river e.g., a sharp bend to the east or where the river flows closest to the road, visitors can ensure that they are at the right site in the absence of navigation tools. Figure 1. Geological context of the Barberton-Makhonjwa Mountains World Heritage Site (BMM WHS: green filled polygon). The site includes 40% of the Barberton Greenstone Belt (BGB: grey filled polygon) and is best explored along a road network called the ‘Genesis Route’ (orange). This route intercepts key lithologies and other attractions. 5 As a proposed geotrail, this section of the R38 between Barberton and Kaapmuiden currently contains no information panels, bins, shops or toilet facilities. However, one of the proposed R38 Geotrail sites (Site 2) is located outside Diggers’ Retreat, which is a hotel and restaurant facility where toilets can be used after purchasing refreshments. 0.1.4. Regional Geology The Barberton Granite Greenstone Terrain (Figure 1) records more than 900 Ma (ca. 3644 to ca. 2740 Ma) of Earth history from the Paleo- to Neoarchean (Kröner and Compston, 1988; Tegtmeyer and Kröner, 1987; Kamo and Davis, 1994; Robb et al., 2006; Byerly et al., 2018). Preserved components of the BGGT indicate at least three cycles of crustal evolution from ultramafic to felsic magmatism (Kohler and Anhaeusser, 2002; Robb et al., 2006; Kröner et al., 2013; 2018). Most rocks that constitute the BGGT are magmatic and occur either as volcanic components of the Barberton Greenstone Belt (BGB) supracrustal sequence or as intrusive layered ultramafic sequences or plutons of various compositions (Viljoen and Viljoen, 1969; Robb et al., 2006; Kohler, 2003). Ovoid and circular granitoid plutons of the region surround the BGB (Anhaeusser, 1976; Anhaeusser and Robb, 1981; Robb et al., 1986). The supracrustal rock sequence has been divided into three lithostratigraphic units. These units are, from oldest to youngest, the Onverwacht, Fig Tree and the Moodies Groups. They have Figure 2. Selected roadside exposures along the R38 Geotrail. 6 corresponding thicknesses of >10 000 m, ca. 1 800 m and ca. 3 000 m, respectively (Byerly et al., 2018). 0.1.4.1. Stratigraphic units The Barberton greenstone belt (BGB) has been categorised into three tectonostratigraphic blocks which, from south to north, are the Songimvelo, Umuduha and Kaap Valley blocks (Byerly et al., 2018). Though consisting of similar lithostratigraphic units, these terrains have distinct stratigraphy and tectonic histories, generally becoming younger towards the north (Lowe and Byerly, 2007). This study area lies on the Kaap Valley block of the BGB. The 3550-3270 Ma Onverwacht Group of the Kaap Valley Block consists of ultramafic to mafic volcanics and their serpentinised equivalents (Anhaeusser, 1976; Lowe and Byerly, 2007; Byerly et al., 2018). Minor occurrences of intermediate to felsic volcanics also occur, with the former often found to have ages overlapping with those of nearby intermediate granitoids (Agangi et al., 2018). Cherts and silicified clastic units form a volumetrically minor but conspicuous and stratigraphically important component (Lowe and Byerly, 2007; Hofmann et al., 2013). Conformably overlying the Onverwacht Group is the 3260-3225 Ma Fig Tree Group, mostly composed of argillite, greywacke, various chemical sediments and intermediate volcanics. It Figure 3. Simplified geological map of the R38 Geotrail in the context of the Barberton Greenstone Belt. Modified after Visser et al., (1956) and Anhaeusser et al. (1981). Stratigraphic column modified after Byerly et al., 2018. 7 contains several meteorite-impact spherule beds (Lowe et al., 2014; Jones and Kisters, 2018). The Bien Venue Formation that forms part of the upper Fig Tree Group defined in the northeastern Barberton Greenstone Belt is dominated by quartz-muscovite schists (Kohler and Anhaeusser, 2003). Stratigraphically, the Bien Venue Formation is overlain by felsic schists and agglomerates of the Schoongezicht Formation that are in a transitional contact with the Moodies Group (Byerly et al., 2018). The 3224-3214 Ma Moodies Group is comprised of terrigenous sediments interlayered with iron-rich shales, jaspilites, rhyodacitic ash-fall tuffs and a regional basaltic unit (Visser et al., 1956; Anhaeusser, 1976; Eriksson, 1978; Heubeck and Lowe, 1994). 0.1.4.2. Deformation The present fabric of the Barberton Greenstone Belt (BGB) is a result of multistage deformation during the deposition of BGB supracrustal deposits (Visser et al., 1956; Anhaeusser, 1966; 2019; De Ronde and De Wit., 1994; Heubeck and Lowe, 1994; Byerly, 1999; Kisters et al., 2003; Van Kranendonk, 2011; Byerly et al., 2018; Dziggel and Kisters, 2019). In the northern BGB, the earliest deformation event (D1) produced a series of cleavage- less, doubly-plunging folds of a northeast-southwest trend (Ramsay, 1963; Visser et al., 1956) understood to have resulted from gravity-induced sagging of BGB supracrustals (Anhaeusser, 1966). The folds verge to the north and northwest, with the southern limb overturned and invariably thicker than the northern limb due to tectonic truncation by high-angle faults (Anhaeusser, 1976; Byerly et al., 2018). The high-angle faults coincide with poorly preserved anticlinal structures between regional synclines (Anhaeusser, 1976). The former were considered to have initially been low-angle thrust faults separating synclines in a fold-and- thrust belt during D1 shortening (De Wit and De Ronde, 1994). These faults were reactivated as strike-slip faults during events subsequent to D1 (Anhaeusser, 1976). The second deformation produced slaty cleavage of a northeast-southwest trend at an oblique angle to bedding and fold axes of D1 (Ramsay, 1963). However, the similar orientation of D1 and D2 principal strain axes may indicate a single event of progressive deformation (Ramsay, 1963). The third event in Ramsay (1963) is considered to be northeast-southwest shortening which Anhaeusser (1963, 1966, 1972, 1975) associated with the emplacement of the Kaap Valley Tonalite (Anhaeusser, 1966; Gay, 1969). This third event of shortening was coincident with the refolding of the Eureka and Ulundi Synclines and their detachment folding about a northwest-southeast axis (Anhaeusser, 1968; 1972; 1975; 2019). A final deformation event is indicated by sub-vertical shortening of D2 slaty cleavage to produce sub-horizontal crenulation 8 lineations, kink bands as well as conjugate and chevron folds (Anhaeusser, 1963, 1968, 1975; Viljoen, 1963). However, after noting the consistent juxtaposition of Moodies and Fig Tree sequences over Fig Tree and Onverwacht sequences in the northern BGB, Byerly et al. (2018) proposed a northwards thrusting event as the final deformation event. Figure 4. Chronological order (red numbers) of characteristic structures of main deformation events that affected the northern Barberton Greenstone Belt. Redrawn from Anhaeusser (1975). 9 0.1.5 Present Study 0.1.5.1 Aims To identify roadside outcrops in the northern Barberton Greenstone Belt about which existing and new geological information with an appeal to laypersons would be generated and further research would be inspired. To produce written and visual content for science outreach, geoheritage awareness and geotourism development in the Barberton region. 0.1.5.2 Objectives To undertake a literature review, collate data and to generate new information for sites with unresolved key features. Detailed accounts of objectives are listed in individual chapters. 0.1.5.3. Methods: Science outreach and thesis components To fulfil science objectives of the funding agency, the geoheritage of the Barberton region was translated to laypersons through T.V., radio, a media article and a talk at a local university where c. 300 tourism and heritage students were in attendance (Appendix A). General to all sites is the synthesis of existing information, the digitisation of map sections relevant to each site and redrawing of figures. New figures were drawn where new information was produced. Due to travel restrictions and limited lab access resulting from the COVID-19 pandemic, original lab-intensive studies meant for some sites were substituted by comparative field studies of other sites. This was to confirm apparent characteristic features and lithological associations observed in R38 outcrops. Unless otherwise stated in individual chapters, thin sections of characteristic samples of each site were produced from oriented samples photographed within the sampled outcrop. Sample composition and fabrics were described using a petrographic microscope. Photomicrographs were created in the Wits Imaging Lab using the Olympus BX41 and BX53M photographic microscopes connected to the Olympus DP74 imaging system (Fig. 5 A). All bulk chemical analyses were carried out on samples trimmed of weathered surfaces and subsequently processed using a jaw crusher and disc mill at the Wits Geosciences Earthlab. Major element analysis of samples was conducted using the Panalytical Axios x-ray spectrometer to measure x-ray fluorescence. Elements were determined using the Norish Fusion technique (Norrish and Hutton, 1969) together with in-house correction procedures. 10 Major elements were fused using Johnson Matthey Spectrolflux 105 at 1000 ֯ C and raw data corrected using in-house software. Standard calibrations were made using synthetic oxide mixtures and international standard rocks as well as in-house controls. Sample weight used was 0.35 gm and flux weight 2.0 gm. Calibration standards were primary International Reference Materials USGS series (USA) and NIM series (South Africa). Precision was taken as 1% for elements in abundance of greater than 5% by weight, and 5% for elements in abundance less than 5%. Standard calibrations used BCR2, NIM-P, NIM-D, W-2, NIM-S, GSP-2, NIM-N, NIM-G, PCC1, BHVVO2, AGV2, G2 DTS1. Bulk chemical analyses of trace elements of samples were conducted using the Phillips PW2404 x-ray spectrometer. Samples were pressed into pellets using a Moviol solution binder. Standardisation was carried out using International Reference Materials USGS series (USA) and NIM series (South Africa). Precision was determined on the basis of counting time and was taken as 5% for elements in abundance greater than 100ppm, and 10% for elements in abundance between 10 and 100 ppm. Elemental maps were generated using a Bruker Tornado M4 micro-X-ray fluorescence analyser at the Friedrich-Schiller University of Jena, Germany. Selected parts of thin sections were carbon coated using graphite and irradiated with an electron beam. The ratio of counts per second to applied electrical potential from the electron beam was plotted against the applied electrical potential to generate characteristic peaks of phases within the analysed sample. Peaks on a spectrum of reference materials were matched with peaks on a spectrum of the analysed sample to identify phases. Confirmation of phases and the characterisation of advancely altered components was achieved using the liberation analysis option of the VEGA3 X64 TESCAN TIMA 1.7.0 (Fig. 5 B). scanning electron microscope. In particular, the energy-dispersive X-ray spectroscopy component was used together with the VEGA3 X64 TESCAN TIMA 1.7.0. software within the Wits Automated Mineralogy Lab of the School of Geosciences at the University of the Witwatersrand, Johannesburg. Beam conditions during liberation analyses on the VEGA3 X64 TESCAN SEM were 25 kV at accelerating voltage and an emission current of 43 µA. Working distance of 15 mm, a specimen beam current of 17.68 nA, spot size of 590.00 nm and an absorption current of < 1 pA were selected. The scanning mode was set at resolution mode, the heating at 47 %, the filament live time was at 431 hours and the gun pressure was at 1.6x10-3 Pa while the column pressure was at 5.5 x10-3 Pa. Acquisition of high-resolution maps was conducted at a pixel spacing of 2 microns, x-ray counts of 1000 and an imaging speed of 4 microseconds. 11 Raman spectra were produced from four spots on a freshly cut, unpolished surface of outcrop sample MJS were examined on a HORIBA LabRam Raman instrument (Fig. 6) at the Department of Geosciences, Friedrich-Schiller University of Jena, Germany. The microscope magnification was 20x and a laser wavelength of 532 nm was used for 9 seconds and two runs and at 5 % laser intensity. Acquired spectra were found to lie on curves which were flattened by simple curve fitting. Spectra also showed (poorly defined) periodicities of ca. 90 cm-1 between 300 and 1300 cm-1. These deficiencies may be due to weak electric current changes, effects of room lighting or fluctuations in laser intensity. Corrected spectra were compared to those proposed by the HORIBA KnowItAll Spectral library and the public RRUFF Raman spectra database. Figure 5. Instruments used for analyses. (A): Olympus BX53M petrographic microscope and DP740 digital camera (top component) used to generate photomicrographs. (B): TESCAN integrated mineral analysis (TIMA) instrument. Insert shows carbon-coated thin sections mounted on a stage. 12 Figure 6. Analytical and data-processing instruments. (A) Sample MJS mounted on a stage prior to analysis. (B): Hardware coupled with spectrometer for data acquisition, processing, and export. 13 Chapter 1: Kaap Valley Tonalite (Crumbly Granite) in a Quarry Way in at 25°45'0.28"S; 31° 3'12.07"E 1.1 Site description Weathered outcrops of the Kaap Valley Tonalite are exposed in a municipal sandpit where tonalite fragments and sand are extracted for various uses in construction and road maintenance (Fig. 1.1). Where intact, outcrops of the tonalite occur as medium- to coarse-textured domes within masses of friable weathered material. Northwest-trending veinlets of epidote are common within the quarry. These are of a similar trend to dolerite dykes that invade the tonalite in the area between Barberton and Mbombela (Fig. 1. 3). The various mafic enclaves in the tonalite will be described in later sections of this chapter. As a public location, this site may be accessed by other people at any given time. Care should be taken to discourage harm or theft by accessing the site in groups and locking or hiding away valuable items in vehicles. Figure 1.1 Main features of Site 1. (A): Simplified geological map of the R38 Geotrail showing the context of Site 1 in the Barberton Greenstone Belt (BGB). (B): A detailed view of the black rectangle in (A) shows the location of Site 1 in the Kaap Valley Tonalite at the western margin of the BGB. (C) View of Site 1 looking southeast. Maps modified after Visser et al. (1956) and Anhaeusser et al. (1981). 14 1.2 Aspects of the site that may appeal to laypersons during site visits Solid outcrops of the Kaap Valley Tonalite occur in and around Barberton. At the sandpit (Fig. 1.1), tectonic fabrics and fractures in the marginal zones of the tonalite can be observed. These contribute to an understanding of the emplacement history of the pluton. They may be highlighted for their significance concerning deductions made about the emplacement history of the pluton. Visitors will also see well-preserved outcrops of the tonalite found within cores of friable plutonic rock on the quarry floor. Such exposures provide examples of weathering processes that form part of the rock cycle. Fundamentally, concepts of weathering and variable resistance to erosion are easily seen here because the drastic change of slope at margins of the valley is very apparent. The weathering of the quartz-poor tonalite into a near-circular valley also contrasts with the resistance of the high hills formed by quartz-rich phases of the Nelspruit Batholith at the end of the R38 Geotrail. On a more advanced level, coarse crystals making up the tonalite in the sand pit will be a discussion point on the control of cooling rate on crystal size and shape. Various mafic enclaves and dykes (Fig. 1.2) can be used to discuss the usefulness of inclusions and cross-cutting relations in establishing relative ages of rocks. Figure 1.2 Kaap Valley Tonalite exposures at selected public and recreational localities in and around Barberton. (A): Extensive tonalite outcrops are a common occurrence in school yards and playgrounds. (B): A picnic location c. 16km WSW of Barberton (Tsekwane/Tiguan Pools) allows relative dating from the use of cross-cutting relationships between a basaltic dyke that crosscuts mineral layering and hosts angular tonalite fragments. 15 1.3 Geology 1.3.1 Background 1.3.1.1 Kaap Valley Tonalite Archean cratons contain granitoids of varying compositions (Moyen et al., 2010; Halla, 2018). These granitoids broadly fall into two main categories, the Trondhjemite-Tonalite-Granodiorite (TTG) series and the Granodiorite-Monzogranite-Syenogranite (GMS) series. During the Archean, granitoid magmatism often occurred episodically, and often produced composite TTGs (Halla, 2018). Further studies of TTG geochemistry led to the identification of two subcategories of TTGs, those with low Y and HREE, and those with high Y and HREE. These subgroups respectively correspond to deep and shallow crustal melting of oceanic crust with and without garnet in the residual phase (Halla, 2018). In the Barberton-Granite-Greenstone-Terrane (BGGT), TTGs constitute the oldest remaining granitic bodies of the region and were formed during three main magmatic episodes (Condie and Hunter, 1975; Anhaeusser and Robb, 1981; Kamo and Davis, 1994; Agangi et al., 2018; Kröner et al., 2018). The youngest of these magmatic episodes produced TTGs in the western and northern margins of the BGB during a period of rapid crustal thickening between 3250- 3225 Ma (Kamo and Davis, 1994; Byerly et al., 2018). It is during this period that the Kaap Valley Tonalite was emplaced. The tonalite is a medium- to coarse-grained (0.5-6mm) plagioclase-rich granite (albite-to- andesine) mainly with hornblende and lesser biotite as mafic phases (Robb et al., 1986). It is more mafic than granitoids of the region (Anhaeusser, 1972; Robb and Anhaeusser, 1980; Robb et al., 1986). Fine- to medium-grained tonalitic dykes with no apparent chill margin occur in central parts of the pluton (Robb et al., 1986), while feldspar porphyries, quartz and carbonate veins commonly occur at and near pluton margins (De Ronde et al., 1988; Faure, 1989). The higher hornblende content in the tonalite compared to other granites, was initially ascribed to magma contamination by hornblende xenocrysts from ultramafic country rocks (Condie and Hunter, 1975). However, complete assimilation of mafic-ultramafic country rocks, or mixing of rhyolitic and mafic-ultramafic magma were considered by Robb et al. (1986) to be appropriate based on petrographic and geochemical data. The most precise age determined for the tonalite is a U-Pb age of 3227±1 (Kamo et al., 1990). Subsequently, Ar-Ar ages between 3250 and 3035 Ma were determined from hornblende and 16 biotite crystals from central and marginal zones of the tonalite (Layer et al., 1992). The upper limit of the 3250-3035 Ma age spectrum corresponds to crystallisation of the tonalite after semi-solid diapiric emplacement to shallower crustal levels (Robb et al., 1986). Crystallisation was soon followed by a pre-3214 Ma magnetic reversal proposed by Layer et al. (1996) to possibly be the oldest preserved magnetic reversal on Earth. The younger ages of the 3250- 3035 Ma spectrum represent hydrothermal events presumably linked to the c.3.1 Ga intrusion of the Nelspruit Batholith and associated gold mineralisation of the northern part of the Greenstone Belt (Layer et al., 1992). Once solid, the tonalite was affected by northwest-trending shear zones (Robb et al., 1986; Faure, 1989). These shear zones became sites for the emplacement of sub-economic gold- bearing quartz-chlorite-limonite veins with gold grades ranging from 0.03 to 0.9 ppm (Faure, 1989). Two later generations of pre-Transvaal basaltic and gabbroic dykes with U-Pb baddeleyite ages of c.2965-2967 Ma and c. 2662-2686 Ma intrude the Kaap Valley Tonalite (Olsson et al., 2010). The older of these is the northwest trending Badplaas Dyke Swarm which occurs in a c.80 by 100 km2 area between Eswatini and the Nelspruit region. It is correlated with lavas of the Nsuze Group (Olsson et al., 2010). Some warm-water springs of the Badplaas region occur where dykes of this swarm intercept sheared Archean plutons (Visser et al., 1956). The younger swarm is known as the Rykoppie Dyke Swarm (Olsson et al., 2010). It extends over a c.50 by 100 km2 area within the Archean basement part of the Barberton region and is correlated with the upper Ventersdorp Supergroup. The Rykoppie Dyke Swarm formed in a rift basin that created accommodation space for the deposition of Transvaal Group sediments (Olsson et al., 2010). 1.3.1.2 Enclaves According to previous workers cited in Barbarin and Didier (1991) and Barnes et al. (2021), mafic magmatic enclaves may be genetically related or unrelated to the enclosing granitic host. Enclaves may form through: - chemical alteration of xenoliths (Bower, 1922); - the disaggregation of synplutonic mafic dykes (Roddick and Armstrong, 1959); - partial melting of restite material to produce the enclosing magma (White and Chappell, 1977); 17 - magma mixing where granitic magma is recharged by incoming mafic magma transporting undercooled, (semi) consolidated fragments to be later enclosed in the granitic magma (Eichelberger, 1978; Gamble, 1979; Barnes et al., 1986) or - the incorporation (into granitic magma) of fragments of a more mafic, earlier phase crystallised from an evolving magma chamber (Pabst, 1928; Bateman et al., 1963). It has been widely accepted that TTG magma is generated through 15-30 % partial melting of hydrous ultramafic-mafic rocks (Holloway and Burnham, 1972; Moyen et al., 2010). This, according to Beard et al. (2004) occurs through incongruent melting of anhydrous mafic minerals in an assemblage of, for example, orthopyroxene-clinopyroxene-calcic-plagioclase- FeTi oxides to produce trondhjemites or tonalites comprised of hornblende-biotite-sodic plagioclase-quartz. In this way, preserved pyroxenitic restites may be associated with the resulting TTGs bodies. Less enigmatic associations of mafic fragments and TTGs are seen in angular mm- to km-scale enclaves which are understood to be rafts stoped into TTG magma (Anhaeusser and Robb, 1981; Robb et al., 1986). These enclaves are abundant near TTG margins due to the emplacement of most Barberton regional TGGs directly against ultramafic and mafic sequences of the Onverwacht Group (Anhaeusser, 1969; 1976; Robb and Anhaeusser, 1981). The tonalite margins may therefore contain enclaves of both obvious and contentious origin. According to Robb et al. (1986), the enclave population of the tonalite is made up of large and small fragments of serpentinites and amphibolites. The larger fragments were regarded as xenoliths and smaller fragments which were partially digested and found to have reacted with the enclosing tonalite were regarded as comagmatic phases. Comparative geochemical analyses by Faure (1989) found that hornblende from enclaves in the tonalite and that from the host tonalite overlapped in chemical composition. The challenges in distinguishing the origins of enclaves were highlighted in Barbarin and Didier (1991). These authors described how the isotopic equilibration of enclaves with the enveloping melt may lead to a loss of the original isotopic fingerprint of enclaves. However, studies that include detailed comparisons of texture and chemical evolution of zoned plagioclase crystals from enclaves and the enclosing host allow for better assessments of enclave origins (e.g., Barnes et al., 2021). Distinct phases of the tonalite (Fig. 1.3) were identified in different spatial and topographic zones of the pluton (Robb et al., 1986). The dominant of these phases contains hornblende and 18 lesser biotite as primary mafic minerals. Less common was a phase that contained hornblende as the only mafic phase in low topographic zones of the tonalite whereas the latter was in high topographic zones (Robb et al., 1986). In contrast, field observations and petrographic analyses by Faure (1989) contended that the tonalite was originally homogeneous. The heterogeneous occurrence of biotite in the tonalite was attributed to complete chloritisation of biotite within the original biotite-hornblende assemblage due to higher susceptibility of biotite to chloritisation (Faure, 1989). Studying 900 points on 78 tonalite samples, Faure (1989) provided a modal average of 58.1% primary plagioclase, 22% quartz, 5.8-20% hornblende and 1-10% biotite with 10.1% secondary chlorite, 2.0% epidote and 0.9-40% of secondary calcite in the samples studied. 1.3.1.3 Margins Except for the western margin which is overlain by Proterozoic Transvaal Group metasedimentary rocks, the margins of the Kaap Valley Tonalite are adjacent to sheared metavolcanics of the Onverwacht Group (Anhaeusser, 1966, 1969). These sheared metavolcanics are locally associated with gold deposits that have been mined for more than a Figure 1.3 Distribution map of different tonalite phases of the Kaap Valley Tonalite. Also shown are NW-trending and roughly E-W-trending dyke swarms. Map redrawn from Robb et al. (1986). 19 century (Anhaeusser, 1966, 1969, 1976, 2013). Unlike the 3.1 Ga potassic granites of the Barberton region, the tonalite margins are not discordant intrusive bodies (Anhaeusser, 1966, 1969, 1972). Instead, interleaving contact zones comprised of concordant, alternating layers of sheared tonalite and metavolcanic country rocks are observed within an amphibolite-to- greenschist metamorphic aureole of less than 5 km (Anhaeusser, 1966, 1969; Robb et al, 1986; Faure, 1989; Dziggel et al., 2006). According to Anhaeusser (1966, 1969) and Faure (1989), lithologies on either side of the tonalite margin present a strong contact-parallel mineral foliation and lineation that dips or plunges southwest towards the core of the pluton at certain localities (Fig. 1.4). These fabrics decrease in intensity away from the contact zone, shallowing towards the core of the pluton suggesting that the margins of the tonalite were semi-solid during emplacement into the metavolcanics of the Onverwacht Group resulting in metasomatised schists (Anhaeusser, 1966, 1969). In this way, margins of the tonalite were sheared and recrystallised as they steepened and caused shearing of metavolcanic country rocks (Anhaeusser, 1969; Robb et al., 1986) and - according to Faure (1989) - produced a cataclasite zone in the southeastern contact of the tonalite. Both the sub-vertical and sub-horizontal strain within the Barberton Greenstone Belt was attributed to the shallow diapiric emplacement of the tonalite (Viljoen, 1963; Anhaeusser, 1968, 2019). Within the contact aureole of the tonalite are pluton-parallel extensional shear zones that were responsible for the downward displacement of Onverwacht metavolcanics (Anhaeusser, 1976). Beyond the contact aureole, the effects of the emplacement of the tonalite caused sub- horizontal northeast-southwest shortening and detachment folding of the inner Eureka and Ulundi folds about a northwest-southeast fold axis. Progressive shortening led to northwestwards thrusting of the refolded Eureka and Ulundi folds during D3 northeast- southwest shortening (Anhaeusser, 1976, 2019). This deformation was bound by the Barbrook fault in the southeast and the Lily Fault in the northwest (Anhaeusser, 1976, 2019). Figure 1.4 Schematic diagram of the interleaved contact between the Kaap Valley Tonalite and metasomatised volcanics of the Onverwacht Group. This outcrop can be accessed at Site 1 of the R40 Barberton-Makhonjwa Geotrail. Note the southwest-dipping foliation in the country rocks that also occurs in the tonalite. Modified after Faure (1989). 20 1.3.2 Present study 1.3.2.1 Aim and Objectives To aim of this research was to make field descriptions of enclaves and their contacts with the tonalite and to describe textural and compositional characteristics of sampled enclaves and their margins. 1.3.2.2 Methods Where possible, characteristic enclaves were viewed in three dimensions and were described in the field in terms of composition and texture, size, nature of contacts and orientation with respect to tonalite foliation. Representative samples of enclave variants within the host tonalite were cut into thin sections for petrographic study. The focus of descriptions was composition, identity, and texture of minerals on either side of the enclave-tonalite contacts. To verify enclave descriptions made at Site 1 of the R38 Geotrail, enclaves from three other sites that are easily accessible were also described. These include a river pavement from the Caledonia Siding, outcrops on a school playground in the Barberton Township and a river pavement at Tsekwane/Tiguan Pools that are respectively c.10 km north, c. 2 km NNW and c. 16 km WSW of the Barberton Central Business District. All sites studied occur within a hornblende-tonalite phase of the Kaap Valley Tonalite identified by Robb et al. (1986) in the southeastern margin of the pluton (Fig. 1.3). Elemental maps of representative samples of the tonalite and of a sub-angular mafic enclave with sharp margins were produced from complementary offcuts of blocks from which thin sections were cut. 21 1.4 Results 1.4.1 Field observations Enclaves were more rare (<5 in a 10 by 10 m2 area) at the school playground locality than in the three other localities situated closer to tonalite-BGB contact zones so enclaves are only described from Site 1 and two river pavements: Caledonia Siding and Tsekwane/Tiguan Pools. The enclave population observed mainly contains 5 to 50 cm long, isolated enclaves that are either well-rounded or angular. Rare enclave clusters (Fig. 1.6 I) are comprised of 5 to 40 cm enclaves that do not have mutually complementary margins suggestive of an origin through brecciation of a single mafic dyke or larger enclave (e.g., disaggregated mafic synplutonic dykes of Barnes et al. 2021). Instead, enclave clusters are comprised of individual angular to sub-angular enclaves ranging in hornblende abundance from 40 to 70 %. No consistent link was observed between enclave size or shape and hornblende content within the enclave clusters, or the enclave population studied. All enclaves observed were comprised of at least 20% hornblende and had no apparent rims and or internal enclaves e.g. “double enclaves” of Barnes et al. (2021). Internal structures include quartzo-feldspathic dykelets that occur in both circular and angular enclaves (Fig. 1.6 A-C); as well as schistosity in angular variants. The long axis of angular enclaves and their Figure 1.5 Exposures of the Kaap Valley Tonalite at localities that are easily accessed by locals. (A): Site 1 quarry showing sampling locations (yellow pins) and other general features to see (red pins). (B): Localities accessed for enclave descriptions. 22 internal mineral foliation are parallel or sub-parallel to tonalite foliation. In contrast, unsheared circular enclaves do not show any internal alignment or rotation of crystals into alignment with the foliation of the host rock (Fig. 1.6 D, H). Contacts of the enclaves range from sharp to gradational. Unless partially resorbed, rounded variants mostly have sharp contacts, and the angular varieties exhibit sharp and gradational contacts, sometimes even within different ends of the same enclave. While some enclaves are conspicuous due to a high content of hornblende, some are barely noticeable due to having similar mineral proportions to the host tonalite as well as diffuse margins (Fig. 1.6 C). Coarse (2-5 mm), isolated feldspar crystals occur heterogeneously within angular and circular enclaves commonly without a systematic increase in modal abundance from margins to cores. In other enclaves, feldspar content is uniform throughout the enclave. The latter are mainly composed of chloritised hornblende, 30-50 % plagioclase and <20% quartz. 23 Figure .1.6 Representative mafic enclaves of the Kaap Valley Tonalite form Site 1 of the R38 Geotrail, Tsekwane Pools (TP) and Caledonia Siding (CS). Sample names are included near enclaves, see appendix for other samples. Scale bar has 1 cm segments. (A-C): Enclaves invaded by quartzo-feldspathic dykelets include angular enclaves (A), sub-rounded enclaves (B) and inconspicuous, tabular enclaves (C). Other enclaves are circular (D-E) or angular (F) mafic enclaves with uniformly distributed, coarse feldspar crystals. (G-I): Non-uniform distribution of feldspar in mafic enclaves is observed in enclaves with all or some diffuse margins and more mafic cores (G-H), or discontinuous internal zones of higher feldspar content (I). (J): A rare cluster of enclaves of varying hornblende proportions. 24 1.4.2 Petrography Samples studied were predominantly of the hornblende-rich tonalite phase. Texturally, the tonalite is coarse-grained and mainly comprised of euhedral to subhedral crystals of albite and hornblende 2-8 mm in size. The groundmass contains quartz, sericite, subhedral apatite and anhedral zoisite. Where present, euhedral to subhedral biotite is entrained in a weakly developed foliation that anastomoses around angular enclaves. Euhedral accessory apatite, magnetite and sphene are common inclusions in both hornblende and biotite. Hornblende in both the tonalite and enclaves is chloritised. Except for fractured enclaves, the main difference in angular enclaves and the tonalite host was the lower quartz content in enclaves (Fig. 1.6). Figure 1.7 PPL photomicrographs of representative samples of the tonalite and enclaves from Site 1, sample numbers are shown in the bottom left corners. (A): Kaap Valley Tonalite. (B-C): Sharp (B) and diffuse contacts of a single hornblende- albite-quartz enclave in a biotite-bearing phase of the tonalite. Note how the quartz- filled diagonal fractures in the lower half of the thin section have transformed part of the enclave into a less mafic phase. (D): Gradational boundary between the tonalite and enclave where the only difference apparent is the higher abundance of chloritised hornblende in the enclave. (E-F): Sharp tonalite-enclave contact where the enclave is a finer grained and quartz-poor version of the host tonalite. 25 1.4.3 Elemental Maps Element distribution maps of the representative samples of the Kaap Valley Tonalite (NP-001) and a sharp margin of a representative mafic enclave show the following. Magnesium, calcium, iron, titanium, sulphur, chromium and rare copper occur in hornblende. Pyrite is highlighted where iron and sulphur co-exists while ilmenite is highlighted where titanium and iron clusters occur in association with hornblende. Albite crystals coincide with discrete zones of sodium, aluminium and lesser calcium-which is less abundant in the feldspar than the adjacent hornblende crystals. Potassium is concentrated at albite margins and in cores sericitised crystals. Figure 1.8 Photomicrographs of hornblende textures in representative samples of the Kaap Valley Tonalite from Site 1. PPL, XPL labels are in photomicrographs produced using plane-polarised or cross-polarised light, respectively. (A): Magnetite exsolution in chloritised hornblende (Hbl-Chl). Note the microfault (yellow star) (B): Mainly unchloritised biotite (Bt) along margins of a hornblende (Hbl) enclave that lacks biotite. (C): Mainly unchloritised biotite adjacent to chloritised hornblende. (D): Quartz (Qtz) growth along hornblende cleavage planes. Note the alignment of hornblende cleavage with fold axes of kinked albite (Ab) and quartz (Qtz) in the tonalite sector of enclave NP-206. 26 Figure 1.9 Element distribution maps of a representative sample of the tonalite (NP-001) shows an association of albite-forming elements (Na, Al, Ca and lesser K from potassic alteration). Fe, S, Ti and K occur in quartz within interstices of hornblende mineral clusters and Cu is associated with some sulphur clusters. Mafic minerals coincide with Mg, Fe and Cr-rich zones and are associated with discrete sulphur-rich zones. 27 Figure 1.10 Element distribution maps of a representative sample of a fine-grained mafic enclave (NP-204A) shows an association of albite-forming elements (Na, Ca and lesser K). Fe, S, Ti and K occur in quartz that occupies interstices of hornblende mineral clusters while Cu is associated with some sulphur clusters suggesting traces of chalcopyrite. Mafic minerals coincide with Mg, Fe and Cr-rich zones. 28 1.5 Discussion Different enclaves of the Kaap Valley Tonalite reveal the following. Some enclaves viewed in the field were inconspicuous due to a similar texture and composition as the host tonalite. They were only identified because they have a slightly higher proportion of hornblende. Such enclaves appear to be a more mafic, possibly an earlier phase of the tonalite, or they may have undergone an expulsion of felsic components due to compression prior to solidification (e.g., Barnes et al., 2021). These inconspicuous enclaves may be angular or circular. The angular variants may be fragments of earlier-crystallised layers of the tonalite, whereas the circular variants may represent unmixed globules or undercooled fractions of basaltic magma introduced to tonalite magma during magma recharge (e.g., Barnes et al. 2021 and references therein). However, the presence of brittle fractures even in rounded enclaves (e.g., Fig. 1.6 B) might indicate thermochemical or mechanical rounding of once-angular enclaves that were fractured prior to being incorporated into tonalite magma. Fracturing of the enclaves facilitated mineralogical equilibration with the tonalite magma. This is seen in the higher feldspar content in enclave margins compared to cores and in highly fractured sectors of deformed enclaves (Fig. 1.6 G-I). The occurrence of sub-angular enclaves of lower hornblende content within a few centimetres to similar-sized, subangular enclaves of a higher hornblende content (Fig. 1.6 I) suggests an origin of enclaves from different sources. According to Flinders and Clemens (1996), methods of enclave transportation may include magmatic flow that may facilitate clustering of enclaves from unrelated sources. Petrographic studies of sampled margins of enclaves and the host tonalite revealed the following. The only biotite observed occurred in samples NP-204A and B which are respective thin sections of sharp and gradational margins found at different boundaries of a non-uniform enclave (Fig. 1.6 A). These samples contained clusters of euhedral biotite that were aligned with hornblende cumulates at the tonalite-enclave margin in the tonalite sector of the sample. No biotite was found in the enclave sector of the sample (Fig. 1.8 B, C). According to Faure, (1989), this lack of biotite in the Kaap Valley Tonalite may indicate more advanced chloritisation of biotite in the fractured enclave. However, in the biotite-bearing thin sections studied (NP-204A and B), chloritised hornblende occurred adjacent to biotite that is generally not chloritised (Fig. 1.8 B, C), possibly indicating a composition that does not support chloritisation. Alternatively, the enclave may be a fragment of a less evolved, biotite-free, 29 earlier fraction of the tonalite enclosed in a younger, biotite-bearing, more evolved, shallow- level fraction of the tonalite as suggested by Robb et al. (1986). The chloritisation and formation of epidote from mafic components occurs throughout the tonalite-enclave samples. This alteration was found to be more advanced in (poorly) mineralised gold-bearing shear zones of the tonalite (Faure, 1989). Elemental maps show the association of sphene, pyrite (±chalcopyrite) and magnetite with hornblende, and presence of sodium and lesser calcium in albite, indicating a composition towards oligoclase. This albite-oligoclase content of the feldspars in tonalite is in agreement with previous descriptions by Robb et al. (1986) and Faure (1989). 1.6 Conclusion Mineralogical similarities between the enclave and host tonalite suggest mineralogical equilibration of angular mafic xenoliths with the tonalite and the partial digestion of early- crystallised phases of the tonalite. Microfaults and kink-folding of mafic and felsic minerals in the tonalite-side of the tonalite-enclave samples of this study (Fig. 1.8) indicates post- crystallisation brittle deformation. 30 Chapter 2: Verdite at Diggers’ Retreat 2.1 Site description Site 2 is a roadside stop outside a hotel and restaurant called Diggers’ Retreat (Fig. 2.1). It is situated near the intersection of the R38 Geotrail and the road to Consort Mine at 25°40'43.60"S; 31° 4'47.69"E Rocks to be seen here are in a cocopan container filled with green verdite in a grey-green host rock. Also in the container are samples of talc and talc-carbonate schist in serpentinite (see talc description in the Chapter on site 4). Both verdite and talc were transported with permission- from the Handsup Verdite Mine and the Scotia Talc Mine after the mines closed in 1999 and 2019 respectively. A lot of parking space is available around the cocopan on either side of the road. Figure 2.1. Main features of Site 2. (A): Simplified geology of the R38 road in the context of the Barberton Greenstone Belt. Maps modified after Visser et al. (1956) and Anhaeusser et al. (1981). (B): a detailed view of the area outlined in (A) shows the location of site 2 northeast of the Handsup Layered Ultramafic Complex (white star) where verdite was once mined. (C): View of the turn off to site 2. (D): Cocopan at site 2 filled with verdite samples from the Handsup Mine and talc samples from Scotia Talc Mine. 31 2.2 Aspects of verdite and Site 2 that may interest laypersons Verdite (a variety of fuchsite), buddstone (a mottled or layered green stone related to verdite that is sometimes called “South African Jade”) and other colourful semi-precious ornamental stones were mined in the Barberton area from alteration zones of ultramafic lithologies. On the way to site 2, waste rock of the Handsup Mine will be an obvious feature on the western side of the R38 road about 9 km north of Barberton. This old mine is one of many historical mines of the region where colourful minerals of chromium were extracted. The Handsup quarry is visible from the turn-off to Site 2 and from the backyard of Digger’s Retreat. Visitors could be told about barbertonite (Mg6Cr2(OH)16[CO3]·4H2O), a colourful chromium hydrotalcite aggregate named after the town of Barberton (Frondel, 1941). It was discovered in 1941 within the Barberton region, in a layered ultramafic host of the Kaapsehoop Asbestos Mine c. 30 km northwest of Barberton (Frondel, 1941). Barbertonite is also known as stichtite- 2H (Mills et al., 2011) as it is a polytype of, and commonly occurs with stichtite (Ashwal and Cairncross, 1997; Mills et al., 2011). It occurs as pink, lilac or purple masses within altered chrome-rich serpentinites (Ashwal and Cairncross, 1997; Mills et al., 2011. Figure 2.2 Hand specimen of barbertonite (stichtite-2H), a lilac-to-purple chromium hydrotalcite mineral. It is a polytype of, and usually co-occurs with Stichtite-3R (Mills et al., 2011). 32 Within Digger’s Retreat there is a display of described rock samples of selected local lithologies. They include banded ironstones of the Fig Tree Group that provided sub-economic iron oxide deposits which have been mined in the region since pre-colonial times (Fig 2.4, Myburgh, 1949). Figure 2.4 Local Swati women carrying red ochre from Dumaneni, a pre-colonial iron mine at the northeastern margin of the Barberton Greenstone Belt. Image sourced from Myburgh (1949). Fig. 2.3 An example of related geological features that occur at different scales. The structural fabric near the Handsup verdite quarry can be used to show visitors how features present clues used by geologists to make plausible reconstructions. 33 2.3 Geology 2.3.1 Background 2.3.1.1 Handsup Layered Ultramafic Complex Situated at the southeastern margin of the Jamestown Schist Belt (JSB), the Handsup Layered Ultramafic Complex occupies a tectonically complex zone affected by multiple deformation episodes. These include the regional shortening events of the BGB, possibly associated with the forceful shallow emplacement of the Stentor Pluton (c. 3250 Ma) in the northeast, the emplacement of the Kaap Valley Tonalite (c. 3227 Ma) in the south, and the intrusion of Nelspruit granites in the north (c. 3109 Ma) (Anhaeusser, 2019). The Handsup Layered Ultramafic Complex occurs as a sub-vertical anticline detached in a top- to-the-northwest sense along the Albion Fault (Anhaeusser, 1969, 2019). It is made up of cyclic compositional layering that may include a basal dunitic layer, overlain respectively by a harzburgitic, websteritic, gabbroic and marginally anorthositic gabbro layers. Unlike similar layered complexes at the northeastern and southwestern margins of the BGB, lithologies of the Handsup Layered Ultramafic Complex are generally advancely altered. Common alteration includes serpentinisation, carbonatisation, steatisation, uralitisation, chloritisation, saussuritisation, sericitisation, propylitisation and silicification (Anhaeusser, 1969, 2019). Layered ultramafic complexes of the BGB intruded only into sequences of the Onverwacht Group (Anhaeusser, 2006; Hofmann et al., 2021). In the northern margin of the BGB, they intrude volcanosedimentary sequences of the Weltevreden Formation which is diachronous with uppermost Onverwacht Group strata of the southern BGB (Hofmann et al., 2021). A Sm- Nd age of c. 3286 and Re-Os age of c. 3266 were determined by Lahaye et al. (1995) and Connolly et al. (2011) respectively. No direct age has been determined from the Handsup Layered Ultramafic Complex. However, Pb-Pb ratios of ultramafic and mafic lithologies from the adjacent Mundt’s Concession Layered Ultramafic Complex defined an age of c. 3244 Ma (Dupré and Arndt, 1990). More recently, a direct age of 3 244 ± 11 Ma was determined from a gabbroic unit of the same Layered Ultramafic Complex (Hofmann et al., 2021). If the Handsup and Mundt’s Concession Layered Ultramafic Complexes constitute a continuous layered sequence as proposed by Anhaeusser (1969), the abovementioned ages may be regarded as age estimates for lithologies of the Handsup Layered Ultramafic Complex. 34 2.3.1.2 Verdite Verdite is a variety of fuchsite schist found in altered serpentinites (Anhaeusser, 1963). According to Burns and Burns (1975) and Treloar (1987), the green colour in verdite results from a substitution of Cr3+ for Al3+ in octahedral sites of muscovite. Muscovite Cr3+ substitution into octahedral Al3+ site Fuchsite True verdite is described as a green semi-precious ornamental stone with a hardness range between 3 and 4, and a specific gravity range between 2.70 and 2.87 (Harding and Jobbins, 1984; Groeneveld, 1975). Compositionally, it is described as a schist composed mainly of fuchsite (chrome-muscovite) and varying proportions of rutile, quartz, albite, chlorite, talc, and corundum (Anhaeusser, 1969, 1975; Groeneveld, 1975; Antenen, 1991). The name ‘verdite’ in Wybergh (1932) and Anhaeusser (1963) was used to describe a soft, green, fine-grained, fuchsite-rich alteration product of mafic-to-ultramafic serpentinite. According to Antenen (1991), the name was derived from ‘verde’, the Spanish, Portuguese or Italian word for green. ‘Verde’ (Kerrich et al., 1987), verd antique or ‘African Jade’ (Hobbs, 1982) as well as serpentinite (Anhaeusser, 1984) are other terms which have been used in literature to describe verdite. Except for serpentinite/verd-antique, these various names for verdite all described a soft, green, fuchsite-rich schist with abundant rutile and varying proportions of quartz, feldspars and margarite, an Al-rich muscovite (Anhaeusser, 1963, 1975; Groeneveld, 1975; Harding and Jobbins, 1984). Verdite from the Barberton region has been known since 1907 before it was discovered in Zimbabwe where it is mined as a by-product of corundum production. Zimbabwean reserves have outlasted South African counterparts, the majority of which were confined to the northern margin of the Barberton Greenstone Belt (Groeneveld, 1975; Antenen, 1991). Best occurrences in the Barberton region, such as the one mined at the prominent Handsup quarry visible from the R38 road 12 km north of Barberton, were found within some multiply deformed and metasomatised Archean ultramafic sequences. According to Groeneveld (1975), three varieties of verdite were identified in quarries of the Barberton Greenstone Belt. The first type was pale green in colour, of a granular texture, compact and hard. This variety was found in a gradational contact with a white rock with the same texture. The second type was a dark green, foliated, much softer variety. The third, a 35 green and white brecciated variety with white fragments often drawn out into thin bands folded together with the verdite. The latter variety is a semi-precious decorative stone locally known as buddstone. The three variants were found to grade into each other (Groeneveld, 1975). Later work by Antenen (1991) identified sub-economic verdite deposits reported to be altered mafic xenoliths sheared together with the encompassing 3.1 Ga potassic Nelspruit Batholith. No paragenetic sequence or metamorphic grade was described due to the ultramylonitic fabric of samples studied by the same author. A number of workers have investigated the composition of verdite within the Barberton Greenstone Belt, the Mashishimala deposit in the Murchison Greenstone Belt to the north, as well as the ‘O’Briens Claims’ and the ‘Concessions’ deposits found in the 2.73 Ga Salisbury Greenstone Belt of Zimbabwe (Partridge, 1936; Schreyer et al., 1981; Kerrich et al., 1987; Antenen, 1991). In all of these deposits, all samples were dominated by chrome-muscovite with a whole-rock concentration of 0.2-4 wt % Cr (Table 1), >25wt. % alumina and >7wt. % potash with abundant rutile and lesser chromite and magnetite as main opaque mineral phases. Various lithological associations of verdite occurrences were reported in attempts to characterize verdite deposits. It was initially proposed that verdite occurrences were linked to Cr-metasomatism of felsic schists (Hutton, 1942). Partridge (1936) proposed a genetic-and sometimes-spatial link of verdite to granitic bodies (Partridge, 1936). In contrast, an association of verdite occurrences with sheared contacts between altered serpentinised pyroxenites and metagabbros were noted by Anhaeusser (1963, 1972). In support of the latter, Groeneveld (1975) reported a close association of talc-chlorite lenses and verdite deposits in shear zones near contacts of a serpentinised metagabbro. This linked verdite formation to intrusive ultramafic lithologies which have been multiply metasomatised, initially through serpentinisation, then steatisation to produce talc, and later K-, Al-, Cr and Si- metasomatic processes of “verditisation” under conditions that are only partially understood. At a later stage, Pearton (1981) proposed a sequence of alteration events whereby K-, Al-, Cr- and Si- metasomatism associated with verdite petrogenesis, was subsequent to the carbonitisation of mafic/ultramafic wallrocks. In this way, carbonates were understood to form part of the composition of verdite (Anhaeusser, 1969; Groeneveld, 1975; Antenen, 1991; Ward, 1999, Cairncross, 2004). In later years, more verdite occurrences were discovered in the area north of the BGB. These deposits were described as northwest-trending, sheared ultramafic xenoliths within the 36 Nelspruit Batholith (Antenen, 1991; Cairncross and Dixon, 2007). A total of three deposits occurring from ca. 5 km (Nelspruit quarry) up to 30 km north of the northern margin of the BGB were identified (Antenen, 1991). Figure 2.5 Background of the old Handsup Mine. (A) Polished verdite ornament sourced from the quarry. (B) View of the unrehabilitated verdite quarry visible from the R38 road situated within the Jamestown Schist Belt. (C) Modified geological map of the Handsup Layered Ultramafic Complex (Anhaeusser, 1975) superimposed on a Google Earth image. White rectangle mark shows that the Handsup verdite quarry is situated at a lithological boundary between a pyroxenitic serpentinite and a metagabbro associated with a pyroxenite. 37 2.3.2 Present study 2.3.2.1 Research question Verdite from the Barberton Greenstone Belt has been described by several workers, however, aspects of the mineralogy, occurrence and paragenetic model remain unclear. 2.3.2.2 Aims This work aims to characterise, and propose a paragenetic model of, the best quality verdite from the Barberton Greenstone Belt and from a widely unknown, possible extension of the Jamestown Schist Belt located in Nelspruit (Fig. 2.10). 2.3.2.3 Methods 2.3.2.3.1 Sampling The Handsup Layered Ultramafic Complex (LUC) was selected for sampling due to the occurrence of high quality verdite. Due to the rarity of in-situ verdite within the quarry, samples thereof were collected from the quarry dump. Care was taken to collect representative variants in order to link composition of samples to physical features observed. Verdite host rocks were sampled together with intrusive bodies proximal to the only remaining verdite-bearing alteration zone still in-situ within the quarry. Less altered equivalents of the host rocks were also sampled from outside the quarry along strike from those bearing verdite within the quarry. This was to identify assemblages only present in altered verdite-bearing host rocks. Additional samples of verdite and the host rocks were obtained from a different geological context which lies in Nelspruit at a location ca. 19 km north of the Handsup LUC verdite (Fig. 2.10). This was done to assess the validity of a petrogenetic and paragenetic model proposed for the Barberton verdite sampled from the Handsup LUC. 2.3.2.3.2 Petrography Labelled samples were cut into thin sections. Phase maps of thin sections and elemental maps of thin section off-cuts were produced to confirm petrographic observations made on fine grained, and highly altered samples of this study. 38 2.3.2.3.3 Bulk Chemical Analysis X-ray fluorescence spectrometry was used to determine major and trace element compositions of selected samples. 39 2.4 Results 2.4.1 Field Occurrence: Handsup quarry At the Handsup quarry, a NW-trending shear zone at the centre of the quarry intercepts north- northeast trending strata of a westerly dip. From the quarry entrance in the southeast to the only remaining in-situ verdite in the northwestern part of the quarry are asbestos-bearing serpentinites (Sample HUVH) that are interlayered with thin (<25cm) grey chert layers. Some of these serpentinites contain disseminated grey-white millimetre-scale masses, are multiply deformed and in some areas within the quarry, occur as thin shards known as “pencils” formed by the intersection of two cleavage planes. Verdite occurs in southwards-plunging lenses within a grey-brown, sub-vertically oriented chlorite-epidote schist (Fig. 2.6 B C, D). Disseminated within the schist are white, millimetre- scale ovoid masses of albite that either occur as individual masses or lenticular aggregates in northwest-plunging tension gashes. Sheared and altered equivalents of these are altered to sericite ± kaolinite possibly from recent weathering. Sub-vertical quartz-feldspar porphyritic dykes occur at both margins of the verdite-bearing chlorite schist and are the likely source of the sericite ± kaolinite in the tensional fractures of the chlorite-epidote schist. A few tens of centimetres northeast of the verdite lenses, the verdite-bearing sericite-chlorite-epidote schist and an altered, southwest-dipping quartz-feldspar porphyritic dyke are truncated by a harzburgite dyke. A thicker quartz-feldspar porphyry dyke occurs ca. 1m south of the verdite zone (Fig. 2.6 A) and is offset in a top-to-the southeast sense at the top of the verdite-bearing zone. In situ verdite observed invariably occurs as lenses between the chlorite-epidote schist and films of a powdery kaolinite-sericite schist (Fig 2.6 B, C). A few tens of centimetres towards the northeast of the verdite lenses, the verdite-bearing sericite-chlorite-epidote schist and an altered, southwest-dipping quartz-feldspar porphyritic dyke are truncated by a harzburgite dyke. A thicker quartz-feldspar porphyry dyke occurs ca. 1m south of the verdite zone (Fig 2.6 A, B, sample CDWV) and appears to be offset in a top-to-the southeast sense at the top of the verdite-bearing zone. At the northeastern edge of the verdite-bearing zone is a layer-parallel zone of carbonated serpentinite (sample HUT). However, no obvious zones of pervasive carbonate development were observed in the quarry or vicinity. 40 2.4.2 Petrography: Handsup quarry The chlorite schist host rock of verdite (HUTH/CHWVH) is comprised of chlorite groundmass/oikocrysts that envelop kinked, ovoid albite masses that sometimes exhibit a Maltese Cross extinction (bottom right part of Fig. 2.7 A). The chlorite crystals are partially replaced by muscovite. Fine aggregates of chromite are disseminated throughout the sample, occurring as inclusions in albite crystals as well as in the muscovite-rich groundmass (Fig. 2.7 B and E). Tensional fractures of different generations cut through the schist. The earliest is a feldspar-rich filled phase now occurring as muscovite-rich lineaments at a slight angle to Figure 2.6 Lithological context of the remaining in-situ verdite pods in the Handsup Quarry. (A): Verdite lenses in a sub-vertical array of southwest-plunging tension gashes. Labels refer to sample names except “ALT” which represents sericitised and kaolinitised feldspar porphyry. (B-C): The verdite lenses occur within earlier sub-vertical tension gashes that plunge towards the northwest. (D): Illustrated diagrams show the development of tension gashes mentioned above. The deformation sequence is based on features of the intrusive quartz-feldspar porphyritic dyke (CDWV) and the orientation of tension gashes. See Appendix B, Figure 1. 41 schistosity (centre of Fig. 2.7 A). The second generation is characterised by spinel-bearing fractures that offset the layer-parallel felsic veins. The latter are kinked sub-horizontally in a top-to-the northwest shear sense (Fig. 2.7 B-C). 42 Though verdite from Zimbabwe and the Limpopo Province of South Africa may contain red spots due to the presence of red rubies (Schreyer et al., 1981), the colours of Handsup verdite are mainly different shades of green, white, and yellow. These colourful layers are respectively made up of chrome muscovite, stringers of albite ovoid masses or lenses, and clusters of euhedral rutile crystals associated with sericitised albite layers. The chrome-muscovite which dominates the samples studied occurs as radial or cross-shaped aggregates with no preferred orientation, but where deformed, strong mineralogical layering characterizes the samples. Three main textures that were observed are described below and shown in Fig. 2.8. Type 1 A sample of this variant is the closest to true verdite (NP-004). It is the softest sample and is vividly green. It consists of interlaminations of distinct, fine-textured mineral layers. The thickest and dominant layers are composed of chrome-muscovite intergrown with chlorite. Thinner layers include very fine sericite-feldspar-quartz layers and layers of rutile aggregates responsible for the characteristic yellow streaks in verdite hand specimens. Very rarely are albite masses and lenses preserved in this variant. The few observed have been heavily sericitised or resorbed into the micaceous groundmass and are better preserved within microfold hinges and along boundaries of mineral layers (Fig 2.8, H-I). Late-stage veinlets of albite cut across mineral layers. Opaque minerals were too fine to identify. Type 2 Samples of this variant are dominated by lenticular aggregates of albite with chrome-muscovite rims separated by micro-shear zones filled with coarse, interlocking muscovite crystals. These Figure 2.7 Characteristics of the chlorite-muscovite-albite schist host of verdite in the Handsup quarry. Sample names are in bottom-left corners and PPL, XPL and RL labels are in photomicrographs produced using plane-, cross-polarised or reflected light, respectively. (A): Subvertical muscovite-rich lenses with albite crystals at kinks of an opaque vein. (B-C): Photomicrographs of disseminated clusters of chromite in a micaceous groundmass cut by a vein kinked along the orientation of mica fish. Magnetite crystals occur in dilational jogs of the kinked vein. (D): Syntaxial feldspar vein kinked normal to vein opening is cut obliquely by a thin opaque vein. (E-F): Bladed chromite crystals growing at margins of advancely altered ovoid crystals of albite. (G): Steatised and carbonated serpentine in the flanks of the verdite- bearing zone. (H): Advancely sericitised feldspar porphyry. Cr-Ms: chrome-muscovite, Ms: muscovite, Chl: chlorite, Ab: albite, Spl: spinel (chromite/magnetite), Srp: serpentine, Tlc: talc, Mgs: magnesite, Ser: sericite, Fsp: feldspar, Rt: rutile. 43 lenticular aggregates of albite are set within chlorite partially replaced by randomly-oriented muscovite (Fig. 2.8.C-E). In sheared samples, the felsic lenses are subdivided into shear bands with margins lined by sillimanite (Fig. 2.8 I). No sillimanite was found in the chrome- muscovite zone which lies in a transitional contact with the sillimanite-bearing albite lenses. Aggregates of euhedral rutile occupy fractures and cleavage planes of albite. Clusters of chromite and magnetite are disseminated in the fine-grained, vividly green chrome-muscovite groundmass and show no preferred orientation except when sheared into alignment with the schistosity of the sample (Fig. 2.8 C). Unsheared clusters of these oxides are associated with pale-coloured muscovite (Fig. 2.8 C). Figure 2.8 Mineral composition of the verdite-bearing samples and host rocks. Sample names are in the top right corners and PPL and XPL labels are in photomicrographs produced using plane-polarised or cross-polarised light, respectively. (A-B): Folded mineral layers in true verdite (Type 1). (C): Layered sample of incomplete verdite-formation shows more chrome- muscovite in a sheared sector of the same fuchsitic schist. (D-F): Lens-bearing verdite (Type 2) at the interface of a chloritised serpentinite and an intrusive, porphyritic albite ± quartz dyke with sillimanite. (F). (G-I): Chloritised serpentinite in-filled by muscovite along tension fractures. This sample contains chrome-muscovite at margins of albite crystals that occur within lenses filled by porphyritic albite ± quartz fluid. (I). Ab: albite, Spl: spinel (chromite/magnetite), Cr-Ms: chrome-muscovite, Ep: Epidote, Srp-Chl: chloritised serpentine, Sil: sillimanite, Chl: chlorite, Rt: rutile. 44 2.4.3 Field Occurrence: Nelspruit mafic ridge Additional samples of verdite and host rocks (Fig. 2.9) were collected in Nelspruit (Fig. 2.10) for comparison. At this locality, verdite lenses within quarry A in Nelspruit occur within a mafic ridge, in a schistose foliation that is strike-parallel to a northwest-trending shear zone. The shear zone cuts through a sub-vertical chlorite schist with interfolial veins of white quartz either encrusted with goethite or bearing prismatic voids lined with hematite. The chlorite schist in the southwestern quarry wall is folded sub-horizontally and the fold amplitudes become larger towards the southeast (f1 in Fig. 2.11 A-D). The larger structure of the folded quarry wall is here proposed to be a southeast-plunging antiformal syncline with z- folds forming the southeastern wall of the quarry. A proposed interpretative model (Fig. 2.11 D) depicts the proposed structural context of verdite in quarries A and B where verdite was once mined. Quartz ± feldspar veins intrude verdite in both quarry A and B. In quarry A, verdite occurs between an altered leucogranite and the chlorite schist. The contact is part of a sub-vertical, top-to-the-southeast, transpressional shear zone that runs parallel to the schistose foliation of the chlorite schist. The altered leucogranite is slightly sheared (Fig. 2.9 A) and marks the northwestern margin of the quarry. At the southeastern edge, the same chlorite schist is bound to the south by a northwest-trending fault zone occupied by a cohesive fault rock (Fig. 2.9 G) which is in turn bound to the west by boulders of massive cryptocrystalline white quartz with voids (quartz-feldspar porphyry?) that occupies at least 100 m2 of grassland. No contact was observed between the white quartz and the adjacent verdite zone or with the closest potassic granite which lies c. 100 m southwest of the quarry. Texturally, the fault rock at the southern extent of quarry A contains sub-rounded to sub- angular, partially digested, boulders and cobbles comprised of alternating layers of white quartz ± feldspar and pale green verdite (Fig. 2.9 F). These clasts represent a semi-precious carving stone known locally as buddstone (Anhaeusser, 1969, 2013). Buddstone clasts in the fault rock are aligned with the northwest-trend of the verdite-bearing shear zone in the quarry. A similar lithology was sampled at quarry B where advanced silicification and compression had not occurred (Fig 2.9. G). The latter could be separated into the verdite matrix (NBN4A) and powdery cobbles of sericite-rich nodules in a verdite groundmass (sample NBN4B/C). These 45 cohesive faultrocks at Quarry B only occur as boulders close to sparse outcrops of the mafic ridge which, at that locality, occurs as sparse, metre-scale massive bodies of pale to dark green chrome muscovite. The latter are invaded by veins of pinkish, euhedral quartz veins with prismatic voids (Fig. 2.9 E). Figure 2.9 Lithologies in Nelspruit verdite quarry A and B (images E and G) labelled with sample names. Fuchsite is used here to describe siliceous verdite. (A): Fuchsite at the interface of the chlorite schist verdite host and a massive, foliation-parallel quartz-feldspar granitoid. (B): Layer-parallel quartz vein with prismatic voids (e.g. yellow star) lined with iron oxides. (C): Sub-vertical verdite sample. (D): One of the common layer-parallel felsic veins interlayered with verdite in the layer-parallel shear zone. (E): Sample of chlorite schist with incipient verdite formation (green in yellow circle) near veins of euhedral K-feldspar and quartz. (F): An apparent clast of banded fuchsite and quartz ± feldspar (buddstone) in a matrix of dark fuchsite schist. (G): Unsilicified angular clasts of massive fuchsite and sericite aggregates in a dark green fuchsite schist. 46 Figure. 2.10 Widely unknown verdite deposits 6 km south of Nelspruit (Quarry A: 26 32 30.36 S; 30 57 52.50 E. Quarry B: 25 32 21.30 S; 30 57 46.26 E). (A): Simplified map of the Barberton Granite Greenstone Belt with two yellow circles marking the context of verdite deposits studied. The southern circle surrounds the Handsup Layered Ultramafic Complex of the BGB, and the northern circle marks the location of a verdite-bearing mafic ridge (quarries A and B) of Onverwacht Group affinity (B): Two verdite deposits sampled during the present study (quarries A and B) are separated by apparent dextral displacement. (C) Wider context of the verdite-bearing ridge marked with locations of verdite quarries reveals a similarly-trending ridge to the east. 47 Figure 2.11 Nelspruit verdite quarry A. (A): Verdite lenses (green ovals) occur within a foliation- and strike-parallel northwest-trending shear zone cutting through a chlorite schist. (B): Close up view of metre-scale fold in the quarry wall (f1) in image (A). (C): Close-up image of left section of image (A) shows the merging of f1 folds (ii) into a larger fold (i) towards the right edge of the image. (D): Z-folds on the eastern margin of quarry may (Images A-C) suggest that the verdite-bearing shear zone occurs in the eastern limb of a northwest- trending antiform. 48 2.4.4 Petrography: Nelspruit mafic ridge The main difference between the host rocks to the verdite in Handsup Quarry and Nelspruit quarries is the higher iron content in the Nelspruit chlorite schist which also contains a higher proportion of iron oxides and hydroxides at contacts with invading quartz veins. The Nelspruit verdite samples are also dominated by chrome-muscovite and an increasing proportion of preserved subhedral albite from quarry A to quarry B. Albite in this context is mostly sheared into layers. Where unsheared and preserved as distinct crystals, advanced alteration to muscovite is predominant (Fig. 2.12, C-E). The alteration of feldspar and the resorption of andalusite porphyroblasts into a groundmass of muscovite and partially sericitised feldspar appear to be the source of muscovite in the samples. Unlike samples from Handsup, the alteration of chlorite by muscovite-precipitating fluids was not observed in the Nelspruit verdite samples as all samples were completely composed of muscovite-after-feldspar interlayered with or surrounded by chrome-muscovite. As observed in Handsup verdite, euhedral rutile also occurs in association with rarely preserved albite. Figure 2.12 Main minerals related to the formation of Nelspruit verdite from quarry A and B. Sample numbers at top right corner PPL and XPL labels are in photomicrographs produced using plane-polarised or cross-polarised light, respectively. (A): Photomicrographs of the Fe- rich chlorite schist verdite host. (B): Dextrally sheared verdite with rarely-preserved feldspar crystals (insert). Quarry B samples show that the angular clasts in a fuchsite matrix (Fig. 2.9 G) are comprised of nodular zones of muscovite aggregates (red stars) in a chrome-muscovite groundmass (C); and that resorbed andalusite occurs in some fuchsite schists of that quarry (D). Fe-OH: iron hydroxides, Chl: chlorite, And: andalusite, Ms: muscovite, Spl: spinel (chromite/magnetite), Cr-Ms: chrome-muscovite, Rt: